2022-annual-markets-report.pdf

ISO-NE PUBLIC

2022 Annual Markets Report

Internal Market Monitor

JUNE 5, 2023

2022 Annual Markets Report page ii

ISO-NE PUBLIC

Document Revision History

Date

Version

Remarks

6/5/2023

Original

Initial Posting

2022 Annual Markets Report page iii

ISO-NE PUBLIC

Preface/Disclaimer

The Internal Market Monitor (IMM) of ISO New England (ISO) publishes an Annual Markets

Report (AMR) that assesses the state of competition in the wholesale electricity markets

operated by the ISO. The 2022 Annual Markets Report covers the ISO’s most recent operating

year, January 1 to December 31, 2022. The report addresses the development, operation, and

performance of the wholesale electricity markets administered by the ISO and presents an

assessment of each market based on market data, performance criteria, and independent

studies.

This report fulfills the requirement of Market Rule 1, Appendix A, Section III.A.17.2.4, Market

Monitoring, Reporting, and Market Power Mitigation:

The Internal Market Monitor will prepare an annual state of the market report on market trends and the

performance of the New England Markets and will present an annual review of the operations of the New

England Markets. The annual report and review will include an evaluation of the procedures for the

determination of energy, reserve and regulation clearing prices, Net Commitment-Period Compensation costs

and the performance of the Forward Capacity Market and Financial Transmission Rights Auctions. The review

will include a public forum to discuss the performance of the New England Markets, the state of competition,

and the ISO’s priorities for the coming year. In addition, the Internal Market Monitor will arrange a non-public

meeting open to appropriate state or federal government agencies, including the Commission and state

regulatory bodies, attorneys general, and others with jurisdiction over the competitive operation of electric

power markets, subject to the confidentiality protections of the ISO New England Information Policy, to the

greatest extent permitted by law.

This report is being submitted simultaneously to the ISO and the Federal Energy Regulatory

Commission (FERC) per FERC order:

The Commission has the statutory responsibility to ensure that public utilities selling in competitive bulk power

markets do not engage in market power abuse and also to ensure that markets within the Commission’s

jurisdiction are free of design flaws and market power abuse. To that end, the Commission will expect to

receive the reports and analyses of a Regional Transmission Organization’s market monitor at the same time

they are submitted to the RTO.

This report presents the most important findings, market outcomes, and market design

changes of New England’s wholesale electricity markets for 2022. The executive summary gives

an overview of the region’s wholesale electricity market outcomes, the important market issues

and our recommendations for addressing these issues. It also addresses the overall

competitiveness of the markets, and market mitigation and market reform activities. Sections 1

through Section 8 include more detailed discussions of each of the markets, market results,

analysis and recommendations. A list of acronyms and abbreviations is included at the back of

the report.

ISO New England Inc. Transmission, Markets, and Services Tariff (ISO tariff), Section III.A.17.2.4, Market Rule 1, Appendix A,

“Market Monitoring, Reporting, and Market Power Mitigation”, http://www.iso-ne.com/static-

assets/documents/regulatory/tariff/sect_3/mr1_append_a.pdf.

FERC, PJM Interconnection, L.L.C. et al., Order Provisionally Granting RTO Status, Docket No. RT01-2-000, 96 FERC ¶ 61, 061

(July 12, 2001).

2022 Annual Markets Report page iv

ISO-NE PUBLIC

A number of external and internal audits are also conducted each year to ensure that the ISO

followed the approved market rules and procedures and to provide transparency to New

England stakeholders. Further details of these audits can be found on the ISO website.

All information and data presented are the most recent as of the time of writing. The data

presented in this report are not intended to be of settlement quality and some of the underlying

data used are subject to resettlement.

In case of a discrepancy between this report and the ISO New England Tariff or Procedures, the

meaning of the Tariff and Procedures shall govern.

Underlying natural gas data are furnished by the Intercontinental Exchange (ICE):

Underlying oil and coal pricing data are furnished by Argus Media.

See https://www.iso-ne.com/about/corporate-governance/financial-performance

 
Annual Markets Report    page ii 
  ISO-NE PUBLIC 
Contents 
Preface/Disclaimer......................................................................................................................................................................... iii 
Contents ............................................................................................................................................................................................ ii 
Figures ............................................................................................................................................................................................... ii 
Tables................................................................................................................................................................................................. v 
Executive Summary ........................................................................................................................................................................ 6 
IMM Market Enhancement Recommendations ..................................................................................................................13 
Section 1 Overall Market Conditions ........................................................................................................................................18 
1 Wholesale Cost of Electricity ............................................................................................................................................18 
2 Supply Conditions ...............................................................................................................................................................20 
2.1 Generation and Capacity Mix ..................................................................................................................................20 
2.2 Generation Fuel and Emissions Costs ....................................................................................................................27 
2.3 Generator Profitability ..............................................................................................................................................31 
3 Demand Conditions............................................................................................................................................................35 
3.1 Energy Demand ..........................................................................................................................................................35 
3.2 Reserve Requirement and Margin ..........................................................................................................................38 
3.3 Capacity Market Requirements...............................................................................................................................40 
4 Imports and Exports (External Transactions) ................................................................................................................42 
Section 2 Market Structure and Competitiveness Assessment  ..........................................................................................44 
1 Energy Market.....................................................................................................................................................................45 
1.1 C4 Concentration Ratio for Generation .................................................................................................................45 
1.2 C4 Concentration Ratio for Load .............................................................................................................................46 
1.3 Residual Supply Index and the Pivotal Supplier Test...........................................................................................47 
1.4 Day-Ahead Price-Cost Markup ................................................................................................................................49 
1.5 Real-Time Price-Cost Markup (Lerner Index)........................................................................................................51 
1.6 Real-Time Economic Withholding...........................................................................................................................52 
2 Energy Market Mitigation .................................................................................................................................................53 
2.1 Mitigation Outcomes and Drivers...........................................................................................................................54 
2.2 Energy Market Mitigation Assessment ..................................................................................................................58 
3 Forward Capacity Market..................................................................................................................................................63 
4 Forward Capacity Market Mitigation ..............................................................................................................................66 
4.1 Supplier-Side Market Power  ....................................................................................................................................66 
4.2 Buyer-side Market Power (Minimum Offer Price Rule) ......................................................................................69 
5 Financial Transmission Rights Market ............................................................................................................................70 
6 Ancillary Services ................................................................................................................................................................72 

 
Annual Markets Report    page iii 
  ISO-NE PUBLIC 
6.1 Forward Reserve Market ..........................................................................................................................................72 
6.2 Regulation Market .....................................................................................................................................................74 
Section 3 Day-Ahead and Real-Time Energy Market .............................................................................................................76 
1 Energy Prices .......................................................................................................................................................................76 
1.1 Day-Ahead and Real-Time Energy Prices ...............................................................................................................77 
1.2 Fast-Start Pricing: Impact on Real-Time Outcomes .............................................................................................81 
1.3 Energy Price Convergence........................................................................................................................................85 
2 Drivers of Energy Market Outcomes...............................................................................................................................88 
2.1 Generation Costs........................................................................................................................................................89 
2.2 Supply-Side Participation..........................................................................................................................................96 
2.3 Reserve Adequacy Analysis Commitments ...........................................................................................................98 
2.4 Load and Weather Conditions .............................................................................................................................. 100 
2.5 Demand Bidding...................................................................................................................................................... 103 
2.6 Load Forecast and Market Implications.............................................................................................................. 105 
2.7 System Events During 2022................................................................................................................................... 109 
2.8 Reliability Commitments and Posturing ............................................................................................................. 118 
2.9 Transmission Congestion....................................................................................................................................... 122 
2.10 Marginal Resources .............................................................................................................................................. 125 
3 Net Commitment Period Compensation (Uplift) ....................................................................................................... 127 
4 Demand Response Resources ....................................................................................................................................... 131 
4.1 Energy Market Offers and Dispatch under PRD ................................................................................................ 131 
4.2 Energy Market Compensation under PRD.......................................................................................................... 133 
Section 4 Virtual Transactions and Financial Transmission Rights .................................................................................. 135 
1 Virtual Transactions ........................................................................................................................................................ 135 
1.1 Virtual Transaction Volume .................................................................................................................................. 136 
1.2 Virtual Transaction Profitability ........................................................................................................................... 137 
1.3 Virtual Transactions and Price Convergence ..................................................................................................... 141 
2 Financial Transmission Rights........................................................................................................................................ 142 
2.1 FTR Volume .............................................................................................................................................................. 142 
2.2 FTR Funding.............................................................................................................................................................. 144 
2.3 FTR Profitability ....................................................................................................................................................... 145 
Section 5 External Transactions .............................................................................................................................................. 149 
1 External Transactions...................................................................................................................................................... 150 
1.1 External Transaction Volumes .............................................................................................................................. 150 
1.2 External Transaction Offer Composition ............................................................................................................ 154 

 
Annual Markets Report    page iv 
  ISO-NE PUBLIC 
1.3 External Transaction Uplift (NCPC) Payments  ................................................................................................... 156 
2 Coordinated Transaction Scheduling ........................................................................................................................... 158 
2.1 CTS Performance..................................................................................................................................................... 158 
2.2 CTS and ISO Scheduling Efficiency ....................................................................................................................... 162 
2.3 CTS and Participant Bidding Opportunities ........................................................................................................ 165 
Section 6 Forward Capacity Market ....................................................................................................................................... 168 
1 Review of Seventeenth Forward Capacity Auction (FCA)......................................................................................... 168 
1.1 Qualified and Cleared Capacity ............................................................................................................................ 169 
1.2 Results and Competitiveness................................................................................................................................ 170 
1.3 Results of the Substitution Auction (CASPR)...................................................................................................... 171 
2 Forward Capacity Market Outcomes  ........................................................................................................................... 172 
2.1 Capacity Resource Mix ........................................................................................................................................... 176 
2.2 Pay-for-Performance Outcomes .......................................................................................................................... 180 
2.3 Secondary Forward Capacity Market Results .................................................................................................... 181 
Section 7 Ancillary Services...................................................................................................................................................... 183 
1 Real-Time Operating Reserves ...................................................................................................................................... 184 
1.1 Real-Time Operating Reserve Payments ............................................................................................................ 185 
1.2 Real-Time Operating Reserve Prices and Frequency........................................................................................ 187 
2 Forward Reserves ............................................................................................................................................................ 191 
2.1 Market Requirements ............................................................................................................................................ 191 
2.2 Auction Results........................................................................................................................................................ 194 
2.3 FRM Payments......................................................................................................................................................... 195 
3 Regulation ......................................................................................................................................................................... 197 
3.1 Regulation Prices..................................................................................................................................................... 197 
3.2 Regulation Payments ............................................................................................................................................. 198 
3.3 Regulation Requirements, Resource Mix, and Performance .......................................................................... 199 
Section 8 Market Design or Rule Changes ............................................................................................................................ 202 
1 Major Design Changes Recently Implemented .......................................................................................................... 202 
1.1 Transmission Cost Allocation to Network Customers with Behind-the-Meter Generation ...................... 202 
1.2 Competitive Transmission Solicitation Enhancements .................................................................................... 203 
1.3 Extended-Term Transmission Planning Tariff Changes.................................................................................... 203 
2 Major Design or Rule Changes in Development or Implementation for Futur e Years ....................................... 204 
2.1 Incorporate Solar into Do-Not-Exceed Dispatch ............................................................................................... 204 
2.2 Inventoried Energy Program................................................................................................................................. 204 
2.3 FERC Order 2222, Distributed Energy Resources .............................................................................................. 205 

 
Annual Markets Report    page v 
  ISO-NE PUBLIC 
2.4 Competitive Capacity Markets without a Minimum Offer Price Rule (MOPR)............................................ 206 
2.5 Resource Capacity Accreditation (RCA) in the Forward Capacity Market (FCM)  ........................................ 206 
2.6 Day-Ahead Ancillary Services Initiative............................................................................................................... 207 
3 Additional Notable Studies ............................................................................................................................................ 207 
3.1 New England's Future Grid Initiative................................................................................................................... 207 
Acronyms and Abbreviations .................................................................................................................................................. 209 
 
 

 
2022 Annual Markets Report    page ii 
  ISO-NE PUBLIC 
Figures 
Figure 1-1: Wholesale Costs ($ billions and $/MWh) and Average Natural Gas Prices ....................................................18 
Figure 1-2: Average Output and Share of Native Electricity Generation by Fuel Type .....................................................21 
Figure 1-3: Average Capacity by Fuel Type ...............................................................................................................................22 
Figure 1-4: Capacity Factor by Fuel Type...................................................................................................................................23 
Figure 1-5: Average Native Electricity Generation and Load by State .................................................................................24 
Figure 1-6: Average Age of New England Generator Capacity by Fuel Type (2018 - 2022) .............................................25 
Figure 1-7: Generator Additions, Retirements, and FCM Outcomes....................................................................................26 
Figure 1-8: Average Fuel Prices by Quarter and Year..............................................................................................................27 
Figure 1-9: New England vs. Henry Hub and Marcellus Natural Gas Prices ........................................................................28 
Figure 1-10: Annual Estimated Average Costs of Generation and Emissions .....................................................................31 
Figure 1-11: Estimated Net Revenue for New Gas -fired Generators ...................................................................................33 
Figure 1-12: Average Hourly Load by Quarter and Year .........................................................................................................36 
Figure 1-13: Load Duration Curves .............................................................................................................................................37 
Figure 1-14: Average Quarterly Weather-Normalized Load with Energy Efficiency and Solar Impacts ........................38 
Figure 1-15: Average System Reserve Requirements  .............................................................................................................39 
Figure 1-16: System Reserve Margin, Peak Load, and Available Capacity...........................................................................40 
Figure 1-17: ICR, NICR, Local Sourcing Requirements, and Maximum Capacity Limits.....................................................41 
Figure 1-18: Average Hourly Real-Time Pool Net Interchange by Quarter and Year ........................................................42 
Figure 2-1: Real-time System-wide Supply Shares of the Four Largest Firms.....................................................................46 
Figure 2-2: Real-time System-wide Demand Shares of the Four Largest Firms .................................................................47 
Figure 2-3: System-wide Residual Supply Index Duration Curves.........................................................................................49 
Figure 2-4: Hourly Real-time Economic Withholding During On-Peak Hours .....................................................................53 
Figure 2-5: Energy Market Mitigation ........................................................................................................................................55 
Figure 2-6: Potential Economic Withholding, General Threshold Energy ...........................................................................59 
Figure 2-7:  Supply Controlled by Pivotal Suppliers on June 13, 2022, During HE 18 (Interval 2) ...................................60 
Figure 2-8: Generator Capacity Lacking a CSO by Asset Type................................................................................................61 
Figure 2-9: LMP-Based and Marginal Cost Reference Prices .................................................................................................62 
Figure 2-10: Capacity Market Residual Supply Index, by FCA and Zone ..............................................................................64 
Figure 2-11: Overview of Suppliers, Pivotal Supplier, and Capacity Margin, by Zone.......................................................65 
Figure 2-12: General Static De-list Bid Summary Statistics, by Key Milestone Action (FCAs 13 – 17)............................68 
Figure 2-13: Reviewable Offer Request Summary Statistics, by Key Milestone Action (FCAs 13 – 17) .........................69 
Figure 2-14: Average FTR MWs Held per Hour by Top Four FTR Holders by Year and Period.........................................71 
Figure 2-15: Average Regulation Market Requirement and Available Capacity, 2022 .....................................................74 
Figure 2-16: Average Regulation Requirement and Residual Supply Index ........................................................................75 
Figure 3-1: Annual Simple Average Hub Price ..........................................................................................................................78 
Figure 3-2: Simple-Average Hub and Load Zone Prices, 2022 ...............................................................................................79 
Figure 3-3: Load-Weighted and Simple Average Hub Prices, 2022.......................................................................................80 
Figure 3-4: Day-Ahead Load-Weighted Prices ..........................................................................................................................81 
Figure 3-5: Reserve Prices in the Pricing vs. Dispatch Software on December 12, 2022..................................................85 
Figure 3-6: Average Annual Day-Ahead Price Premium at the Hub and Average Day-Ahead Hub LMP ........................86 
Figure 3-7: Average Hourly Day-Ahead to Real-Time Hub Price Differences, 2022...........................................................87 
Figure 3-8: Estimated Generation Costs and On-Peak LMPs .................................................................................................89 
Figure 3-9: Annual Average Natural Gas Price-Adjusted LMPs  .............................................................................................91 
Figure 3-10: Estimated Average  Cost of RGGI CO
2
 Allowances and Contribution of Emissions to Energy Production 
Costs .................................................................................................................................................................................................92 
Figure 3-11: Contributions of Emissions Cost to Energy Production Costs .........................................................................94 
Figure 3-12: Massachusetts GWSA Auction Results ................................................................................................................95 

 
2022 Annual Markets Report    page iii 
  ISO-NE PUBLIC 
Figure 3-13: Day-Ahead and Real-Time Supply Breakdown by Hour Ending, 2022 ...........................................................97 
Figure 3-14: Priced and Unpriced Supply vs. Real-Time LMP, March 16, 2022 ..................................................................98 
Figure 3-15: Average RAA Generator Commitments and the Day-Ahead Energy Gap .....................................................99 
Figure 3-16: Average Demand and LMP by Hour, 2022....................................................................................................... 101 
Figure 3-17: Seasonal vs. Five-Year Average Temperatures ............................................................................................... 102 
Figure 3-18: Day-Ahead Cleared Demand as a Percentage of Real-Time Load by Bid Type ......................................... 104 
Figure 3-19: Day-Ahead Cleared Demand as Percentage of Real -Time Load by Bid Type by hour, 2022 .................. 104 
Figure 3-20: Components of Day-Ahead Cleared Demand as a Percentage of Total Day-Ahead Cleared Demand  . 105 
Figure 3-21: ISO Day-Ahead Load Forecast Error by Time of Year..................................................................................... 107 
Figure 3-22: Impact of BTM Solar on Load Forecast Error .................................................................................................. 108 
Figure 3-23: Price Difference and Forecast Error Relationship .......................................................................................... 109 
Figure 3-24: Pricing, Demand and the Reserve Margin during System Events in 2022 ................................................. 110 
Figure 3-25: LMP Duration Curves for Top 1% of Real-Time Pricing Hours...................................................................... 113 
Figure 3-26: Actual vs. Forecasted Load on May 21 ............................................................................................................. 114 
Figure 3-27: Real-Time Prices on May 21, 2022 .................................................................................................................... 114 
Figure 3-28: December 24 Event Timeline ............................................................................................................................. 115 
Figure 3-29: Day-Ahead vs. Real-Time Net Import Deviations by Interface..................................................................... 116 
Figure 3-30: Differences between Hourly Real-time and Day-Ahead Generation Obligations .................................... 117 
Figure 3-31: Average Hourly Energy Output from Reliability Commitments, Peak Load Hours................................... 119 
Figure 3-32: Day-Ahead and Real-Time Average Out-of-Rate  Energy from Reliability Commitments, Peak Load 
Hours, 2022.................................................................................................................................................................................. 120 
Figure 3-33 Monthly Postured Energy and NCPC Payments............................................................................................... 121 
Figure 3-34: Average Day-Ahead Hub LMP, Congestion Revenue Totals and as Percent of Total Energy Cost ........ 123 
Figure 3-35: New England Pricing Nodes Most Affected by Congestion, 2022 ............................................................... 124 
Figure 3-36: Day-Ahead Marginal Resource by Transaction Type ..................................................................................... 126 
Figure 3-37: Real-Time Marginal Resource by Transaction Type ....................................................................................... 127 
Figure 3-38: Total Uplift Payments by Year and Category................................................................................................... 129 
Figure 3-39: Economic Uplift by Sub-Category ...................................................................................................................... 130 
Figure 3-40: Demand Response Resource Offers in the Real-Time Energy Market ....................................................... 132 
Figure 3-41: Demand Response Resource Dispatch in the Real -Time Energy Market ................................................... 133 
Figure 3-42: Energy Market Payments to Demand Response Resources......................................................................... 134 
Figure 4-1: Average Hourly Submitted and Cleared Virtual Transaction Volumes by Time of Day, 2022  .................. 136 
Figure 4-2: Average Annual Gross and Net Profits for Virtual Transactions .................................................................... 138 
Figure 4-3: Virtual Transaction Volumes and Price Convergence ...................................................................................... 141 
Figure 4-4: Average FTR MWs in Effect per Hour by Year ................................................................................................... 143 
Figure 4-5: FTR Funding and Congestion Revenue Fund Components by Year ............................................................... 144 
Figure 4-6: FTR Costs, Revenues, and Profits......................................................................................................................... 146 
Figure 4-7: FTR Profits and Costs for FTRs Sourcing from Roseton  ................................................................................... 148 
Figure 5-1: Hourly Average Day-Ahead and Real-Time Pool Net Interchange ................................................................ 151 
Figure 5-2: Real-Time Net Interchange at Canadian Interfaces .......................................................................................... 152 
Figure 5-3: Real-Time Net Interchange at New York Interfaces ......................................................................................... 153 
Figure 5-4: Transaction Types by Market and Direction at Canadian Interfaces (Average MW per hour) ................ 154 
Figure 5-5: Transaction Types by Market and Direction at New York Interfaces (Average MW per hour) ................ 155 
Figure 5-6: CTS Outcome Summary ......................................................................................................................................... 160 
Figure 5-7: CTS Unused Capacity, 2022 .................................................................................................................................. 162 
Figure 5-8: Price Sensitivity of Offered CTS Transactions .................................................................................................... 166 
Figure 6-1: Qualified and Cleared Capacity in FCA 17 .......................................................................................................... 169 
Figure 6-2: System-wide FCA 17 Demand Curve, Prices, and Quantities ......................................................................... 170 
Figure 6-3: FCA Clearing Prices in the Context of Market Rule Changes  .......................................................................... 173 

 
2022 Annual Markets Report    page iv 
  ISO-NE PUBLIC 
Figure 6-4: Forward Capacity Auction Clearing Prices ......................................................................................................... 174 
Figure 6-5: FCM Payments and FCA Volumes by Commitment Period............................................................................. 175 
Figure 6-6: Capacity Mix by Fuel Type..................................................................................................................................... 176 
Figure 6-7: New Generation Capacity by Fuel Type.............................................................................................................. 179 
Figure 6-8: FCA Import De-list MW.......................................................................................................................................... 180 
Figure 6-9: Total PfP Payments by Resource Type (December 2022) ............................................................................... 181 
Figure 6-10: Traded Volumes in Reconfiguration Auctions................................................................................................. 182 
Figure 7-1: Ancillary Service Costs by Product....................................................................................................................... 184 
Figure 7-2: Real-Time Reserve Payments  ............................................................................................................................... 186 
Figure 7-3: Impact of Fast-Start on Reserve Payments ........................................................................................................ 187 
Figure 7-4: Frequency and Average of Non-Zero Reserve Prices ....................................................................................... 188 
Figure 7-5: Average Real-Time Reserve Prices for all Pricing Intervals ............................................................................. 189 
Figure 7-6: Reserve Constraint Penalty Factor Activation Frequency ............................................................................... 190 
Figure 7-7: Forward Reserve Market System-wide Requirements .................................................................................... 192 
Figure 7-8: Aggregate Local Forward Reserve (TMOR) Requirements.............................................................................. 193 
Figure 7-9: Forward Reserve Prices by FRM Procurement Period ..................................................................................... 194 
Figure 7-10:  FRM Payments and Penalties by Year ............................................................................................................. 195 
Figure 7-11: Regulation Payments ........................................................................................................................................... 199 
Figure 7-12: Average Hourly Regulation Requirement, 2022 ............................................................................................. 200 
Figure 7-13:  Regulation Resource Mix ................................................................................................................................... 200 
 
   

 
2022 Annual Markets Report    page v 
  ISO-NE PUBLIC 
Tables 
Table 2-1: Residual Supply Index and Intervals with Pivotal Suppliers (Real-time) ...........................................................48 
Table 2-2: Day-Ahead Price-Cost Markup, % ............................................................................................................................50 
Table 2-3: Real-Time Price Cost Markup, % ..............................................................................................................................51 
Table 2-4: Offer RSI in the FRM for TMNSR (system-wide) and TMOR (zones)..................................................................73 
Table 3-1: Fast-Start Pricing Outcome Summary, 2022 ..........................................................................................................82 
Table 3-2: Annual Average On-Peak Implied Heat Rates and Spark Spreads .....................................................................90 
Table 3-3: Average, Peak and Weather-Normalized Load................................................................................................... 100 
Table 3-4: OP-4 and M/LCC 2 Event Frequency..................................................................................................................... 111 
Table 3-5: Frequency of Negative Reserve Margins (System Level) .................................................................................. 112 
Table 3-6: Most Frequently Binding Interface Constraints in the Day-Ahead Market, 2022........................................ 125 
Table 3-7: Energy and Uplift Payments................................................................................................................................... 128 
Table 4-1: Top 10 Most Profitable Locations for Virtual Supply......................................................................................... 139 
Table 4-2: Top 10 Most Profitable Locations for Virtual Demand ..................................................................................... 140 
Table 4-3: Top 10 Most Profitable FTR Paths in 2022 .......................................................................................................... 146 
Table 5-1: NCPC Credits at External Nodes ............................................................................................................................ 156 
Table 5-2: Summary of CTS Flow Outcomes  .......................................................................................................................... 159 
Table 5-3: Forecast Error in CTS Solution ............................................................................................................................... 163 
Table 5-4: Profit Scenarios, 2022 ............................................................................................................................................. 164 
Table 6-1: Generating Resource Retirements over 50 MW ................................................................................................ 178 
Table 7-1: Regulation Prices...................................................................................................................................................... 198 
Table 8-1: Market Design or Rule Changes ............................................................................................................................ 202 
 
 
 
 

2022 Annual Markets Report 6 ISO New England Inc.

ISO-NE PUBLIC

Executive Summary

The 2022 Annual Markets Report by the Internal Market Monitor (IMM) at ISO New England

(ISO) addresses the development, operation, and performance of the wholesale electricity

markets. The report presents an assessment of each market based on market data and

performance criteria. In addition to buying and selling wholesale electricity day-ahead and in

real time, the participants in the forward and real-time markets buy and sell operating reserve

products, regulation service, financial transmission rights, and capacity. These markets are

designed to ensure the competitive and efficient supply of electricity to meet the energy needs

of the New England region and secure adequate resources required for the reliable operation of

the power system.

To provide readers with additional background on our markets, this year, we have published a

supporting document, “An Overview of New England’s Wholesale Electricity Markets: A Market

Primer”. We expect that this document will be a useful in helping readers understand the

fundamental concepts and mechanics of our markets.

In this executive summary, we provide an overview and assessment of key market trends,

performance, and issues. We follow this with a consolidated list of recommended enhancements

to the market design and rules from this and prior IMM reports.

---------------------------------

The capacity, energy, and ancillary service markets performed well and exhibited competitive

outcomes in 2022. The day-ahead and real-time energy prices reflected changes in underlying

primary fuel prices, electricity demand and the region’s supply mix.

We saw record high energy prices in 2022; the annual average day-ahead price of $86/MWh

was almost 90% higher than last year, and was the highest since the market was implemented

in 2003. We need to go back to the 2005 and 2008 to find comparable pricing levels, before the

shale gas boom in the United States. Energy prices continued to be driven by the market price of

natural gas, which at $9.32/MMBtu was more than double last year’s price, and was the highest

average price since 2008.

Natural gas prices were driven by a combination of market conditions and events at an

international and national level, in addition to regional New England winter issues. The

Russian-Ukrainian conflict exacerbated global uncertainty in natural gas inventories and prices,

and caused a significant uptick in international demand for Liquefied Natural Gas (LNG). Prices

at major trading hubs in the United States were higher due to international LNG demand and

higher US demand. In New England, periods of sustained cold weather led to increased demand

on a constrained pipeline system, and along with reduced injections from LNG terminals,

resulted in very high gas and electricity prices during the first quarter of the year. Oil-fired

generation made a more significant contribution to meeting the region’s energy demand than

recent years, by displacing natural gas generation and mitigating some of the energy price

impacts on days with extremely high gas prices.

While no major reliability issues occurred in 2022, in late December, the region experienced

only its second capacity scarcity condition in five years. This was due to lower imported energy

and generator outages during bitterly cold weather conditions at the tail end of Winter Storm

Elliott, which had a greater impact on other regions and markets, notably PJM. In this instance,

2022 Annual Markets Report 7 ISO New England Inc.

ISO-NE PUBLIC

the region was short of meeting its reserve requirement for about 1½ hours, but there was

sufficient supply to meet system load. High real-time energy and reserve scarcity prices and

capacity performance credits (peaking at over $6,300/MWh, combined) provided strong

market signals and incentives for supply resources to respond and correct the deficit.

High–level market statistics for the five-year period covered in this report are presented below.

At a Glance: High-level Market Statistics

Demand (MW)

2018 2019 2020 2021 2022

% Change

'22 to '21

Sparkline

Load (avg. hourly) 14,095 13,614 13,309 13,566 13,576 0%

Weather-normalized load (avg. hourly)

[a]

13,725 13,558 13,279 13,419 13,472 0%

Peak load (MW) 26,024 24,361 25,121 25,801 24,780 -4%

Generation Fuel Costs ($/MWh)

[b]

Natural Gas 38.72 25.48 16.34 36.07 72.57 101%

Coal 54.52 40.58 37.82 67.77 144.87 114%

No.6 Oil 127.73 130.89 89.42 138.21 221.17 60%

Diesel 187.55 173.55 112.07 184.50 331.99 80%

Hub Electricity Prices: LMPs ($/MWh)

Day-ahead (simple avg.) 44.14 31.22 23.31 45.92 85.56 86%

Real-time (simple avg.) 43.54 30.67 23.37 44.84 84.92 89%

Day-ahead (load-weighted avg.) 46.88 32.82 24.57 48.30 91.36 89%

Real-time (load-weighted avg.) 46.85 32.32 24.79 47.34 91.07 92%

Estimated Wholesale Costs ($ billions)

Energy 6.0 4.1 3.0 6.1 11.7 92%

Capacity 3.6 3.4 2.7 2.3 2.0 -11%

Uplift (NCPC) 0.07 0.03 0.03 0.04 0.05 49%

Ancillary Services 0.1 0.1 0.1 0.1 0.1 127%

Regional Network Load Costs 2.3 2.2 2.4 2.7 2.8 2%

Total Wholesale Costs 12.1 9.8 8.1 11.2 16.7 49%

Supply Mix

[c]

Natural Gas 40% 39% 42% 45% 45% 0%

Nuclear 25% 25% 22% 22% 23% 0%

Imports 17% 19% 20% 16% 14% -2%

Hydro 7% 7% 7% 6% 6% 0%

Other

[d]

5% 5% 5% 5% 4% -1%

Wind 3% 3% 3% 3% 3% 0%

Solar 1% 1% 2% 2% 3% 0.8%

Coal 1% 0% 0% 0% 0.3% -0.20%

Oil 1% 0% 0% 0% 1.5% 1.34%

2022 Annual Markets Report 8 ISO New England Inc.

ISO-NE PUBLIC

Energy costs drove an overall increase in wholesale costs due to high natural gas prices,

comprising over two thirds of total costs

The total wholesale cost of electricity in 2022 was $16.7 billion, the equivalent of $140 per

MWh of load served.

These costs were at their highest level over the last five years and were

considerably higher than the 2021 total of $11.2 billion. This 50% increase (or $5.5 billion) was

driven by higher energy costs. Also, with the exception of capacity costs (down by $0.22 billion),

every other component of the wholesale cost of electricity increased in 2022.

Wholesale Costs ($ billions and $/MWh) and Average Natural Gas Prices

Energy costs continued to comprise the largest share of wholesale costs, at 70%, a significant

jump from a 55% share in 2021. At their highest level since 2008, energy costs totaled $11.7

billion, up 92% (or $5.6 billion) on 2021 costs. With natural gas prices increasing by over 100%

year-over-year, day-ahead LMPs averaged $86.56/MWh, up 86% (or by $40.64/MWh) on 2021.

In addition, while there were increases in energy costs in each quarter, Quarter 1 (Q1)

accounted for 32% of the total annual change. Q1 2022 saw the highest natural gas prices since

Q1 2014, with gas generation costs exceeding $100/MWh on average. This resulted in an

The wholesale cost of electricity comprises energy, uplift, ancillary services and transmission costs.

[a] Weather-normalized results are those that would have been observed if the weather were the same as the long-term

average.

[b] Generation costs are calculated by multiplying the daily fuel price ($/MMBtu) by the average standard efficiency of

generators for each fuel (MMBtu/MWh)

[c] Capacity cost in 2022 includes the Mystic cost-of-service costs of $0.17 billion.

[d] Provides a breakdown of total supply, which includes net imports. Note that section 2 provides a breakdown of native

supply only.

[e] The "Other" fuel category includes landfill gas, methane, refuse and steam

denotes change is within a band of +/- 1%

Sparkline: Green = High Point, Red - Low Points

$6.0

$4.1

$3.0

$6.1

$11.7

$3.6

$3.4

$2.7

$2.3

$2.0

$2.3

$2.2

$2.4

$2.7

$2.8

$12.1

$9.8

$8.1

$11.2

$16.7

$12

$15

$12

$15

$18

2018 2019 2020 2021 2022

Natural Gas ($/MMBtu)

Costs ($ billion)

Billions

Energy NCPC Ancillary Services Capacity Regional Network Load Natural Gas Price

$49

$34

$26

$51

$98

$29

$23

$19

$17

$18

$20

$23

$98

$82

$69

$94

$140

$30

$60

$90

$120

$150

2018 2019 2020 2021 2022

$ per MWh of Load

2022 Annual Markets Report 9 ISO New England Inc.

ISO-NE PUBLIC

average day-ahead LMP of $115.23/MWh, also the highest since Q1 2014. This upward pressure

on LMPs from natural gas prices was attenuated by more frequent in-merit oil generation than

prior years.

Net Commitment Period Compensation (NCPC), or uplift, costs remained relatively low at just

$53 million, or 0.5% of total energy payments. There were few out-of-market commitments or

dispatch decisions impacting NCPC. In fact, the vast majority of uplift (95%) uplift was paid to

resources committed and dispatched in economic merit order, with the remaining 5% (just $1.1

million) required to meet the costs of out-of-merit reliability commitments. There were

significantly fewer out-of-merit local reliability commitments needed to address impactful

transmission outages.

Capacity, which comprised just 12% of total wholesale costs, continued to decline as the market

maintains surplus capacity over the system’s capacity requirement. Costs totaled $2 billion and

were down by 10% (or $0.22 billion) on 2021. The costs were a function of lower combined

clearing prices in the twelfth and thirteenth Forward Capacity Auctions (FCAs 12 and 13).

When compared against 2021 capacity costs, lower auction clearing prices more than offset

supplemental payments under the Cost of Service (CoS) agreement during 2022 for the Mystic 8

and 9 generators ($166 million in 2022) that began in summer 2022.

Natural gas prices drive high wholesale energy prices in New England

Natural gas prices in New England increased significantly in 2022, with an average price of

$9.32/MMBtu, up from $4.62/MMBtu in 2021. This price increase was driven by local market

conditions, New England's reliance on a constrained natural gas transmission system, and

events at a national and international level, particularly the Russian-Ukrainian conflict.

New England vs. Henry Hub and Marcellus Natural Gas Prices

Prices at supply basins also increased, with Henry Hub averaging $6.38/MMBtu and Marcellus

averaging $5.46/MMBtu. New England's natural gas traded at a premium of $2.94/MMBtu and

$3.86MMBtu compared to Henry Hub and Marcellus, respectively. During Q1 2022, New

England's natural gas price reached its highest quarterly price since 2014, at $13.86/MMBtu,

$0.00

$2.00

$4.00

$6.00

$8.00

$10.00

$12.00

$14.00

$16.00

$18.00

$20.00

Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4

2018 2019 2020 2021 2022

Natrual Gas Price ($/MMBtu)

New England Marcellus Henry Hub

2022 Annual Markets Report 10 ISO New England Inc.

ISO-NE PUBLIC

due to decreased LNG injections and increased reliance on interstate pipeline gas. However, the

highest daily natural gas prices of the year were in Q4. While the quarterly natural gas price

averaged $8.41/MMBtu, a cold snap from December 24 to December 27 drove the average daily

natural gas price to $35.37/MMBtu over the four days. These were the highest daily averages

since Q4 2018.

The trend of decreasing load may have reached an inflection point

Net Energy for Load (NEL) averaged 13,576 MW per hour in 2022, which was comparable to

2021. On a weather-adjusted basis, load has been declining, which reflects the long-term trend

of increased energy efficiency (EE) and, more recently, the adoption of behind-the-meter (BTM)

solar generation. However, weather-normalized load increased slightly over the past two years.

While it is difficult to attribute this directly to any particular driver, this change is consistent

with the ISO forecast that average load will increase each year with the continued adoption of

electricity-fueled transportation and electric heating.

In 2022, EE reduced weather-normalized annual average load by an estimated 2,538 MW (by

16%), which was a 2% decrease (40 MW) compared to 2021. This is in line with the ISO’s

expectation that EE will decline over time due to rising costs of eligible EE measures and the

associated baselines used to calculate claimable savings. By contrast, BTM solar generation

reduced weather-normalized load by 426 MW (by ~3%) which was a 15% increase (57 MW)

compared to 2021, and is expected to continue this upward trend in future years.

Net interchange with neighboring control areas continued to decrease

In 2022, net interchange (or net imports) averaged 1,914 MWs per hour, an 11% (or 231 MW)

decrease compared to 2021, and the lowest amount over the past five years. Net imports met

just 14% of New England’s electric demand, compared to up to 19% during the 2018-2020

period.

Net imports fell at the New York North and New Brunswick interfaces. At New York North,

average net interchange decreased by 189 MW per hour compared to 2021 (401 MW vs. 590

MW). The retirement of the Indian Point nuclear generator in New York in April 2021 and

increased planned transmission outages in New York led to increased congestion and higher

energy prices relative to New England prices. At the New Brunswick interface, average net

interchange decreased by 68 MW (228 MW vs. 296 MW) compared to 2021 due to the extended

outage of a nuclear generator in New Brunswick, which has since returned to service.

Price formation is generally robust but we recommend a review of reserve pricing under

the fast start pricing rules

Large volumes of unpriced (or fixed) supply can have important implications for pricing

outcomes because it increases the likelihood of low or negative prices. We expect this impact to

become more prevalent as additional capacity from renewable generation (e.g., wind, solar)

with low marginal costs enter the energy markets. At present, we generally find that energy

price formation is robust under current levels of unpriced supply. Further, as more low

marginal cost generation participates in the wholesale market, we would expect to see a market

response in terms of more price-responsive supply, particularly with more energy storage

devices joining the market; otherwise, there is a higher risk of energy prices not covering short-

run production costs.

2022 Annual Markets Report 11 ISO New England Inc.

ISO-NE PUBLIC

The Fast Start Pricing rules in the real-time energy market are working as intended by better

reflecting the production cost of fast-start generators in LMPs, and reducing uplift costs.

However, we have observed frequent non-zero reserve pricing in scenarios when resources’

dispatch instructions were not impacted by the reserve constraint and the system had a surplus

of reserves.

Due to tradeoffs presented by the separation of the dispatch and pricing software, the ISO chose

a pricing optimization methodology that minimizes false negatives (i.e., no reserve pricing when

there is a physical reserve constraint binding) but allows false positives (i.e., reserve pricing

when there is not a physical reserve constraint binding). This was an intentional decision when

fast-start pricing was implemented, however, the frequency in which we have observed reserve

pricing when there is not a physical reserve constraint binding has exceeded our expectations.

In 2022 alone, $13.7 million in reserve payments were made when there was a reserve surplus

(over half of the $26.9 million in total reserve payments during the year), despite dispatch being

unaffected by the reserve constraint.

We recommend that the ISO revisits reserve pricing mechanics under fast-start pricing to

address the frequency of non-zero reserve pricing when there is a physical reserve surplus.

Low levels of structural market power and mitigations in the energy market, but not in the

Forward Reserve Auction

In 2022, the market concentration among the four largest firms controlling supply in the real-

time market was in line with the values of the last five years, at 44%. There was an increase in

the amount of time when structural market power was evident in real-time because of less

reserve supply, leading to slightly lower total operating reserve margins. There was at least one

pivotal supplier in a quarter of all hours. However, operating reserve margins remained

relatively high on average compared to the underlying requirement, and mitigation remained

relatively infrequent, which tempers any significant concerns regarding the exercise of market

power.

Markups, which estimate the impact of above-cost bidding on clearing prices, were close to zero

or negative in the real-time and day-ahead markets. In 2022, the quantity of withheld economic

capacity was relatively low (below ~ 2%) and generally in line with levels seen in past years.

While the mitigation process for the energy markets has functioned reasonably well in recent

years, we have identified a number of potential issues with the current rules that we

recommend the ISO and stakeholders consider. For example, the current thresholds allow for

considerable latitude in supply offer levels over competitive benchmarks (300% and 50%);

these thresholds have been in place for many years with little empirical support. With the

benefit today of a rich history of supply offer data and robust simulation tools to measure

impact and incentives, an assessment should consider whether the current thresholds

adequately limit the exercise of market power under a plausible set of system scenarios.

Another area of concern is with the structural competitiveness of the Forward Reserve Auction

(FRA). Forward Reserve Market (FRM) auction prices for Ten Minute Non-Spinning Reserve

(TMNSR) have frequently been below $2,000/MW-month over the last few years. However, the

most recent auctions (Summer 2022 and Winter 2022-23) have seen increasing price levels and

have not been structurally competitive. The Summer 2022 auction, in particular, had TMNSR

prices of $7,386/MW-month, reflecting a significant increase in participant offer prices. While

2022 Annual Markets Report 12 ISO New England Inc.

ISO-NE PUBLIC

we reviewed participant offers in the Summer 2022 auction and did not find evidence of the

exercise of market power, we remain concerned about the structural competitiveness of this

auction and the lack of market power mitigation measures. We note, and support, the ISO’s

plans to sunset this market, coinciding with the implementation of day-ahead ancillary service

in early 2025.

Low capacity costs continue as older fossil-fueled generators are replaced with renewable

resources

Capacity market outcomes remained relatively constant in the past two auctions, driven by an

unchanged FCA 16 and 17 (rest-of-pool) clearing price of $2.59/kW-month. Since FCA 10,

capacity prices have decreased 63% ($4.44/kW-month) and Net ICR has decreased 11% (3,846

MW), resulting in a record low $0.9 billion in projected capacity payments for CCP 2026-2027.

Prices are reflective of continued surplus supply conditions and recent reductions in the

capacity requirement.

The capacity market has been structurally competitive at the system level over the past five

auctions and competitiveness indices have further improved in the past two auctions due to

reductions in NICR, along with the absence of major retirements. In FCA 17, Southeast New

England (SENE) was not modelled as an import constrained zone due to a decrease in the zonal

load forecast and an increase in the import capability limit into the zone. This also enhanced the

overall structural competitiveness of the capacity market.

New entrants to New England have varied over the past seven auctions, with gas and energy

efficiency resources making up a majority of new capacity additions from FCA 10-13. Since FCA

13, renewable technologies, typically sponsored-policy resources, have made up 56% of all new

generator additions. Major retirements have comprised nuclear, oil-, and gas-fired resources,

and an additional 1,100 MW of new gas-fired resources have been terminated since FCA 10 for

failing to meet critical development milestones.

From FCA 10-16, the volume of cleared import capacity has fluctuated between 1,000 MW to

1,500 MW across the four external ties in the auction, with most capacity at the New York AC

lines (50%) and Phase II (30%). However, cleared import capacity was down significantly in the

most recent auction, FCA 17, at just 567 MW, comprising 390 MW at New York and 177 MW at

New Brunswick. In FCA 17, we observed the largest percentage of import capacity de-listing

(85%). Despite an import limit of 1,600 MW, import resources cleared at the lowest level over

the five-year period; a key factor driving reduced import capacity is the expected value of

capacity in the neighboring control area.

December 2022 saw the first Pay-for-Performance event since 2018. The December 2022 event

lasted for 17 five-minute intervals and resulted in the transfer of $35.9 million from under-

performing resources to over-performing resources. Import and nuclear resources received the

most performance credits. While cold weather led to forced outages for some of the gas fleet,

the leading cause of gas resource under-performance was high gas prices. Spot natural gas

prices climbed to over $30/MMBtu, resulting in gas generation being pushed out of economic

merit order and, instead, relatively cheaper oil generation was scheduled in the day-ahead

energy market.

Once system conditions tightened in real-time and system-wide LMPs

More information on fuel prices in Q4 2022 can be found in Section 1.2.2.

2022 Annual Markets Report 13 ISO New England Inc.

ISO-NE PUBLIC

increased significantly, operating limitations prevented many gas-fired resources from coming

online in time to contribute to meeting the load and reserve requirement.

IMM Market Enhancement Recommendations

One of the IMM’s key functions is to recommend rule changes to enhance the performance of

the markets. In practice, we communicate our recommendations through our reports,

particularly our quarterly markets performance reports, and through comments filed with

FERC on proposed rules changes.

The table below summarizes the IMM’s recommended market enhancements, first showing

issues with an “open” status, followed by recently closed issues. Recommendations included in

this report for the first time are identified as “*NEW*”. A hyperlink is provided to the document

in which the recommendation was first put forward, along with the IMM's priority ranking of

each recommendation.

The priority ranking (High, Medium or Low) considers the potential market efficiency gains, as

well as the potential complexity and cost of implementing each recommendation. High priority

recommendations may deliver significant market efficiency gains, with the benefit outweighing

the cost of implementing them. At the other end of the scale, low priority recommendations are

not intended to indicate low importance, but rather issues which may not have as significant

long-term efficiency gains (compared to high priority recommendations) and/or may be very

costly to implement.

Recommendation

When made

Status

Priority

Ranking

2023-1

*NEW* Review energy mitigation thresholds

and reference level methodologies, eliminate

mitigation exemptions for non-capacity

resources, and extend mitigation to export-

constrained area. Market power mitigation

rules need to strike a reasonable balance

between producer and consumer interests, and

in turn prescribe adequate threshold tests to

determine when market monitors override

generators supply offers. The IMM has

identified a number of potential rule

improvements to better serve the mitigation

function.

1. Review of the current energy mitigation

thresholds that apply to instances of system-

wide and local market power. The current

thresholds allow for considerable latitude in

supply offers levels over competitive

benchmarks (300% and 50%) and have been in

place for many years with little empirical

support.

2. Eliminate the energy offer mitigation

exemption for non-capacity resources in the

day-ahead energy market.

3. Extend the scope of offer mitigation to cover

the potential exercise of market power in

export-constrained areas.

4. Review the methodologies for determining

reference levels, which are used to evaluate if

2022 AMR

New Recommendation – the

IMM will work with the ISO

to assess the

implementation

requirements for this

project.

Medium

2022 Annual Markets Report 14 ISO New England Inc.

ISO-NE PUBLIC

Recommendation

When made

Status

Priority

Ranking

an offer is competitive (the “conduct test”).

Currently, reference levels can be based on

marginal cost, or historical fuel-adjusted

accepted supply offers or LMPs. We have

observed instances in which the latter two

methodologies produce unreasonably high

reference levels.

2023-2

*NEW* Review reserve pricing mechanics

under fast-start pricing. Under current fast-

start pricing rules, we have observed frequent

non-zero reserve pricing in scenarios when

resources’ dispatch instructions were not

impacted by the reserve constraint and the

system had a surplus of reserves. Due to

tradeoffs presented by the separation of the

dispatch and pricing software, the ISO chose a

pricing optimization methodology that

minimizes false negatives (no reserve pricing

when there is a physical reserve constraint

binding) but allows false positives (reserve

pricing when there is not a physical reserve

constraint binding). This was an intentional

decision when fast-start pricing was

implemented, however, the frequency in which

we have observed reserve pricing when there is

not a physical reserve constraint binding has

exceeded the frequency in which we expected

these scenarios to occur, and the cost of

reserve payments in these intervals warrants

additional consideration of other solutions.

2022 AMR

IMM and ISO to assess the

implementation

requirements for this

project.

Medium

2022-1

Incentive rebuttal component of proposed

Buyer-side Mitigation Rules. The ISO's

proposed buyer-side mitigation rules will allow

a Project Sponsor to demonstrate a lack of

incentive through a Net Benefits Test to avoid

mitigation of a below-cost supply offer from

certain resources. The IMM has recommended

that removing the incentive rebuttal provision

from the proposal would make the buyer-side

mitigation review more predictable and

capable of being administered more reliably

and with less subjectivity.

Filed

Comments

with FERC on

MOPR

Elimination

and Buyer-side

Mitigation

Rules (Apr

2022)

The ISO /NEPOOL proposal

was approved by FERC. The

IMM will keep this

recommendation under

review, which will be

informed by implementation

experience.

Medium

2021-1

Develop Offer Review Trigger Price (ORTP) for

co-located solar/battery facilities

Under the current rules, the ORTP for a co-

located battery and solar project is based on

the weighted average of the individual

technologies. This results in a value that is

below the true “missing money” for the

combined resource, allowing such resources to

offer in at prices below competitive levels

Filed

Comments

with FERC on

ORTP

Recalculation

(Apr 2021

The value of this

recommendation is low in

the context of the

elimination of MOPR in FCA

19. The IMM will reassess

this recommendation

pending the outcome of the

MOPR elimination proposal.

Low

2022 Annual Markets Report 15 ISO New England Inc.

ISO-NE PUBLIC

Recommendation

When made

Status

Priority

Ranking

without review and mitigation, and

undermining the protections put in place by the

minimum offer price rule (MOPR). In our

opinion, a bottom-up calculation is preferable

because it accurately represents the

constraints that co-located solar/battery

facilities face and results in a more precise cost

estimate.

2020-1

Reference level flexibility for multi-stage

generation Given that recommendation 2017-1

below is not part of the ISO’s work plan, and is

unlikely to be developed for some time, we

recommend related changes that could be

made to the market power mitigation function

in the meantime. We believe these changes will

be less resource-intensive and complex to

adopt, compared to incorporating multi-stage

generation modeling into the day-ahead and

real-time market and systems software.

However, it is not a replacement of the above

recommendation. The recommendation is to

provide generators with the ability to

dynamically select their active or planned

configuration and to adjust reference levels to

be consistent with their operating costs and

their supply offers. This will address the current

ris k of false positive and negative errors in

mitigation, given the potentially high costs

differences between configurations. It may also

eliminate a potential deterrent to generators

from offering configurations to avoid the risk of

mitigation, which may ultimately be more cost

effective to consumers.

Winter 2020

QMR (May

2020)

Not in the scope of the ISO’s

current work plan.

Medium

2018-1

Unoffered Winter Capacity in the FCM The

IMM is concerned that generators may be

contracting at, or close to, their maximum

capacity (i.e., their winter qualified capacity), as

determined by the ISO, even though that

capacity is not deliverable in certain months

given expected ambient temperatures. The

IMM recommends that the ISO review its

existing qualification rules to address the

disconnect between the determination of

qualified capacity for two broad time horizons

(summer and winter), the ability of the

generators to transact on a monthly basis, and

the fluctuations in output capability based on

ambient conditions. A possible solution would

be for the ISO to develop more granular (e.g.,

monthly) ambient temperature-adjusted

qualified capacity values, based on forecasted

temperatures and the existing

output/temperature curves that the ISO

currently has for each generator.

Fall 2018 QMR

(Mar 2019)

While this recommendation

remains open it may need to

be reviewed by the IMM in

the context of the design

effort to revise the

methodology for calculating

qualified capacity (the

resource capacity

accreditation project).

Medium

2022 Annual Markets Report 16 ISO New England Inc.

ISO-NE PUBLIC

Recommendation

When made

Status

Priority

Ranking

2017-1

Treatment of multi-stage generation Due to

the ISO’s current modeling limitations, multi-

stage generator commitments can result in

additional NCPC payments and suppressed

energy prices. This issue was first raised by the

external market monitor, Potomac Economics.

[1]The IMM recommends that the ISO consider

improvements to its current approach to multi-

stage generator modeling. Two possible

options are: a. Expanding the current pseudo-

combined cycle (PCC) rules- Consider whether

to make PCC rules a mandatory requirement

for multi-stage generators through proposed

rule changes, or b. Adopt multi-configuration

resource modeling capability- More dynamic

approach to modeling operational constraints

and costs of multiple configurations.

Fall 2017 QMR

(Feb 2018)

Not in the scope of the ISO’s

current work plan.

Medium

2016-1

Improving price forecasting for Coordinated

Transaction Scheduling (CTS): There is a

consistent bias in the ISO’s internal price

forecast at the New York North interface, which

may reduce the effectiveness of CTS. To date,

biases in ISO-NE and NYISO forecasts have been

in opposite directions, which increase the price

spread between the markets relative to actual

spreads, and may produce inefficient tie

schedules. ISO-NE should assess the causes of

biases in the price forecast and assess how the

accuracy of the forecast can be improved. ISO-

NE should periodically report on the accuracy

of its price forecast at the NYISO interface, as

well as the differences between the ISO-NE and

NYISO price forecasts.

2016 AMR

(May 2017)

The IMM will continue to

assess and report on the

price forecasting issue. The

ISO is also periodically

reporting on the forecast

accuracy. Future

improvements are not in the

scope of the ISO’s current

work plan.

High

2016-2

Analyzing the effectiveness of Coordinated

Transaction Scheduling: ISO-NE should

implement a process to routinely access the

NYISO internal supply curve data that is used in

the CTS scheduling process. This data is an

important input into the assessment of the cost

of under-utilization and counterintuitive flows

across the CTS interface.

2016 AMR

(May 2017)

Related to the item above

(Improving price forecasting

for CTS). Not in the scope of

the ISO’s current work plan.

Medium

2015-1

Corporate relationships among market

participants: The ISO develop and maintain a

database of corporate relationships and asset

control that allows for accurate portfolio

construction for the purpose of identifying

uncompetitive participation, including the

potential exercise of market power and market

manipulation.

Q2 2015 QMR

(Oct 2015)

The project is not in the

scope of the ISO's current

work plan. The IMM will

continue to rely on a

combination of internal data

and its own market research

to s atisfy its monitoring

needs.

Medium

2015-3

Pivotal supplier test calculations: The ISO,

working in conjunction with the IMM, enhance

the real-time energy market mitigation pivotal

supplier test to include (1) ramp-based

accounting of supply recognizing the

differences between energy and reserve

products and (2) participant affiliations.

2015 AMR

(May 2016)

IMM and ISO to assess the

implementation

requirements for this

project.

Medium

2022 Annual Markets Report 17 ISO New England Inc.

ISO-NE PUBLIC

Recommendation

When made

Status

Priority

Ranking

2015-2

Forward reserve market and energy market

mitigation: The ISO develop and implement

processes and mechanisms to resolve the

market power concerns associated with

exempting all or a portion of a forward reserve

resource’s energy supply offer from energy

market mitigation.

Q2 2015 QMR

(Oct 2015)

The IMM will evaluate

revising or eliminating

mitigation exemptions for

FRM resources to resolve

the market power issues.

[The ISO is proposing to

sunset the FRM with the

implementation of DASI in

Q1 2025]

Low

2013-1

Limited energy generator rules: The ISO

modify the market rules as necessary to ensure

that the use of the limited-energy generator

(LEG) provisions in both the day-ahead and

real-time markets are restricted to instances

when the availability of fuel is physically

limited.

2013 AMR

(May 2014)

Further analysis required by

the ISO to assess whether

specific rule or procedure

improvements are

appropriate. The IMM will

continue to monitor the use

of the limited-energy

generation provision and

address any inappropriate

us e on a case-by-case basis.

Low

2010-1

NCPC charges to virtual transactions: The ISO

develop and implement processes and

mechanisms to reduce NCPC charges to virtual

transactions (to better reflect the NCPC cost

causation principle) in response to the

historical decline in virtual trading activity. A

reduction in NCPC charges to virtual

transactions will likely improve day-ahead

scheduling by adjusting expectations of real-

time conditions.

2010 AMR (Jun

2011)

The ISO had planned to

review this issue as part of

the conforming changes

related to the Energy

Security Improvements

Project.

Medium

2022 Annual Markets Report 18 ISO New England Inc.

ISO-NE PUBLIC

Section 1

Overall Market Conditions

This section provides an overview of the key trends in the wholesale markets. It covers the

underlying supply and demand conditions behind those trends, and provides important context

to the market outcomes discussed in more detail in subsequent sections of this report.

1.1 Wholesale Cost of Electricity

This section provides a high-level overview of wholesale electricity costs over the past five

years, focusing on the significant factors that influenced changes in different cost categories,

particularly the fluctuations in the natural gas prices.

Key Takeaways

In 2022, the wholesale market cost of electricity totaled $16.7 billion, an increase of $5.5 billion

(or 50%) compared to 2021 costs ($11.2 billion). The increase was largely due to higher energy

costs, which were close to record levels, driven by higher natural gas prices. Natural gas prices

rose by about 100% between 2021 ($4.62/MMBtu) and 2022 ($9.28/MMBtu). Natural gas-fired

generators continued to be the single largest resource type in terms of installed capacity and

energy output, making up 52% of total native generation. As such, natural gas prices continued

to be the primary driver of energy, ancillary services and NCPC costs.

The relationship between wholesale electricity costs and the price of natural gas is evident in

Figure 1-1 below. This figure shows the breakdown of the wholesale costs of electricity over the

past five years, along with average natural gas prices.

Figure 1-1: Wholesale Costs ($ billions and $/MWh) and Average Natural Gas Prices

$6.0

$4.1

$3.0

$6.1

$11.7

$3.6

$3.4

$2.7

$2.3

$2.0

$2.3

$2.2

$2.4

$2.7

$2.8

$12.1

$9.8

$8.1

$11.2

$16.7

$12

$15

$12

$15

$18

2018 2019 2020 2021 2022

Natural Gas ($/MMBtu)

Costs ($ billion)

Billions

Energy NCPC Ancillary Services Capacity Regional Network Load Natural Gas Price

$49

$34

$26

$51

$98

$29

$23

$19

$17

$18

$20

$23

$98

$82

$69

$94

$140

$30

$60

$90

$120

$150

2018 2019 2020 2021 2022

$ per MWh of Load

2022 Annual Markets Report 19 ISO New England Inc.

ISO-NE PUBLIC

Energy costs are incurred by market participants with load obligations (typically Load Serving

Entities or LSEs) in the day-ahead and real-time energy markets. This component accounted for

the largest share of wholesale electricity costs in each of the past five years. It comprised more

than 70% of total costs in 2022, its highest share over this time.

Energy costs totaled $11.7 billion, the highest total since 2008 ($12.0 billion). Energy costs

increased substantially, by $5.6 billion, or 92%, compared to the 2021 total of $6.1 billion.

Natural gas prices more than doubled, rising to $9.28/MMBtu (up by 101%) and were the

primary driver of the change in energy costs.

However, the upward pressure of natural gas on

energy costs was attenuated by relatively cheaper oil prices during winter cold spells,

particularly at the end of the year. This is the primary reason that the energy cost increase was

slightly lower than the gas price change. Oil-fired generators set price for a larger share of

electricity demand in 2022 compared to 2021 (4% vs. 1%).

Regional network load (RNL) costs, or transmission costs, includes transmission owners’

recovery of infrastructure investments, maintenance, operating, and reliability costs. This

accounted for a large share (17%) of total costs in 2022. Transmission and reliability costs were

$2.8 billion in 2022, $46 million (2%) more than 2021 costs. The change was due to a slight

increase in infrastructure costs, which make up the majority (around 95%) of RNL costs.

Capacity costs comprise payments to supply resources in the Forward Capacity Market, and

accounted for 12% of wholesale costs in 2022, decreasing by 10%, or $0.22 billion.

Lower

auction clearing prices in FCA 12 (2021/22) and FCA 13 (2022/23) more than offset

supplemental payments under the Cost of Service (CoS) agreement for the Mystic 8 and 9

generators ($166 million in 2022) that began in Summer 2022. Capacity clearing prices peaked

in FCA 9 (2018/19) at $9.55/kW-month, and have since declined in each subsequent auction

through FCA 14 (2023/24). Clearing prices in FCA 12 ($4.63/kW-month) reflected lower

installed capacity requirements and increased surplus due to new entry. In FCA 13, the clearing

price ($3.80/kW-month) decreased further due to a large influx of lower-priced, new capacity

displacing higher-priced existing capacity.

Ancillary service costs include payment to supply resources for providing operating reserves

and regulation services, and totaled $123 million in 2022, $69 million more than 2021 costs.

Unless otherwise stated, the natural gas prices shown in this report are based on the weighted average of the

Intercontinental Exchange next-day index values for the following trading hubs: Algonquin Citygates, Algonquin Non -G,

Portland, Tennessee gas pipeline Z6-200L, Tennessee North gas, Tennessee South gas, and Maritimes and Northeast. Next-

day implies trading today (D) for delivery during tomorrow’s gas day (D+1). The gas day runs from hour ending 11 on D+1

through hour ending 10 on D+2.

The highest daily average natural gas prices of 2022 occurred at the end of the year, and oil-fired generators were

frequently in merit from December 23 to December 29. If we exclude these dates from the annual gas price average, we

see a 93% increase in gas prices, which is in line with the 92% increase in energy prices.

The annual figure is the sum of the monthly Total RNL Costs as reported in the ISO’s Monthly Regional Network Load Cost

Reports, available at: https://www.iso-ne.com/markets-operations/market-performance/load-costs/

In 2022, capacity costs also include supplemental payments under the Cost of Service Agreement to the Mys tic 8 and 9

generators, which were retained for their fuel security attributed for power years 2022/23 and 2023/24. Supplemental

payments are in addition to base FCM payments at the auction clearing prices.

2022 Annual Markets Report 20 ISO New England Inc.

ISO-NE PUBLIC

Both reserve and regulation costs increased in 2022 primarily due to the increase in energy

prices (which, in general, increased the opportunity costs of providing these products).

Net Commitment Period Compensation (NCPC costs, or uplift, covers supply resource

productions costs not recovered through energy prices. NCPC totaled $53 million in 2022, an

increase of 49% compared to $35 million in 2021. NCPC costs comprised a small component of

total energy costs at just 0.5%. Higher payments to resources committed in economic merit (up

$23.2 million) were offset by fewer local reliability commitments and payments (down by $5.7

million). The increase in economic payments was in-line with higher production costs

associated with higher gas prices.

1.2 Supply Conditions

This section provides a macro-level view of supply conditions across the wholesale electricity

markets and describes how conditions have changed over the past five years. The first

subsection (1.2.1) covers the New England generation mix, the second subsection (1.2.2) covers

fuel and emissions market prices, and the third subsection (1.2.3) provides estimates of

generator profitability.

Key Takeaways

Natural gas generation continued to account for the largest share (52%) of native electricity

generation, twice as much as the second largest fuel type (nuclear). In 2022, natural gas prices

averaged $9.32/MMBtu, moving upward from the record lows in 2021. A large factor for this

increase was the start of the conflict between Russia and Ukraine, which exacerbated global

uncertainty in natural gas inventories and prices.

These higher natural gas prices led to larger spark spreads, which drove a significant increase in

gas-fired generators’ net revenues in 2022. While oil prices also saw year-over-year increases,

oil generation increased in 2022 as it displaced some gas generation during times of elevated

gas prices (e.g., Winter 2022). Emissions costs were relatively low compared to fuel costs in

2022 (less than 10%), but these costs have grown in recent years.

1.2.1 Generation and Capacity Mix

An overview of New England’s native generation and capacity mix by fuel type, location, and

age, is presented below.

The composition of the system supply portfolio provides context to

the relationship between fuel and wholesale prices, as well as emerging operational challenges.

In 2022, there was slightly more wholesale solar (increased installed capacity) and oil

generation (generation during the winter and peak summer hours) compared to 2021. Nuclear

The ancillary services total presented here does not include blackstart and voltage costs. Those costs are included in the

RNL category. The costs of the winter reliability program are included in the 2018 total; this program ended in February

2018.

Section 1.2.1 focuses on native generation. Section 5 provides an explanation of interchange between New England and

the bordering control areas.

2022 Annual Markets Report 21 ISO New England Inc.

ISO-NE PUBLIC

and gas-fired generation displayed the highest capacity factors in 2022, emphasizing the

region’s high utilization of, and dependence on, these two fuels for electricity production.

Average Generator Output by Fuel Type

Energy production by fuel type has exhibited minor changes on average as illustrated in Figure

1-2 below. The chart shows a breakdown of hourly average native generation (in MW per hour)

by fuel type, along with percent share of each fuel.

Figure 1-2: Average Output and Share of Native Electricity Generation by Fuel Type

Native generation increased slightly in 2022 (by 249 MW compared to 2021), balancing out a

reduction in net imports, which decreased by 231 MW, primarily at the New York North

interface (see Section 1.4 for more detail). Natural gas generation continued to account for the

largest share (52%) of native electricity generation, twice as much as the second largest fuel

type (nuclear).

There was a relatively significant increase in oil generation in 2022 compared to prior years,

producing 211 MW per hour compared to 26 MW per hour in 2021. The largest increase

occurred during cold winter days, when higher gas prices meant oil-fired generators (including

dual-fuel generators) were more frequently in economic merit relative to gas-fired generators.

To a lesser extent, fast-start oil generators operated more often on summer days in 2022 with

high loads and tight system conditions.

While the share of native generation is small (roughly equivalent to hydro generation), solar

and wind production remain a key focus of energy policies impacting New England’s energy

footprint. State and federal policies have driven additional wholesale (front-of the-meter) solar

energy production and wind energy production.

Solar and wind accounted for 4% of native

generation in 2018 and increased each year, up to 7% in 2022.

Section 1.3.1 discusses the impact of solar generation on load from both behind- and front-of-the-meter solar.

49%

48%

52%

53%

52%

30%

27%

26%

2,000

4,000

6,000

8,000

10,000

12,000

14,000

2018 2019 2020 2021 2022

Average Hourly Output (MW)

Gas Nuclear Hydro Other Wind Solar Coal Oil

% Share

Average Output

2022 Annual Markets Report 22 ISO New England Inc.

ISO-NE PUBLIC

Capacity by Fuel Type

Capacity by generator fuel type in Figure 1-3 below shows the breakdown of total capacity

under contract in the FCM. Total capacity is significantly higher than actual generation and

demand due to the need to have sufficient operating reserves to account for uncertainty in

supply availability.

Figure 1-3: Average Capacity by Fuel Type

In total, average capacity in New England decreased 4% year-over-year, from 30,200 MW in

2021 to 28,900 MW in 2022, driven by steady decreases in the demand for capacity; the net

installed capacity requirement (Net ICR).

No single fuel type saw a significant change in capacity share in 2022. Natural gas generation

continues to make up the most capacity of any fuel source. Combined, gas-fired and dual-fuel

(oil/gas-fired) generators accounted for over 62% (about 18,000 MW) of total average

generator capacity in 2022. Coal resources have seen a steady decline in capacity share due to

the retirement of Bridgeport Harbor 3 in June 2021 (383 MW).

The underlying data to determine resource fuel type changed in the 2019 AMR. The change was reflected across all five

years. With this change, more resources were identified as dual-fuel. This shifted resources out of the gas category into the

dual-fuel category. Additional, this AMR introduced two new categories for dual-fuel generators: combined-cycle (CC) and

other.

Coal category includes generators capable of burning coal and dual-fuel generators capable of burning coal and oil.

“Other” category includes active capacity demand response, landfill gas, methane, refuse, solar, battery s torage, steam,

and wood.

30.1%

30.2%

31.5%

32.4%

32.5%

15.8%

17.3%

16.8%

16.4%

16.9%

12.2%

13.3%

13.9%

12.9%

12.7%

13.1%

11.6%

10.6%

11.0%

11.6%

10.6%

10.0%

9.9%

10.7%

10.5%

11.3%

10.9%

10.7%

10.6%

10.3%

3.0%

2.9%

2.8%

2.0%

1.5%

5,000

10,000

15,000

20,000

25,000

30,000

35,000

2018 2019 2020 2021 2022

Capacity (MW)

Gas Dual - CC Dual - Other Nuclear Hydro Oil Coal Battery Storage Other Solar Wind

2022 Annual Markets Report 23 ISO New England Inc.

ISO-NE PUBLIC

Capacity Factors

A capacity factor is the percentage of a generator’s capacity being utilized and is calculated as

the ratio of a resource type’s average hourly output over their total capacity supply obligation

(CSO).

The individual capacity factors are aggregated by fuel type and shown in Figure 1-4

below.

Figure 1-4: Capacity Factor by Fuel Type

In 2022, nuclear generators and gas-fired generators saw modest increases to their capacity

factors. Nuclear generation capacity continued to be the most utilized generation source,

generally producing at their maximum output when available. Gas-only generators saw the

largest increase in capacity factors year-over-year, up from 41% in 2021 to 45% in 2022,

making up for flower net imports.

The inset graph shows capacity factors for combined-cycle

(CC) gas-fired generators categorized by age tranche. As expected, newer and more efficient CCs

have a higher capacity factor. Most system capacity is in the 10+ year category, which saw a

slight increase in overall utilization in 2022.

Since 2020, increases in the average natural gas price has made oil and coal generation more

economic in New England. While coal-fired generator capacity comprises a very small

proportion of total system capacity, there was a modest increase in their capacity factors in

2022, up from 1% in 2020 to 8% in 2021 and 2022. Oil-fired generators provided the lowest

capacity factor at 2%, but their production increased significantly year-over-year, from an

hourly average of 8 MW in 2021 to 65 MW in 2022.

The methodology for capacity factor calculation changed for this AMR. The capacity factor is now calculated as the ratio

of average hourly MW production over the average hourly CSO, aggregated by fu el type. For example, if there was an

hourly average of 5,000 MW of gas generation for a given year with an average gas resource CSO of 10,000 MW, the yearly

capacity factor for the fuel type would equal 50%. The new methodology includes production MWs fro m resources

without an active CSO.

More information on net import trends in New England can be found in Section 1.4.

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

Nuclear Hydro Gas Dual - CC Dual - Other Coal Oil

Capacity Factor

2018 2019 2020 2021 2022

Average total 2022

CSO for each category

(in MW)

3,300

3,000

400

2,800

9,400

4,900

3,700

20%

40%

60%

80%

Less than 10 years More than 10 years

Capacity Factors - Combined-

Cycle by Asset Age

2018 2019 2020 2021 2022

1,900

12,000

2022 Annual Markets Report 24 ISO New England Inc.

ISO-NE PUBLIC

Generation by State

A breakdown of energy production and consumption within each state and aggregated across

the ISO-NE market is shown in Figure 1-5 below.

Darker shaded bars show state load, while

lighter shaded bars show state generation. The red and blue bars simply show the difference

between production and consumption; the red bars illustrate net imports into each state, and

the blue bars net exports out of the state.

Figure 1-5: Average Native Electricity Generation and Load by State

Given their larger populations, Massachusetts and Connecticut are the largest consumers and

producers of electricity within the six-state footprint. Massachusetts, the state with the most

load, consumed an average of 3,690 MW per hour more than it generated in 2022, up from

3,200 MW per hour in 2018. The gap between load and generation was driven by a decrease in

generation from two existing combined-cycle generators due to relatively expensive fuel input

costs. Connecticut generated an average of 1,600 MW per hour more than it consumed in 2022,

up from 1,050 MW per hour in 2018. New gas-fired generators built in Connecticut over the

past five years accounted for the majority of new generation in the state.

The total ISO-NE bar summarizes two key trends. First, average native load in New England fell

by 530 MW per hour compared to 2018. This is largely due to the impact of energy efficiency

and behind-the-meter solar generation, which are discussed in Section 1.3.1 below. Second,

New England continues to be a net importer of power, although its reliance on imports fell. In

2022, 14% (or 1,910 MW per hour) of New England’s electricity demand was met by energy

imported from neighboring jurisdictions. This was the lowest level of net imports over the five-

year period; more detail on this trend is provided in Section 1.4 below.

The s tate breakdown shows native energy production and consumption within each state; it does not include imports

into the state from neighboring jurisdictions

The green bar for ISO-NE illustrates losses as energy flows through the system.

Net imports in this context are not necessarily from neighboring jurisdictions outside of New England (New York or

Canada), but refer to any imports from outside the state.

Note: MW values are rounded to the nearest 10 MW.

2,000

4,000

6,000

8,000

10,000

12,000

14,000

16,000

MA CT NH ME RI VT ISO-NE

MW per Hour

Native Load Native Generation

Red bar represents

net-import state

(MA = 3,690 MW)

Blue bar represents

net-export state

(CT = 1,600 MW)

New England is a

net-importer of

about 1,910 MW

2,000

4,000

6,000

8,000

10,000

12,000

14,000

16,000

MA CT NH ME RI VT ISO-NE

MW per Hour

Native Load Native Generation

Red bar represents

net-import state

(MA = 3,200 MW)

Blue bar represents

net-export state

(CT = 1,050 MW)

New England is a

net-importer of

about 2,460 MW

New England has

about 230 MW of

losses per hour

Drop in native load of 530 MW

2018

2022

2022 Annual Markets Report 25 ISO New England Inc.

ISO-NE PUBLIC

Average Age of Generators by Fuel Type

The age profile of the generation fleet in New England captures how the supply mix is evolving

and can be indicative of potential changes in the future. As generators age, they require

increased maintenance and upgrades to remain operational, thereby increasing costs. Older

coal- and oil-fired generators in New England also face other market dynamics, including higher

compliance costs associated with public policies intended to reduce greenhouse gas emissions.

Compared with coal- and oil-fired generators, new natural gas-fired generators are cleaner,

more efficient, and generally have lower fuel and emission costs. As a result, recent investments

have been in new natural gas-fired generators, wind turbines, and solar panels. Most

retirements include older nuclear, coal- and oil-fired generators.

The average age, in years, of New England’s generation fleet is illustrated in Figure 1-6 below.

If there were no retirements or new generation, we would expect each colored line to increase

by one year as generators age. Either an influx of new generators or a retirement of old

generators can cause a decline in average age. Data labels above the bars show average total

FCM capacity in 2022 by fuel type.

Figure 1-6: Average Age of New England Generator Capacity by Fuel Type (2018 - 2022)

The average age of New England’s generators in 2022 ranged from one and a half years (battery

storage) to 60 years (hydro), with a weighted-average total system age of 37 years. Oil

generation comprises most capacity considered by the ISO to be “at risk of retirement”, with an

average age approaching 50 years.

Oil generation plays an important role in the supply mix

during winter cold spells when natural gas is priced high and pipeline capacity is limited.

Age is determined based on the generator’s first day of commercial operation. The values are weighted by the max net

output for each generator within the fuel type.

The “Other” category includes landfill gas, methane, refuse, steam, and wood.

For an overview of generator retirements see https://www.iso-ne.com/about/what-we-do/in-depth/power-plant-

retirements#:~:text=More%20than%205%2C200%20MW%20of,be%20retiring%20in%20coming%20years.

Hydro Coal Oil Nuclear Dual -

Other

Other Dual - CC Gas Wind Solar Battery

Storage

Age in Years

3,041

438

2,981

3,339

807

4,883

9,385

182

Average total 2022

CSO for each category

(in MW)

3,671

2022 Annual Markets Report 26 ISO New England Inc.

ISO-NE PUBLIC

Wind, solar, and battery storage remain the newest generation fuel types; all three groups of

generators had an average age below 10 years. These technology or fuel categories will

comprise an increasing share of supply over the coming years, playing a big part in meeting the

states’ decarbonization goals.

Generation Additions and Retirements

Generator additions and retirements in each Forward Capacity Auction (FCA), beginning with

Capacity Commitment Period 9 (CCP 9, 2018/19) are shown in Figure 1-7 below. Future

periods are years for which the FCA has taken place, but the capacity has yet to be delivered or

retired. The FCA clearing prices (for existing rest-of-system resources) are also shown for

further context.

Figure 1-7: Generator Additions, Retirements, and FCM Outcomes

In the past eight primary auctions, capacity additions from new generation (9,906 MW) have

significantly outpaced capacity withdrawals (2,690 MW) from retiring generation. Cumulative

net additions to capacity have reached 7,215 MW since FCA 9.

While the surplus amounts (red line) fluctuate between commitment periods, the additions of

cheaper, more efficient new capacity, as well as limited retirements continue, to drive down

clearing prices. In FCA 17, new generation increased significantly year-over-year; the largest

types of new capacity consisted of battery storage projects (400 MW) and solar projects (124

MW).

Existing generation nearing the end of their economic life, and facing high environmental costs,

is incentivized to retire when the capacity price (and associated revenue) falls below their net

going forward costs. However, retirements were quite low in the recent auction, FCA 17,

totaling just 22 MW, a 244 MW (91%) decrease from FCA 16.

It is important to note that a significant amount of existing and new gas capacity was removed

from the capacity market, which is not reflected in the above graph. This totaled over 2,500

MW, effectively reducing the actual net capacity additions since FCA 9 to 4,685 MW. First, the

(2,000)

(1,000)

1,000

2,000

3,000

4,000

5,000

6,000

7,000

8,000

9 10 11 12 13 14 15 16 17

(2018-19) (2019-20) (2020-21) (2021-22) (2022-23) (2023-24) (2024-25) (2025-26) (2026-27)

Capacity Commitment Period (Years)

New Generation Retired Generation

Surplus/Deficit Net Cumulative Additions since FCA 9

Future Periods

1.00 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00

$/kW

-mo

Clearing Price

2022 Annual Markets Report 27 ISO New England Inc.

ISO-NE PUBLIC

retirement of Mystic 8 and 9 (1,413 MW) was suspended in FCA 13 as the ISO retained the units

for reliability for two years.

Second, two large gas resources, Burrillville Energy Center (485

MW) and Killingly Energy Center (632 MW), had their capacity supply obligations (CSOs)

terminated in FCA 10 and 13, respectively.

1.2.2 Generation Fuel and Emissions Costs

Input fuel costs and combustion engines’ operating efficiencies are major drivers of electricity

prices. In 2022, average prices increased for all major fossil fuels as follows:

 Natural gas: $9.28/MMBtu (up 101%)

 No. 2 oil: $22.69/MMBtu (up 72%)

 No. 6 oil: $21.06/MMBtu (up 60%)

 Coal: $14.50/MMBtu (up 113%).

The annual (top) and quarterly (bottom) trends in fuel prices are shown in Figure 1-8 below.

Figure 1-8: Average Fuel Prices by Quarter and Year

The delayed retirement of the two resources was not captured in FCA 15 retirement data as the resources never entered

the auction. More information on the Mystic 8 and 9 retention can be found in Section 6.2.1.

The terminated resources appeared as new capacity in the FCAs they clear in, but did not appear as FCA retirements

post-termination as they were not qualified to participate in any more primary auctions. More information on the

Burrillville Energy Center and Killingly Energy Center terminations can be found in Section 6.2.2.

$10

$15

$20

$25

$30

Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4

2018 2019 2020 2021 2022

Average fuel price ($/MMBtu)

Quarterly Prices

Natural Gas Coal (LS) No. 6 Oil (0.3%) No. 2 Oil

$10

$20

$30

2018 2019 2020 2021 2022

Annual Prices

2022 Annual Markets Report 28 ISO New England Inc.

ISO-NE PUBLIC

Natural Gas

In 2022, natural gas prices averaged $9.32/MMBtu, continuing an upward trend in prices from

record lows in 2020.

Prices increased by $4.70/MMBtu, or 102%, compared to 2021

($4.62/MMBtu). The significant year-over-year increase was due to a combination of market

conditions and events at a national and international level, as well as high New England winter

gas prices due to a constrained natural gas transmission system.

Higher nationwide prices due to international LNG demand and higher US demand

Given the region’s lack of native production, New England’s gas prices are influenced by gas

prices at supply basins. Figure 1-9 below compares annual average prices in New England to

prices at Henry Hub and the Marcellus trading hub.

Figure 1-9: New England vs. Henry Hub and Marcellus Natural Gas Prices

For the second year in a row, prices increased significantly at the two major supply basins due

to increased domestic natural gas demand and LNG export demand. There was a 6% increase in

US demand for natural gas along with increased international demand for LNG exports in all

quarters of 2022.

Prices at Henry Hub averaged $6.38/ MMBtu, up by $2.56/MMBtu (67%)

New England natural gas prices averaged $2.10/MMBtu in 2020, the lowest natural gas prices since at least 1999.

While Henry Hub is the predominant pricing benchmark in the United States, over the last several years, prices in the

Marcellus region have often traded below the Henry Hub price due to the prevalence of cheap shale gas. Additionally, the

geographical proximity between New England and the Marcellus region provides a stronger relationship between prices,

particularly during times when New England pipelines are unconstrained.

Average daily natural gas consumption for energy increased from 84 billion cubic feet per day in 2021 to 89 billion cubic

feet per day in 2022, per Table 1 of EIA’s Short Term Energy Outlook data.

$0.00

$2.00

$4.00

$6.00

$8.00

$10.00

$12.00

$14.00

$16.00

$18.00

$20.00

Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4

2018 2019 2020 2021 2022

Natrual Gas Price ($/MMBtu)

New England Marcellus Henry Hub

2022 Annual Markets Report 29 ISO New England Inc.

ISO-NE PUBLIC

year-over-year, and $5.46/MMBtu at Marcellus, up by $2.56/MMBtu (88%), with the Henry

Hub price reaching its highest quarterly price since 2008.

The Russian-Ukrainian conflict exacerbated global uncertainty in natural gas inventories and

prices. With the start of the conflict in late February 2022, US exports of liquefied natural gas

(LNG) rose to a record high of 11.9 Bcf in March 2022, with the quarterly LNG exports up 24%

year-over-year (up from 9.27 Bcf in Q1 2021 to 11.51 Bcf in Q1 2022).

The significant increase

in LNG exports was primarily driven by the Northwestern Europe LNG prices increasing

substantially to $30.08/MMBtu in Q1 2022 (compared to $9.67/MMBtu in Q1 2021).

In 2022, the price spread between New England and the two supply basins increased markedly

and was at its highest level, in dollar terms, over the five-year period. New England natural gas

traded at a premium of $2.94/MMBtu (or 46%) and $3.86/MMBtu (or 71%) compared to Henry

Hub and Marcellus, respectively. The premium in New England prices was far more pronounced

during the winter months (in Q1 and Q4 2022), when constrained natural gas pipelines during

cold periods, leads to elevated prices. The price premium during Q2 and Q3 were relatively

small and in line with prior year ranges.

Larger increase in New England natural gas prices due to interstate pipeline constraints

During the colder winter months, New England typically relies on LNG imports to fulfill the

additional residential heating demand for natural gas. In Q1 2022, LNG injection into New

England pipelines decreased 46% year-over-year as global LNG prices outpaced New England

interstate pipeline prices.

Without as much pipeline sendout from New England’s LNG

terminals, generators relied heavily on interstate pipeline gas, which drove up the Q1 spot

market price significantly. New England’s Q1 average natural gas price increased 150% year-

over-year, from $5.55/MMBtu in Q1 2021 to $13.86/MMBtu in Q1 2022, the highest quarterly

natural gas price since 2014.

In the warmer months of Q2 and Q3 2022, residential natural gas demand decreased

significantly and the spot price for natural gas in New England remained in line with US hub

prices. The average natural gas price in New England increased 181% ($2.58/MMBtu to

$7.24/MMBtu) year-over-year in Q2 and 96% ($3.93/MMBtu to $7.71/MMBtu) year-over-year

in Q3, in line with price increases seen at the Marcellus and Henry hubs.

As New England temperatures dropped in Q4, residential heating demand placed upward

pressure on the spot natural gas price. The quarterly natural gas price averaged $8.41/MMBtu

in Q4 2022, a $1.97/MMBtu (31%) increase from Q4 2021. A cold snap affected New England

from December 24 to December 27 and led to an average daily natural gas price of

$35.37/MMBtu over the four days, the highest daily averages since Q4 2018.

Historical Henry Hub spot prices can be found in Table 5b of EIA’s Short Term Energy Outlook data.

More information on March LNG exports can be found in the April 2022 Short Term Energy Outlook by EIA.

In-depth analysis on LNG prices can be found in the gas market review in the Winter 2023 Quarterly Markets Report.

The average forward price for the Algonquin pipeline was $18.73/MMBtu in winter 2022, much less than Northwest

Europe’s price of $30.08/MMBtu. More information on the quarterly comparison of pipeline and LNG forward prices can

be found in the gas market review in the Winter 2023 Quarterly Markets Report.

In Q2, the Marcellus and Henry hubs saw a 260% and 157% year-over-year increase in price, respectively. In Q3, the

Marcellus and Henry hubs saw a 97% and 87% year-over-year increase in price, respectively.

2022 Annual Markets Report 30 ISO New England Inc.

ISO-NE PUBLIC

Oil

In 2022, No. 2 Oil increased 72% year-over-year ($13.21/MMBtu to $22.69/MMBtu) and No. 6

Oil prices increased 60% year-over-year ($13.17/MMBtu to 21.06/MMBtu). The significant

increases in price were primarily driven by higher demand Q1 leading to decreased oil storage

levels entering the colder Q4 months.

The larger increase in gas prices in 2022 led to greater

displacement of gas generation with lower-priced oil generation, with oil generators providing

almost 2% of total native electricity generation, up from 0.2% in 2021.

Coal

In 2022, coal prices increased by 113% year-over-year ($6.69/MMBtu to $14.50/MMBtu) due

in part to increased worldwide coal generation.

Coal generation remains the least active

thermal generation source in New England, however, providing only 0.3% of total native

generation in 2022.

Emission Prices

Emissions allowances are secondary drivers of electricity production costs for fossil fuel-fired

generators, but have been trending up. State regulations require some generators to purchase

emissions allowances, which are incorporated into generator offers and their mitigation

reference levels.

New England has two carbon-reducing cap-and-trade programs that influence electricity

prices:

1. The Regional Greenhouse Gas Initiative (RGGI), covering generators in all New England

states, and

2. The Electricity Generator Emissions Limits (EGEL) under the Global Warming Solutions

Act (referred to as the MA GWSA program below), covering only Massachusetts

generators.

These programs aim to make the environmental cost of CO

explicit in dollar terms so that

energy producers consider it in their production decisions. Market prices for CO

credits affect

the total energy costs of fossil fuel-fired generators and, in turn, energy market prices.

Consequently, existing fossil fuel-fired generators are incentivized to maintain or improve their

operating efficiency while newly constructed generation facilities are incentivized to construct

high efficiency generators to minimize generator production/operating costs.

The average cost of emissions by generator fuel type for each program in the context of short-

run fuel costs is illustrated in Figure 1-10.

More information on Q1 oil price trends can be found on EIA’s website here. More information on Q4 oil price trends

can be found on EIA’s website here.

More information on 2022 international coal trends can be found in Coal 2022 Report by the International Energy

Agency.

Both of these programs are discussed in more detail in Section 3.2.1.

310 CMR 7.74: Reducing CO

Emissions from Electricity Generating Facilities (https://www.mass.gov/guides/electricity-

generator-emissions-limits-310-cmr-774)

2022 Annual Markets Report 31 ISO New England Inc.

ISO-NE PUBLIC

Figure 1-10: Annual Estimated Average Costs of Generation and Emissions

The cost of emissions is still relatively low compared to fuel costs (less than 10%), but has

grown in recent years and is increasingly impactful on energy prices. In 2022, the average

estimated costs of the RGGI program increased 41% for most fossil fuel-fired generators year-

over-year: natural gas ($4.36/MWh to $6.15/MWh), coal ($9.85/MWh to $13.89/MWh), No. 6

oil ($8.73/MWh to $11.69/MWh), and No. 2 oil ($9/MWh to $12.70/MWh).

Since natural gas

generators set price for most of load (~80% in 2022), one would expect that the impact on

energy prices will be most closely related to their CO

cost.

The average estimated costs of the Massachusetts GWSA program increased 19% from 2021 to

$4.23/MWh for the average natural gas combined cycle. This was largely due to expectations of

tighter conditions in future years. Therefore, in 2022, a typical combined cycle gas generator

located in Massachusetts incurred CO

costs of over $10/MWh ($6.15 + $4.23)/MWh and would

be expected to incorporate this cost into its energy market supply offer (along with fuel costs of

approximately $72/MWh and other operating and maintenance costs). The impact of this

program on New England energy prices is more challenging to determine, given that it applies

to one state only.

1.2.3 Generator Profitability

With high natural gas prices leading to large spark spreads in 2022, gas-fired generators’ net

revenues increased significantly. As a result, our estimates of the average energy and ancillary

revenues for gas-fired generators over the past five years increased to levels that exceed the

capacity market benchmarks for new entry.

IMM standard generator heat rates and fuel emission rates are used to convert $/ton CO

prices to $/MWh generation

costs. The Massachusetts EGEL program began in 2018, but 2018 costs are excluded due to limited available market

information regarding the value of allowances resulting in varied bid prices. The MA GWSA costs are a trade-weighted

average of auction clearing prices and secondary trades for a given year. MA GWSA was removed for coal because there

are currently no coal generators affected by the EGEL program.

This was due to a variety of factors discussed in Section 3.2.1.

$40

$80

$120

$160

$200

$240

$280

Natural Gas

Coal

No.6 Oil

No.2 Oil

Natural Gas

Coal

No.6 Oil

No.2 Oil

Natural Gas

Coal

No.6 Oil

No.2 Oil

Natural Gas

Coal

No.6 Oil

No.2 Oil

Natural Gas

Coal

No.6 Oil

No.2 Oil

2018 2019 2020 2021 2022

Fuel and Emission Costs ($/MWh)

Fuel Costs RGGI MA GWSA

2022 Annual Markets Report 32 ISO New England Inc.

ISO-NE PUBLIC

New generator owners rely on a combination of net revenue from energy and ancillary service

markets and forward capacity payments to cover their fixed costs. Revenue from the Forward

Capacity Market (FCM) is a critical factor considered when moving forward with the

development of a new project. Developer expectations for minimum capacity revenues are

based on the cost of the project (CONE, or cost of new entry) and their expectations for net

revenue from the energy and ancillary services markets.

This section presents estimates of the net revenues that hypothetical new gas-fired generators

(combined cycle (CC) and combustion turbine (CT)) could have earned in the energy and

ancillary services markets in each of the previous five years. In addition to providing a basis for

the amount of revenue required from the capacity market to build a new generator, this section

also highlights the incremental revenue that could be earned from dual-fuel capability and

evaluates participation in the Forward Reserve Market (FRM) for a combustion turbine

generator.

The analysis is based on simulations of generator scheduling under an objective that maximizes

net revenue while enforcing operational constraints, i.e., ramp rates, minimum run and down

times, and economic limits.

The simulation model also includes a Regional Greenhouse Gas

Initiative (RGGI) cost for every short ton of CO

emitted.

Figure 1-11 shows the result of the simulations.

Each stacked bar represents revenue

components for a generator type and year. A combined cycle generator is shown in green and a

combustion turbine generator that participates in the FRM market is shown in blue. The

simulation produces base revenue (energy and ancillary services (AS)) and incremental dual-

fuel revenue numbers for 2018-2022.

Estimates of future year’s base revenue, dual-fuel

revenue, and FRM revenue are simple averages of these numbers. For all years, the FCA revenue

numbers shown are calculated using the actual payment rates applied to calendar years.

The s imulation uses historical market prices, which implies that the generator’s dispatch decisions do not have an

impact on day-ahead or real-time energy prices. Results should be considered in the high range for potential revenue

estimates because this analysis does not account for forced outages (which should be infrequent for a new generator).

In the model, the RGGI cost for each year is the average auction clearing price for RGGI allowances in that year. RGGI

Auction data is available at https://www.rggi.org/auctions/auction-results/prices-volumes.

The Gross CONE figures for the CC and CT gas-fired generators reflect Net CONE values of $10.50/kW-month and

$6.33/kW-month with the difference between gross and net figures attributed to net revenue from energy and ancillary

service sales.

Incremental dual-fuel energy revenue is earned by the generator when running on its second fuel type.

2022 Annual Markets Report 33 ISO New England Inc.

ISO-NE PUBLIC

Figure 1-11: Estimated Net Revenue for New Gas-fired Generators

When compared with 2021, the simulation results show 2022 net energy revenues increased by

approximately 100% for the combined cycle generator and approximately 190% for the

combustion turbine that participates in the FRM. These large year-over-year increases were

driven by a moderate increase in capacity utilization and significantly higher spark spreads,

which increased by 87% from the prior year.

Unlike 2020 and 2021, dual-fuel capability

added significant revenue to both technology types with CT generators earning an additional

$0.66/kW-month and CC generators adding $1.42/kW-month to net revenue this year. This was

driven by relatively colder temperatures in January and February than in the previous two

winters, which resulted in more frequent oil-fired generation.

In addition, a very high clearing

price for TMNSR in the Summer 2022 Forward Reserve Market (FRM) Auction resulted in a

300% increase in FRM revenue when compared with the prior year.

Overall, the results show that if future market conditions remain similar to the previous five

years, owners of new gas-fired combined cycle generators could expect net revenues (not

including capacity payments) to average $5.23/kW-month, which increases to $5.69/kW-month

for generators with dual-fuel capability. Under the same conditions, new combustion turbines

could expect net revenue earnings from $3.43/kW-month for single-fuel generators to

$3.67/kW-month for generators with dual-fuel flexibility. With higher capacity factors,

combined cycle generators can benefit more often from dual-fuel capability than peaking CT

generators, but both technologies can expect significant revenue gains during periods when

high demand for natural gas pushes its price above the oil price.

A combustion turbine generator can also participate in the FRM where off-line reserves are

procured prior to the reserve season. A forward reserve resource receives revenue from the

forward reserve auction, but it foregoes real-time reserve payments and, in most hours where

Section 3.2.1 of this report discusses spark spreads in more detail.

See: https://www.iso-ne.com/static-assets/documents/2022/05/2022-winter-quarterly-markets-report.pdf

See: https://www.iso-ne.com/static-assets/documents/2022/11/2022-summer-quarterly-markets-report.pdf

2018 2019 2020 2021 2022 2023 2024 2025

$10

$15

$20

$/kW

-mo

Series7 Axis

CT Gross CONE

(FCA18) = $10.59

CC Gross CONE

(FCA18) = $14.61

Future Periods

Base revenue

Incremental dual-fuel

FRM revenue

FCA revenue

Combined Cycle

Combustion Turbine

2022 Annual Markets Report 34 ISO New England Inc.

ISO-NE PUBLIC

the energy price is within a normal range, also foregoes energy revenue since it will be held in

reserve. When the energy price is very high, as in the case of a scarcity event, the forward

reserve resource may be dispatched for energy and would then receive net revenue (above

variable cost) for those high-priced periods. Taking the average of FRM payment rates over the

past five years, a new combustion turbine that is designated as an FRM resource could earn

$0.81/kW-month more net revenue than the same resource might accumulate in the real-time

market alone. Overall, participation in the FRM results in greater net revenue than non-

participation in all five years where these revenues have been observed.

For the next Forward Capacity Auction (FCA 18), the most recent CONE revisions approved by

FERC establish net revenue components of $4.12/kW-month and $4.27/kW-month for

combined cycle and combustion turbine generators respectively.

When compared with these

benchmark estimates, the simulations show that average revenues for both the CC generator

and the CT generator that participates in the FRM are sufficient to meet these expectations.

Regardless, these revenue numbers are clearly insufficient to support new entry without the

addition of capacity payments to cover the fixed costs of a new gas-fired generator.

In recent years, capacity prices have remained low reflecting a system that is long on capacity.

Consequently, total revenues from the energy and capacity markets appear insufficient to

incent either type of gas-fired generator to enter the region’s energy market. In fact, New

England has not added a new gas-fired generator since 2019 (FCA 13).

These revenue components include “Pay-for-Performance” (PFP) revenue which this study does not.

Note that CONE benchmarks are produced from financial and engineering studies that estimate the cost of adding

green-field generators. In practice, the cost of new entry for a generator may be lower than the current CONE benchmarks

for a number of reasons. In particular, when new generating units are built on existing generation sites or when there are

material additions to the capacity of an existing operational plant, the presence of existing infrastructure tends to lower

fixed costs.

2022 Annual Markets Report 35 ISO New England Inc.

ISO-NE PUBLIC

1.3 Demand Conditions

This section presents trends and assesses underlying drivers of demand for the primary

wholesale market products: energy, operating reserve and capacity. The first subsection (1.3.1)

reviews changes in energy demand, and quantifies the impact of weather, energy efficiency, and

behind-the-meter solar generation. The second subsection (1.3.2) provides an overview of

system and local reserve requirements. The final subsection (1.3.3) reviews the amount of

capacity needed to meet the region’s reliability standard.

Key Takeaways

Similar average weather conditions between 2021 and 2022 resulted in comparable average

wholesale energy demand (13,576 MW in 2022 vs 13,566 MW in 2021). In general, however,

wholesale load continued to trend downward due to increases in energy efficiency and

behind-the-meter solar generation. In 2022, energy efficiency reduced annual average load

by an estimated 2,538 MW (by 15%) and behind-the-meter solar generation reduced annual

average load by 426 MW (by 3%).

The average system reserve requirements in 2022 were similar to prior years. The one

exception was the 10-minute spinning reserve requirement, which declined modestly from

previous years as a result of an operational decision by the ISO due to improved resource

performance.

Capacity (Net ICR) needed to meet the system’s reliability objective has generally declined

over recent capacity auctions, primarily due to lower projected peak load forecasts. The Net

ICR for FCA 17 (2016/17 delivery) was 30,305 MW, representing a 4% decrease from the Net

ICR for FCA 16 (31,645 MW).

1.3.1 Energy Demand

New England wholesale electricity demand, otherwise known as Net Energy Load (NEL)

, has

exhibited a downward trend in recent years, largely due to the impact of state policies

promoting the installation of energy efficiency measures, as well as the increase in the behind-

the-meter generation. However, load levels are projected to increase significantly over the next

decade due to the electrification of the heating and transportation sectors. Loads are projected

to average over 16,000 MW per hour by 2031, almost 20% above 2022 levels.

Wholesale electricity demand in 2022 similar to 2021

In 2022, load was within 0.1% of 2021 wholesale electricity demand (13,576 MW in 2022 vs

13,566 MW in 2021), largely due to similar weather conditions. Annual and quarterly average

loads from 2018 to 2022 are shown in Figure 1-12 below.

NEL is net of (excludes) electricity demand that it met by “behind-the-meter” generation, including photovoltaic

generation, not participating in the wholesale market. It also excludes pumped-storage demand since pumped-storage

facilities are energy neutral.

See ISO-NE’s 2022 CELT Report, s ection 1.5.2, at https://www.iso-ne.com/system-planning/system-plans-studies/celt

2022 Annual Markets Report 36 ISO New England Inc.

ISO-NE PUBLIC

Figure 1-12: Average Hourly Load by Quarter and Year

With the exception of 2020, when load was diminished due to the COVID-19 pandemic in Q1

and Q2, 2022 annual load was lower than previous years, reflecting annual variations in

weather and the continued growth in energy efficiency and behind-the-meter solar

generation.

At a quarterly level, load in each quarter of 2022 was within a narrow (4%) range

of 2021 quarterly load levels. Higher winter loads in Q1 were offset by lower loads during

spring (Q2) and fall (Q4) driven partially be an increase in behind-the-meter solar electric

production.

Distribution of load levels in-line with prior years

New England’s system load over the last five years is shown as load duration curves in Figure

1-13 below. A load duration curve depicts the relationship between load levels and the

frequency that load levels occur. The inset graph highlights the 5% of hours with the highest

load levels for each year.

The comparatively high average load in 2018 was driven by cold temperatures in Q1 and warmer tem peratures in Q3.

10,000

11,000

12,000

13,000

14,000

15,000

16,000

17,000

18,000

Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4

2018 2019 2020 2021 2022

Annual Average

Quarterly Average

2022 Annual Markets Report 37 ISO New England Inc.

ISO-NE PUBLIC

Figure 1-13: Load Duration Curves

The 2022 load duration curve has a similar overall profile to prior years (with the exception of

2018), showing no significant differences between the frequencies of various load levels.

Looking at periods of high load, and typically tighter supply/demand conditions, the inset graph

focuses on the top 5% of hourly load. Peak load continues to be concentrated during summer

months due to increased air-conditioning demand. Therefore, weather differences tend to

explain annual variations during the top 5% of hours. The highest peak loads, especially in the

top 1% of all hours, occurred due to hot weather at the end of July and beginning of August

2022. On August 4, load peaked at 24,780 MW with a temperature of 92⁰F.

Controlling for weather effects, wholesale demand continued to decrease due to energy efficiency

and behind-the-meter solar generation

The trend in estimated gross energy demand, wholesale market demand and the contribution of

energy efficiency and behind-the meter solar is presented below.

In addition to the trend

lines, Figure 1-14 also shows the average quarterly breakdown of each category

Weather-normalized load adjusts observed load for the effects of weather, leap year and non-holiday weekdays.

Energy Efficiency is based on aggregated performance of installed measures on end-use customer facilities that reduce

the total amount of electrical energy needed, while delivering a comparable or improved level of end-use service. Such

measures include, but are not limited to, the installation of more energy efficient lighting, motors, refrigeration, HVAC

equipment and control s ystems, envelope measures, operations and maintenance procedures, and industrial process

equipment. Energy Efficiency and Demand Response Distributed Generation (DG) measures are aggregated to On -Peak

and Seasonal-Peak resources. Performance of DG accounts for only 5% of energy efficiency performance.

Behind-the-meter s olar production estimates provided by ISO New England’s system planning department.

8,000

10,000

12,000

14,000

16,000

18,000

20,000

22,000

24,000

26,000

28,000

0% 20% 40% 60% 80% 100%

Hourly Load (MW)

2018 2019 2020 2021 2022

2022 Annual Markets Report 38 ISO New England Inc.

ISO-NE PUBLIC

Figure 1-14: Average Quarterly Weather-Normalized Load with Energy Efficiency and Solar Impacts

Weather-normalized gross load (dashed purple line), which shows load without the effects of

energy efficiency (EE) and behind-the-meter solar, has continued to grow. Weather-normalized

net load (solid blue line), the remaining demand satisfied through the wholesale energy market,

trended downward over the past five years.

The gap between gross and net load is the combined impact of energy efficiency (green bar

area) and behind-the-meter solar generation (gold bar area). Greater energy efficiency and

behind-the-meter solar generation have typically helped offset increases in gross load, causing

weather-normalized wholesale load to fall. Energy efficiency has a greater effect during Q1 and

Q4, while behind-the-meter solar generation has a greater impact during Q2 and Q3.

In 2022, energy efficiency reduced weather-normalized annual average load by an estimated

2,538 MW, a 2% decrease (40 MW) compared to 2021. This is in line with the ISO’s expectation

that EE will decline over time due to rising costs of eligible EE measures and the associated EE

baselines used to calculate claimable savings. Behind-the-meter solar generation reduced

weather-normalized annual average load by 426 MW (by ~3%) or nearly 14% of estimated

installed capacity (3,170 MW). The 426 MW average load reduction was a 15% increase (57

MW) compared to 2021.

1.3.2 Reserve Requirement and Margin

In New England, there are three distinct reserve requirements determined by the North

American Electric Reliability Corporation (NERC) and Northeast Power Coordinating Council

Inc. (NPCC). Figure 1-15 shows the system total 10-minute requirement (purple), which

includes both system 10-minute spin (blue) and non-spin reserves.

The system total 30-

minute requirement (green) contains the total 10-minute and 30-minute requirements.

Requirements are based on the two largest contingencies on the system (commonly known as

There are also 30-minute local reserve requirements that are not shown or discussed below. These requirements bind

infrequently, and showed no discernable trends during the reporting period.

1,000

2,000

3,000

4,000

5,000

6,000

7,000

8,000

9,000

10,000

8,000

9,000

10,000

11,000

12,000

13,000

14,000

15,000

16,000

17,000

18,000

1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4

2018 2019 2020 2021 2022

Energy Efficiency & Solar

Impacts (MW)

Energy Efficiency Behind-the-Meter Solar

Weather-Normalized Net Load Weather-Normalized Gross Load

Weather-normalized

load has declined...

Trendlines

for Gross

Load, Net

Load, and

EE+ Solar

...as energy efficiency

and solar have grown.

2022 Annual Markets Report 39 ISO New England Inc.

ISO-NE PUBLIC

first- and second-contingencies). The size of the contingencies, and therefore requirements,

have not varied significantly on an annual average basis.

Figure 1-15: Average System Reserve Requirements

The average 10-minute spin requirement was 456 MW in 2022, down 12% from 2021 (515

MW). The requirement was notably lower due to an operational change that reduced the

percentage of the ten-minute reserve requirement that must be spinning from 31% to 25% on

May 31, 2022.

The total 10-minute (1,667 MW) and total 30-minute requirements (2,458 MW)

were similar to prior years reflecting changes to the largest system contingency. In particular,

Phase II, a 2,000 MW direct current tie line connecting the Hydro-Quebec control area to New

England, was the largest contingency in 87% of hours of the past two years. Planned reductions

in 2021 and 2022 reduced the average size of Phase II as the largest contingency, which

reduced total 10-minute and 30-minute requirements.

The reserve margin measures additional available capacity over the load and reserve

requirements.

On average, there has been a healthy surplus of reserves capability on the

system compared to the requirements, and relatively few instances where the system

experienced a deficiency of capacity or reserves. These instances are included in the discussion

of interesting system events in section 3.2.8.

The annual average margins for 10-minute spin reserves (red line), total 10-minute reserves

(purple line), and total 30-minute reserves (blue line) are shown in Figure 1-16 below. The

margins are equal to the amount of reserves provided in excess of the corresponding reserve

requirement. The bars represent annual average load (gray bar) and available capacity (orange

The operational decision to change this percentage stemmed partly from an enhancement of the Energy Management

System (EMS) that led to more accurate accounting of reserves.

The reserve margin is the difference between available capacity and demand. The equations below illustrate this

relationship: i. 



 







  



   . Equation i. is

equivalent to:   



    or     

500

1,000

1,500

2,000

2,500

3,000

2018 2019 2020 2021 2022

Reserve Requirement (MW)

10-minute spin Total 10-minute Total 30-minute

2022 Annual Markets Report 40 ISO New England Inc.

ISO-NE PUBLIC

bar) during the peak hour of each day.

Combined, these bars show the difference between

load and available capacity, which is when reserve margins are typically at their lowest.

Figure 1-16: System Reserve Margin, Peak Load, and Available Capacity

Slightly lower year-over-year total 30-minute and total 10-minute reserve margins reflect

changes in the generation mix over the course of 2022. Both margins fell by about 200 MW,

primarily due to the decline in the total 10-minute reserve margin.

The lower margin was an

outcome of the dynamics between lower net interchange (231 MW less) and the corresponding

increase in native generation. More native energy generation reduced reserve capability and

contributed to lower margins. The 10-minute spin margin increased by 22 MW (15%) in 2022

because the reserve requirement declined after operational tool enhancements.

1.3.3 Capacity Market Requirements

The Installed Capacity Requirement (ICR) is the amount of capacity needed to meet the region’s

1-in-10 year reliability standard.

When developing the target capacity to be procured in the

Forward Capacity Auction (FCA), the ISO utilizes a Net ICR, which is the amount of necessary

capacity after accounting for tie benefits with Hydro-Quebec. Due to transmission limitations,

there are also local sourcing requirements (LSR) for import constrained areas and maximum

capacity limits (MCL) for export-constrained areas.

Trends in system capacity requirements, ICR and Net ICR, between 2018 and 2027 are shown in

Figure 1-17 below. The system ICR and Net ICR are represented as line series. LSRs (positive

bars) and MCLs (negative bars) are also shown.

Available capacity is the generation capacity that can be delivered within a 30 minute period: 













  



The decline of the total 10-minute requirement is reflected in the total 30-minute margin, since the total 30-minute

requirement is inclusive of the 10-minute requirement.

The ICR requirements are designed such that non-interruptible customers can expect to have their load curtailed not

more than one day every ten years.

2022 Annual Markets Report 41 ISO New England Inc.

ISO-NE PUBLIC

Figure 1-17: ICR, NICR, Local Sourcing Requirements, and Maximum Capacity Limits

The Net ICR for FCA 17 (2026/27) was 30,305 MW, down by 1,340 MW (4%) from FCA 16.

Since FCA 13, Net ICR has steadily declined due in part to decreases in the ISO’s expected load

forecast and increases in external tie benefits.

Specifically in FCA 16, the load forecast

decreased over 1,000 MW due to a revision in the reconstitution process of passive demand

response resources.

At a local level, Southeast New England (SENE) was the only import-constrained zone from FCA

10 to 16.

In FCA 17, SENE was not modelled as an import-constrained zone due to

improvements to the local transmission network and decreases in the zonal load forecast. The

Maine and Northern New England (NNE) capacity zones were modeled as separate export-

constrained capacity zones in recent auctions with similar MCL values year-over-year.

The ISO publishes Net ICR amounts and related values for all auctions here.

More information on the FCA 15 demand response reconstitution changes can be found in ISO filing to FERC in

September 2020.

Southeast New England consists of the NEMA/Boston, Southeastern Massachusetts, and Rhode Island load zones.

(15,000)

(10,000)

(5,000)

5,000

10,000

15,000

20,000

25,000

30,000

35,000

40,000

9 10 11 12 13 14 15 16 17

2018-19 2019-20 2020-21 2021-22 2022-23 2023-24 2024-25 2025-26 2026-27

Forward Capacity Auction Number (Commitment Period)

Connecticut NEMA/Boston SEMA/RI

Southeast New England Maine Northern New England

Net ICR ICR

Future Periods

2022 Annual Markets Report 42 ISO New England Inc.

ISO-NE PUBLIC

1.4 Imports and Exports (External Transactions)

This section summarizes the level of external transactions with neighboring control areas in

Canada and New York over the last five years. External transactions are covered in more detail

in Section 5.

Key Takeaways

Net interchange (imports minus exports) from Canada and New York averaged about 1,900

MW per hour, meeting 14% of New England demand. Like last year, the level of net interchange

continued to decline. The significant driver for this decline was an increase in exports to the

New York market, particularly over the New York North interface.

The retirement of the Indian Point nuclear generator in New York in April 2021 and increased

planned transmission outages in New York led to increased congestion and higher energy

prices relative to New England prices. These system and market dynamics have resulted in an

increase in export transactions over the New York North interface, approaching 400 MW over

the last two years. With the reduction in net interchange from New York, imports from the

Canadian provinces comprised a higher share (over 90%) of total net interchange in 2022.

New England tends to import high volumes of power from its neighboring regions, particularly

during the colder winter months when energy prices in the region can be elevated as a result of

constrained natural gas supplies. Seasonal variations in net interchange can be seen in Figure

1-18 below which shows the hourly average real-time system-wide net interchange value by

quarter and year. The black line series illustrates each year’s hourly average net interchange,

while the colored circles show the average net interchange for each specific quarter.

Figure 1-18: Average Hourly Real-Time Pool Net Interchange by Quarter and Year

Net interchange from Canada and New York averaged about 1,900 MW in 2022, representing its

lowest level over the five-year period. This marks the second year in a row that average annual

500

1,000

1,500

2,000

2,500

3,000

3,500

Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4

2018 2019 2020 2021 2022

2022 Annual Markets Report 43 ISO New England Inc.

ISO-NE PUBLIC

net interchange decreased, and the 2022 value is nearly 30% less than the value observed in

2020 (2,680 MW). The primary driver for the reduction in net interchange in recent years is the

increase in exports over the New York North interface, which went from 487 MW per hour on

average in 2020 to 874 MW per hour on average in 2022. One factor behind this increase in

exports to New York is the retirement of the Indian Point nuclear generator in New York in

April 2021, which was followed by planned transmission work in the region during 2022. As a

result of this decline in net interchange with New York, a significantly higher share of New

England’s net interchange came from the Canadian provinces in 2022 (93%) than from New

York (7%).

From a seasonal perspective, Q1 2022 was the only quarter that saw a year-over-year increase

in net interchange. This was due to increased imports from New Brunswick and decreased

exports over Northport-Norwalk and Cross Sound Cable, the two interfaces connecting Long

Island and New England. Meanwhile, from Q2 2022 through Q4 2022, net interchange

decreased year-over-year due to generator outages or retirements and increased transmission

work. In Q2 2022, net interchange decreased by 13% (or 242 MW) due to decreased net

interchange from New Brunswick. In April 2022, a 660 MW nuclear generator in New

Brunswick began an extended planned outage. The generator’s outage led to reduced imports

as it typically accounts for over 40% of New Brunswick’s native generation. This outage

continued into July, and additionally contributed to the year-over year decrease in net

interchange during Q3 2022.

In Q3 2022, net interchange averaged 273 MW at New York North, a 55% (or 339 MW)

decrease compared to Q3 2021. This was the lowest level of net interchange at New York North

for any Q3 over the prior five years. Transmission work and congestion continued to impact net

interchange in Q4 2022. In Q4 2022, average net interchange at New York North decreased by

280 MW year-over-year (412 MW vs. 692 MW). New York prices averaged $4.86/MWh higher

than New England compared to $3.01/MWh higher in Q4 2021. The increased price was due to

increased congestion in New York, with the congestion component of the New York LMP at New

York North increasing from $16.53/MWh to $34.97/MWh year-over-year.

2022 Annual Markets Report 44 ISO New England Inc.

ISO-NE PUBLIC

Section 2

Market Structure and Competitiveness Assessment

This section assesses the level of competition in the wholesale electricity markets in New

England. Competition is important because it ensures that the prices consumers pay and that

producers receive for the products are the result of competitive forces and not unduly

influenced by the exercise of market power. In situations where the electricity markets are

unable to achieve competitive outcomes, market power mitigation controls may be necessary.

The IMM performs reviews across various ISO electricity markets to identify these situations

and limits their impact through the mitigation process.

This section is organized in the following order:

 (2.1) Energy market

 (2.2) Energy market mitigation

 (2.3) Forward capacity market (FCM)

 (2.4) Forward capacity market mitigation

 (2.5) Financial transmission rights (FTR) market

 (2.6) Ancillary services markets, which includes the forward reserve market (FRM) and

the regulation market

In general, we find that ISO New England capacity, energy, and ancillary service markets

exhibited competitive outcomes in 2022. Across the majority of these markets, metrics

indicated that levels of competition were generally sufficient to limit opportunities where

market power could be exercised. However, this was not the case for the Summer 2022 FRM

auction, which was found to be structurally uncompetitive.

The IMM reviewed conduct in this

auction in some detail and we were unable to conclude that the exercise of market power had

occurred. Additionally, market concentrations across the different markets generally remained

in historical ranges and mitigation was relatively infrequent.

To further strengthen market safeguards, the IMM recommends that the ISO consider a number

of changes related to energy market mitigation. These changes include (1) reviewing the

current thresholds for system-wide and local market power in the energy market; (2)

eliminating the energy offer mitigation exemption for non-capacity resources in the day-ahead

energy market; (3) extending mitigation to include resources in export-constrained areas; and

(4) reviewing the reference level calculation methodology for energy offers. These

recommendations are discussed in more detail in Section 2.2.2.

Importantly, the IMM does not have defined mitigation authority for certain markets, including the regulation and

Forward Reserve Markets.

See Section 2.6.1 for further discussion of the Summer 2022 auction competitiveness.

2022 Annual Markets Report 45 ISO New England Inc.

ISO-NE PUBLIC

2.1 Energy Market

This section presents a number of metrics on the structure and competitiveness of the energy

markets. Specifically, we present and interpret the following:

 high-level market concentration measures for the supply side (2.1.1) and demand side

(2.1.2) of the real-time market,

 supply-side structural market power tests in the real-time market (2.1.3): pivotal

supplier test (PST) and the residual supply index (RSI),

 supply offer cost mark-up metrics in the day-ahead (2.1.4) and real-time(2.1.5) markets,

and

 degree of economic withholding in the real-time market (2.1.6).

Key Takeaways

In 2022, the market concentration among the four largest firms controlling supply in the real-

time market was in line with the values of the last five years, at 44%. The four largest load-

serving entities served 53% of total load in real time, the lowest share of electricity purchased

over the five-year reporting period.

There were more instances of structural market power in real-time because of slightly lower

total operating reserve margins due to less supply. However, operating reserve margins

(surplus) remained relatively high on average compared to the underlying requirement. There

were more five-minute intervals with pivotal suppliers in 2022 than in 2019-2021. Additionally,

the average RSI in the real-time energy market fell from 106.0 in 2022 to 104.6 in 2021. Despite

the lower RSI and higher frequency of pivotal suppliers in 2022, mitigation was still relatively

infrequent and did not indicate any significant concerns.

Markups in the real-time and day-ahead markets were close to zero or negative during all years

of the reporting period. In 2022, withheld economic capacity was relatively low (below ~ 2%)

and generally in line with levels seen in past years.

2.1.1 C4 Concentration Ratio for Generation

This subsection presents an analysis of supplier market concentration among the four largest

firms controlling generation and scheduled import transactions in the real-time energy market.

The measure, termed the “C4,” is useful in understanding the general trend in supply

concentration as companies enter, exit, or consolidate control of supply serving the New

England region over time.

As shown in Figure 2-1 below, the C4 value for 2022 increased

slightly from the prior year, to 44%.

The C4 is the simple sum of the percentages of system-wide market supply provided by the four largest firms in on-peak

hours of the year and accounts for affiliate relationships among suppliers.

This C4 analyses for both supply and demand does not account for market participants with both load and generation

positions. These firms generally have less incentive to exercise market power. Any spot market actions that would tend to

2022 Annual Markets Report 46 ISO New England Inc.

ISO-NE PUBLIC

Figure 2-1: Real-time System-wide Supply Shares of the Four Largest Firms

The C4 values of the last five years range between 42-44%, indicating low levels of system-wide

market concentration, particularly because the market shares are not highly concentrated in

any one company. In 2022, the total on-peak supply of generation and imports was about

66,400 GWh, of which about 29,000 GWh (44%) came from the four largest suppliers.

The C4

line indicates a tight range over the past five years, and one company maintained a dominant

share of on-peak supply, with the split among the top four suppliers remaining stable.

2.1.2 C4 Concentration Ratio for Load

The C4 for load measures the market concentration among the four largest load-serving entities

(LSEs) in the real-time energy market and is presented in Figure 2-2 below.

raise prices to benefit their generation would come at a cost to their load position. Any actions that w ould suppress prices

to benefit their load would come at a cost to their generation position.

On-peak hours last from HE 8 to HE 23 on Monday to Friday.

The C4 load metric also accounts for any affiliations among different LSE and includes on-peak hours only.

44%

43%

42%

44%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

2018 2019 2020 2021 2022

Supplier 1 Supplier 2 Supplier 3

Supplier 4 Other Suppliers C4

2022 Annual Markets Report 47 ISO New England Inc.

ISO-NE PUBLIC

Figure 2-2: Real-time System-wide Demand Shares of the Four Largest Firms

In 2022, the real-time load obligation (RTLO), or the amount of electricity purchased, was

65,479 GWh.

Overall, the four largest LSEs served 49% (31,913 GWh) of total load, the lowest

share of electricity purchased over the five-year period. The red C4 line shows that the total

load share of the four largest LSEs decreased in 2022 and falls 4% lower than 2018 levels. The

decrease was largely due to one participant divesting a large share of its generation and load

into an independent company.

The observed C4 values presented above indicate relatively low levels of system-wide market

concentration. The above figure shows that individual shares are not highly concentrated in any

one company. Additionally, there is no evidence to suggest that LSEs exhibit energy market

bidding behavior that would suppress prices. On average over 100% of demand clears in the

day-ahead market and the aggregate demand curve is relatively price-insensitive around

expected LMPs (see Section 3.2.5 on Demand Bidding).

2.1.3 Residual Supply Index and the Pivotal Supplier Test

This section examines opportunities for participants to exercise market power in the real-time

market using two metrics: the pivotal supplier test (PST) and the residual supply index (RSI).

Both of these widely used metrics identify instances when the largest supplier has market

power.

The RSI represents the amount of demand that the system can satisfy without the

largest supplier’s available energy and reserves. If the value is less than 100, the largest

supplier is needed to meet demand, and could potentially exercise market power (if permitted).

The firms labeled “LSE 1”, “LSE 2” and so on are not necessarily the same LSE across all years; these are generic labels for

the top four firms during a given year.

Real-time load obligation is measured as all end-use wholesale load in the ISO New England region, along with all

exports. The difference between this number and the real-time generation obligation should equate to energy losses.

In this report, the RSI and pivotal s upplier tests are calculated using supply, load, and reserve requirement data from the

ISO’s real-time market software. This differs from the calculation methodology of previous AMRs, which used the results

and inputs of the real-time pivotal supplier test conducted by the mitigation software process.

When the RSI exceeds 100, there is sufficient supply available to meet demand without any generation from the largest

supplier. In this case, no individual supplier is pivotal and sufficient competition exists in the market.

53%

55%

57%

56%

49%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

2018 2019 2020 2021 2022

LSE 1 LSE 2 LSE 3 LSE 4 Other LSEs C4

2022 Annual Markets Report 48 ISO New England Inc.

ISO-NE PUBLIC

Further, if the RSI is less than 100, there is at least one pivotal supplier. Conversely, when the

RSI exceeds 100, there is enough supply available to meet demand without any generation from

the largest supplier. In this case, no individual supplier is pivotal and sufficient competition

exists in the market.

The data used to calculate the RSI come from the ISO’s real-time market software (the Unit

Dispatch System, or UDS). Based on these data, the RSI for an interval t is calculated as follows:











 











 



Pivotal suppliers are identified for every five-minute pricing interval by comparing the real-

time supply margin

to the sum of each participant’s total supply that is available within 30

minutes.

When a participant’s available supply exceeds the supply margin, they are

considered pivotal. The number of five-minute intervals with at least one pivotal supplier are

divided by the total number of five-minute intervals in each year to obtain the percentage of

intervals with pivotal suppliers.

The average RSI for all five-minute real-time pricing intervals and the percentage of five-minute

intervals with pivotal suppliers are presented in Table 2-1 below.

Table 2-1: Residual Supply Index and Intervals with Pivotal Suppliers (Real-time)

Year

% of Intervals With At

Least 1 Pivotal Supplier

RSI

2018

30.7%

103.6

2019

14.7%

106.4

2020

16.6%

106.9

2021

18.0%

106.0

2022

24.9%

104.6

There were more five-minute intervals with pivotal suppliers in 2022 than in 2019-2021, but

fewer than in 2018. By this measure, this indicates that suppliers faced relatively less

competition in 2022 than during the three previous years. The increase in the number of

intervals with at least one pivotal supplier was driven by lower total 30-minute reserve

margins in 2022.

Reserve margins have fluctuated throughout the reporting period for several reasons, including

generator outages, resource additions or retirements, and changes in the reserve requirement.

In 2022, the average total 30-minute reserve surplus was 2,888 MW, down by about 200 MW

from 3,091 MW in 2021. The decrease was primarily driven by changes in net imports. In 2022,

The real-time supply margin measures the amount of available supply on the system after load and the reserve

requirement are satisfied. It accounts for ramp constraints and is equal to the Total30 reserve margin: 







+ [Net Interchange] - - [ ]

This is different from the pivotal supplier test performed by the mitigation software, which does not consider ramp

constraints when calculating available supply for each participant. Additionally, the mitigation software determines pivotal

suppliers at the hourly level.

2022 Annual Markets Report 49 ISO New England Inc.

ISO-NE PUBLIC

average hourly net imports decreased by 231 MW compared to 2021. As a result, native

generation provided more energy and fewer reserves.

Duration curves that rank the average hourly RSI over each year in descending order are

illustrated in Figure 2-3 below. The figure shows the percent of hours when the RSI was above

or below 100 for each year. An RSI below 100 indicates the presence of at least one pivotal

supplier.

Figure 2-3: System-wide Residual Supply Index Duration Curves

Like the pivotal supplier statistics, Figure 2-3 shows that there was a lower availability of

competitive supply in 2022 and 2018 compared to 2019-2021. The RSI was above 100 in 75%

of real-time pricing intervals in 2022, which was lower than the 2021 result (82%). Despite the

lower RSI and higher frequency of pivotal suppliers in 2022, mitigation was still relatively

infrequent and did not indicate any significant concerns.

2.1.4 Day-Ahead Price-Cost Markup

In a perfectly competitive market, all market participants’ energy supply offers would equal

their marginal costs. The energy component of the LMP would then be set by the supply offer or

demand bid on the margin. However, in practice, participants can raise their supply offers

above marginal costs. Uncompetitive offers priced above marginal cost can distort prices and

impact generator commitment decisions, leading to inefficient market outcomes. Though the

IMM administers mitigation rules in the energy market to prevent the exercise of market

power, participants are allowed to increase their offers within a certain threshold before

mitigation is applied.

The price-cost markup estimates the divergence of the observed market outcomes from the

ideal scenario in which all energy supply is offered at marginal cost. The results provide insight

on how uncompetitive offer behavior impacts the day-ahead energy market.

Additionally, there was a 43 MW reduction in offline reserves from two generators that shed their CSOs for FCA 13 (J une

2022 – May 2023). Section 1.3.2 provides additional information on reserve margin trends.

100

110

120

130

140

150

0% 20% 40% 60% 80% 100%

RSI

2018

2019

2020

2021

2022

2022 RSI lower than that of 2019-2021

across most ranked observations

No pivotal supplier in 75% of

real-time intervals in 2022

2022 Annual Markets Report 50 ISO New England Inc.

ISO-NE PUBLIC

To calculate price-cost markup, the IMM simulated the day-ahead market clearing using two

scenarios:

 Scenario 1 is an offer case that uses actual day-ahead energy market supply offers

submitted by market participants.

 Scenario 2 is a marginal cost case that assumes all market participants offered at an

estimate of their short-run marginal cost.

The price-cost markup is then calculated as the percentage difference between the annual

generation-weighted LMPs for the offer case and the marginal cost case simulations:

 





 







 

- 



is the annual generation-weighted LMP that results from actual generator

offers, and

- 



is the annual generation-weighted LMP that would occur if generator

offers were replaced with their respective marginal costs.

A larger price-cost markup means that a larger component of the LMP is the result of inflated

supply offers.

The annual day-ahead price-cost markup values from the simulation are shown in Table 2-2

below.

Table 2-2: Day-Ahead Price-Cost Markup, %

Year

Price-Cost Markup

2018

-4.0%

2019

-2.4%

2020

0.9%

2021

-0.6%

2022

-1.8%

The IMM uses the PROBE, or “Portfolio Ownership and Bid Evaluation,” simulation model for this analysis. The softwa re

simulates the day-ahead LMP-based market clearing. See http://www.power-gem.com/PROBE.html. This is a more

dynamic approach than calculating the difference between a static offer price and marginal cos t. Rather, this approach re-

runs the market optimization process with both as-offered and competitive (marginal cost) supply curves, and calculates

the difference in the resulting LMPs.

The marginal costs estimates are based on underlying variable cost data and generator heat rate parameters used in the

calculation of IMM reference levels. Reference levels are calculated pursuant to Appendix A to Market Rule 1 of the ISO

tariff and are used in market power mitigation analyses to represent a competitive offer. Where a good estimate of

marginal cost does not exist the marginal cost is set equal to the supply offer. Some differences between estimated and

actual marginal costs are to be expected.

Prior to this report, this metric used a different methodology to estimate marginal costs. This is why the values in this

report are lower than values for the same years in previous reports.

2022 Annual Markets Report 51 ISO New England Inc.

ISO-NE PUBLIC

In 2022, the price-cost markup for the day-ahead energy market remained low (below the

strictest mitigation threshold of 10%) at -1.8%. This indicates that the average marginal

resource offered below its marginal costs, and that offers deviating from marginal cost

decreased the generation-weighted day-ahead energy market price by approximately 1.8%.

This result is similar to that of prior years, and is consistent with normal year-to-year variation

given modeling and estimation error.

This indicates that competition among suppliers in the

day-ahead market limited their ability to inflate LMPs by submitting offers above marginal cost.

In this assessment, we reviewed price-cost markup values at an hourly level and compared the

peak load hour price-cost markup with the forecasted supply margin at peak. Comparing these

attributes provides insight into whether participants take advantage of tight system conditions,

when market power tends to be more of a concern, by increasing offer markups during those

times. There was no meaningful correlation between the price-cost markup and the supply

margin in 2022, indicating that the day-ahead market remained competitive even when the ISO

expected supply margins to be low and market power was present.

2.1.5 Real-Time Price-Cost Markup (Lerner Index)

This analysis used different methods for the day-ahead and real-time price cost markup

calculations. Recall that the day-ahead analysis above was performed by calculating generation-

weighted LMPs for each scenario using simulation software. The real-time analysis calculated

(load-weighted) LMPs by creating supply curves for 1) available generation

by offer price and

2) available generation by marginal cost estimate, and then intersecting real-time demand with

each. The real-time price cost markup is presented in Table 2-3 below.

Table 2-3: Real-Time Price Cost Markup, %

Year

Price-Cost Markup

2018

-8.9%

2019

-7.4%

2020

-3.1%

2021

0.2%

2022

-1.7%

Annual load-weighted markups were close to zero or negative during all years of the reporting

period. Consequently, average generator offers were close to or below marginal costs,

indicating that competition limited participants’ abilities to inflate real-time prices.

Gas-fired generator offer behavior had the largest impact on annual markup values. Gas-fired

generators may offer at values lower than their estimated marginal costs for multiple reasons,

Note that the IMM’s estimates of marginal cost are an approximation of actual marginal costs, and the simulations used

to calculate the price-cost markup are subject to modeling differences when compared to the market model the ISO runs

for the day-ahead market.

Available generation is equal to on-line generation plus generation capacity that can come on-line within 30 minutes. It

comes from on-line generators (both long lead-time and fast-start) and off-line fast-start generators.

2022 Annual Markets Report 52 ISO New England Inc.

ISO-NE PUBLIC

including managing gas burn to nominations or offering lower output levels below their fuel

price adjusted reference level.

Renewable generator offer behavior also tend to reduce

markups, but to a lesser extent. Though it is common for renewable generators to offer below

marginal cost, these generators are typically smaller and/or export-constrained.

Differences between the real-time and day-ahead price cost markups values are due to several

factors, including: 1) differences in the methodologies used to calculate the price cost markup in

each respective market; 2) modeling differences between the day-ahead and real-time energy

markets; and 3) real-time events that the day-ahead market did not anticipate.

2.1.6 Real-Time Economic Withholding

Economic withholding occurs when suppliers offer above marginal cost in order to prevent

some generation that would otherwise be economic from clearing, which in turn raises the

market price. The quantity that does not clear as result of suppliers offering above cost is

considered economically withheld.

We estimate the economically withheld MWs for each generator in every real-time interval as

the difference between:

a) the quantity that was economic (i.e., the sum of MWs where marginal cost ≤ LMP) and,

b) the actual quantity offered (i.e., the sum of MWs where offer price ≤ LMP).

In cases where the quantity offered exceeds the quantity that was economic, the withheld MWs

are set to zero (i.e., withheld MWs cannot be negative). This analysis considers only non-fast-

start generators that are online and all fast-start generators (online or offline), and it does not

assess potential withholding by offline, non-fast-start generators.

Hourly economic withholding (as a percent of capacity) during on-peak hours in each of the

past five years is summarized in Figure 2-4 below. Note the curves depict the distribution of

hourly withholding, where the widest sections of each curve represent the most-frequently

observed levels of withholding.

Results are broken down for two groups: combined withholding

by the top four generating companies (those with the largest share of generation) versus all

others.

The most extreme negative markup values occurred during the winter months, when generators are more likely to have

fuel price adjustments (FPAs) in place. FPAs are applied to the entire output curve and replace the default gas index value,

but gas-fired generators often use FPAs to reflect the price of incremental gas needed at higher output levels, which might

be higher than the market value of liquidating gas that they have already purchased.

For example, if the LMP is $30/MWh and a participant offered 50 MW at $45/MWh but had a $20/MWh marginal cost,

then those 50 MW would be considered economically withheld. The IMM cost-based reference level is used as the

generator’s marginal cost. The calculation accounts for ramp rate limitations and fast-start generators’ startup and no-load

costs.

2022 Annual Markets Report 53 ISO New England Inc.

ISO-NE PUBLIC

Figure 2-4: Hourly Real-time Economic Withholding During On-Peak Hours

In 2022, economic withholding was relatively low across both groups (generally below 2%) and

generally in line with levels seen in past years. Although not presented in the figure, levels of

economic withholding did not increase when reserve margins where low, suggesting that

suppliers were largely unable or did not attempt to take advantage of tight system conditions

by economically withholding.

2.2 Energy Market Mitigation

The IMM reviews energy market supply offers for generators in both the day-ahead and real-

time energy markets. This review minimizes opportunities for participants to exercise market

power.

Under certain conditions, the IMM will mitigate generator supply offers. Mitigation

results in a participant’s financial parameters for a generator supply offer (i.e., start-up, no load,

and segment energy offer prices) being replaced with “reference” values. The reference values

are estimated and maintained by the IMM; these values are used in mitigation to reduce impacts

on energy market pricing (LMPs) and uplift payments (NCPC) from participant offers that

appear to overstate a generator’s operating costs.

In this section, we review the frequency and underlying drivers of the various forms of energy

market mitigation in the day-ahead and real-time energy markets (2.2.1). We also provide an

overview of a number of issues we identified through our ongoing assessment of mitigation rule

performance, and we make a number of recommendations on design enhancements (2.2.2).

This review of supply offers is automated (along with the offer mitigation process), and occurs within the ISO’s energy

market software.

Seven mitigation types utilize mitigation tests and are used in ex-ante supply offer mitigation. An eighth mitigation type

for dual-fuel generators is performed after-the-fact, when a dual-fuel generator burns a low-priced fuel but s ubmits supply

offers based on a higher-cost fuel.

2022 Annual Markets Report 54 ISO New England Inc.

ISO-NE PUBLIC

Key Takeaways

Energy market mitigation is an infrequent occurrence. In 2022, only 930 asset hours of

mitigation occurred, compared to 1.3 million asset hours of generator operation that were

subject to potential mitigation. Over the past five years, mitigation instances have been

relatively stable, fluctuating between 874 and 990 asset hours per year. Mitigation

frequencies are low because: (1) energy market mitigation requires that up to three tests

(structure, conduct, impact) be failed, prior to the imposition of offer mitigation; and (2) the

tests typically have relatively tolerant thresholds for triggering a test failure. Note that the

summary data for mitigation outcomes in Figure 2-5 below utilizes data that have been

scaled relative to asset hours subject to mitigation.

2.2.1 Mitigation Outcomes and Drivers

An indication of mitigation frequency, relative to opportunities to mitigate generators, is

illustrated in Figure 2-5 below.

It compares asset hours of structural test failures for

dispatch and commitment (depending on mitigation type) to asset hours of mitigations. To

provide additional context, the values in the figure represent multiples of one percent of total

asset hours subject to potential mitigation.

As set hours refer to the commitment and operation hours of a generator. For example, a generator (asset) committed

for reliability for a 12-hour period would represent 12 asset hours of commitment. If that asset were mitigated upon

commitment, then 12 asset hours of mitigation would occur. For constrained areas, if 10 assets were located in an import-

constrained area for two hours, then 20 asset hours of s tructural test failures would have occurred. If a pivotal supplier has

seven assets and is pivotal for a single hour, then seven hours of structural test failures would have occurred for that

supplier; however, more than one supplier may be pivotal during the same period (especially during tighter system

conditions), leading to a larger numbers of structural test failures than for other mitigation types. Manual dispatch energy

commitment data indicate asset hours of manual dispatch (i.e., the asset hours when these generators are subject to

commi tment). Finally, Start-up and no-load commitment hours are not shown because mitigation hours equal

commitment hours.

Because the general threshold commitment and constrained area commitment conduct tests resulted in only two

instances of mitigation during the review period, those mitigation types have been omitted from the figure. The structural

test failures associated with each mitigation type are the same as for the respective general threshold energy and

constrained area energy structural test failures. Ex-post, dual-fuel mitigation also is not summarized in the graphs, since

the process for applying that mitigation does not involve conduct, structural and market impact tests. However, we do

provide results for this mitigation type in the discussion provided below.

2022 Annual Markets Report 55 ISO New England Inc.

ISO-NE PUBLIC

Figure 2-5: Energy Market Mitigation

On average, approximately 1.2 million asset hours of ISO-committed generation are subject to

the IMM’s mitigation rules each year.

In 2022, the total asset hours reached 1.3 million asset

hours, with approximately 85,000 asset hours (6.6%) failing structural tests. Mitigation asset

hours represented a very small fraction of potential asset hours subject to mitigation. For

example, real-time reliability commitment mitigation totaled just 231 asset hours for 2022,

equaling 0.02 of asset hours scaled to 1% in the figure.

Overall, mitigation occurs very infrequently relative to the initial triggers for potential

mitigation. The highest frequency of mitigation occurs for reliability commitments (0.02% in

2022, light blue and orange shading) driven by the frequency of local reliability commitments

and the relatively tight conduct test thresholds (10% supply offer/reference level threshold).

General threshold (pivotal supplier) mitigation and constrained area mitigation (green, dark

blue, and yellow shading) have had the lowest mitigation frequency, just 0.01% of asset hours

in 2022. Both of these mitigation types have relatively tolerant conduct test and market impact

test thresholds, reducing the likelihood of mitigation given a structural test failure.

Reliability commitment mitigation: Reliability commitments primarily occur to satisfy local

reliability needs, and are generally due to routine transmission line outages, outages facilitating

The asset hours subject to mitigation are estimated as on-line generators with a dispatchable range above eco min.

Each such on-line generator during a clock hour represents one asset hour of generation potentially subject to energy

market mitigation.

2022 Annual Markets Report 56 ISO New England Inc.

ISO-NE PUBLIC

upgrade projects, or localized distribution system support.

While real-time reliability

commitments declined in 2022, mitigation asset hours increased (from 164 to 231), compared

to 2021, as a result of two generators that were needed for localized transmission needs being

frequently mitigated. Overall, reliability mitigations in the day-ahead market have declined

significantly since 2019 (from 587 asset hours to 46 asset hours in 2022).

This decrease

resulted from both a decline in reliability commitment asset hours (decline from 3,765 to 522

asset hours) and of mitigated offers in Maine and SEMA-RI (decline of 540 to 30 asset hours),

the two areas with the highest frequency of reliability upgrades.

Start-up and no-load (SUNL) commitment mitigation: This mitigation type addresses grossly

over-stated commitment costs (relative to reference values), which could otherwise result in

very high uplift.

SUNL mitigations occur very infrequently and may reflect a participant’s

failure to update energy market supply offers as fuel prices fluctuate – particularly natural gas.

Generators associated with just three participants accounted for 84% of these mitigations.

There were just 183 asset hours of SUNL mitigation in 2022.

Constrained area energy (CAE) mitigation: With relatively tolerant conduct and market impact

test thresholds, the frequency of this mitigation is low relative to the frequency of structural

test failures. Over the five-year period, only 91 asset hours occurred in the real-time energy

market and only 162 asset hours in the day-ahead energy market. The frequency of structural

test failures follows the incidence of transmission congestion and import-constrained areas

within New England.

The 2022 structural test failures totaled 15,000 asset hours. In fall 2022,

a constraint in Connecticut resulted in a large number of structural test failures during the

month of November (i.e., 10,000 asset hours). Both planned and unplanned transmission

outages caused this constraint. Although this constraint caused an 11-fold increase in structural

test failures (compared to summer 2022), only a total of 23 asset hours of mitigation occurred

during fall 2022.

General threshold energy (GTE) mitigation: This mitigation type has the lowest frequency of any

mitigation type, and has the most tolerant conduct test and market impact thresholds. GTE

mitigation occurred for only 175 asset hours in 2022. This happened despite of the highest

frequency of structural test failures (i.e., pivotal supplier asset hours) for any mitigation type.

As expected, structural test failures tend to occur for lead market participants with the largest

portfolios of generators. Five participants accounted for 74% of the structural test failures over

the review period.

This mitigation category applies to most types of “out-of-merit” commitments, including local first contingency, local

second contingency, voltage, distribution, dual-fuel resource auditing, and any manual commitment needed for a reason

other than meeting system load and operating reserve constraints. See Market Rule 1, Appendix A, Section III.A.5.5.6.1.

Reliability commitments are typically made in the day-ahead energy market and carry over to the real-time energy

market. Hence, day-ahead reliability commitments account for approximately 69% of the reliability commitment asset

hours in the real-time energy market.

The conduct test for this mitigation type compares a participant’s offers for no-load, start-up and incremental energy

cost up to economic minimum to the IMM’s reference values for those same parameters. It uses a very high conduct test

threshold (200% applied to the start-up, no-load, and offer segment financial parameters).

Day-ahead energy market structural test failures are not being reported at this time. This results from questions about

some of the source data for these failures. We expect to report on these structural test failures in future reporting.

2022 Annual Markets Report 57 ISO New England Inc.

ISO-NE PUBLIC

Manual dispatch energy (MDE) mitigation: Manual dispatch energy mitigation can occur when a

generator is manually dispatched by the ISO in the real-time market. Like reliability

commitment mitigation, manual dispatch energy mitigation has a relatively tight conduct test

threshold (10%), and had comparable mitigation hours in 2022. There were 1,780 asset hours

of manual dispatch and 251 asset hours of mitigation. Manual dispatch is relatively infrequent

in the real-time energy market, with typically fewer than 2,000 asset hours occurring each year

and primarily (85%) involving combined-cycle generators.

Dual-fuel ex-post mitigation: Dual-fuel mitigations occur relatively infrequently, typically fewer

than 100 asset hours of mitigation per year. In 2022, only five asset hours of dual-fuel

mitigation occurred; this happens when a dual-fuel generator’s reference level is set to the

highest-cost fuel but the generator actually operates using its lower cost fuel.

This is consistent with manual dispatch frequently occurring i n the context of (1) regulation service provided to the real-

time energy market and (2) the need for relatively flexible generators to be positioned away from the market software-

determined dispatch to address short-term issues on the transmission grid.

2022 Annual Markets Report 58 ISO New England Inc.

ISO-NE PUBLIC

2.2.2 Energy Market Mitigation Assessment

The IMM routinely assesses how the energy market mitigation rules and processes are

functioning.

This is done with the perspective of relying on competition to drive efficient

prices and incentives, and intervening only when markets are structurally uncompetitive and

participants may have the ability and incentive to exercise market power and inflate prices

above competitive levels. The market power mitigation rules need to strike a reasonable

balance between producer and consumer interests, and in turn prescribe adequate threshold

tests to determine when market monitors override generators supply offers.

Recommendation

We have identified a number of areas of potential improvement to the current rules and we

recommend for ISO and stakeholder consideration. In particular, IMM recommends that the

following four issues be addressed:

1. Review of the current mitigation thresholds that apply to instances of system-wide

and local market power. The current thresholds allow for considerable latitude in

supply offers levels over competitive benchmarks (300% and 50%) and have been in

place for many years with little empirical support. With the benefit today of a rich

history of supply offer data and robust simulation tools to measure impact and

incentives of withholding, the assessment would consider whether the current

thresholds adequately limit the exercise of market power under a plausible set of

system scenarios.

2. Eliminate the energy offer mitigation exemption for non-capacity resources in the

day-ahead energy market.

3. Extend the scope of offer mitigation to cover the potential exercise of market power

in export-constrained areas.

4. Review the methodologies for determining reference levels, which are used to

evaluate if an offer is competitive (the “conduct test”). Currently, reference levels can

be based on marginal cost, or historical fuel-adjusted accepted supply offers or LMPs.

We have observed instances in which the latter two methodologies produce

unreasonably high reference levels. In particular, we recommend that only marginal

cost-based reference levels be relied on for generators that have robust cost

estimates with the IMM.

Mitigation Thresholds

The current mitigation rules potentially allow for economic withholding in the ISO’s energy

markets because of relatively generous mitigation thresholds. The current mitigation rules rely

on up to three tests, before mitigation is applied to a generator’s energy market supply offer: a

structural test, a conduct test and an impact test.

The External Market Monitor also has responsibility for assessing the quality and appropriateness of the ISO’s supply

offer mitigation activities. See Market Rule 1, Appendix A, Section 2.2.(d).

The IMM believes that a review of the threshold levels for both the conduct and impact tests would be appropriate, to

ensure that opportunities for uncompetitive behavior have been minimized.

2022 Annual Markets Report 59 ISO New England Inc.

ISO-NE PUBLIC

As an example of our concern with the current thresholds, we have summarized instances of

potential economic withholding for “general threshold energy” (GTE) in the real-time energy

market. Figure 2-6 provides summary statistics for pricing intervals during 2022, when at least

one supplier was pivotal (i.e., failed the mitigation structural test) and pivotal supply was

offered at three different conduct test threshold levels: 10% above the marginal cost of supply,

50% above the marginal cost of supply, and the current conduct test threshold (i.e., the

minimum of 300% or $100 above the marginal cost of supply).

Figure 2-6: Potential Economic Withholding, General Threshold Energy

In 2022, the potentially-withheld capacity during these intervals averaged between 100 and

600 MW, depending on the month. The potential withholding exceeded 1,000 MW during at

least one 5-minute interval in every month. At the increased conduct test threshold levels,

conduct test failures declined significantly. At the 50% conduct test threshold, potentially-

withheld megawatt levels decline by 40-50% in most months. At the current conduct threshold

level, average potential withholding is quite low, ranging from approximately 10 MW to 200

MW.

Further, we highlight a pricing interval in June 2022, when potential economic withholding

under the GTE mitigation rules exceeded 2,000 MWs in the real-time energy market. During this

interval, the system supply margin was 24 MW (indicating an extremely constrained supply

position, relative to demand, in the real-time energy market). Figure 2-7 summarizes the

capacity offered by pivotal suppliers that was priced at more than 10%, 50% and 300% in

excess of the supply’s marginal cost.

500

1,000

1,500

2,000

2,500

3,000

3,500

4,000

Jan 22

Feb 22

Mar 22

Apr 22

May 22

Jun 22

Jul 22

Aug 22

Sep 22

Oct 22

Nov 22

Dec 22

Jan 22

Feb 22

Mar 22

Apr 22

May 22

Jun 22

Jul 22

Aug 22

Sep 22

Oct 22

Nov 22

Dec 22

Jan 22

Feb 22

Mar 22

Apr 22

May 22

Jun 22

Jul 22

Aug 22

Sep 22

Oct 22

Nov 22

Dec 22

Max Withheld MW Avg Withheld MW Min Withheld MW

10% Threshold

50% Threshold

Actual Threshold

2022 Annual Markets Report 60 ISO New England Inc.

ISO-NE PUBLIC

Figure 2-7: Supply Controlled by Pivotal Suppliers on June 13, 2022, During HE 18 (Interval 2)

More than 10 pivotal suppliers had energy market offers that were in excess of 10% above

marginal cost. Nine of these suppliers had more than 100 MW of supply priced in excess of this

level, with one pivotal supplier offering more than 800 MWs priced above this level. These

offers on a weighted average basis exceeded marginal cost by 43%. At the 50% threshold, seven

suppliers had energy market offers more than 50% above marginal cost, constituting 600 MW

of supply. At the 300% threshold, no supplier failed the conduct test threshold. Given these

results, we are concerned that the current mitigation thresholds may not provide adequate

safeguards against the exercise of market power in the ISO’s energy markets.

Supply Offers Not Associated with a Capacity Supply Obligation

Under the economic withholding rules, offer mitigation does not apply to non-capacity

resources in the day-ahead energy market, and only applies up to (and including) the energy

segment corresponding to the capacity supply obligation (CSO) for resources with partial

capacity obligations. All components of the supply offer are evaluated for economic withholding

in the real-time market. We are concerned that the omission of mitigation rules in the day-

ahead market for resource capacity lacking a CSO creates an opportunity for economic

withholding. In November 2022, we outlined the rationale for ensuring reasonably

comprehensive mitigation rules:

Under FERC’s market-based rate approach, prices resulting from market-based rates

are just and reasonable if they are a result of competitive markets among participants

that lack market power and/or are mitigated. In other words, market prices must be

competitive in order to be just and reasonable. Because electricity markets are more

susceptible to the exercise of market power from suppliers than many other markets

due to the relative inelasticity of demand, mitigation measures are relied upon to limit

the exercise of market power. Effective rules that mitigate or deter the exercise of

100

200

300

400

500

600

700

800

900

Withheld MWs, 10% CT Threshold Withheld MWs, 50% CT Threshold

Withheld MWs, Actual Conduct Test System Supply Margin

2022 Annual Markets Report 61 ISO New England Inc.

ISO-NE PUBLIC

market power are critical to assuring competitive outcomes and just and reasonable

rates.

Currently, almost 10 thousand MWs of active capacity (as measured by winter claimed

capability) does not have a CSO associated with it. Figure 2-8 summarizes this capacity by asset

type.

Figure 2-8: Generator Capacity Lacking a CSO by Asset Type

With the exception of photovoltaic capacity which currently is non-dispatchable, these capacity

types (totaling approximately 7,200 MW) use supply offer pricing to determine dispatch in the

ISO’s energy markets. These supply offers can be used by participants with market power to

economically withhold capacity, irrespective of whether the capacity is associated with a CSO.

The successful exercise of market power (for capacity with or without a CSO) would result in

uncompetitive market outcomes that result in market prices that are not just and reasonable.

Export Constrained Mitigation

Constrained area energy (CAE) mitigation only applies to import-constrained areas. CAE

mitigation does not address the potential for economic withholding within export-constrained

areas. There may be instances when export-constrained areas allow for the exercise of market

power by a participant. These instances occur when there is insufficient competition behind

the export constraint to result in participant supply offers reflecting competitive pricing levels;

instead, offer pricing may reflect an estimate of higher price levels (LMPs) available outside of

the export-constrained area, resulting in elevated pricing behind the export constraint. This

behavior is uncompetitive, undermines locational pricing signals, and raises costs to consumers

above the level expected in a competitive market.

See IMM’s memorandum re "Current Market Rules on Physical Witholding" November 30, 2022. The memorandum was

provided to the NEPOOL Markets Committee.

500

1,000

1,500

2,000

2,500

3,000

2022 Annual Markets Report 62 ISO New England Inc.

ISO-NE PUBLIC

Reference Pricing Methodology

The ISO’s current methodology for setting reference prices for incremental energy offers can

result in values that do not reasonably approximate a competitive offer. The current reference

price methodology uses three methods for estimating reference prices: accepted offer, LMP,

and marginal cost.

The reference level is biased upward by the methodology providing a

preference for accepted offer and LMP-based reference values, as long as they are greater than

the marginal cost reference value. The accepted offer and LMP pricing methodologies utilize

historical data for setting reference prices, based either on a generator’s accepted offers during

the prior 90 days or LMPs during operating hours over the prior 90 days. For generators that

may be called on-line during periods with high energy market prices and otherwise operate

very infrequently, the current reference pricing methodology can result in significant

distortions in reference pricing over the subsequent 90 days after operation.

Figure 2-9 provides an example of distorted reference pricing after the December 24, 2022

system event. The figure shows the actual reference prices used for a generator on March 11,

2023 in the real-time energy market, relative to its marginal cost pricing level for that day.

Figure 2-9: LMP-Based and Marginal Cost Reference Prices

As indicated in the figure, the reference prices (blue bars) are equal to the marginal cost

reference levels (orange line) in most hours. However, in hours 17-19, the reference levels are

based on the LMP methodology. The LMP-based reference prices continue to reflect the very

high LMPs that occurred during the system event from almost three months prior, and range

from $500 to $1,000/MWh. The LMP-based reference prices are not reflective of the generator’s

marginal cost of operation or other measures of potential opportunity costs of operation for

March 11. Such distortions in reference levels create opportunities for participants to

economically withhold capacity, whenever high energy market prices or high fuel prices from

prior periods result in overstated accepted-offer and LMP-based reference levels.

See Market Rule 1, Appendix A, Section 7.2.

200

400

600

800

1,000

1,200

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

$/MWh

Hour of Day

Reference Price Marginal Cost Reference Price

2022 Annual Markets Report 63 ISO New England Inc.

ISO-NE PUBLIC

2.3 Forward Capacity Market

In this section, we review the structural competitiveness of the Forward Capacity Market (FCM)

using two metrics: the Residual Supply Index (RSI) and Pivotal Supplier Test (PST).

Key Takeaways

The results of two complementary competitiveness metrics - the residual supply index and the

pivotal supplier test - indicate that, over the past five auctions, the New England capacity market

has been structurally competitive at the system level. The capacity market was most competitive

entering FCAs 14, 16, and 17, with an RSI of over 100% (no pivotal suppliers).

Reductions in the Net Installed Capacity Requirement due to lower peak forecasts and higher tie

benefits, along with the absence of major retirements over the past two FCAs, has resulted in a

higher capacity margins heading into the auction. Consequently, structural competitive indices

have improved. Further, in FCA 17 SENE was not modelled as an import constrained zone due to

a decrease in the zonal load forecast and an increase in the import capability limit into the zone.

This enhanced the overall competitiveness of the capacity market.

In FCA 17, there were no pivotal suppliers and the residual supplier index was 102%. A supplier

would have needed a portfolio of over 3,600 MWs to be pivotal.

Similar to the PST and RSI calculated for the energy market, we account for affiliations between

suppliers to reflect all capacity under a supplier’s control.

For each Forward Capacity Auction

(FCA), we consider the qualified capacity of existing resources prior to the auction given the

difficultly of predicting intra-auction new supply behavior. Both metrics are calculated for the

import-constrained zone (Southeastern New England or SENE), in addition to the larger system.

For the purposes of market power mitigation, de-list bids from a pivotal supplier above the

dynamic de-list bid threshold may be subject to mitigation.

This process ensures that

suppliers do not withdraw capacity from the auction at uncompetitive prices to raise the FCA

clearing price in a way that may benefit the remainder of their portfolio.

The PST the RSI rely on the following inputs:

 Capacity requirements – both at the system level (Net Installed Capacity requirement, or

Net ICR) and the import-constrained area level (Local Sourcing Requirement, or LSR).

The Net ICR and LSR change from year to year.

 Capacity zone modelling – different capacity zones are modelled for different FCAs

depending on the quantity of capacity in the zone and transmission constraints.

As defined in Section III.A.23.4 of the Tariff, for the purposes of this test, “the FCA Qualified Capacity of a supplier

includes the capacity of Existing Generating Capacity Resources, Existing Demand Resources, Existing Import Capacity

Resources, and New Import Capacity Resources (other than (i) a New Import Capacity Resource that is backed by a single

new External Resource and that is associated with an investment in transmission that increases New England’s import

capability; and (ii) a New Import Capacity Resource associated with an Elective Transmission Upgrade).”

Note that there are certain conditions under which capacity is treated as non-pivotal. These conditions are described in

Section III.A.23.2 of the Tariff.

2022 Annual Markets Report 64 ISO New England Inc.

ISO-NE PUBLIC

 The total quantity of existing capacity – a value driven by retirements from existing

resources and additions from new resources (which become existing resources in

subsequent years.

 Supplier-specific portfolios of existing capacity – values that can change year-over-year as

a result of mergers, acquisitions, divestitures, affiliations, resource performance, etc. To

avoid providing supplier-specific data, these are not described in any detail in this

document, but should be taken into account when considering the analysis.

Residual Supply Index Results

The RSIs for the system and for each import-constrained zone over the past five FCAs are

shown in Figure 2-10 below.

Figure 2-10: Capacity Market Residual Supply Index, by FCA and Zone

The primary outcomes from FCA 17 were: 1) the system-wide RSI was above 100% for the

second auction in a row; and 2) the SENE zone was not modelled.

The RSI in FCA 17 was 102%, which indicates that Net Installed Capacity requirement (Net ICR)

could still be satisfied without the largest supplier. There was no pivotal supplier at the system

level. Existing capacity remained relatively stable between FCA 16 and FCA 17, while Net ICR

decreased from 31,645 MW in FCA 16 to 30,305 MW in FCA 17. The decrease was driven by

lower peak load forecasts as well as greater net tie benefits from transmission across the New

York interface, which had the effect of reducing reliance on capacity resources to meet load.

There were also no significant retirements in FCA 16 and 17. The larger system margins meant

the suppliers would have to be much larger in FCA 17 than in FCA 16, which was not the case.

There were no zonal RSI values (red) in FCA 17. The Southeastern New England zone was

removed mainly due to a decrease in the zonal load forecast and an increase in the import

capability limit into the zone.

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

110%

FCA 13 FCA 14 FCA 15 FCA 16 FCA 17 FCA 13 FCA 14 FCA 15 FCA 16 FCA 17

System-Wide SENE

Residual Supply Index (Percent)

System-Wide SENE

N/A

2022 Annual Markets Report 65 ISO New England Inc.

ISO-NE PUBLIC

Pivotal Supplier Test Results

The number of suppliers and pivotal suppliers within each zone over the past five FCAs are

presented in Figure 2-11 below. To provide additional insight into the approximate portfolio

size needed to be pivotal, the figure also presents the margin by which capacity exceeded or fell

below the relevant capacity requirement. For example, consider the SENE capacity zone in FCA

14. The amount of existing capacity exceeded the local sourcing requirement (LSR), resulting in

a capacity margin of 1,105 MW (right axis – blue marker). Consequently, only suppliers with a

portfolio greater than 1,105 MWs in this zone were pivotal in FCA 14. Of the 53 suppliers in

SENE in FCA 14 (left axis – yellow bar), only four (highlighted in yellow) were pivotal.

Figure 2-11: Overview of Suppliers, Pivotal Supplier, and Capacity Margin, by Zone

At the system level, the capacity margin remained high over the past five FCAs. In FCA 17, a

supplier needed a portfolio of over 3,600 MWs to be pivotal, resulting in no pivotal suppliers.

The increase in the system capacity margin from roughly 1,900 MWs in FCA 15 to over 3,600

MWs in FCA 17 was driven largely by a significant decrease in net ICR. For FCA 15, in SENE, the

amount of capacity was less than the LSR, resulting in a capacity margin of approximately -711

MWs. Consequently, every supplier located in SENE of every portfolio size was pivotal. As

covered above, the SENE did not need to be modelled in FCA 17 as an import-constrained zone,

which improved the overall structural competitiveness of the FCM.

When the market is not structurally competitive, buyer- and supplier-side mitigation rules are

in place to prevent the potential exercise of market power. While a pivotal designation may

indicate the ability to influence clearing prices, a de-list bid is necessary to exercise it. In FCA

17, there were 25 MWs of de-list capacity subject to the pivotal supplier test.

100

Since there

100

Static and retirement de-list bid capacity that is 1) below the FCA starting price and 2) had not been withdrawn prior to

the auction, is shown here. A static de -list bid is entered in the auction as a sealed bid and indicates the minimum price at

which an existing capacity resources seeks to retain a capacity supply obligation. Static de-list bids belonging to a pivotal

suppler are subject to IMM mitigation if the bid is deemed uncompetitive. Dynamic de-list bids are entered during the

auction below a given threshold and are not subject to Tariff prescribed market power tests and mitigation. Retirement

(1,000)

(500)

500

1,000

1,500

2,000

2,500

3,000

3,500

4,000

100

120

140

160

13 14 15 16 17 13 14 15 16 17

System-Wide SENE

Capacity Margin (MWs)

Suppliers (Count)

Pivotal Suppliers (LA) Suppliers (LA) Margin (RA)

N/A

2022 Annual Markets Report 66 ISO New England Inc.

ISO-NE PUBLIC

were no pivotal suppliers in FCA 17, none of the de-list bids were part of a pivotal portfolio.

Therefore, participant-submitted de-list bid prices were used in the FCA. This is discussed in

the next section.

2.4 Forward Capacity Market Mitigation

In this section, we provide a summary of the mitigation measures employed in the FCM

between FCA 13 and FCA 17. The first subsection (2.4.1) looks at supplier-side mitigation for

existing resources. The second subsection (2.4.2) covers buyer-side mitigation, namely the

Minimum Offer Price Rules (MOPR) for new resources.

101

Key Takeaways

There were no significant retirement or permanent de-list bid reviews by the IMM in FCAs 16

or 17. However, older fossil-fueled resources used dynamic de-lists to exit the FCAs for one

year at a time. There was a significant retirement in FCA 15, when Mystic 8 and 9 were allowed

to retire after a two-year reliability retention for fuel security.

There have been a notable number of static de-list bids (85% of a capacity) denied by the IMM

over the past five years. There were no pivotal suppliers in FCA 17, therefore no de-list bids

were mitigated.

There were no Offer Review Trigger Price (ORTP) challenges in FCA 17 from renewable

resources. We attribute this to lower ORTP values for batteries and solar resources and the re-

introduction of the renewable technology resource (RTR) exemption that allowed new

renewable resources to qualify for the FCA irrespective of their ORTP.

2.4.1 Supplier-Side Market Power

The IMM reviews certain de-list bids to determine if they are consistent with a resource’s

avoidable costs of a capacity supply obligation; i.e., its net going forward operating and capital

costs, including expected capacity performance payments, risk premium, and opportunity costs.

While there are a variety of de-list bid types, only a few require review by the IMM prior to the

auction. Those include: general static de-list bids, import and export bids, retirement de-list

bids and permanent de-list bids.

102

Retirement and permanent de-list bids

Retirements and permanent de-list bids are removed from the capacity market – permanently –

when the FCA clearing prices falls below the bid value. All bids for capacity exceeding 20 MW

are subject to an IMM cost review and potential mitigation. For the past two auctions, FCA 16

and 17, there has not been any significant retired or permanently delisted capacity, despite low

and permanent de-list bids (>20 MW) are subject to a net benefits test, whereby the potential impact on clearing prices

and the overall portfolio position is assessed.

101

For more information on changes to the MOPR rule, see Section 8.

102

The term “general” is used to differentiate between other types of static de-list bids, including ambient air static de-list

bids and ISO low winter static de-list bids, which are not subject to IMM review.

2022 Annual Markets Report 67 ISO New England Inc.

ISO-NE PUBLIC

capacity clearing prices. Resource owners may be waiting to assess how future market rules

changes (removal of MOPR, Resource Capacity Accreditation (RCA), and Day-Ahead Ancillary

Services(DASI)) will impact qualified capacity, going forward costs, and ultimately future FCA

prices.

A substantial amount of capacity was retired in FCA 15 and reviewed by the IMM. The Mystic 8

and 9 resources (1,400 MW) will retire after a two-year reliability retention for fuel security for

periods covering FCA 13 and 14 (up to June 2024). The IMM reviewed the retirement bids for

these resources for FCA 13, which were mitigated down; however, they were subsequently

compensated under the terms of a Cost of Service agreement, rather than at their mitigated

retirement bid. Two other smaller retirements did take place (CDECCA and West Springfield 3),

accounting for approximately 160 MW.

Static de-list bids

The IMM reviews all static de-list bids during the FCA qualification process, and issues a

determination denying or accepting the bid value. Participant can take various actions after the

IMM determination, including accepting the bid price, reducing or withdrawing the bid.

Ultimately, only denied bids of pivotal suppliers are mitigated; the pivotal supplier test is

performed closer to the auction date.

Summary statistics for static de-list bids from FCA 13 through FCA 17 as well as the path the

bids took from the time of initial submittal to the auction are provided in Figure 2-12 below.

Note that all de-list bid prices are megawatt-weighted averages.

103

Price calculations are not presented for new import capacity resources because, depending on the circumstances, the

direction of the price difference can vary for price-quantity pairs within the same supply offer. Consequently, the resulting

price difference summary statistics are less meaningful.

2022 Annual Markets Report 68 ISO New England Inc.

ISO-NE PUBLIC

Figure 2-12: General Static De-list Bid Summary Statistics, by Key Milestone Action (FCAs 13 – 17)

104

For FCA 13 through FCA 17, the IMM reviewed 48 general static de-list bids from 11 different

lead participants, totaling roughly 5,900 MW of capacity (an average of 1,200 MW per

auction).

105

Generation resources accounted for 99% of the total capacity and 44 of the 48

general static de-list bid submissions. Four demand response resources made up the remaining

44 MW of the total capacity. The IMM denied approximately 38% of the general static de-list

bids (18 bids or 85% of de-list MW capacity), generally finding that the submitted bids were

either inconsistent with the resource’s net going forward costs or were not sufficiently

supported.

106

Roughly 60% of bids were accepted by the IMM without any changes (left box, second level). Of

the static de-list bids that were denied, many were voluntarily withdrawn or the bid price

further reduced prior to the auction. For resources that were denied and went to the auction

(box furthest to the right, third level), the weighted-average price of denied static de-list bids

was $7.36/kW-month less than the market participant’s originally submitted price.

The number of static delist bids has decreased significantly in previous years. In 2013, the IMM

received over 3,200 MWs of static de-list bids; since then, we have not received over 1,000 MWs

in any year (and less than 500 MWs in FCA 17). Concurrently, the FCA rest-of-pool clearing

prices have fallen below the dynamic de-list bid threshold in each year. When this occurs,

existing resources are allowed to submit dynamic de-list bids without a review by the IMM.

With high levels of surplus capacity, resource owners may: 1) deem the cost of review by the

104

All MW values are rounded to the nearest hundred.

105

A resource with a static de-list bid in each of the three auctions would be counted three times in the MW total;

however, the associated lead participant is only counted once.

106

If a supplier is pivotal, the IMM-determined value is entered into the auction; if not, the participant-submitted bid is

entered. The mitigation only takes effect if the supplier is deemed pivotal, an evaluation that is done some months after

the cost review process in completed.

2022 Annual Markets Report 69 ISO New England Inc.

ISO-NE PUBLIC

IMM greater than the risk of not being able to dynamically de-list; and 2) see an opportunity to

leave the auction without an IMM review.

As discussed above, only de-list bids belonging to pivotal suppliers are mitigated. There were

active de-list bids from pivotal suppliers in FCA 13 only; the four other auctions did not have

any de-list bids from pivotal suppliers. In FCA 13, the denied de-list bids for three resources

(628 MW) were mitigated in the auction. There were no pivotal suppliers in FCA 17, therefore

no de-list bids were mitigated.

2.4.2 Buyer-side Market Power (Minimum Offer Price Rule)

For FCAs 13 through 17, the IMM reviewed 423 new supply offers

107

from participants

requesting to offer below the offer review trigger price (ORTP).

108

These offers came from 59

different participants and totaled 19,200 MWs of qualified capacity, of which about 10,000 MW

(~52%) entered the auction.

109

Non-emitting resources inclusive of battery storage, solar and

wind made up 71% (11,200 MWs) of 19,200 MWs. Demand response resources accounted for

5% (1,200 MW) of total capacity reviewed and import resources accounted for 12% (2,400

MW).

Summary statistics for resources requesting to offer below their respective ORTP in FCAs 13

through 17 are provided in Figure 2-13 below.

110

Note that all offer prices are megawatt-

weighted averages.

Figure 2-13: Reviewable Offer Request Summary Statistics, by Key Milestone Action (FCAs 13 – 17)

107

Note that the count does not capture all unique resources. If a resource was mitigated in FCA 11 and did not clear, it

could return in FCA 12 and would be captured twice in the count.

108

Note that this total does not include supply offers from new import capacity resources without transmission

investments, which are discussed in the supplier-side market power section.

109

A resource with a new supply offer in each of the three auctions would be counted three times in the MW total. In

addition, where FCA qualified capacity does not exist for a resource (e.g., the proposal was withdrawn or denied), the

summer capacity from the resource’s show-of-interest is used instead. Consequently, the presented total overstates the

actual capacity.

110

All MW values a re rounded to the nearest hundred.

423 New Supply Offers

19,200 MW

$1.97/kW-mon

Accepted

78 New Supply Offers

2,500 MW

$1.61/kW-mon

Denied

Submitted

58 New Supply Offers

7,300 MW

Submitted: $1.77/kW-mon

IMM-Determined: $6.50/kW-mon

Withdrawn

To Auction

287 New Supply Offers

9,500 MW

Submitted: $2.22/kW-mon

IMM-Determined: $6.79/kW-mon

30 New Supply Offers

2,000 MW

$1.47/kW-mon

48 New Supply Offers

500 MW

$2.16/kW-mon

Withdrawn

To Auction

345 New Supply Offers

16,800 MW

Submitted: $2.03/kW-mon

IMM-Determined: $6.66/kW-mon

2022 Annual Markets Report 70 ISO New England Inc.

ISO-NE PUBLIC

The IMM mitigated approximately 82% (345) of new supply offers it reviewed, or

approximately 87% (16,800 MW) of new supply capacity.

111

Similar to supplier-side mitigation,

the degree of MOPR mitigation can be measured by the relative increase in the offer floor price

imposed by the IMM. The mitigation process (box furthest to the right, second level) resulted in

an average increase in offer price of $4.57/kW-month (from a submitted price of $2.22/kW-

month to an IMM-determined price of $6.79/kW-month).

In FCA 17, the IMM did not receive Offer Review Trigger Price (ORTP) challenges from any solar

or battery resources. The ORTPs for solar declined from $1.381/kW-month in FCA 16 to

$0/kW-month, due to updated renewable energy credit prices that reduced reliance on FCM

payments. Battery storage fell from $2.601/kW-month to $0.79/kW-month due to reduced

capital costs year-over-year, and high expected revenues based on higher energy futures than

previously assumed. In addition to lower ORTP’s for renewables, the ISO reintroduced the

renewable technology resource (RTR) exemption in FCA 17. Total new qualified capacity under

the RTR exemption reached the maximum allowance of 300 MW in FCA 17. RTR exempt

resources may not have challenged their ORTP, since they cleared irrespective of the auction

price.

112

2.5 Financial Transmission Rights Market

In this section, we look at the concentration of Financial Transmission Rights (FTRs). In this

context, market concentration refers to the extent to which FTR MWs are concentrated among

participants.

Key Takeaways

The ownership of FTRs continued to be relatively highly concentrated in 2022 with the top

four participants holding 66% of FTR MWs in the on-peak period and 69% in the off-peak

period. There were 34 unique FTR holders in both the on-peak and off-peak periods in 2022,

which marked the lowest levels of participation in the previous five years.

Anticipating what a competitive level of FTR ownership looks like is complex as it is unlikely to

match system-level load-serving or generation ownership percentages. This is because there

are not clear commercial reasons for all market participants to hold FTR positions (e.g.,

participants that serve load or own generation in unconstrained areas). Even those participants

that could benefit may have risk preferences that favor exposure to day-ahead congestion over

managing that exposure with the purchase of a financial instrument. Further, FTR market

design permits the purchase of FTRs for financial speculation, so many FTR holders may have

no load or generation position at all.

The concentration of FTR MWs among market participants in 2022 was similar to prior years.

The average amount of FTRs held per hour by the top four participants with the most MWs each

111

Note that the value does not capture all unique capacity. In other words, if a 100 MW photovoltaic (PV) resource was

mitigated in FCA 16 and did not clear, it could return in FCA 17 and would be captured as 200 MW.

112

See Section 6 of the AMR for more information.

2022 Annual Markets Report 71 ISO New England Inc.

ISO-NE PUBLIC

year is shown in Figure 2-14 below.

113

This figure also shows the number of different

participants that held FTRs each year (indicated by the number above each stacked column).

This information is broken down separately for the on-peak and off-peak periods.

Figure 2-14: Average FTR MWs Held per Hour by Top Four FTR Holders by Year and Period

The top four participants held 66% of on-peak FTR MWs and 69% of off-peak FTR MWs in

2022. The concentration ratio of the top four FTR holders has stayed fairly stable over the

reporting period, ranging between 58%-69%. However, the percentage of FTRs held by the

largest FTR holder has trended downward over the reporting period. The largest FTR holder

held 30% of on-peak and off-peak FTR MWs in 2018, but held only 20% of on-peak and off-peak

FTR MWs in 2022. The total number of unique FTR holders fell to its lowest level of the

reporting period in 2022 with only 34 unique participants in both the on-peak and off-peak

periods. This is down modestly from the range of 35-45 different participants in the previous

four years.

113

The firms labeled “Participant 1,” “Participant 2” and s o on are not necessarily the same companies across all years;

these are generic labels for the top four firms during a given year.

5,000

10,000

15,000

20,000

25,000

30,000

35,000

40,000

45,000

Off On Off On Off On Off On Off On

2018 2019 2020 2021 2022

Average MW per Hour

Participant 1 Participant 2 Participant 3

Participant 4 Other Participants # - Number of Participants

2022 Annual Markets Report 72 ISO New England Inc.

ISO-NE PUBLIC

2.6 Ancillary Services

In this section, we review the competitiveness of the Forward Reserve Market (FRM) auctions

and the regulation market. The first subsection (2.6.1) provides RSI results for the last 10 FRM

auctions. The second subsection (2.6.2) reviews available regulation capacity relative to the

regulation requirement and indicates the RSI for 2022.

Key Takeaways

Four FRM auctions during the review period were structurally uncompetitive at the system-

level for TMNSR; three auctions were structurally uncompetitive for total thirty reserves. In all

cases, variations in requirements and offered supply for each product resulted in RSI values

below 100 (a structurally competitive result). In one instance, TMNSR supply in the Summer

2022 auction, the RSI declined significantly with a large reduction in offered supply.

At the zonal level, FRM TMOR auctions in the NEMA zone were structurally uncompetitive in

2018. After 2018, none of the modeled zones in the FRM auctions needed to procure supply.

The regulation market was structurally competitive in 2022, with available supply significantly

exceeding the regulation requirement and with no supplier controlling enough supply to

potentially have market power.

2.6.1 Forward Reserve Market

The competitiveness of the FRM is assessed using the Residual Supply Index (RSI) and is based

on FRM offer quantities by participant and the operating reserve requirements in each

auction.

114

The heat map provided in Table 2-4 below shows the offer RSI for TMNSR at a

system level and for TMOR at a zonal level. The colors indicate the degree to which structural

market power was present; red is associated with low RSIs, white with moderate RSIs, and

green with high RSIs. Dark red indicates that structural market power was present, while dark

green indicates that there was ample offered supply without the largest supplier.

114

The RSI for TMNSR is computed at a system-level based on the total quantity of TMNSR offers across all reserve zones,

excluding the largest TMNSR offer quantity by a single market participant. The RSI for TMOR is computed similarly for each

reserve zone with a non-zero TMOR local reserve requirement. Given that the TMNSR quantity also satisfies the TMOR

requirement, the TMNSR offer quantity in a zone is included in the total TMOR offer quantity within that zone.

2022 Annual Markets Report 73 ISO New England Inc.

ISO-NE PUBLIC

Table 2-4: Offer RSI in the FRM for TMNSR (system-wide) and TMOR (zones)

Procurement

Period

Offer RSI

TMNSR

(System-

wide)

Offer RSI

Total

Thirty

(System-

Wide))

Offer RSI

TMOR

(SWCT)

Offer RSI

TMOR

(CT)

Offer RSI

TMOR

(NEMA)

Summer 2018

112

108

438

N/A

Winter 2018-19

127

N/A

Summer 2019

N/A

Winter 2019-20

120

118

N/A

Summer 2020

N/A

Winter 2020-21

102

115

N/A

Summer 2021

108

N/A

Winter 2021-22

110

116

N/A

Summer 2022

N/A

Winter 2022-23

109

112

N/A

At the system level, four (out of the ten) auctions had RSI values below the structurally-

competitive level for TMNSR, TMOR or both.

TMNSR RSI values were below structurally-competitive levels in four of the five summer

periods. In Summer 2019, the decline in TMNSR RSI resulted from a slightly increased

requirement and a medium-sized supplier not participating in that auction. The Summer 2020

TMNSR results likewise had an increased requirement (up an additional 4% compared to

Summer 2019), coupled with a small net reduction in supply offers (approximately 2%

compared to the prior summer). The Summer 2021 RSI improved somewhat compared to the

Summer 2020 RSI, with a small increase in supply and a small reduction in the requirement. In

Summer 2022, the requirement did not change significantly from the prior summer auction, but

supply in the auction declined by approximately 800 MW; the change in supply resulted in the

decline in the RSI.

System-wide total thirty RSI values were not structurally-competitive for the Summer 2019,

2020 and 2022 auctions. In the 2019 and 2020 auctions, the RSI estimates were only slightly

below the competitive level, reflecting slightly reduced supply and slightly increased reserve

requirements in those auctions (relative to the other system-wide total thirty auctions). In

Summer 2022, a small increase in the requirement and an approximately 200 MW reduction in

supply offers compared to Summer 2021 resulted in the RSI decline. The decline in total thirty

supply in Summer 2022 was the result of numerous supply offer changes (both increases and

reductions in supply) compared to the prior summer auction.

Considering the TMOR RSI at the zonal level, only the NEMA/Boston zone had an RSI less than

the structurally-competitive level. For the Summer 2018 and Winter 2018-19 auctions, every

participant that offered forward reserve supply in NEMA/Boston was needed to meet the local

requirement; in these auctions, every supplier for that zone had market power.

2022 Annual Markets Report 74 ISO New England Inc.

ISO-NE PUBLIC

A Further Discussion of Summer 2022 Auction Competitiveness

As noted earlier in this subsection, the Summer 2022 auction for the TMNSR product was not

structurally competitive. The FRM is not subject to offer price mitigation and it is often

structurally uncompetitive, i.e., demand is fixed and supply is limited. However, the IMM

reviews each auction to assess whether the auction results are consistent with the results we

would expect from a competitive process, i.e., did participants that had market power make

competitive offers that reflected their anticipated costs? In evaluating the competitiveness of

the Summer 2022 auction, we reviewed potential physical and economic withholding and found

no evidence of the exercise of market power.

 Bid in supply was consistent with prior auctions including auctions when there was no

pivotal supplier.

 Offer prices appeared to reflect a plausible range of anticipated costs of providing the

service. Note, offer prices can vary significantly depending on the participant generation

portfolio and expectation of market conditions. The FRM auction offer cap (set at

$9,000/MW-month) suggests that a large range of auction offers may fall within a

reasonable range. The reasonableness of auction offers depends on factors such as

expectations about reserve revenue (which is sensitive to the likelihood of incurring

reserve constraint penalty factor reserve pricing), foregone energy market net revenue,

and the likelihood of incurring FRM failure-to-reserve and failure-to-activate penalties.

2.6.2 Regulation Market

We reviewed the competitiveness of the regulation market by examining market structure and

resource abundance. The abundance of regulation resources, and relatively unconcentrated

control of that supply, implies that market participants had little opportunity to engage in

economic or physical withholding in 2022. For these reasons, we find that the regulation

market was competitive in 2022. Figure 2-15 below indicates the regulation requirement

relative to available supply.

Figure 2-15: Average Regulation Market Requirement and Available Capacity, 2022

200

400

600

800

1,000

1,200

1,400

1,600

1,800

2,000

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

Hour of Day

Average Available Capacity/Hour (MW) Average Requirement/Hour (MW) [Right Axis]

2022 Annual Markets Report 75 ISO New England Inc.

ISO-NE PUBLIC

On average, during every hour of the day, available supply far exceeds the regulation

requirements. However, an abundance of available supply alone is not a dispositive indicator of

market competitiveness, as one - or a small number of suppliers - could control the available

supply and seek to exercise market power.

The RSI provides a better indicator of the structural competitiveness of the regulation market.

As shown in Figure 2-16, the regulation requirement (right axis) and RSI (left axis) are

inversely correlated (the lower the requirement the higher the RSI).

Figure 2-16: Average Regulation Requirement and Residual Supply Index

In 2022, the lowest hourly average RSI did not fall below 1,000%, implying that, on average, the

system had the capability to serve ten times the regulation requirement without the largest

regulation supplier, even in the hours with the greatest regulation requirements.

100

120

140

160

180

500

1,000

1,500

2,000

2,500

3,000

3,500

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

Requirement MW

RSI %

Hour of Day

Average RSI/Hour (%) [Left Axis] Average Requirement/Hour (MW) [Right Axis]

2022 Annual Markets Report 76 ISO New England Inc.

ISO-NE PUBLIC

Section 3

Day-Ahead and Real-Time Energy Market

This section covers key trends in energy market outcomes. In the first section (3.1), energy

prices are analyzed across a number of dimensions. The second section (3.2) dives into the

major factors behind recent energy market outcomes, looking at such factors as input costs,

load and weather conditions, and notable system events. The subsequent section (3.3) looks at

net commitment period compensation (NCPC) payments, which are the payments that

resources receive when energy prices are too low to cover their production costs. The final

section (3.4) summarizes demand response resource (DRR) participation in the energy

markets.

Day-ahead and real-time energy prices increased significantly from 2021 levels. This reflected a

large increase in underlying primary fuel prices, most notably natural gas. The average Hub

price was $85.56/MWh in the day-ahead market, up by 86% on 2021. This increase was

consistent with the 101% increase in natural gas prices that occurred as a result of the conflict

between Russia and Ukraine, in addition to local pipeline limitations that New England often

experienced during the winter months.

In addition to generation costs, numerous other factors played roles in determining the energy

market outcomes in 2022. While weather conditions in 2022 were generally comparable to

those observed in 2021, several notable events had outsized impacts. Chief among them was

Winter Storm Elliot, which moved through New England on December 24, 2022, and led to a

system capacity shortage. Transmission congestion and out-of-market actions taken to

maintain system reliability only slightly impacted energy market outcomes in 2022, as

incidence of both were relatively low. Meanwhile, DRRs continued to be infrequently

dispatched in 2022 as a result of high offers prices.

NCPC payments totaled $52.9 million in 2022, an increase of $17.0 million (48%) compared to

2021. However, payments as a percentage of total energy payments remained low, and

decreased from 0.6% in 2021 to 0.5% in 2022, the lowest percentage level over the past five

years. One notable NCPC outcome was the reduction in payments for resources committed for

local reliability (i.e., LSCPR NCPC payments); these payments decreased from $6.8 million in

2021 to $1.1 million in 2022.

Importantly, one area that we recommend the ISO assess further is related to the fast-start

pricing rules that were implemented in the real-time energy market in March 2017. We found

that there were significant periods of non-zero pricing (and payments) during times when there

were physical surpluses of reserves. This topic is discussed in more detail in Section 3.1.2.

3.1 Energy Prices

This section evaluates and discusses energy prices across a number of dimensions, including by

energy market (i.e., day-ahead and real-time), time-of-day, and location. These dimensions

provide useful context for understanding differences in energy prices (LMPs) over the review

period. The first subsection (3.1.1) summarizes energy market pricing over a five-year period,

reviews price separation for 2022 for load zones, and examines load-weighted LMPs, which

provide an indication of the effective prices that load-serving entities pay for energy. The

second subsection (3.1.2) estimates the impacts of fast-start pricing rules on LMPs and other

2022 Annual Markets Report 77 ISO New England Inc.

ISO-NE PUBLIC

market outcomes. Finally, the third subsection (3.1.3) examines the extent to which prices

converged across the day-ahead and real-time energy markets, which is a useful barometer of

market efficiency.

Key Takeaways

Energy market prices increased significantly in 2022, primarily as a result of increased natural

gas prices. Average LMPs in 2022 were $85.56/MWh in the day-ahead energy market and

$84.92/MWh in the real-time energy market; this compares to day-ahead and real-time

average LMPs in 2021 of $45.92/MWh and $44.84/MWh, respectively.

Zonal prices in 2022 exhibited modest degrees of separation in the real-time and day-ahead

energy markets. Comparing Hub and zone LMPs, the average absolute differences were

$0.68/MWh in the day-ahead energy market and $0.70/MWh in the real-time energy market, a

difference of less than 1.0% in each market. This reflects relatively low levels of both

transmission losses and congestion on average throughout ISO New England’s control area.

Load-weighted LMPs, which reflect effective prices paid by load-serving entities, tend to follow

seasonal patterns reflecting high-load and high fuel price periods. The impacts were evident in

2022 as the summer period (with higher loads) and the winter periods (with elevated fuel

prices) exhibited the highest prices. The day-ahead energy price at the Hub exceeded the real-

time price by an average of $0.64/MWh in 2022. However, this difference was consistently

larger on average during the middle part of the day.

3.1.1 Day-Ahead and Real-Time Energy Prices

Energy prices at the Hub and the eight New England load zones are presented below, with an

assessment of price levels and differences across a number of dimensions: time-of-use (e.g.,

peak, off-peak hours), market, location and load-weighted compared to simple average prices.

Hub prices by time-of-use and market

First, Figure 3-1 shows simple average Hub prices in the day-ahead and real-time market for

three time tranches: all hours, peak and off-peak hours.

115

On-peak periods are weekday hours ending 8 to 23 (i.e., Monday through Friday, excluding North American Electric

Reliability Corporation (NERC) holidays); the off-peak period encompasses all other hours.

2022 Annual Markets Report 78 ISO New England Inc.

ISO-NE PUBLIC

Figure 3-1: Annual Simple Average Hub Price

Prices in 2022 were at their highest level since Standard Market Design was implemented 19

years ago in 2003 – price levels exceeded $80/MWh in only one other year (2008). In 2022, the

simple annual average Hub price (in all hours) was $85.56/MWh in the day-ahead market and

$84.92/MWh in the real-time market, a significant increase compared to 2021; up 86% in the

day-ahead market and by 89% in the real-time market.

These price changes are consistent with observed market conditions, including input fuel costs,

load levels, and generator operations. The primary driver of energy prices, natural gas prices,

increased significantly in 2022, by over 100%, and gas generators set price for almost 80% of

load in the real-time energy market.

Pricing by time-of-day (i.e., on-peak and off-peak) in 2022 exhibited the same trend; on-peak

prices increased by 78% in the day-ahead market and 80% in the real-time market, while

average off-peak prices increased by 96% in the day-ahead market and 100% in the real-time

market, respectively.

On average, day-ahead and real-time Hub prices were comparable indicating that the day-ahead

market performed reasonably well in reflecting expected real-time outcomes. Differences

ranged from $0.38/MWh (0.5%) for the off-peak period to $1.81/MWh (2.0%) for the on-peak

period.

Prices by load zone and market

At the zonal level, locational price differences were relatively small in 2022 in both the day-

ahead and real-time energy market, as shown in Figure 3-2.

Off-Peak, 79.80

Off-Peak, 80.18

On-Peak, 92.17

On-Peak, 90.36

All Hours, 85.56

All Hours, 84.92

100

2018 2019 2020 2021 2022 2018 2019 2020 2021 2022

Day-Ahead . Real-Time

$/MWH

Off-Peak On-Peak All Hours

2022 Annual Markets Report 79 ISO New England Inc.

ISO-NE PUBLIC

Figure 3-2: Simple-Average Hub and Load Zone Prices, 2022

The relatively small price differences among the load zones and with the Hub were the result of

modest levels of both marginal losses and congestion. The average absolute difference between

the Hub price and load zone prices was $0.68/MWh in the day-ahead energy market and

$0.70/MWh in the real-time energy market, a difference of less than 1% in each market.

The Connecticut and Maine load zones had the lowest overall average prices in the region in

2022. Connecticut’s prices averaged $1.50/MWh (1.8%) lower than the Hub price in the day-

ahead market and Maine’s prices averaged $1.85/MWh (2.2%) lower than the Hub price in the

real-time market. Most of the difference in average prices between load zones and the Hub

resulted from the imputed cost for transmission losses that is included in the LMP; losses

represented about 74% of the price difference in the day-ahead market and 91% in the real-

time market.

Conversely, the NEMA pricing zone had the highest average prices in the day-ahead and real-

time markets. NEMA’s average day-ahead and real-time prices were slightly higher than the

Hub, by $0.52/MWh and $0.50/MWh, respectively. While NEMA is import-constrained at times,

with the transmission network limiting the ability to import relatively less expensive power

into the load zone, losses represented the bulk of the price difference between the Hub and

NEMA: 100% in both the day-ahead and real-time markets, respectively. The average

congestion component in NEMA was quite low in both markets for 2022.

Consumer prices: load-weighted prices

Compared to simple-average prices presented above, load-weighted prices are a better

indicator of average prices that load-serving entities (LSEs) pay for energy.

116

The amount of

energy consumed in the markets can vary significantly by hour. Load-weighted prices reflect

116

While a simple-average price weights each energy market price equally across the day, load-weighting reflects the

proportion of energy consumed in each hour: load-weighted prices give greater weight to high-load consumption hours

than to low-load consumption hours, with each hour being weighted in proportion to total consumption for th e entire day.

100

Hub

Maine

New Hampshire

Vermont

Connecticut

Rhode Island

SE Mass

WC Mass

NEMA Boston

Hub

Maine

New Hampshire

Vermont

Connecticut

Rhode Island

SE Mass

WC Mass

NEMA Boston

Day-Ahead . Real-Time

$/MWH

Line = HUB Price

2022 Annual Markets Report 80 ISO New England Inc.

ISO-NE PUBLIC

the increasing cost of satisfying demand during peak consumption periods when higher

demand necessitates the commitment and dispatch of more expensive generators. Because of

this, load-weighted prices tend to be higher than simple average prices.

The average load-weighted prices were $91.36 and $91.07/MWh in the day-ahead and real-

time markets in 2022, respectively. Monthly load-weighted and simple average prices for 2022

are provided in Figure 3-3.

Figure 3-3: Load-Weighted and Simple Average Hub Prices, 2022

As expected, load-weighted average prices were higher than simple-average prices in 2022. The

differences range from approximately 1% to 9%, depending on the month and energy market

(day-ahead or real-time). These price differences reflect the variability in load over the course

of a day, which is typically a function of temperature and business/residential consumption

patterns. For example, hours with low electricity consumption tend to occur overnight, when

business and residential activity is low and summer cooling needs are minimal. We are also

observing relatively low wholesale loads during some mid-day periods, as the result of

increased behind-the-meter solar generation reducing retail loads.

Monthly day-ahead load-weighted prices across load zones over the past five years are shown

in Figure 3-4 below. The figure illustrates significant monthly variability in LMPs, particularly

during winter months with fuel price volatility.

100

120

140

160

180

J F M A M J J A S O N D J F M A M J J A S O N D

Day-Ahead . . Real-Time

$/MWH

Simple Avg Monthly Simple Avg Annual

Load-Wgt Monthy Load-Wgt Annual

2022 Annual Markets Report 81 ISO New England Inc.

ISO-NE PUBLIC

Figure 3-4: Day-Ahead Load-Weighted Prices

Load-weighted energy prices by load zone from 2018 to 2022 indicate a pattern that varies

considerably by year and by month, but typically not by load zone. Very high pricing occurred in

January 2022, consistent with colder temperatures and high natural gas prices. Winter periods

with high fuel prices and summer months with elevated load variability typically have the

highest load-weighted prices; a similar trend applies to the real-time market. Notably in 2021,

high winter gas prices and relatively high fall gas prices resulted in those periods having the

highest energy prices during the year.

3.1.2 Fast-Start Pricing: Impact on Real-Time Outcomes

This section details the impacts of fast-start pricing rules on market outcomes in 2022. We

found that fast-start pricing impacts were similar to prior analyses, and is generally meeting the

design’s key objective of improving real-time price formation by better reflecting the

production costs of flexible, fast-start resources in energy prices.

117

However, our assessment of

the impact on reserve pricing shows that there are significant periods of non-zero pricing (and

payments) during times when the reserve constraint is not impacting the physical dispatch of

resources and there is a physical surplus of reserves. We recommend that the ISO assess this

issue.

The following table compares a number of actual and estimated counterfactual market

outcomes. The column labeled Fast-Start Pricing details actual pricing and settlement outcomes.

The column labeled Non-Fast-Start Pricing provides estimates of counter-factual outcomes if

fast-start pricing had not been implemented.

117

See Section 5.5 of the Summer 2017 Quarterly Markets report for detail on fast-start pricing:

https://www.iso-ne.com/static-assets/documents/2017/12/2017-summer-quarterly-markets-report.pdf

$46.88

$32.82

$24.57

$48.30

$91.36

100

120

140

160

180

J F M A M J J A S O N D J F M A M J J A S O N D J F M A M J J A S O N D J F M A M J J A S O N D J F M A M J J A S O N D

2018 2019 2020 2021 2022

$/MWH

Hub Maine New Hampshire Vermont

Connecticut Rhode Island SE Mass WC Mass

NEMA/Boston Annual Average Hub Gas Cost

Average annual gas cost

Average annual LMP

Average Monthly LMP

2022 Annual Markets Report 82 ISO New England Inc.

ISO-NE PUBLIC

Table 3-1: Fast-Start Pricing Outcome Summary, 2022

Market Outcome

Fast-Start

Pricing (Actual

Outcomes)

Non Fast-Start

Pricing

(Counterfactual

Outcomes)

Difference

System LMP ($/MWh)

118

$84.52

$78.84

$5.68 (7%)

Real-Time Energy Payments ($, Millions)

119

$208.3

$181.7

$26.5 (15%)

NCPC Payments ($, Millions)

120

$29.9

$44.5

-$14.6 (-33%)

Reserve Prices ($/MWh)

121

$2.54

$1.08

$1.47 (136%)

Reserve Payments ($, Millions)

122

$26.9

$9.0

$17.9 (199%)

Percent of Intervals with Reserve Pricing (%)

13.5%

7.5%

6.0% (81%)

Intervals Fast-Start Resource Marginal

123

26.4%

10.0%

16.4% (165%)

To summarize the key takeaways, the fast-start pricing approach:

 resulted in a higher frequency of fast-start resources setting price,

 increased the average annual system LMP by 7% and real-time energy payments from

load by 15% over the course of 2022,

 decreased real-time NCPC paid to generators and ARDs by 33%

124

, and

 had a substantial impact on reserve pricing and payments, with non-zero pricing

occurring in 81% more intervals. Overall, average reserve prices were 136% higher in

the fast-start-pricing case than in the non-fast-start-pricing case and payments were

200% higher.

118

The s ystem LMP shown here is the energy component of the LMP in each interval.

119

The estimation of energy payments is calculated using generation-weighted zonal LMPs, as opposed to load-weighted,

due to data limitations. Additionally, generation that do not set price for any load are removed and real-time load

deviations are only considered for locations and customers with physical load (i.e., exports and day-ahead demand that

does not correspond to physical load are excluded). The actual value of real-time payments to these locations and

customers is $210.6 million.

120

NCPC payments included in this analysis are Commitment-Out-Of-Merit (COOM), Dispatch-Out-Of-Merit (DOOM), and

Rapid Response Pricing Opportunity Cost (RRPOC) payments for generators and asset-related demand resources (ARDs).

Due to data limitations, LMPs were not available during each generator’s full ramp time so estimated payments are slightly

higher than actual payments. Actual payments (i.e., not based on IMM estimates like the data shown in the table) in 2022

were $29.8 million.

121

These reserve prices represent the average reserve price in every interval – including $0/MWh reserve price intervals.

122

The netting of real-time payments for a participant’s forward reserve market obligations is not accounted for in the

reported reserve payments. For more information on the impact of fast-start pricing on reserves, see Section 7.1.

123

This metric represents the percentage of intervals in which at least one fast-start generator that was marginal (i.e., set

price).

124

Breaking down the reduction further, fast-start pricing reduced commitment-out-of-merit and dispatch-out-of-merit

NCPC to generators that did not recover their costs when following ISO dispatch instructions by 50%. The decrease was

offset by an increase in Rapid Response Pricing Opportunity Cost (RRPOC) NCPC.

2022 Annual Markets Report 83 ISO New England Inc.

ISO-NE PUBLIC

Reserve Pricing under Fast-Start Pricing

Reserve prices are intended to:

 offset lost opportunity costs when a resource is selected to serve as reserve capacity

instead of producing electricity in real-time,

125

and

 compensate market participants with on-line and fast-start generators for the increased

value of their product when the reserve constraint is binding (economically, when

reserves become scarce).

Since fast-start pricing was implemented in 2017, we have observed frequent non-zero reserve

pricing in scenarios when resources’ dispatch instructions were not impacted by the reserve

constraint and the system had a surplus of reserves. In 2022,

 there was a TMSR surplus 45% of the time there was a positive TMSR price (about

516/1,151 hours),

 there was a TMNSR surplus 65% of the time there was a positive TMNSR price (about

11/16 hours), and

 there was a TMOR surplus 86% of the time there was a positive TMOR price (about

18/21 hours).

126

In 2022 alone, $13.7 million in reserve payments were made when there was a reserve surplus

(over half of the $26.9 million in total reserve payments during the year), despite dispatch being

unaffected by the reserve constraint.

The separation of pricing and dispatch under fast-start pricing rules introduced reserve pricing

challenges. Physically (i.e., in the dispatch software), generation up to EcoMin must be

producing energy and generators cannot be dispatched below EcoMin to provide reserves;

these reserves are unattainable. Fast-start pricing rules relax fast-start units’ EcoMin in the

pricing software to give the appearance of a larger dispatchable range, which allows fast-start

units to set price more often. However, the ISO’s fast-start pricing design does not allow the

expanded dispatchable range (i.e., the range between 0 MW and EcoMin) to provide reserves

125

Because the pricing software frequently generates LMPs that are higher than the dispatch software, there are often

cases in which resources are incentivized to increase their output in the presence of the higher prices but no units have

been impacted by the reserve constraint. The IMM does not believe these units have been “selected to serve as reserve

capacity instead of producing electricity” because physically their dispatch has not been impacted by the reserve

constraint. In principle, RRPOC NCPC, not reserve pricing, is the mechanism through which the market should compensate

these resources. RRPOC NCPC is designed to compensate units when faced with higher fast-start pricing-driven prices,

while reserve prices are designed to compensate units for foregoing the energy price when the reserve constraint is

binding.

126

Throughout this section, reserve prices represent incremental reserve prices for each individual product. Because

reserve prices are cascaded (TMSR MWs are paid the TMOR price + the incremental TMNSR price + the incremental TMSR

price), we first separated each price into their components. When there was a physical reserve surplus for a given product,

we set that component =$0. More complex methodologies, taking into account different permutations of pricing and

physical outcomes yielded similar results. The number of hours reported here are the number of 5-minute intervals

divided by 12. Percentages do not match number of hours for all three products due to rounding error.

2022 Annual Markets Report 84 ISO New England Inc.

ISO-NE PUBLIC

for pricing purposes because these reserves are physically unattainable.

127

This methodology

produces tradeoffs:

 Pro: It ensures the system does not appear to have more reserves than are physically

attainable and reserve prices are always produced when the reserve constraint is

impacting physical dispatch.

 Con: Under certain circumstances, the system can appear to have less reserves than are

physically available at no cost to the system, thus it over-values reserves in many

scenarios when the reserve constraint is not impacting physical dispatch and the system

is operating with a reserve surplus. This is explained in the following paragraph.

When there is $0/MWh reserve pricing, dispatchable offers that are priced higher than the LMP

are not dispatched because they are not “in-the-money” and provide reserves without

additional compensation—they are essentially providing reserves ‘for free.’ Often, in pricing,

when fast-start units’ EcoMins are relaxed the energy between 0 MW and EcoMin is relatively

expensive compared to other units on the system and the energy moves towards the top of the

supply stack. The ISO’s methodology dictates that for pricing purposes this available energy

cannot provide reserves “for free” because these reserves are physically unattainable. In some

cases, the pricing model must take reserves from lower-cost asset offers that are permitted to

provide reserves to meet the modeled reserve requirement, and take energy from these higher-

priced fast-start unit offers that are not permitted to provide reserves because they are

physically unattainable. This dynamic gives the appearance of a reserve opportunity cost (the

pricing model is producing a result where the lower-cost assets would like to be producing

energy at the higher LMP but are designated for reserves) and, in turn, produces reserve prices,

although physically the reserve constraint is not impacting physical dispatch and there is a

physical reserve surplus.

To summarize, due to tradeoffs presented by the separation of the dispatch and pricing

software, the ISO chose a methodology that minimizes false negatives (no reserve pricing when

there is a physical reserve constraint binding) but allows false positives (reserve pricing when

there is not a physical reserve constraint binding). This was an intentional decision when fast-

start pricing was implemented, however, the frequency in which we have observed reserve

pricing when there is not a physical reserve constraint binding has exceeded the frequency in

which we expected these scenarios to occur, and the cost of reserve payments in these intervals

warrants additional consideration of other solutions.

As an example, Figure 3-5 below, shows reserve clearing prices from the pricing and dispatch

software on December 12, 2022.

128

On this day, the ten-minute-spinning-reserve (TMSR)

reserve constraint penalty factor (RCPF) bound for 85 minutes despite a TMSR surplus. The

pricing software (top orange bars) represents the fast-start pricing (actual) case. The dispatch

software (bottom blue bars) represents the non-fast-start pricing (counterfactual) case.

127

The ISO considered different reserve accounting methodologies for pricing when fast-start pricing was implemented

and chose not to allow physically unattainable reserves between 0 MW and EcoMin to contribute reserves for pricing.

128

Reserve clearing prices from the dispatch software are not used in market settlement.

2022 Annual Markets Report 85 ISO New England Inc.

ISO-NE PUBLIC

Figure 3-5: Reserve Prices in the Pricing vs. Dispatch Software on December 12, 2022

Figure 3-5 shows that fast-start pricing can have a substantial impact on the frequency of non-

zero reserve prices. On December 12, fast-start pricing generated TMSR RCPF pricing for 85

minutes, despite the RCPF binding for only 20 minutes in the counterfactual case. Additionally,

there was non-zero reserve pricing, designed to signal that the optimal dispatch had to be

adjusted to meet the reserve requirement, in five and a half hours, although the physical

dispatch was only adjusted in less than one hour.

Recommendation

We recommend that the ISO revisits reserve pricing mechanics under fast-start pricing to

address the frequency of non-zero reserve pricing when there is a physical reserve surplus.

3.1.3 Energy Price Convergence

Price convergence refers to the extent to which prices are equal across the day-ahead and real-

time energy markets. Price convergence can serve as a signal of market efficiency – which, in

this case, means achieving the necessary real-time generator commitments at the lowest

possible cost. In an efficient market, day-ahead (forward) prices should generally reflect

expected real-time (spot) prices.

While day-ahead prices will almost never perfectly match real-time prices in any given hour

(because real-time conditions will usually differ from expectations), one might expect to see

similar average prices between the two markets over a longer period. Consequently, one way to

assess price convergence is to look at the average annual difference between day-ahead and

real-time prices (i.e., the day-ahead price premium).

129

The day-ahead price premium is defined as the day-ahead energy price minus the real-time energy price.

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

Dispatch Software

Hour Ending

Dispatch

Pricing Software

Pricing

Reserve Clearing Price ($/MWh)

2022 Annual Markets Report 86 ISO New England Inc.

ISO-NE PUBLIC

Convergence across all hours

The average day-ahead price premium at the Hub in 2022 remained in line with recent historic

values, although there was an increasing amount of volatility. This can be seen in Figure 3-6

which shows the distribution of the day-ahead price premium at the Hub using a box-and-

whiskers diagram.

130

This figure also shows the average annual day-ahead Hub LMP (orange

line) for 2018–2022.

Figure 3-6: Average Annual Day-Ahead Price Premium at the Hub and Average Day-Ahead Hub LMP

The day-ahead premium at the Hub averaged $0.64/MWh in 2022, a moderate decrease from

2021 when the premium was $1.08/MWh. In percentage terms, the 2022 price premium (0.7%)

was the second lowest in the reporting period. Between 2018 and 2022, the average price

premium was as low as -$0.06/MWh (in 2020) and as high as $1.08/MWh (in 2021).

An increased amount of variability in the day-ahead price premium in 2022 is evident in Figure

3-6. While the widening price ranges might suggest worsening price convergence, it is

important to note that, over time, these percentiles generally track the average day-ahead Hub

LMP (orange series, right axis) and natural gas prices. Since natural gas prices are the primary

drivers of LMPs in New England, average LMPs tend to be higher when natural gas prices in

New England are higher. The average price of natural gas in New England was $9.28/MMBtu in

2022, compared to $4.62/MMBtu in 2021. Similarly, differences between day-ahead and real-

time prices tend to be larger when gas prices are higher. This is because the cost difference

130

The day-ahead price premium is measured on the left axis (“LA”), while the average annual day-ahead Hub LMP is

measured on the right axis (“RA”).

($100)

($80)

($60)

($40)

($20)

$20

$40

$60

$80

$100

($50)

($40)

($30)

($20)

($10)

$10

$20

$30

$40

$50

2018 2019 2020 2021 2022

Day-Ahead LMP ($/MWh)

Day-Ahead Price Premium ($/MWh)

25th to 75th Percentile for Premium [LA] 5th and 95th Percentiles for Premium [LA]

Average Day-Ahead Premium [LA] Average Day-Ahead LMP [RA]

2022 Annual Markets Report 87 ISO New England Inc.

ISO-NE PUBLIC

between two gas-fired generators with different heat rates is greater when gas prices are

higher.

131

Convergence by time of day

Although the annual average day-ahead price premium in 2022 was small, there were notably

higher day-ahead price premiums, on average, during the middle part of the day. This can be

seen in Figure 3-7 below, which shows, by hour, the average day-ahead price premium at the

Hub (blue line). The gray bars show the interquartile range (i.e., the middle 50 percent) of the

day-ahead price premium.

Figure 3-7: Average Hourly Day-Ahead to Real-Time Hub Price Differences, 2022

The average day-ahead price at the Hub exceeded the average real-time price during nine

consecutive hours in 2022 (HE9 to HE17), by as much as $5.85/MWh. One potential factor

behind this daytime premium is the price-suppressing impact of increased solar generation on

real-time load.

132

Much of New England’s solar capacity is “behind-the-meter,” meaning that it

appears in real time in the form of load reduction. To the extent that more solar generation

shows up in real time than anticipated in the day-ahead market, prices in the real-time energy

131

For example, consider two gas-fired generators: Gen A, which is marginal in the day-ahead market, has a heat rate of 10

MMBtu/MWh and Gen B, which is marginal in real-time, has a heat rate of 7 MMBtu/MWh. If the gas price is $5/MMBtu,

the generation cost for Gen A is $50/MWh (10 MMBtu/MWh x $5/MMBtu) and the cost for Gen B is $35/MWh (7

MMBtu/MWh x $5/MMBtu). The difference in generation cost between Gen A and Gen B – and by construction, the

difference between the day-ahead price and the real-time price – is $15/MWh. If the gas price were to increase to

$10/MMBtu, the generation costs for Gen A and Gen B would now be $100/MWh and $70/MWh, respectively, for a day-

ahead premium of $30/MWh. In this example, the increased day-ahead premium ($30/MWh from $15/MWh) is only

reflective of a higher gas price ($10/MMBtu from $5/MMBtu).

132

The growth in solar generation in New England is discussed in more detail in Section 1.3.1.

($25)

($20)

($15)

($10)

($5)

$10

$15

$20

$25

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

Price Difference ($/MWh)

Hour Ending

Interquartile Range Average Difference

2022 Annual Markets Report 88 ISO New England Inc.

ISO-NE PUBLIC

market are likely to be dampened relative to the day-ahead market (and often considerably

so).

133

134

3.2 Drivers of Energy Market Outcomes

This section presents a more detailed assessment of the important contributing market and

operational factors that impact energy market outcomes. For example, as we covered in Section

1 earlier, underlying natural gas prices explain, to a large degree, movements in energy prices.

Here we explore that relationship in more detail. Operational factors, such as load forecast

error or notable system events, provide additional insight into pricing outcomes. The section is

structured as follows:

 Generation costs (3.2.1)

 Supply-side participation (3.2.2)

 Reserve Adequacy Analysis (RAA) commitments (3.2.3)

 Load and weather conditions (3.2.4)

 Demand bidding (3.2.5)

 Load forecast error and market implications (3.2.6)

 System events (3.2.7)

 Reliability commitments and posturing (3.2.8)

 Transmission Congestion (3.2.9)

 Marginal resources (3.2.10)

Key Takeaways

Day-ahead and real-time electricity prices continue to be closely correlated with the

estimated cost of operating a natural gas-fired generator. In fact, with the exception of 2020,

the day-ahead and real-time gas price-adjusted LMPs have been within a narrow band over

the reporting period ($29-$32/MWh), indicating that changes in gas prices are a key driver of

changes in in the LMP. Natural gas was the marginal fuel for 43% of load in the day-ahead

market in 2022 and 79% in the real-time market in 2022.

New England continued to be relatively unconstrained system due to its robust transmission

grid; in 2022 congestion revenue represented only 0.44% of total energy costs.

While average temperatures in 2022 were similar to those in 2021, weather did play an

important role in energy market outcomes at times. One of the most notable events occurred

on December 24 when Winter Storm Elliot moved through New England. The cold weather

led to tight system conditions, which culminated in a capacity deficiency.

Unpriced or out-of-market actions to maintain system reliability can also impact energy

market outcomes. In 2022, the ISO’s reliability commitments averaged only 15 MW per hour.

The level of posturing also remained low.

133

In the day-ahead energy market, it is often virtual supply that takes the place of the expected additional real-time solar

production. See Section 4.1 for more information about this relationship.

134

Section 3.2.2 looks more closely at the differences in the supply mix between the day-ahead and real-time energy

markets.

2022 Annual Markets Report 89 ISO New England Inc.

ISO-NE PUBLIC

3.2.1 Generation Costs

Day-ahead and real-time electricity prices, as well as year-over-year changes, continue to be

closely correlated with the estimated cost of operating a natural gas-fired generator. In 2022,

natural gas-fired generators continued to be the dominant price setters (79% in real-time) and

supply over 50% of native generation. Other fossil fuel-types have a comparatively minimal

impact on overall energy prices. The relationship between electricity prices and generation fuel

costs are shown in Figure 3-8 below, alongside the estimated gross margin of a natural gas-fired

generator, commonly referred to as a spark spread.

135

Figure 3-8: Estimated Generation Costs and On-Peak LMPs

Natural gas generation costs averaged $72.57/MWh (or $9.28/MWh) in 2022, double last year’s

gas costs and the highest since 2008.

136

See Section 1.2 of this report for more information on

the global, national and regional factors impacting 2022 natural gas prices.

The relationship between gas and energy costs varies within each year, especially during the

summer (Q3) when electricity demand is higher. Higher demand typically requires the

operation of less efficient natural gas-fired generators and/or generators that burn more

expensive fuels. During the summer months, efficient natural gas-fired generators earn higher

margins (commonly referred to as spark spreads) compared to other months.

137

Despite higher

Q3 demand (10% higher) compared to Q1, the highest energy costs (the product of demand and

135

Variable generation costs are calculated by multiplying the average daily fuel price ($/MMBtu) by the average standard

efficiency of generators of a given technology and fuel type. Our s tandard heat rates are measured in MMBtu/MWh as

follows: Natural Gas 7.8, Coal – 10.0, No. 6 Oil – 10.7, No. 2 Oil – 11.7. The spark spread is the difference between the on-

peak LMP and the fuel cost of a gas-fired generator with a heat rate of 7.8.

136

Based on a time series of historical Algonquin City Gates pricing data from Bloomberg.

137

During the winter months, coal- and oil-fired generators, as well as imports, can displace natural gas-fired generators in

economic merit order more frequently than in other seasons, as natural gas prices increase due to gas network demand

and constraints. This tends to lessen the impact of higher gas prices on LMPs as more costly gas-fired generators are

pushed out of merit and leads to reduced spark spreads.

$50

$100

$150

$200

$250

$300

$350

$400

$450

Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4

2018 2019 2020 2021 2022

$/MWh

Quarterly Values

Spark Spread DA LMP Natural Gas Coal No.6 Oil No.2 Oil Diesel

$40

$80

$120

$160

2018 2019 2020 2021 2022

Annual Values (excl. oil)

2022 Annual Markets Report 90 ISO New England Inc.

ISO-NE PUBLIC

LMP) were in Q1, totaling $3.7 billion and accounting for 32% of the annual total of $11.7

billion. Energy costs in Q3 totaled $3.4 billion.

Quarter 1 2022 saw the highest natural gas prices since Q1 2014, with gas generation costs

exceeding $100/MWh on average, resulting in average day-ahead LMPs of $115.23/MWh, also

the highest since Q1 2014. Natural gas prices remained high for several long periods,

particularly in January 2022, when prices frequently exceeded $20/MMBtu ($156/MWh

equivalent). More frequent in-merit oil generation (than prior years) attenuated upward price

pressure of natural gas on these higher-priced natural gas days.

138

Industry-standard profitability metrics

Industry-standard profitability metrics for gas-fired generators – implied heat rates and spark

spreads – indicate a significant increase in profitability across most generator efficiency levels

due to high gas prices and their pass-through effect on electricity prices. Table 3-2 shows the

annual average day-ahead on-peak LMP and natural gas price - the key inputs into the implied

heat rate, or breakeven point. For context, a heat rate of 7,800 Btu/kWh represents the average

standard efficiency of the New England fleet of combined cycle natural-gas fired generators, and

a heat rate of 6,381 Btu/kWh reflects the standard efficiency of a new entrant combined cycle

gas-fired generator.

Table 3-2: Annual Average On-Peak Implied Heat Rates and Spark Spreads

Year

Day-Ahead On-

Peak LMP

($/MWh)

Gas Price

($/MMBtu)

Implied Heat

Rate

(Btu/kWh)

Spread ($/MWh) corresponding to Heat Rate (Btu/kWh)

6,381

7,000

7,800

8,000

9,000

10,000

2018

50.11

5.05

9,918

17.87

14.74

10.70

9.69

4.64

(0.41)

2019

34.89

3.32

10,518

13.73

11.67

9.02

8.35

5.04

1.72

2020

26.57

2.12

12,558

13.07

11.76

10.07

9.65

7.53

5.41

2021

51.77

4.60

11,247

22.40

19.55

15.86

14.94

10.34

5.74

2022

92.17

9.07

10,159

34.28

28.66

21.40

19.59

10.52

1.44

The spark spreads for the typical New England gas-fired generator (7,800Btu/kWh) increased

significantly for the second year in a row, up 35% year-over-year ($15.86/MWh to

$21.40/MWh). The higher spark spreads were driven by the increase in gas prices and the pass-

through effect on energy prices.

139

The implied (breakeven) heat rate decreased by 10%

138

More information on New England’s oil generation in January 2022 can be found on EIA’s website here. More

information on oil-fired generator utilization can be found in Section 2.2.

139

For example, assume the implied heat rate was 10,000 MMBtu/MWh in both 2021 and 2022. Given the natural gas

prices, 2021 LMPs would average $46.00/MWh ($4.60/MMBtu x 10,000 Btu/kWh) in 2021 and $90.07/MWh

($9.07/MMBtu x 10,000 Btu/kWh) in 2022. Since we estimate a heat rate of 7,800 Btu/kWh for s tandard efficiency gas-

fired generator, the estimated cost of natural gas-fired generation would be $35.88/MWh ($4.60/MMBtu x 7,800

Btu/kWh) in 2021 and $70.75/MWh ($9.07/MMBtu x 7,800 Btu/kWh) in 2022. This means the spark spreads (or LMP

minus estimated cost of generation) would average $10.12/MWh ($46.00/MWh minus $35.88/MWh) in 2021 and

$19.32/MWh ($90.07/MWh minus $70.75/MWh) in 2022. In this example, the increase in natural gas prices caused the

increase in s park spreads.

2022 Annual Markets Report 91 ISO New England Inc.

ISO-NE PUBLIC

(11,247 Btu/kWh in 2021 to 10,159 Btu/kWh in 2022), indicating that slightly more efficient

gas generation was marginal on average in 2022.

Natural Gas Price-Adjusted LMP

While changes in LMPs have a strong, positive correlation with changes in natural gas prices,

other factors influence LMPs, including supply mix changes, system demand levels, and

unanticipated events, such as forced equipment outages. The gas price-adjusted LMP is a high

level metric used to estimate the impact of these of non-gas price factors on the energy price

and is shown in in the Figure 3-9 below.

140

The solid lines show the gas-price adjusted load-weighted LMPs, while the dashed lines are the

actual unadjusted LMPs for the day-ahead and real-time markets (note that the red and blue

lines overlap due to little average price separation between the markets on a load-weighted

basis).

Figure 3-9: Annual Average Natural Gas Price-Adjusted LMPs

With the exception of 2020, the day-ahead and real-time gas price-adjusted LMPs were within a

relatively narrow band ($29-$32/MWh), indicating that the change in gas prices explains nearly

all of the change in the LMP.

141

In 2022, on a gas price-adjusted basis, day-ahead and real-time

prices decreased slightly by 6% (from $32.60 to $30.65/MWh) and 4% (from $31.90 to

$30.55/MWh), respectively.

140

The gas price-adjusted LMP is derived by dividing the reference year natural gas price (2016) by the current year natural

gas price, then multiplying by the load-weighted LMP.

141

The results for 2020 were somewhat of an outlier. The behavior was largely due to less fixed supply on the system as a

result of increased nuclear generator outages and a nuclear generator retirement. This supply was replaced by more

expensive priced supply from gas-fired generation.

$20

$40

$60

$80

$100

2018 2019 2020 2021 2022

$/MWh

DA Gas-Adjusted LMP RT Gas-Adjusted LMP

RT LMP (Load-Weighted) DA LMP (Load-Weighted)

Reference Year: 2016

Gas-Price

Adjusted LMP

2022 Annual Markets Report 92 ISO New England Inc.

ISO-NE PUBLIC

Regional Greenhouse Gas Initiative (RGGI) Prices

The key driver of emissions costs for generators in New England is RGGI, a marketplace for CO

credits in the Northeast and Mid-Atlantic regions; it covers all six New England states. RGGI

operates as a cap-and-trade system, where fossil fuel-fired generators must purchase emission

allowances equal to their level of CO

emitted over a specific compliance period.

142

The average estimated dollar per MWh costs of CO

emissions and their percent contribution to

total variable production costs are shown in Figure 3-10 below.

143

The line series illustrate the

average estimated cost of emissions allowances by fuel type for the past five years. The bar

series show the proportion of the average energy production costs attributable to CO

emissions costs for each year.

144

145

Figure 3-10: Estimated Average Cost of RGGI CO

Allowances and Contribution of Emissions to Energy

Production Costs

As shown in Figure 3-10 above, the estimated RGGI costs for generators of all fuel types

increased over the period. RGGI allowance prices increased by 41% in 2022 (from $9.56/short

ton in 2021 to $13.48/short ton in 2022). For a typical natural gas-fired generator the average

estimated CO

cost was $6.15/MWh in 2022. This was an increase of $1.79/MWh from 2021.

142

For more information, s ee the RGGI website: https://www.rggi.org/program-overview-and-design/elements

143

Only fuel and CO

emission costs are considered in calculating the variable cost of each generator. In practice,

generators incur other variable operating and maintenance productions costs, but fuel comprises the vast majority of

variable costs. CO

prices in $ per ton are converted to estimated $/MWh using average generator heat rates for each fuel

type and an emissions rate for each fuel.

144

This average CO

cost is an estimated cost using average heat and emission rates. This figure shows the CO

costs

associated with the RGGI program only. Generators in Massachusetts are subject to additional CO

costs from the

Massachusetts GWSA program, which is covered further below.

145

RGGI accounts for nearly all of emissions costs.

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

$11

$12

$14

$15

Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4

2018 2019 2020 2021 2022

Average Contribution of CO2 Emissions to Energy Cost

Average Cost of CO2 Emissions ($/MWh)

Natural Gas No.6 Oil No.2 Oil Coal

Lines: $/MWh cost of

(left axis)

Bars: % contribution of

(right axis)

2022 Annual Markets Report 93 ISO New England Inc.

ISO-NE PUBLIC

The bars in Figure 3-10 show the relative contribution of emissions allowance costs to

generator energy costs. This contribution remained similar for all fuel types in 2022, although

the cost of CO

increased 41%, on average, from the previous year. The impact of these higher

prices on generator costs was diminished by a 101% increase in natural gas prices from

2021 to 2022.

Although prices increased from the previous year, the figure shows that costs flattened over

2022. This was due to RGGI allowance prices remaining around $13/short ton of CO

146

There

are several factors potentially influencing the price of RGGI allowances:

 The Emission Containment Reserve (ECR) continues in the program. Through the ECR,

allowances would be withheld from circulation to secure additional emission reductions

if prices fell below an established price.

147

 The conclusion of the third RGGI program review in 2023, signaling potential further

emission reductions similar to the second program review in 2017.

148

 Futures trading activity and participation by investors increased.

149

 Pennsylvania and North Carolina potentially joining the program.

A wider view of the impact of RGGI CO

allowances on generator production costs is presented

in Figure 3-11 below. The line series in the figure illustrate the quarterly estimated production

costs using the average heat rate for generators of a representative technology type in each fuel

category. The height of the shaded band above each line series represents the average energy

production costs attributable to emissions costs in each quarter.

146

Allowances remained below the Cost Containment Reserve (CCR) trigger price of $13.91/s hort ton in 2022. The CCR is a

quantity of allowances in addition to the cap which are held in reserve. These allowances are sold if prices exceed the

annual level. https://www.rggi.org/sites/default/files/Uploads/Market-Monitor/Quarterly-

Reports/MM_Secondary_Market_Report_2022_Q4.pdf

147

The 2022 ECR trigger price was $6.42/allowance and will rise 7% each year through 2030.

https://www.rggi.org/sites/default/files/Uploads/Market-Monitor/Quarterly-

Reports/MM_Secondary_Market_Report_2022_Q4.pdf

148

https://www.rggi.org/program-overview-and-design/program-review

149

https://www.rggi.org/sites/default/files/Uploads/Market-Monitor/Quarterly-

Reports/MM_Secondary_Market_Report_2022_Q3.pdf

2022 Annual Markets Report 94 ISO New England Inc.

ISO-NE PUBLIC

Figure 3-11: Contributions of Emissions Cost to Energy Production Costs

Figure 3-11 highlights that CO

allowance costs continue to have a relatively small impact on

generator production costs, and consequently, they do not have a noticeable impact on the

economic merit order of generators.

Massachusetts GWSA (310 CMR 7.74)

The Massachusetts CO

cap-and-trade program has been in place since 2018.

150

This program

provides additional requirements to the RGGI program discussed above, thus generators

located in Massachusetts must meet both requirements. Administered by the Massachusetts

Department of Environmental Protection (MassDEP), the program places an annual cap on

aggregate CO

production for the majority of fossil fuel-fired generators within the state.

151

The

cap will be lowered every year until the target annual CO

emission rate is reached in 2050.

152

The regulation requires fossil fuel-fired generators to hold an allowance for each metric ton of

they produce during a year. For the first two years, these allowances were primarily

allocated based on historical emissions levels, but beginning in 2021, allowances were available

for sale through auction only.

153

The program also allows generators to trade emissions

allowances to meet their compliance obligations.

150

310 CMR 7.74: Reducing CO

Emissions from Electricity Generating Facilities

https://www.mass.gov/guides/electricity-generator-emissions-limits-310-cmr-774

151

Participating generators are fossil-fuel generators with a nameplate capacity of 25 MW or more.

https://www.mass.gov/doc/310-cmr-700-air-pollution-control-regulations/download

152

The annual emissions cap for 310 CMR 7.74 will reduce by 223,876 metric tons in each subsequent year, eventually

reaching 1,791,019 metric tons in 2050.

153

For more information about allowance allocations see market monitor quarterly report:

https://www.mass.gov/doc/market-monitor-quarterly-report-2021-q3/download

$50

$100

$150

$200

$250

$300

$350

Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4

2018 2019 2020 2021 2022

Energy Costs ($/MWh)

Coal No.6 Oil No.2 Oil Natural Gas

Shaded area represents

the cost of emissions

2022 Annual Markets Report 95 ISO New England Inc.

ISO-NE PUBLIC

The annual volume of CO

allowances sold at auction and the respective volume-weighted

auction clearing prices for the MA GWSA program are shown in Figure 3-12 below.

154

155

Figure 3-12: Massachusetts GWSA Auction Results

The volume-weighted annual auction clearing prices increased 7% to $8.17/metric ton of CO

2022 from the previous year. The phasing out of direct allowance allocations after 2020 meant

that 2021 was the first year all allowances were offered at auction. The increase in price

reflected expectations of a need for more allowances to fulfill obligations in the future. They

also reflected some participants seeking to meet compliance obligations for the year as electric

load, and thus emissions increased.

156

In general, generators continued to incorporate the cost

of allowances into their energy market supply offers.

157

As the number of available allowances

decreases, prices are expected to rise. If the volume of secondary market transactions remains

low, participants may find it difficult to obtain additional allowances without paying significant

premiums.

158

154

For the 2018, 2019, and 2020 compliance years, MassDEP directly allocated 100, 75, and 50 percent of emissions cap.

Beginning in 2021, MassDEP no longer distributes allowances through direct allocation and all allowances were offered at

auction.

155

There were less allowances at auction in 2021 compared to 2022 because there were more allowances banked going

into that year. Auction volumes are adjusted based on banked allowances. For example, to calculate 2022 allowances to be

distributed, you subtract the 2.7 million banked allowances minus 223,875 (because the number of banked allowances is

over 223,875) from original cap of 8 million to get 5.6 million total allowances.

156

The total CO

emissions for all MA GWSA-affected generators was 6.7 metric tons in 2022, which was more than the

previous year.

157

To incorporate the cost of these allowances into generator reference levels, the IMM uses an adder that values the

allowances based on recent trades and auction results.

158

https://www.mass.gov/doc/market-monitor-quarterly-report-2022-q3/download

$10

$12

2019 2020 2021 2022

Allowances (Millions)

$/Metric Tons of CO₂

Adjusted Auction Volume Volume-Weighted Auction Clearing Price

2022 Annual Markets Report 96 ISO New England Inc.

ISO-NE PUBLIC

3.2.2 Supply-Side Participation

In 2022, unpriced supply made up around 70% of total supply in the energy market, a level

similar to previous years. Unpriced supply consists of offers from suppliers that are willing to

sell (i.e., clear) at any price, and offers that cannot set price. These suppliers may be insensitive

to price for a number of reasons, including fuel and power contracts, hedging arrangements,

unwillingness to cycle (on and off) a generator, or operational constraints. The remaining 30%

of supply is considered priced supply (i.e., it is willing to sell at a specified offer price or higher,

and it is eligible to set price).

There are three categories of unpriced supply: fixed imports, self-scheduled generation, and

generation-up-to economic minimum.

 Fixed imports are scheduled to flow power into New England on the external

interfaces regardless of price.

 Self-scheduled generation is offered into the energy market as must-run generation.

Generators self-schedule at their economic minimum (EcoMin).

159

 Generation-up-to economic minimum from economically-committed generators is

the portion of output that is below EcoMin. For example, if a unit generating 150 MW

has an EcoMin of 100 MW, then its generation-up-to EcoMin is the portion below 100

MW. Generation-up-to EcoMin is ineligible to set price, as the market software cannot

dispatch it up or down.

There are two categories of priced supply: priced native supply and priced imports.

 Priced native supply is energy from generators, demand response resources (DRRs),

and virtual transactions (day-ahead market only) that is dispatched economically (i.e., is

scheduled based on its price).

 Priced imports include price-sensitive imports and up-to-congestion transactions.

160

161

159

The Economic Minimum (EcoMin) is the minimum MW output available from a generator for economic dispatch.

160

Up-to-congestion (UTC) transactions are external contracts in the day-ahead energy market that do not flow if the

congestion charge is above a specified level. Real-time external transactions cannot be submitted as up-to-congestion

contracts. Participants with real-time external transactions are considered willing to pay congestion charges.

161

There are some nuances to the priced imports category in terms of price-setting ability. While priced imports regularly

set price in the day-ahead market, they rarely s et price in real-time market. This is because the tie lines are scheduled in

advance of the delivery interval in real time and are given a small dispatchable range in the real-time dispatch and pricing

algorithm. This prevents the market software from dispatching the tie lines far away from the scheduled amount

determined by the transaction scheduling process.

2022 Annual Markets Report 97 ISO New England Inc.

ISO-NE PUBLIC

An hourly average breakdown of unpriced and priced supply by category for the day-ahead and

real-time markets in 2022 is provided in Figure 3-13 below.

Figure 3-13: Day-Ahead and Real-Time Supply Breakdown by Hour Ending, 2022

On average, unpriced supply made up 70% and 69% of total supply in the day-ahead and real-

time markets, a small decrease compared to 2021 shares (72% and 71%). The decrease was

due to fewer fixed imports and more priced imports in 2022 compared to 2021.

162

Priced

supply averaged 31% of total supply over all hours in real time in 2022, with its share peaking

in hours ending (HE) 18-21 at 33-34%. In both markets, the daily ramp-ups in load are typically

met by additional supply from generation-up-to EcoMin and priced supply. In the day-ahead

market, the share of supply from self-scheduled generation (the largest component of unpriced

supply) and fixed imports was fairly stable over the course of a day. In contrast, in the real-time

market, average hourly self-scheduled generation was higher during midday due to output from

settlement-only solar generators (SOGs). In 2022, hourly SOG output averaged 428 MW, or 8%

of total real-time self-scheduled generation. These smaller generators do not clear in the day-

ahead market because they are not modeled in the market nor centrally dispatched by the ISO

control room.

163

Fewer fixed imports offset the impact of additional self-scheduled generation in

both markets in 2022 compared to 2021. Specifically, the resulting increase in native load

limited the price-suppressing effect of additional self-scheduled generation.

Price formation and unpriced supply

Large volumes of unpriced (or fixed) supply can have important implications for pricing

outcomes because it increases the likelihood of low or negative prices. This will become more

common when combined with significant amounts of additional capacity from renewable

generation (e.g., wind, and solar) with low marginal costs. However, we generally find that

energy price formation is robust under current levels of unpriced supply, with prices reflecting

the marginal input costs of the highest cost resources dispatched to meet demand. Further, as

more low marginal cost generation participates in the wholesale market, one would expect to

162

The metrics in this section focus on the s upply side and therefore do not account for exports, which increased in 2022,

leading to further decreases in net interchange. Section 5 discusses net interchange.

163

SOGs are passive participants in the real-time energy market only.

2,000

4,000

6,000

8,000

10,000

12,000

14,000

16,000

18,000

1 3 5 7 9 11 13 15 17 19 21 23

Hour Ending

Day-Ahead

Self-Scheduled Generation Fixed Imports Generation-up-to Ecomin Priced Imports Priced Native Supply

2,000

4,000

6,000

8,000

10,000

12,000

14,000

16,000

18,000

1 3 5 7 9 11 13 15 17 19 21 23

Hour Ending

Real-Time

Avg.

unpriced

supply:

70%

Avg.

unpriced

supply:

69%

2022 Annual Markets Report 98 ISO New England Inc.

ISO-NE PUBLIC

see a market response in terms of more price-responsive supply, as otherwise there is a higher

risk of energy prices not covering short-run production costs.

The example below illustrates negative pricing at a time when there is a significant amount of

fixed supply, and combined with negative energy supply offers, results in negative LMPs.

164

Figure 3-14: Priced and Unpriced Supply vs. Real-Time LMP, March 16, 2022

In the midday hours of March 16, real-time loads were relatively low. Additionally, wholesale

solar output (largely participating as SOGs) that had not cleared in the day-ahead market was

high, averaging about 1,050-1,875 MW between 10 am and 4 pm. This had two effects: 1) low

loads due to increased behind-the-meter solar output; and 2) more fixed supply from

settlement-only solar generation. As a result, the ISO only had to dispatch a small amount of

priced generation. The small amount of economically dispatched generation had offered into

the market at negative values, resulting in negative prices. The five-minute Hub LMP ranged

from -$129.50 to $0/MWh from 1:10 pm to 3:25 pm, and the hourly price averaged -$9.39 to -

$35.42/MWh from 1 pm to 4 pm.

The combination of low loads and large amounts of unpriced generation can bring about a

sudden drop in prices, to low or even negative levels. However, the overall frequency of

negative real-time prices at the Hub remains relatively low. Negative prices at the Hub occurred

in 0.5% of five-minute real-time pricing intervals in 2022, and in 0.3% of intervals in 2021.

Even in Maine, which tends to have a higher frequency of negative nodal prices at export-

constrained pockets with wind generation, the zonal price was negative in only 0.5% of five-

minute real-time pricing intervals in 2022 and 0.3% of intervals in 2021.

3.2.3 Reserve Adequacy Analysis Commitments

The commitment, dispatch and pricing outcomes in the financial day-ahead market may not

always reflect expected physical real-time conditions. For example, load-serving entities may

164

Unlike the figure above, this figure includes all imports in the fixed supply category for convenient illustrative purposes.

-60

-40

-20

100

120

2,000

4,000

6,000

8,000

10,000

12,000

14,000

16,000

18,000

20,000

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

3/16/2022

$/MWh

Fixed Supply Priced Supply RT_LMP_Mean

This led to low LMPs.

Low midday loads and high solar output resulted

in relatively little priced supply on the system,

and generators with low offers set price.

2022 Annual Markets Report 99 ISO New England Inc.

ISO-NE PUBLIC

clear less demand than the ISO’s load forecast. When this happens, ISO-NE must ensure there is

enough capacity and reserves to meet forecasted real-time load and reserve requirements.

To that end, after the day-ahead market, the ISO performs the Reserve Adequacy Analysis

(RAA) to meet these forecasted capacity and reserve requirements. If the day-ahead market did

not commit enough physical supply to meet the ISO’s forecasted demand and reserve, additional

generator commitments may be required. Conversely, if the day-ahead market satisfies the

expected real-time requirement, additional generators will not need to be committed in the

RAA process.

The difference between the amount of physical cleared generation (i.e., no virtual supply) in the

day-ahead market and the expected real-time load and reserve requirement (the energy gap), is

shown below (left axis) as a box plot in Figure 3-15.

165

166

The figure also includes the average

RAA committed capacity (right axis).

Figure 3-15: Average RAA Generator Commitments and the Day-Ahead Energy Gap

When the energy gap is high, the RAA process may need to commit additional generation to

meet real-time load. However, the energy gap remained fairly low in most hours, and the RAA

process committed less than 10 MW per hour on average. The day-ahead market generally

cleared sufficient supply and online generation capacity to meet the ISO’s load forecast and

reserve requirements. On average in 2022, the load-serving entities cleared more demand in

the day-ahead market than what they consumed in the real-time.

167

Therefore, additional RAA

commitments were not typically necessary as over-clearing of demand in day-ahead market led

to sufficient levels of physical supply.

165

The box plot shows the 25

and 75

percentiles (interquartile range), the median (50

percentile), along with the more

extreme observations at the 5

and 95

percentiles.

166

The Day-Ahead Ancillary Services Initiative (see Section 8) proposes to procure and price this “energy gap”.

167

See Section 3.2.5 for more information on high levels of demand clearing in the day-ahead market.

-30

-20

-10

-1,500

-1,000

-500

500

1,000

1,500

2018 2019 2020 2021 2022

RAA Commitments (MW)

Energy Gap (MW)

Energy Gap (LA) RAA Commitments (RA)

2022 Annual Markets Report 100 ISO New England Inc.

ISO-NE PUBLIC

3.2.4 Load and Weather Conditions

Load is a key determinant of day-ahead and real-time energy prices. Higher loads generally

require costlier generation to be dispatched, resulting in higher prices. Weather, economic

factors, energy efficiency measures, and, increasingly, behind-the-meter photovoltaic

generation drive changes in wholesale electricity load, in terms of both the level and profile of

load (load curve). The electrification of heating and transportation sectors are forecasted to

continue to increase wholesale load and will play a growing role in determining pricing

outcomes.

168

Demand/Load Statistics

Net Energy for Load (NEL) averaged 13,576 MW per hour in 2022, not meaningfully different to

the 2021 average (a 0.1% increase, 11 MW, compared to 2021). New England’s native

electricity load is shown in Table 3-3 below.

169

170

Table 3-3: Average, Peak and Weather-Normalized Load

Year

Load (GWh)

Average Hourly

Load (MW)

Peak Load

(MW)

Weather

Normalized

Load (GWh)

Average

Hourly

Weather

Normalized

Load (MW)

2018

123,472

14,095

26,024

120,560

13,762

2019

119,235

13,614

24,361

118,772

13,558

2020

116,874

13,309

25,121

116,322

13,242

2021

118,758

13,565

25,801

117,551

13,419

2022

118,874

13,576

24,780

118,337

13,508

During the year, load reached a peak of 24,780 MW, which occurred in HE 18 on August 4, when

temperatures reached nearly 92⁰F, the hottest weekday over the five-year period. This was

4.0% (or 1,020 MW) lower than the peak load in 2021.

171

July 24

saw higher temperatures,

reaching just over 94⁰F but it was a weekend day when loads tend to be lower.

New England has had declining weather-normalized load, which reflects the long-term trend of

increased energy efficiency and behind-the-meter solar generation. However, weather-

normalized load increased slightly over the past two years. While it is difficult to attribute this

directly to any particular driver, this change is consistent with the ISO forecast that average

load will increase each year with the continued adoptions of electricity-fueled transportation

and electric heating.

168

For more information on the growth of electrification, see the ISO New England CELT Report.

169

In this analysis, load refers to net energy for load (NEL). NEL is calculated by summing the metered output of native

generation, price-responsive demand and net interchange (imports – exports). It excludes pumped-storage demand.

170

Weather-normalized load estimates what load would be if monthly total heating degree days and cooling degree days

were in line with historical averages. The estimate also factors in differences non -holiday weekdays and leap days.

171

The all-time peak load was 28,130 MW and occurred on August 2, 2006 at HE 15.

2022 Annual Markets Report 101 ISO New England Inc.

ISO-NE PUBLIC

Demand Curves and Energy Prices

The strong connection between energy prices and load is particularly evident over the course of

the operating day. Figure 3-16 below depicts two series of figures. The first shows the average

time-of-day profile for both day-ahead demand and real-time load (solid lines) compared to

day-ahead and real-time LMPs (dashed lines). The second shows how wholesale, settlement-

only and behind-the-meter (BTM) solar generation (yellow areas) compare to the ISO solar

forecast (dashed line). BTM solar generation lowers load required to be met through the

wholesale market. Since load curves and solar production have different shapes depending on

season, the figure is broken into three panels: winter (December, January, and February),

summer (June – August) and rest of year.

Figure 3-16: Average Demand and LMP by Hour, 2022

172

Day-ahead internal demand is equal to fixed demand + price-sensitive demand + virtual demand. This includes pumped-

storage demand and excludes virtual demand at external nodes. Real-time load is the total end-use wholesale electricity

load within the ISO New England footprint.

$20

$40

$60

$80

$100

$120

$140

$160

$180

$200

10,000

11,000

12,000

13,000

14,000

15,000

16,000

17,000

18,000

19,000

20,000

1 3 5 7 9 11 1315 1719 2123

$/MWh

Hou Ending

Summer Months

Day-Ahead Internal Demand Real-Time Load Day-Ahead Hub LMP Real-Time Hub LMP

$20

$40

$60

$80

$100

$120

$140

$160

$180

$200

10,000

11,000

12,000

13,000

14,000

15,000

16,000

17,000

18,000

19,000

20,000

1 3 5 7 9 11 1315 17 19 21 23

$/MWh

Hour Ending

Rest of Year

$20

$40

$60

$80

$100

$120

$140

$160

$180

$200

10,000

11,000

12,000

13,000

14,000

15,000

16,000

17,000

18,000

19,000

20,000

1 3 5 7 9 1113 15 17 19 21 23

$/MWh

Hour Ending

Winter Period

500

1,000

1,500

2,000

2,500

3,000

3,500

4,000

1 3 5 7 9 11 13 15 17 19 21 23

Hour Ending

Summer Period

Behind-the-meter

500

1,000

1,500

2,000

2,500

3,000

3,500

4,000

1 3 5 7 9 11 13 15 17 19 21 23

Hour Ending

Winter Period

500

1,000

1,500

2,000

2,500

3,000

3,500

4,000

1 3 5 7 9 11 13 15 17 19 21 23

Hour Ending

Rest Of Year

2022 Annual Markets Report 102 ISO New England Inc.

ISO-NE PUBLIC

In the winter, high gas prices lead to higher energy prices even with lower energy demand than

the summer months. Winter, like the rest of the year, also has a distinct evening demand peak.

These peaks have been made more prevalent with the increase in solar generation. Winter solar

generation is notably less than the summer and rest-of-year but has still depressed midday

loads. During the summer, load and prices tend to climb throughout the day as temperatures

increase and air conditioning demand rises. Even with solar production reaching over 3,500

MW during air conditioning demand had not allowed for midday dips. The right panel shows

the average load curve for the rest of the year, when load usually has morning and evening

peaks, with a midday dip. In 2022, even with over 3,000 MW of solar generation, the rest of the

year load profile does not have the usual midday dip. Figure 3-16 also shows that the day-ahead

market tends to clear more internal demand than actually materializes in real time, which is

discussed further in Section 3.2.5.

Impact of Weather

Weather is a significant driver of load in New England. In 2022, average quarterly temperature

was consistent with the prior year, leading to consistent load year-over-year. While warmer

weather in the winter typically contributes to lower loads, warmer weather in the summer

leads to higher electricity usage due to increased air-conditioning demand.

173

Quarterly average

and five-year average temperatures for 2018 through 2022 are illustrated in Figure 3-17

below.

174

Figure 3-17: Seasonal vs. Five-Year Average Temperatures

Load levels tend to be more sensitive to temperature changes during Q3 and Q1, due to cooling

and heating demand, respectively. Higher Q3 temperatures and lower Q1 temperatures

(compared to 2021) both had the impact of increasing load. In Q1 2022, temperatures averaged

173

While the system currently peaks in the summer, the system is forecasted to become a winter peaking system as soon

as Winter 2029. For more information see the Pathways Study.

174

Actual New England temperatures are based on weighted hourly temperatures measured in eight New England cities:

Windsor CT, Boston MA, Bridgeport CT, Worcester MA, Providence RI, Concord NH, Burlington VT, and Portland ME.

11,000

11,500

12,000

12,500

13,000

13,500

14,000

14,500

15,000

15,500

16,000

16,500

20°

25°

30°

35°

40°

45°

50°

55°

60°

65°

70°

75°

2018

2019

2020

2021

2022

2018

2019

2020

2021

2022

2018

2019

2020

2021

2022

2018

2019

2020

2021

2022

Q1 Q2 Q3 Q4

Average Quarterly Load (MW)

Temperature (

⁰F)

Five-Year Quarterly Average Temperature Quarterly Average Temperature

Quarterly Average Load

2022 Annual Markets Report 103 ISO New England Inc.

ISO-NE PUBLIC

32⁰F, down 1.0⁰F from Q1 2021 (33⁰F) and just below the five-year average (33⁰F). The slightly

colder temperatures led to a small increase in Heating Degree Days (HDD) (2,985 vs. 2,894),

with a 2% (305 MW per hour) increase in demand. Notable, average temperatures in January

2022 were 4⁰F cooler than January 2021 and average loads were 854 MW higher.

For the remainder of the year, the average temperature was warmer than the five-year average.

The Q2 and Q3 average temperatures were within one degree of the five-year average.

Compared to 2021, Q3 2022 loads increased due to warmer weather. Heat waves in July and

August 2022 increased loads as air-conditioning demand increased.

In the lowest demand period, Q2 and Q4, weather tends not to be a significant driver. In fact,

BTM solar generation has a comparatively larger impact on wholesale load levels during these

quarters. Q2 average load decreased the most from the prior year, down by 3.5% (434 MW). Q4

2022 was 1.3⁰F warmer than the five-year average resulting in lower loads. This was the largest

temperature difference from the five-year-average of any quarter in 2022, and resulted in the

lowest Q4 load over the five-year period.

3.2.5 Demand Bidding

The quantity and pricing of bid-in demand in the day-ahead market has important pricing and

operational implications. For example, generator commitments for the operating day are

determined by the clearing process that matches supply and demand at last cost, which in turn

impact ISO’s reserve adequacy analysis (RAA).

175

In this section, we examine native day-ahead cleared demand (i.e., delivery within the New

England jurisdiction, which excludes exports).

176

Native demand consists of cleared fixed, price-

sensitive, virtual, and pumped-storage demand bids. Day-ahead cleared demand by bid type as

a percentage of real-time load is shown below in Figure 3-18.

177

175

The reserve adequacy analysis (RAA) is conducted after the day-ahead market is finalized and is designed to ensure

sufficient capacity is available at least cost to meet ISO-NE real-time demand, reserve requirements, and regulation

requirements. For more information see Section 3.2.3.

176

Exports are not included as this section focuses on demand participation within New England. Exports are discussed in

Section 1.4 and Section 5.

177

Real-time load is the total end-use wholesale electricity load within the ISO New England footprint. Real-time load is

equal to Net Energy for Load – Losses.

2022 Annual Markets Report 104 ISO New England Inc.

ISO-NE PUBLIC

Figure 3-18: Day-Ahead Cleared Demand as a Percentage of Real-Time Load by Bid Type

In 2022, participants continued to clear more demand in the day-ahead market than they

actually consumed in real time, with a modest average increase compared to 2021 (100.6% vs.

101.1%). However, it should be noted that this does not account for time-of-day.

In Figure 3-19 below day-ahead cleared demand as a percentage of real-time load by bid type is

averaged by hour. Similar to above this figure shows that, on average, during every hour of the

day participants cleared more than 100% of real-time load in the day-ahead market.

Figure 3-19: Day-Ahead Cleared Demand as Percentage of Real-Time Load by Bid Type by hour, 2022

Both “price-sensitive” and virtual demand’s share of cleared demand increased in 2022 to the

highest percentage over the five-year period. Price-sensitive transactions accounted for 36% of

real-time load, a 0.6% increase from 2021 (35.3%); virtual demand transactions increased

20%

40%

60%

80%

100%

2018 2019 2020 2021 2022

Day-Ahead Cleared Demand as a Percentage of Real

Time Metered Internal Demand (%)

Fixed Demand Price-Sensitive Demand

Pumped Storage Demand Virtual Demand

Uncleared in the Day-Ahead Market

20%

40%

60%

80%

100%

01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

Day-Ahead Cleared Demand as a Percentage of Real

Time Metered Internal Demand (%)

Hour Ending

Fixed Demand Price-Sensitive Demand

Pumped Storage Demand Virtual Demand

Uncleared in the Day-Ahead Market

2022 Annual Markets Report 105 ISO New England Inc.

ISO-NE PUBLIC

slightly from 2.7% in 2021 to 2.9% in 2022. In addition, fixed day-ahead cleared demand was

the lowest average over the period, averaging 61.2% of real-time load in 2022, a 0.2% decrease

from 2021 (61.4%).

However, even though price-sensitive demand bids are submitted with a MW quantity and

corresponding price, the majority of bids are priced significantly above the LMP, and are, in

practical terms, fixed demand bids. Lastly, pumped-storage demand accounted for 1.0% of real-

time load, consistent with 2021.

Cleared internal demand bids by price are shown in Figure 3-20 below. The bid prices are

shown on the vertical axis of both graphs. In the first graph the percentage of cleared bids that

were willing to pay at each bid price are shown on the horizontal axis. For example, nearly 97%

of cleared day-ahead demand was willing to pay more than $374.98/MWh, the maximum day-

ahead hub LMP in 2022.

Figure 3-20: Components of Day-Ahead Cleared Demand as a Percentage of Total Day-Ahead Cleared

Demand

Demand in New England continued to be primarily price insensitive. Nearly two-thirds (61%)

of total day-ahead cleared demand was bid as fixed demand, so it clears in the market at any

price. While price-sensitive demand bids have an attached price, the price usually exceeds the

day-ahead LMP. Therefore, price-sensitive demand bids typically clear, accounting for over 36%

of all day-ahead cleared demand. Virtual demand and price-sensitive pumped-storage demand

bids often have lower prices attached to the bid, so they do not clear as often. However, virtual

and pumped-storage demand only account for approximately 4% of cleared demand bids.

3.2.6 Load Forecast and Market Implications

This section focuses on the day-ahead load forecast: the daily forecast made around 9:30 am

that predicts hourly load for the next operating day and is published on the ISO website.

178

This

178

Twice a day, the ISO produces a three-day system load forecast that projects load for the current day and the following

two days. The first forecast is typically released around 4:30 am and the second, and typically final forecast, is published

near 9:30 am.

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

1,000

900

800

700

600

500

400

300

200

100

-100

Bid Price ($/MWh)

Virtual

Demand: 3%

Price Sensitive

Demand: 36%

Fixed Demand:

61% of total

Nearly 97% of cleared demand was

bid at a price over $374.98/MWh - the

maximum day-ahead LMP in 2022.

Pump Storage

Demand: 1%

2022 Annual Markets Report 106 ISO New England Inc.

ISO-NE PUBLIC

forecast is the ISO’s last load projection prior to the close of the day-ahead market offer

window. Although the ISO’s forecast is not a direct input into the day-ahead market, it serves as

an informational tool for participants bidding in the day-ahead market, and generally aligns

well with total day-ahead cleared demand.

179

Additionally, and as mentioned in Section 3.2.3,

the load forecast is used in the RAA process to finalize the ISO’s next day operating plan.

When the market’s (participants collectively) load forecast is higher than actual load, LSE’s will

tend to clear more price-insensitive demand in the day-ahead market, resulting in the clearing

of higher-priced supply than is needed in real-time. Consequently, day-ahead prices will tend to

be higher than real-time prices as the most expensive physical supply is backed down.

Conversely, lower forecasted load tends to have the opposite effect. We see this dynamic play

out frequently in our empirical assessments, and there is a meaningful statistical relationship

between forecast error and pricing as we present below.

Load Forecast Accuracy

Just as the day-ahead market cannot perfectly predict real-time conditions, the ISO load forecast

will inevitably differ from real-time load. Since weather is both a key driver of load and difficult

to predict, real-time load is challenging to forecast. Additionally, increasing amounts of behind-

the-meter solar generation (discussed in more detail below) add to the challenge of accurately

estimating load even over short time horizons.

180

The mean absolute percent error (MAPE) of the ISO’s day-ahead load forecast by time of year is

shown in Figure 3-21 below. Months are partitioned into four groups based on the ISO’s

monthly load forecast goal (shown as dashed lines).

181

179

Load Serving Entities (LSEs) may also rely on their own in-house or third-party forecasting tools to inform their day-

ahead bidding strategy.

180

The ISO has revised the load forecasting goals to account for growing behind-the-meter s olar generation which

increases the volatility of the load forecast.

181

The ISO’s target MAPE goals during the year are 1.8% in January–April and October–December; 2.0% in May and

September, and 2.6% in June–August.

2022 Annual Markets Report 107 ISO New England Inc.

ISO-NE PUBLIC

Figure 3-21: ISO Day-Ahead Load Forecast Error by Time of Year

Figure 3-21 shows that 2022 forecast error and volatility decreased compared to 2021. In 2022,

the MAPE for May & Sep. increased by 0.1%, from 2.1% to 2.2%, remaining above the ISO’s goal.

The MAPE decreased for Jun – Aug. by 0.6%, from 2.5% in 2021 to 1.9% in 2022, falling well

below the ISO’s goal. The other two groups remained constant with 2021 errors. At a monthly

level, the ISO’s forecast missed the monthly goal in five months in 2022, consistent with five

months in 2021.

Impact of Behind-The-Meter Solar

The growth in behind-the-meter

182

(BTM) solar generation in recent years add to the challenge

of forecasting load.

183

For one, it is difficult to estimate the location and installed capacity of

thousands of small-scale solar installations around New England. Second, forecasting solar

generation at a granular level is notoriously difficult with cloud cover or snowfall having

significant impacts.

184

BTM generation reduces wholesale load, and therefore when less solar

generation occurs than anticipated, the ISO may need to commit more expensive generators to

meet higher real-time wholesale load.

The relationship between the daily average BTM solar forecast and the system level load

forecast is shown below in Figure 3-22.

182

With respect to forecasting, the ISO includes settlement-only s olar generation or SOGs in the forecasted impact of BTM.

183

In recent years the ISO has made significant investments to better forecast BTM solar generation.

183

For more

information on ISO New England’s investment in forecasting behind-the-meter solar generation, see

https://www.esig.energy/building-data-intelligence-for-short-term-load-forecasting-with-behind-the-meter-pv/

184

See, for example, https://www.bnl.gov/isd/documents/94838.pdf.

0.0%

0.5%

1.0%

1.5%

2.0%

2.5%

3.0%

3.5%

4.0%

2018

2019

2020

2021

2022

2018

2019

2020

2021

2022

2018

2019

2020

2021

2022

2018

2019

2020

2021

2022

Jan-Apr May & Sep Jun-Aug Oct-Dec

Absolute Percent Forecast Error (%)

Interquartile Range MAPE ISO MAPE Goal

2022 Annual Markets Report 108 ISO New England Inc.

ISO-NE PUBLIC

Figure 3-22: Impact of BTM Solar on Load Forecast Error

Figure 3-22 shows that BTM solar forecast error generally causes a greater load forecast error.

In other words, when there is less BTM solar generation than forecasted, system load is

typically higher than the ISO’s load forecast (top right quadrant) and vice versa (bottom left

quadrant). In 2022, this relationship was stronger than in prior years, suggesting that load

forecast error was more correlated with BTM photovoltaic (PV) forecast error in 2022 than in

prior years. However, this relationship does not always hold as other factors, like temperature,

impact load forecasting.

The Interaction between Forecast Error and Pricing Outcomes in 2022

The ISO’s load forecast error tends to be consistent with the market’s forecast error. This has

important market clearing and pricing implications as discussed previously. The statistical

relationship between average daily load forecast error and price divergence is shown in Figure

3-23 below.

-1000

-800

-600

-400

-200

200

400

600

800

1000

-1,000 -800 -600 -400 -200 0 200 400 600 800 1,000

BTM PV Forecast

- BTM PV Generation (MW)

Real-Time Load - Forecasted Load (MW)

Higher RT Load

Lower BTM PV Generation

Lower RT Load

Higher BTM PV Generation

2022 Annual Markets Report 109 ISO New England Inc.

ISO-NE PUBLIC

Figure 3-23: Price Difference and Forecast Error Relationship

Figure 3-23 illustrates that in 2022 there was a positive correlation between the forecast error

and the price difference between the day-ahead and real-time markets. In other words, when

real-time loads were higher than day-ahead forecasted demand, real-time prices tended to be

higher than day-ahead prices (top right quadrant), and vice versa (bottom left quadrant).

Similarly to above, this relationship was stronger in 2022 than in prior years, suggesting that

load forecast error was more correlated with price differences in 2022 than in previous years.

3.2.7 System Events During 2022

System events, such as unusual conditions resulting from generator outages or load forecast

error, can have a significant impact on energy market outcomes. Two events occurred in 2022

that bear specific mention: the May 21 minimum generation emergency and the December 24

capacity deficiency (shortage event). This section details the frequency of system events and

abnormal conditions over the past five years, and then provides a summary of the May 21 and

December 24 events.

To provide context for the events covered in this section, the graph below overlays their timing

on a time series of daily real-time LMPs and load levels. Also shown is the minimum total

reserve margin (for an hour) for each day.

-50

-40

-30

-20

-10

-1,000 -800 -600 -400 -200 0 200 400 600 800 1,000

RT - DA Price ($/MWh)

Real-Time Load - Forecasted Load (MW)

Higher RT Load

Higher RT Prices

Lower RT Load

Lower RT Prices

2022 Annual Markets Report 110 ISO New England Inc.

ISO-NE PUBLIC

Figure 3-24: Pricing, Demand and the Reserve Margin during System Events in 2022

The following metrics illustrate the frequency of abnormal system conditions and extreme

market outcomes over the past five years:

 Number of OP-4 and M/LCC 2 Events

 Reserve Deficiency Events

 Frequency of Extreme Hub LMPs

Each of these is examined in more detail below.

OP-4 and M/LCC 2 Events

The ISO uses the following established procedures to alert participants and address issues

during times of tight or abnormal system conditions:

 Master Local Control Center Procedure No. 2 (M/LCC 2, Abnormal Conditions Alert)

185

notifies market participants and power system operations personnel when an abnormal

condition is affecting the reliability of the power system, or when such conditions are

anticipated. The ISO expects these entities to take certain precautions during M/LCC 2

events, such as rescheduling routine generator maintenance to a time when it would be less

likely to jeopardize system reliability.

 Operating Procedure No. 4 (OP-4, Action during a Capacity Deficiency)

186

establishes

criteria and guidelines for actions during capacity deficiencies. There are eleven actions

described in the procedure that the ISO can invoke as system conditions worsen.

The number of instances for each type of event during the reporting period is shown in Table

3-4 below.

185

For more information on M/LCC 2, see https://www.iso-ne.com/static-

assets/documents/rules_proceds/operating/mast_satllte/mlcc2.pdf

186

See ISO New England Operating Procedure No. 4 – Action During A Capacity Deficiency, available at

https://www.iso-ne.com/static-assets/documents/rules_proceds/operating/isone/op4/op4_rto_final.pdf

5,000

10,000

15,000

20,000

25,000

100

200

300

400

500

600

Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

Average Hourly Load (MW)

Average LMP ($/MWh)

LMP Load Hourly Reserve Margin, Min Daily Value

-1000

1000

2000

3000

4000

Reserve

Margin (MW)

2022 Annual Markets Report 111 ISO New England Inc.

ISO-NE PUBLIC

Table 3-4: OP-4 and M/LCC 2 Event Frequency

2018

2019

2020

2021

2022

# of OP-4 Events

# of M/LCC 2 Events

The ISO implemented six M/LCC 2 events in 2022, the same amount as in 2021. In our

assessment of the system events, detailed in the quarterly reports, we generally found that the

market performed well during these periods. The following bullet points summarize the high-

level causes and outcomes of each M/LCC 2 event in 2022.

 January 11: The ISO declared M/LCC 2 due to an imminent capacity deficiency caused by

the partial loss of the Phase II interconnection with Hydro Quebec and several

generators tripping offline due to mechanical issues. Five-minute real-time Hub LMPs

peaked during the evening ramp at $276.92/MWh. This was higher than the peak

hourly day-ahead price of $234.86/MWh, which also occurred during the evening ramp.

 January 28-30: The ISO declared M/LCC 2 due to severe weather when Winter Storm

Kenan brought heavy snowfall to the region. Customer outages peaked at approximately

125,000 customers on January 29, which had little effect on total system load. Five-

minute real-time Hub LMPs peaked at $373.97 during the late morning on January 28,

while the hourly day-ahead price peaked at $260.42 during the evening ramp on

January 29.

187

 July 19-24, August 4, and August 8 (three events total): M/LCC 2 events were declared

on these heat wave days due to imminent capacity deficiencies on days with hot

temperatures and high loads. Five-minute real-time Hub LMPs peaked at

$1,036.92/MWh during the evening ramp on July 20 as a result of high reserve prices.

Hourly day-ahead Hub LMPs peaked at $362.74/MWh during the evening ramp on July

22.

188

 December 24: The ISO declared M/LCC 2 due to an imminent capacity deficiency driven

by cold weather, unplanned generator outages, and reductions in net imports. This

resulted in the highest prices of the year, with five-minute real-time LMPs peaking at

$2,816/MWh in the afternoon. This was notably higher that the peak hourly day-ahead

Hub LMP of $315.27.

The only OP-4 event of 2022 occurred on December 24, which also coincided with an M/LCC 2

event. The last OP-4 event before that occurred on September 3, 2018. The latter part of this

section describes this day in detail.

Reserve Deficiency Events

Reserve deficiency events (i.e., periods when there are negative reserve margins) are indicative

of stressed system conditions. In these instances, the system does not have enough available

supply to meet the requirement for an off-line reserve product, and the product is priced at the

Reserve Constraint Penalty Factor (RCPF prices of $1,000-$1,500/MWh) and added to the

187

The Winter 2022 QMR covers the January 11 and January 28-30 M/LCC 2 events in detail, and is available at:

https://www.iso-ne.com/static-assets/documents/2022/05/2022-winter-quarterly-markets -report.pdf

188

The Summer 2022 QMR covers the July and August M/LCC 2 events in detail, and is available at: https://www.iso-

ne.com/static-assets/documents/2022/11/2022-summer-quarterly-markets-report.pdf

2022 Annual Markets Report 112 ISO New England Inc.

ISO-NE PUBLIC

energy price. Below, Table 3-5 shows the number of hours during which each reserve margin

was negative.

Table 3-5: Frequency of Negative Reserve Margins (System Level)

189

Year

Hours of Negative

Total30 Margins

Hours of Negative

Total10 Margins

Hours of Negative

Spinning Reserve

Margins

2018

2.4

0.9

68.1

2019

0.0

25.9

2020

0.0

14.5

2021

0.0

26.8

2022

1.4

0.1

48.1

Overall, these outcomes reflect a system that has had a healthy reserve margin on average with

few periods of system stress in the past few years. The Total30 and Total10 margins (which are

associated with the TMOR and TMNSR products, respectively) fell below zero in 2018 and 2022,

when tight system conditions led to capacity scarcity events. No comparable events occurred

during 2019-2021. Shortages of ten-minute spinning reserves (with an associated RCPF of

$50/MWh) were also more frequent in 2022 compared to 2019-2021, but occurred less

frequently than in 2018. The spinning reserve shortages occurred across 83 days throughout

the year in 2022 due to a variety of factors, such as tight system conditions caused by higher

real-time loads or unplanned outages.

Frequency of Extreme Hub LMPs

High real-time LMPs can also indicate stressed system conditions, as higher-cost generation is

required to meet load and reserve requirements. The duration curves in Figure 3-25 below

show the top 1% of hourly average real-time LMPs ranked from high to low over the past five

years.

189

The calculations in this table come from the LMP calculation processes in the real-time market software. The “Hours of

Negative Total30 Margins” column does not include instances where only the replacement reserve margin is negative,

because those instances are not associated with the high $1,000/MWh and $1,500/MWh RCPFs.

2022 Annual Markets Report 113 ISO New England Inc.

ISO-NE PUBLIC

Figure 3-25: LMP Duration Curves for Top 1% of Real-Time Pricing Hours

This figure shows that 2022 saw the highest real-time Hub LMPs since 2018, reflecting higher

natural gas prices and tight conditions during system events. Of the top 1% of hourly average

real-time Hub LMPs for 2022, 18% occurred on December 24, including the highest hourly price

of the year ($2,254/MWh). There was also a shortage event in 2018 that resulted in high non-

spinning reserve pricing and high LMPs. No comparable events occurred in 2019-2021.

Overview of Notable Events

May 21, 2022: Minimum Generation Emergency

On May 21, the ISO declared a minimum generation emergency event between 12:00 and 14:00.

This abnormal system occurs when the ISO expects that generation and external transactions

will exceed system demand, which could result in high system frequencies and unscheduled

flows of power into neighboring control areas.

190

With limited downward flexibility, the ISO

takes action to reduce supply to emergency minimum levels and prices are administratively set

to negative $150/MWh. This sends strong signals to supply to reduce output, or conversely to

demand to increase consumption (e.g., battery charging, pumped-storage hydro demand).

Actual loads were significantly less than the forecasted load heading into the operating day due

to lower temperatures than expected and sunnier conditions, resulting in more solar

production. The day-ahead market cleared an excess amount of physical generation compared

to what was required in real-time. Hourly forecasted and actual loads are shown in Figure 3-26

below.

190

For additional information on minimum generation emergencies, see:

https://www.iso-ne.com/participate/support/faq/minimum-generation-emergency

-200

200

400

600

800

1,000

1,200

1,400

1,600

1,800

2,000

2,200

2,400

0.0% 0.5% 1.0%

$/MWh

2018 2019 2020 2021 2022

2022 LMPs were higher than those of

2019-2021 across all ranked observations

in the top 1% of hourly RT LMPs.

Highest 2022 value:

LMP = $2254/MWh

Highest 2021 value:

LMP = $375/MWh

2022 Annual Markets Report 114 ISO New England Inc.

ISO-NE PUBLIC

Figure 3-26: Actual vs. Forecasted Load on May 21

During the minimum generation emergency, gas-fired generators were dispatched down and

produced 2,300 to 2,500 MW less in real time compared to the total day-ahead cleared amount.

The control room reduced transactions across Phase II and Highgate, which led to 500 to 1,200

MW fewer net imports during the event. Prices across the system were administratively set at -

$150/MWh. This occurred for two hours between 12:15 and 14:15. Five-minute real-time Hub

and zonal prices are shown in Figure 3-27 below.

Figure 3-27: Real-Time Prices on May 21, 2022

6,000

8,000

10,000

12,000

14,000

16,000

18,000

20,000

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

Hour End

Actual Load Load Forecast

Actual loads 1,151 to 1,785 MW

less than forecast from HE 12-19

-$200

-$150

-$100

-$50

$50

$100

$150

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

Energy LMP ($ per MWh)

Hour End

Maine (ME) New Hampshire (NH)

Vermont (VT) Connecticut (CT)

Rhode Island (RI) Northeast Massachusetts (NEMA)

Southeast Massachusetts (SEMA) Western Central Massachusetts (WCMA)

Hub

2022 Annual Markets Report 115 ISO New England Inc.

ISO-NE PUBLIC

December 24, 2022: Shortage event

On Saturday, December 24, 2022, the ISO experienced its first capacity scarcity conditions (CSC)

in over four years. The ISO initiated emergency procedures in response to tight system

conditions that resulted from multiple factors, including cold weather, unplanned generator

outages, and reductions in net imports. The conditions led to operating reserve deficits and high

energy prices.

A timeline of the ISO’s emergency procedure actions, along with energy prices, is shown in

Figure 3-28 below.

Figure 3-28: December 24 Event Timeline

The following paragraphs detail the factors that contributed to the event and how event

outcomes were reflected in market settlements.

Weather and Load: On the day before the shortage event, Winter Storm Elliott affected the

region and resulted in around 200,000 customer outages. Temperatures declined rapidly

between December 23 and 24, resulting in a steep increase in demand for natural gas for

heating. Despite the load forecast error, real-time load was close to or under day-ahead cleared

demand for most of the day.

Net Imports: During the capacity scarcity condition, real-time imports from neighboring control

areas decreased relative to day-ahead cleared imports. Deviations between day-ahead cleared

and real-time scheduled MWs by interface are illustrated in Figure 3-29 below.

$500

$1,000

$1,500

$2,000

$2,500

$3,000

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

Hour End

$/MWh

M/LCC 2 OP-4 Capacity Scarcity Conditions Day-Ahead LMP Real-Time Hub LMP

OP-4 from 16:30 to 19:00

Capacity scarcity conditions from 16:40 to 18:00

M/LCC 2 from 16:00 to 21:00

2022 Annual Markets Report 116 ISO New England Inc.

ISO-NE PUBLIC

Figure 3-29: Day-Ahead vs. Real-Time Net Import Deviations by Interface

On December 24, there was a substantial decline in real-time net imports compared to the day-

ahead schedule. The deviations were most severe from 7:00 to 11:00 and 13:00 to 18:00, when

there were 910-1,675 MW fewer net imports in real-time. Around the time of the capacity

scarcity conditions in the afternoon, the most substantial reductions in net imports occurred

over the Canadian interfaces due to high demand and transmission system issues there.

Unplanned Generator Outages: The system lost a large volume of generation capacity shortly

before and during the operating day on December 24 due to unplanned generator outages. The

outages were caused by mechanical problems rather than by fuel availability issues. While out-

of-service capacity varied throughout the day based on the start and end times of the outages,

the sum of out-of-service generation that tripped on December 23 and December 24 was about

2,180 MW. Most of the resources that tripped were older generators that are typically only

committed during tight system conditions.

Supply Mix Changes: Supply Mix Changes: The unplanned outages and net import deviations

resulted in changes to the real-time supply mix compared to the day-ahead mix. The breakdown

in differences between day ahead and real-time generation obligations is shown in Figure 3-30

below.

(1,500)

(1,000)

(500)

500

1,000

1,500

1 3 5 7 9 11 13 15 17 19 21 23 1 3 5 7 9 11 13 15 17 19 21 23

Dec 23 Dec 24

New York North Cross Sound Northport-Norwalk

Phase II New Brunswick Highgate

Pool RTM NI dev.

2022 Annual Markets Report 117 ISO New England Inc.

ISO-NE PUBLIC

Figure 3-30: Differences between Hourly Real-time and Day-Ahead Generation Obligations

During several hours on December 24, there was significantly less real-time gas generation

compared to the day-ahead schedule, primarily due to generator trips and a large dual-fuel

generator switching to oil in real-time. Additionally, large decreases in imports occurred

leading up to and during the capacity scarcity conditions.

191

During the shortage event, there

was increased fast-start generation from pumped-storage and oil-fired generators.

Most of the additional generation commitments were made by the market software rather than

by the operators. From HE 17 to HE 18, the real-time market software committed 2,560-2,680

MW of fast-start generation. The system operators manually committed 175-200 MW of

generation during the same period.

LMPs and real-time reserve pricing: Five-minute real-time energy prices reached a peak of

$2,816/MWh at the Hub over the five-minute period beginning at 17:10. High energy prices

were a result of high reserve prices, which were incorporated into the LMP. During the scarcity

event, there were deficits in both the thirty-minute operating reserve (TMOR) requirement and

the ten-minute non-spinning reserve (TMNSR) requirement. These deficits triggered reserve

constraint penalty factors. The TMOR penalty factor ($1,000/MWh) was in effect from 17:40 to

18:00. The TMNSR ($1,500/MWh) penalty factor was also activated for five minutes during this

interval, resulting in a combined reserve price of $2,500/MWh for that duration.

Market settlements-energy market: Day-ahead and real-time energy costs were significant on

December 24, totaling $81.7 million and $5.2 million, respectively. Uplift payments totaled $1.5

million, and were driven by fast-start pricing opportunity cost payments and payments to

resources dispatched above their economic dispatch point between 13:00 and 17:00.

Market settlements-pay for performance rules: Under pay for performance rules, participants

with a capacity supply obligation (CSO) were expected to provide an average of 67% of their

191

Figure 3-30 illustrates s upply side deviations, and therefore shows imports rather than net interchange (which accounts

for exports).

-$2,500

-$2,000

-$1,500

-$1,000

-$500

$500

$1,000

$1,500

$2,000

$2,500

-4,000 MW

-3,000 MW

-2,000 MW

-1,000 MW

0 MW

1,000 MW

2,000 MW

3,000 MW

4,000 MW

1 3 5 7 9 11 13 15 17 19 21 23 1 3 5 7 9 11 13 15 17 19 21 23

Dec 23 Dec 24

LMP ($/MWh)

RT-DA Generation Difference

Gas Pumped Storage Oil

Import Nuclear Hydro

Solar Coal Wood/Refuse

Wind Other Virtual Supply

Demand Response RT-DA Price Difference

2022 Annual Markets Report 118 ISO New England Inc.

ISO-NE PUBLIC

contracted capacity in form of energy or reserves during this event. Deviations from obligated

performance were settled at the performance payment rate of $3,500/MWh during each five-

minute interval of the capacity scarcity condition. Credits and charges totaled $35.9 million

each. Performance varied widely by fuel type. Import transactions and nuclear resources saw

the largest payments, while gas and dual-fuel generators incurred charges.

3.2.8 Reliability Commitments and Posturing

The ISO is required to operate New England’s wholesale power system to the reliability

standards developed by the North American Electric Reliability Corporation (NERC), the

Northeast Power Coordinating Council (NPCC), and in accordance with its own reliability

criteria.

192

To meet reliability standard these requirements, the ISO may commit additional

resources for several reasons, including to ensure that adequate capacity is available in

constrained areas, for voltage protection, and to support local distribution networks. The ISO

also may take manual actions to constrain (posture) resources from operating at the economic

dispatch point determined by the market software, in order to improve system reliability. This

typically occurs in order to maintain adequate reserves from fast-start pumped-storage

generators and to reserve limited fuel oil inventory.

Reliability Commitments

The real-time average hourly energy output (MW) from reliability commitments during the

peak load hours (hours ending 8-23) for 2018 through 2022 is shown in Figure 3-31 below. The

figure also shows whether the commitment decision was made in the day-ahead or real-time

market.

192

These requirements are codified in the NERC standards, NPCC criteria, and the ISO’s operating procedures. For more

information on the NERC standards, see http://www.nerc.com/pa/stand/Pages/default.aspx. For more information on the

NPCC s tandards, see https://www.npcc.org/program-areas/standards-and-criteria. For more information on the ISO’s

operating procedures, see http://www.iso-ne.com/rules_proceds/operating/isone/index.html.

2022 Annual Markets Report 119 ISO New England Inc.

ISO-NE PUBLIC

Figure 3-31: Average Hourly Energy Output from Reliability Commitments, Peak Load Hours

Reliability commitments remain a relatively small component of total system generation, at less

than 0.1%, or just 15 MW, on average. While the decrease in reliability commitments may be

partially explained by reduced transmission work, the primary explanation for the very large

decline is unclear. Reliability commitments were low over the past five years, averaging just 62

MW per hour. Commitments continue to be more common in the day-ahead market as a

percentage of total reliability commitments.

193

The vast majority (96%) of these commitments occurred for local reserve support – known as

Local Second Contingency Reliability Protection (LSCPR).

194

Other support services - Special-

Constraint Resources (SCR) and voltage - accounted for 3% and 1% of reliability commitment

resources, respectively.

195

Most reliability commitments (83%) occurred in Maine and SEMA-RI;

the remainder were in NEMA (5%) and New Hampshire (10%). These reliability commitments

primarily reflected a need for additional on-line generation in areas with transmission upgrades

and outages, to ensure local reliability.

A monthly breakdown of reliability commitments made during 2022 is shown in Figure 3-32

below. The figure shows the out-of-rate energy for reliability commitments during the peak

193

Committing generators in the day-ahead market is more desirable than commitments later in the Reserve Adequacy

Analysis (RAA) process or in real time as day-ahead commitments tend to reduce the risk of s uppressed real-time prices

and higher NCPC payments. If reliability commitments are known in the day-ahead market, the commitment schedules of

other generators can be adjusted to accommodate the reliability commitment. This provides more flexibility than if the

commitment is made later, reducing the risk of having excess inflexible supply online. Excessive generator commitments

can distort prices by removing other generators from the supply s tack and adding fixed energy to the supply stack. The

excess fixed supply could potentially suppress real-time prices and increase NCPC.

194

Local second contingency protection reliability (LSCPR) commitments are made for import-constrained sub-areas, if

necessary, to ensure that the ISO can re-dispatch the system to withstand a second contingency loss within 30 minutes

after the first contingency loss without exceeding transmission element operating limits.

195

An explanation for the reliability commitment types may be found here: https://www.iso-

ne.com/participate/support/glossary-acronyms/

2018 2019 2020 2021 2022

Average Hourly Energy (MW)

Day-Ahead Real-Time

2022 Annual Markets Report 120 ISO New England Inc.

ISO-NE PUBLIC

load hours in 2022, by market and month. Out-of-rate energy includes reliability commitment

output that is offered at a higher price than the LMP, and, therefore, would not likely have been

committed or dispatched in economics.

Figure 3-32: Day-Ahead and Real-Time Average Out-of-Rate Energy from Reliability Commitments, Peak

Load Hours, 2022

Of the roughly 15 MW of average hourly output from generators committed for reliability, only

4 MW (on average) was out-of-rate. Figure 3-32 shows that the greatest amount of out-of-rate

energy from reliability commitments occurred in February, March, and September; these

commitments supported planned transmission outages in Maine and SEMA-RI. Out-of-rate

commitments require uplift payments, to ensure that operating costs are recovered. NCPC

payments to generators providing LSCPR in 2022 totaled just $1.1 million; while this

represented 2% of total uplift payments for the year, it represented just 0.01% of total energy

payments.

Posturing Actions

The ISO may limit the output of potentially in-merit generators to ensure either system-wide or

local reliability, an action known as “posturing”. Posturing generators results in the

preservation of fuel for “limited energy” generators, allowing fuel to be used later in the event

of system contingencies. A generator may provide operating reserves while postured, but is

only available for manual dispatch above the posturing level in the event of a system

contingency. Postured generators are eligible to receive NCPC for any foregone profits that

occurred during the posturing period.

196

See Market Rule 1, Appendix F, Sections 2.3.8 and 2.3.9.

Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

2022

Average Hourly Energy (MW)

Day-Ahead Hourly Out-of-Rate MWs Real-Time Hourly Out-of-Rate MWs

2022 Annual Markets Report 121 ISO New England Inc.

ISO-NE PUBLIC

Postured energy (GWh) and NCPC payments by month are shown in Figure 3-33 below.

197

The

bars indicate the postured energy obtained (the amount of energy constrained down) from

pumped-storage generators and all other types of generators.

198

Figure 3-33 Monthly Postured Energy and NCPC Payments

Pumped-storage generators were frequently postured throughout the year, prior to 2022. In

2022, posturing was less frequent, resulting in just 5 GWh of posturing. In earlier years, the ISO

tended to posture pumped-storage generating units, during certain periods, to preserve

pondage and maintain operating reserves. This activity declined significantly in 2022.

Posturing actions in January 2018 stand out. Except for this month, only pumped- storage

generators have been postured. The posturing in January 2018 involved a number of oil-fired

generators, with limited fuel, during a prolonged cold snap period due to Operator concerns

about the day-to-day availability of natural gas for electric generation. The postured oil-fired

generators were effectively providing back-up electricity supply, in the event of a natural gas

shortage during the cold snap.

As indicated in the figure, NCPC payments to postured generators were quite low in 2022, with

approximately $0.2 million in total payments (accounting for 0.4% of all NCPC payments in

2022). NCPC payments were highest during January 2018, when the cold snap period resulted

in significant posturing of oil-fired generators. While the magnitude of NCPC payments is

generally consistent with the quantity of energy being postured, posturing during very high

energy price periods also can result in high NCPC payments, even when the postured energy

quantity is not extremely large. This occurs because the postured generators forgo the high

197

Postured energy is the amount of energy that is unavailable for economic dispatch, given the posturing action; this

value is used in the s ettlement compensation for the posturing action.

198

Very infrequently, pumped-storage demand (or asset-related demand) is postured. These resources are postured on-

line (in consumption mode) to increase operating reserves. The energy associated with these posturing activities is not

depicted in the figure.

J F M A M J J A S O N D J F M A M J J A S O N D J F M AM J J A S O N D J F M AM J J A S O N D J F M AM J J A S O N D

2018 2019 2020 2021 2022

NCPC Posturing Payments ($ millions)

Postured Generation (GWh)

Pumped-Storage Gen All Other Gen NCPC Payments

2022 Annual Markets Report 122 ISO New England Inc.

ISO-NE PUBLIC

LMPs and must be compensated for lost profits. This is noticeable in September 2018, during a

capacity deficiency period with operating reserve deficiencies and high energy prices.

3.2.9 Transmission Congestion

Levels of transmission congestion are relatively low in New England compared to other ISO

markets due to the significant amount of transmission investment over the past decade or so. In

contrast, transmission costs in New England are relatively high.

199

However, there are some

specific areas on the New England System that regularly experience congestion, which limits

power flowing from relatively cheaper supply.

This section examines transmission congestion in New England over the last five years.

Subsequently, this section explores the most congested areas within New England in 2022. It

concludes by listing the most frequently binding transmission constraints in New England in

2022.

Congestion Revenue

A feature of New England’s locational energy market is that load pays for energy at the price

where it is consumed and generation is paid for energy at the price where it is produced. This

means that, when there is a binding transmission constraint, load will often pay more for the

energy that it consumes than generation receives for the energy that it produces. Congestion

revenue represents the difference between what load pays for energy and what generation

receives for energy that happens as a result of transmission congestion.

200

Over the last five years, congestion revenue has been small relative to total energy market

payments and it has generally moved in line with the price of energy. This can be seen in Figure

3-34 below. The purple bars represent the day-ahead congestion revenue and the green bars

represent the real-time congestion revenue. The percentages in the figure are the total

congestion revenue each year (i.e., the day-ahead congestion revenue plus the real-time

congestion revenue) expressed as a percent of total energy market costs. This figure also

depicts the annual average day-ahead Hub LMP (blue line).

201

199

In its 2021 Assessment of the ISO New England Electricity Markets, Potomac Economics, the External Market Monitor

for ISO-NE, showed that ISO-NE had the lowest congestion rate in 2021 ($0.38/MWh) of ERCOT, MISO, PJM, and NYISO,

but had the highest transmission rate (nearly $22/MWh) of these other RTOs. See Figure 2 in its 2021 Assessment:

https://www.potomaceconomics.com/wp-content/uploads/2022/06/ISO-NE-2021-SOM-Report_Full-Report_Final-

Clean.pdf

200

See Section III.3.2.1 (i) of ISO-NE Market Rule 1 for an exact definition of day-ahead and real-time congestion revenue.

201

The day-ahead and real-time congestion revenue totals are measured on the left axis (“LA”), while the average annual

day-ahead Hub LMP is measured on the right axis (“RA”).

2022 Annual Markets Report 123 ISO New England Inc.

ISO-NE PUBLIC

Figure 3-34: Average Day-Ahead Hub LMP, Congestion Revenue Totals and as Percent of Total Energy Cost

Total day-ahead and real-time congestion revenue was $51 million in 2022, which represented

only 0.44% of total energy costs, and was much lower than the prior four years in relative

terms. The majority of the congestion revenue is in day-ahead market (as opposed to the real-

time market) where the vast majority of load and generation clear. The day-ahead congestion

revenue totals tend to be strongly correlated with the average day-ahead LMP, although this

relationship weakened in 2022. This was largely driven by a less constrained transmission

system in 2022.

Congested Areas in New England

The New England nodes most affected by transmission congestion in the day-ahead market in

2022 are shown in Figure 3-35 below.

202

The colors of the nodes indicate the average day-ahead

congestion component in 2022. Blue dots represent locations with negative average day-ahead

congestion components – the darker the blue, the lower the congestion component (i.e., more

negative). Locations that are “upstream” of a binding transmission constraint have a negative

congestion component.

203

Meanwhile, red dots represent locations that had a positive average

day-ahead congestion component.

204

Locations that are “downstream” of a binding constraint

have a positive congestion component.

202

This figure only includes nodes that had an average day-ahead congestion component of greater than or equal to

$0.15/MWh or less than or equal to -$0.15/MWh in 2022 in order to better highlight the constrained areas.

203

More specifically, a negative congestion component occurs when a location has a positive shift f actor to a binding

constraint. Conversely, a positive congestion component occurs when a location has a negative shift factor to a binding

constraint. In simple terms, a shift factor measures how an injection of energy at a location impacts the flow of energy

over a transmission constraint.

204

In 2022, there were no locations with average day-ahead congestion components greater than or equal to 0.15/MWh

and, therefore, there are no red dots in this figure.

1.07%

0.80%

0.97%

0.82%

0.44%

($10)

$10

$20

$30

$40

$50

$60

$70

$80

$90

($10)

$10

$20

$30

$40

$50

$60

$70

$80

$90

2018 2019 2020 2021 2022

Day-Ahead LMP ($/MWh)

Congestion Revenue ($ millions)

Day-Ahead Congestion Revenue [LA] Real-Time Congestion Revenue [LA]

% - Percent of Total Energy Cost Mean Day-Ahead LMP [RA]

2022 Annual Markets Report 124 ISO New England Inc.

ISO-NE PUBLIC

Figure 3-35: New England Pricing Nodes Most Affected by Congestion, 2022

Many of the congested areas in New England in 2022 were relatively small geographic areas

where transmission capacity limited the ability to export power to the rest of the system.

Several areas in Figure 3-35 have been highlighted and each of them is discussed in detail

below:

1) Northern Maine: This area has a relatively high concentration of intermittent

(predominantly wind) generators and is also where the New England system

interconnects to the New Brunswick control area (i.e., imports from New Brunswick

flow into this area).

2) Northern New Hampshire and Vermont: Similar to northern Maine, northern New

Hampshire and Vermont are areas with relatively high concentrations of wind

generation. Additionally, northern Vermont receives the power imported from the

Hydro-Québec control area over the Highgate tie line.

3) Southwestern Connecticut: This region also experienced negative congestion pricing

at times in 2022. Multiple high capacity and efficient natural gas-fired generators have

been added in this region in recent years resulting in increased in-merit supply (CPV

Towantic, Bridgeport Harbor). At times, the 115-kV system can limit the export of low-

cost power out of this region to the rest of the system, especially when nearby

transmission lines are taken out of service for repair or upgrade work.

4) New York: The lines connecting New York and New England frequently reach their

limit during periods when there are large spreads between power prices in New

England and New York (e.g., some winter months when New England’s gas

2022 Annual Markets Report 125 ISO New England Inc.

ISO-NE PUBLIC

infrastructure can become constrained resulting in higher New England prices) or when

there are reductions in the interface limit.

The Most Frequently Binding Interface Constraints

The 10 interface constraints that bound most frequently in the day-ahead market in 2022 are

listed in Table 3-6 below.

205

Also included in the table is the average marginal value ($/MWh) of

each constraint when it bound in the day-ahead market in 2022.

206

Lastly, this table includes a

location column, which places the constraints in the areas defined in Figure 3-35.

Table 3-6: Most Frequently Binding Interface Constraints in the Day-Ahead Market, 2022

Constraint Name

Constraint

Short Name

% of Hours

Binding

Average Marginal

Value of Constraint

($/MWh)

Location

Keene Road Export

KR-EXP

15.5%

-$33.12

Sheffield + Highgate Export

SHFHGE

14.3%

-$12.23

New York - New England

NYNE

5.9%

-$24.22

Whitefield South + GRPW

WTS+GR

4.9%

-$18.59

Orrington - South

ORR-SO

4.7%

-$14.88

Inner Rumford Export

IRMF-E

2.1%

-$32.16

New England West-East

NE_WE

1.3%

-$11.95

Sheffield Wind Generation

SHEF

1.0%

-$66.93

Tiverton Generation

TIVRTN

1.0%

-$14.29

Kingdom Wind Generation

KCW

0.9%

-$132.57

The most frequently binding interface constraint in 2022 was the Keene Road Export (KR-EXP)

interface, which bound in 15.5% of hours in the day-ahead market. The only other constraint

that bound in more than 10% of hours in the day-ahead market was the Sheffield + Highgate

Export (SHFHGE) interface, which bound in 14.3% of hours. Several of the interface constraints

listed bound in just one percent of hours or less, indicating that the day-ahead energy market in

New England was relatively unconstrained in 2022.

3.2.10 Marginal Resources

Offers and bids that are physical or financial in nature can set clearing prices in the day-ahead

market, including virtual transactions, price-sensitive demand bids, price-responsive demand,

205

Interface constraints are typically sets of transmission elements whose power flows are jointly monitored for voltage,

stability, or thermal reasons. They can often have a larger impact on congestion revenue when they bind than individual

transmission elements because they likely affect more load and generation.

206

The marginal value of a binding transmission constraint indicates the change in the system production costs if the limit

of the transmission constraint were increased by one MW for one hour. For example, a marginal value of -$10/MWh

indicates that system production costs could be reduced by $10 if the limit of the transmission constraint were increased

by one MW for one hour. The more negative the marginal value of the binding transmission constraint, the more the

system production costs could be reduced if the constraint were relaxed.

2022 Annual Markets Report 126 ISO New England Inc.

ISO-NE PUBLIC

asset-related demand, generator supply offers, and external transactions. Figure 3-36 illustrates

the percentage of load for which each transaction type was marginal over the past five years.

Figure 3-36: Day-Ahead Marginal Resource by Transaction Type

In the 2022 day-ahead market, natural gas (43%), virtual transactions (29%), and external

transactions (18%) continued to set price for a majority of load (90%), which is comparable

across the five years. Of note in 2022, was the increase in oil and virtual transactions setting

price, and the corresponding reduction in marginal gas generation. Oil generation was in merit

more frequently during the winter when natural gas prices were high, leading oil generators to

set price for 3% of load. The increase in marginal virtual transactions was due to the increase in

virtual supply transactions on days with high solar output. Since the majority of wholesale solar

clears in the real-time market only, it can be profitable for virtual supply to take solar’s place in

the day-ahead market. These offers often set price, as they are typically priced near the margin

to maximize their opportunity to arbitrage lower real-time prices.

In the real-time market, only physical assets can set price: generators, pumped-storage demand,

demand response resources, and external transactions. Real-time marginal resources are

typically generators (predominantly natural gas-fired generators) and pumped-storage

demand. The real-time marginal fuel mix over the past five years is shown in Figure 3-37 below.

23%

27%

24%

25%

29%

51%

48%

52%

43%

17%

19%

22%

17%

18%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

2018 2019 2020 2021 2022

% of Load

Virtuals Gas Externals

Pumped Storage Oil Coal

Wind Price-Sensitive Demand Hydro

Demand Response Other

2022 Annual Markets Report 127 ISO New England Inc.

ISO-NE PUBLIC

Figure 3-37: Real-Time Marginal Resource by Transaction Type

Other than generators offering on oil during the winter, the mix of resources that set price in

2022 was very similar to 2021. Natural gas continued to set price for the majority of load

(79%), while pumped-storage units (both generators and demand) remained the second largest

marginal resource (15%).

207

Because pumped-storage generators are often online during peak

load conditions and priced close to the margin, they can set price frequently. Oil generators set

price for 4% of load in the real-time market, the highest over the five-year period. The price-

setting intervals are concentrated in January 2022, when very high natural gas price during cold

periods resulted in more oil generation capacity being in economic merit and setting price.

The remaining transaction types were marginal for less than 3% of load in 2022. Although wind

generators are frequently marginal for a small geographic area, they are usually marginal for

only a small share of total system load (1% in 2022).

208

3.3 Net Commitment Period Compensation (Uplift)

This section provides an assessment of Net Commitment Period Compensation (NCPC)

payments. In the marginal cost-based energy market, prices will not always cover the full

production cost as offered by cleared supply resources, and therefore a make whole payment is

required in the form of NCPC.

207

Pumped-storage generation and demand have different operational and financial incentives. Pumped-storage

generators (supply) tend to operate and set price in on-peak hours when electricity prices are generally higher. Pumped-

storage demand have lower offers and typically consume energy and set price in off-peak hours, when it is generally

cheaper to pump water. In 2022, pumped-storage generation set price about 10% of the time and pumped-storage

demand set price about 5% of the time.

208

Wind generators are often located in export-constrained areas and can only deliver the next increment of load in a

small number of low-load locations. Wind generators often cannot set price outside of the constrained area they are

operating in because the transmission network that moves electricity out of their constrained area is at maximum

capacity.

73%

75%

82%

83%

79%

19%

20%

15%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

2018 2019 2020 2021 2022

% of Load

Gas Externals Pumped Storage

Oil Coal Wind

Hydro Demand Response Other

2022 Annual Markets Report 128 ISO New England Inc.

ISO-NE PUBLIC

In this section, we review NCPC in the context of total energy payments and provide a

breakdown of payments across market category, which can provide some insight into specific

drivers.

Key Takeaways

NCPC totaled $52.9 million in 2022, a $17.0 million increase (47%) from $35.9 million in 2021.

However, NCPC payments represented only 0.5% of energy payments in 2022, which was the

lowest level over the five-year reporting period. NCPC per MWh of load served totaled

$0.44/MWh compared to energy payment of $98/MWh.

The vast majority of NCPC payments in 2022 (95%) were paid to cover the operating costs of

resources committed and dispatched in economic merit order (“economic” NCPC category).

Payments to resources committed for local reserve support decreased from $6.8 million in 2021

to $1.1 million in 2022 (local second-contingency protection resource, or LSCPR, NCPC category).

This was due to significantly fewer local reliability commitments due to fewer impactful

transmission outages.

NCPC payments in the context of the energy market payments

The level of NCPC payments relative to energy payments can be a useful high-level gauge of the

energy market’s ability to cover the as-offered production costs of ISO- or market-scheduled

resources; high levels of NCPC could be symptomatic of price formation issues or gaps in

market products or design.

In 2022, NCPC payments remained relatively low compared to energy payments. Total NCPC

payments (dollars) tend to be positively correlated with energy market prices, as was the case

in 2022, but in relative terms (percentage of energy payments) NCPC declined. Energy and

NCPC payments are summarized in Table 3-7 below.

Table 3-7: Energy and Uplift Payments

2018

2019

2020

2021

2022

Energy Payments ($ millions)

$6,041

$4,105

$2,996

$6,099

$11,698

NCPC Payments ($ million)

$70.30

$30.60

$25.95

$35.94

$53.06

NCPC in $/MWh

$0.57

$0.25

$0.22

$0.30

$0.44

NCPC as % Energy Payments

Day-Ahead NCPC

0.4%

0.3%

0.1%

Real-Time NCPC

0.7%

0.4%

0.5%

0.3%

Total NCPC as % Energy Costs

1.2%

0.7%

0.9%

0.6%

0.5%

In 2022, energy payments totaled $11.7 billion, the highest in the five-year reporting period,

and nearly double the 2021 total. As covered in earlier sections, higher natural gas prices were

the key driver of energy payments, which also doubled year over year.

2022 Annual Markets Report 129 ISO New England Inc.

ISO-NE PUBLIC

NCPC payments totaled $52.9 million, about 50% higher ($17.0 million) than 2021 NCPC of

$35.9 million in 2021. This was due to an almost 90% increase ($23.2 million) in uplift

payments to resources cleared in economic merit order to meet system-wide or first-

contingency protection (“economic” NCPC payments). This is in line with higher production

costs associated with higher gas prices. As discussed further below, the increase in economic

NCPC was offset by lower payments for local reliability.

The table also shows the downward trend in NCPC payments as a percentage of energy costs; in

2022 this was the lowest level over the five-year period at just 0.5%. Most of this share

comprised NCPC in the real-time market, which at $53 million was over 70% of total NCPC.

NCPC Payments by Category

Uplift payments mostly cover the production costs of generators committed and dispatched in

economic merit order (economic or first contingency), as can be seen in Figure 3-38 below. The

inset table shows the percent share of total uplift for each category by year.

Figure 3-38: Total Uplift Payments by Year and Category

Economic payments made up the bulk (95% or $ 50.0 million) of NCPC, and increased by $23.2

million (87%). This was consistent with the increase in energy costs and explains the overall

annual increase in NCPC. All other categories of NCPC payments either dropped or remained

the same compared to 2021.

Local second-contingency protection payments decreased by $5.7 million (85%), from $6.8

million in 2021 to $1.1 million in 2022. This decrease was driven by the lowest level of

reliability commitments required to support local system contingency needs over the past few

years.

209

Distribution reliability payments and voltage payments were consistent with 2021 payments.

Distribution payments totaled $1.1 million in both 2021 and 2022. All of these payments were

209

See Section 3.2.8 for further details on reliability commitments and posturing actions.

$70.3

$30.6

$26.0

$35.9

$52.9

100

2018 2019 2020 2021 2022

$ Millions

Economic Second Contingency Distribution Voltage Generator Audits

Total annual NCPC

Category 2018 2019 2020 2021 2022

Economic 72% 64% 76% 75% 95%

Second Contingency 21% 24% 16% 19% 2%

Distribution 1% 3% 2% 3% 2%

Voltage 4% 2% 2% 0% 0%

Generator Audits 2% 7% 4% 2% 1%

2022 Annual Markets Report 130 ISO New England Inc.

ISO-NE PUBLIC

made in real time to two oil-fired generators on Cape Cod that were committed to support

distribution reliability in the SEMA load zone in January and July through September 2022.

In 2022, there were no voltage reliability commitments that necessitated uplift. Uplift paid to

generators conducting ISO-initiated audits totaled $0.6 million in 2022, down slightly (by $0.2

million) on 2021 costs in this category.

A closer look at “economic” NCPC

There are several subcategories within “economic” NCPC that cover a range of make-whole

payments to cover production costs as well as economic costs (or lost opportunity costs) not

recovered through energy prices. A breakdown of economic uplift by year and by sub-category

is shown in Figure 3-39 below.

Figure 3-39: Economic Uplift by Sub-Category

“Out-of-merit” payments primarily cover the commitment costs of economic scheduled

resources. These payments continued to make up the majority of total economic payment at

$31.1 million, increasing by $13.2 million (74%) from the prior year.

Dispatch lost opportunity cost (DLOC) and rapid response pricing opportunity cost (RRP OC)

both increased significantly in 2022. Even though each sub-category made up less than 20% of

economic uplift, they increased by 98% and 130%, respectively. This was due to more periods

of tight system conditions, when resources were compensated for following dispatch

instructions when faced with higher prices. Posturing payments were the only economic sub-

category that decreased in 2022. This was primarily driven by a significant reduction in ISO-

issued posturing instructions to pumped-storage generators (as discussed in Section 3.2.8).

$50.6

$19.4

$19.6

$26.5

$47.3

2018 2019 2020 2021 2022

$ in Millions

Out-of-Merit Posture External Dispatch LOC RRP OC Total

Total Annual Economic NCPC

2022 Annual Markets Report 131 ISO New England Inc.

ISO-NE PUBLIC

3.4 Demand Response Resources

This section summarizes outcomes and provides insights related to the Price-Responsive

Demand (PRD) program, which integrates demand response resources into the day-ahead and

real-time energy markets.

210

The first subsection (3.4.1) analyzes how demand response

resources participate in the day-ahead and real-time energy markets. The second subsection

(3.4.2) quantifies the level of energy and NCPC payments to demand response resources since

the implementation of PRD.

Key Takeaways

In 2022, participation in the PRD program followed trends observed since the initial

implementation in 2018. Most PRD resources primarily served as capacity and operating reserve

resources available for dispatch at very high offer prices: 81% of PRD capacity was offered at the

energy market offer cap of $1,000/MWh in 2022; on average, 96% of offers have been priced

above $200/MWh since the program’s implementation. Given offer prices, dispatch of these

resources averaged just 7.2 MW per hour in the day-ahead energy market and 9.3 MW per hour

in the real-time energy market in 2022.

PRD resources also provided operating reserves in 2022, averaging 0.3 MW per hour of ten-

minute reserves and 272 MW per hour of thirty-minute reserves. With low dispatch levels and

infrequent thirty-minute reserve pricing in 2022, energy revenues totaled just $4.3 million in the

day-ahead energy market, while energy and reserve revenues totaled $3.7 in the real-time energy

market and NCPC payments totaled $0.3 million for both markets.

3.4.1 Energy Market Offers and Dispatch under PRD

Under the Price-Responsive Demand (PRD) program, over 600 MWs of demand response

resources participate in the day-ahead and real-time energy markets. By virtue of their high

offer prices, most demand resources essentially function as capacity deficiency resources,

providing of energy and 30-minute operating reserves in the real-time energy market only

when prices are extremely high (~$1,000/MWh).

211

Figure 3-40 below indicates hourly

demand reduction offers in the real-time energy market, by offer price category for segment

energy offers since the implementation of PRD in 2018.

210

This was done in order to comply with FERC Order 745 (Demand-Response Compensation in Organized Wholesale

Energy Markets). Prior to June 1, 2018, demand response resources participated in the ISO’s energy markets (1) as

emergency resources activated during OP-4 system conditions (i.e., a capacity deficiency) in the real-time market and (2)

through the Transitional Price-Responsive Demand (TPRD) Program in the day-ahead market.

211

Because these resources primarily function as a source of operating reserves and are dispatched at slightly higher levels

(on average) in the real-time energy market, this section uses real-time offer and dispatch data to illustrate these

resources’ participation in the ISO’s energy markets.

2022 Annual Markets Report 132 ISO New England Inc.

ISO-NE PUBLIC

Figure 3-40: Demand Response Resource Offers in the Real-Time Energy Market

As indicated in the figure, most offers continue to be priced at the energy market offer cap of

$1,000/MWh; 81% of offered capacity, on average, in 2021 and 2022. In most hours, only the

lower tiers of offered capacity ($200/MWh or less) have a reasonable likelihood of being

dispatched in the real-time energy market; these offers did not exceed 10% of offered demand

reduction capacity in any hour of 2021 or 2022, and averaged just 3% of offered capacity.

212

Given the pattern of offer prices for PRD, the ISO dispatches relatively small quantities in the

energy markets. Figure 3-41 below illustrates the hourly dispatch of Demand Response

Resources (DRRs) in the real-time energy market, relative to resources’ offered reductions and

hourly energy prices since the implementation of PRD in 2018.

212

Energy prices in the real-time market exceeded $200/MWh in just 5.4% of pricing intervals in 2022 and 1.3% over the

review period.

2022 Annual Markets Report 133 ISO New England Inc.

ISO-NE PUBLIC

Figure 3-41: Demand Response Resource Dispatch in the Real-Time Energy Market

213

The maximum hourly quantity of demand response capacity dispatched in the real-time energy

market was 117.8 MW in 2022 (during the December 24 shortage event) and 75.7 MW in 2021.

While demand resources were dispatched frequently in the real-time market – in 50% of hours

in 2022 and 52% of hours in 2021 – the dispatch level was very small, averaging just 9.3 MW

per hour in 2022 and 6.4 MW per hour in 2021.

As noted earlier, DRRs also provide a source of operating reserves in the real-time energy

market. DRRs are considered fast-start capable, if those capabilities have previously been

demonstrated

214

In 2022, DRRs provided only 0.3 MW per hour, on average, of ten-minute

operating reserves

215

, but provided substantially more in thirty-minute operating reserves

(TMOR), averaging 272 MW per hour. In 2021, ten-minute reserve designations were not

substantially different, equaling 0.4 MW on average; thirty-minute operating reserves (TMOR)

for 2021 averaged 299 MW per hour – 9% more than in 2022.

3.4.2 Energy Market Compensation under PRD

Demand Response Resources (DRRs) have received relatively modest energy market

compensation during the review period. This results from low dispatch rates in the energy

market and infrequent TMOR pricing in the real-time energy market. When dispatched, DRRs

are eligible to receive uplift payments. NCPC provides additional compensation to resources

when energy market revenues are insufficient to cover as-offered operating costs in the day-

213

The right vertical axis (LMPs) has been truncated to improve the figure’s legibility. During tight system conditions, real -

time LMPs have exceeded $500/MWh on s everal occasions. The truncation obscures the magnitude of those prices, which

have reached as high as $2,677/MWh.

214

To be designated during the operating day as providing thirty-minute fast-start reserves, a DRR must offer certain

operating constraints consistent with fast-start operation. These operating constraints are: total s tart-up time (including

notification time) of less than or equal to 30 minutes, minimum time between reductions and a minimum reduction time

of less than or equal to 1 hour, and a “claim 30” (30-minute reserve capability) greater than 0.

215

While DRRs can provide ten-minute reserves, that s ervice requires interval metering with granularity of one minute or

less, to be able to provide either non-synchronized (TMNSR) or s ynchronized reserves (TMSR).

2022 Annual Markets Report 134 ISO New England Inc.

ISO-NE PUBLIC

ahead and real-time energy markets. Figure 3-42 indicates energy and NCPC payments by

month since the implementation of PRD in 2018.

Figure 3-42: Energy Market Payments to Demand Response Resources

As indicated in the figure, both NCPC payments and energy market payments have been

relatively small, since the implementation of PRD in June 2018.

216

In 2022, total energy

payments were $8.3 million; energy payments to DRRs in 2021 were $2.7 million.

217

The

increase in total energy market payments to DRRs in 2022 (compared to 2021) is consistent

with the small increase in average dispatch and the more significant increases in day-ahead and

real-time LMPs that occurred in 2022.

218

The high winter and summer month payments in 2022

reflect months with elevated LMPs, resulting from high fuel prices (winter months) and

seasonally high loads (summer). The December 2022 real-time energy payments also reflect

payments to DRRs during the system event on December 24, 2022. Energy market LMPs on that

day reached $1,000 to $2,000 per MWh on an hourly basis.

216

Energy market payments include payments for MWh provided to satisfy the energy market’s energy and reserve needs

(labelled “DA Energy” and “RT Energy and Reserves” in the figure) and uplift payments when energy and reserve revenues

are insufficient to cover all of the costs of providing energy and reserves (labelled “DA NCPC” and “RT NCPC” in the graph).

217

This is in contrast to capacity market compensation in 2022, which totaled about $27 million. The capacity market is

discussed in more detail in Section 6.

218

Energy market payments to DRRs represent a very small component of overall energy market payments, which were

$11.7 billion for all resources in 2022.

-0.5

0.0

0.5

1.0

1.5

2.0

2.5

Jun

Jul

Aug

Sep

Oct

Nov

Dec

Jan

Feb

Mar

Apr

May

Jun

Jul

Aug

Sep

Oct

Nov

Dec

Jan

Feb

Mar

Apr

May

Jun

Jul

Aug

Sep

Oct

Nov

Dec

Jan

Feb

Mar

Apr

May

Jun

Jul

Aug

Sep

Oct

Nov

Dec

Jan

Feb

Mar

Apr

May

Jun

Jul

Aug

Sep

Oct

Nov

Dec

2018 2019 2020 2021 2022

Energy Market Payments ($ millions)

DA NCPC RT NCPC DA Energy RT Energy and Reserves

2022 Annual Markets Report 135 ISO New England Inc.

ISO-NE PUBLIC

Section 4

Virtual Transactions and Financial Transmission Rights

This section discusses trends in two important financial instruments that are available to

participants in New England’s wholesale electricity markets: virtual transactions and financial

transmission rights (FTRs). While both instruments can be used by participants to manage

certain risks associated with the actual production or consumption of energy, they can also be

used for completely speculative purposes. Virtual transactions are covered in Section 4.1 and

FTRs are discussed in Section 4.2.

Although the average volume of virtual transactions has remained fairly constant over the past

five years, the use of virtual transactions by time of day is evolving to match changing grid

conditions. Most notably, this is occurring with virtual supply, which is clearing in larger

volumes during the daytime hours. On the other hand, FTR use by participants is generally

declining. This may be due to lower perceived basis risk by market participants, as the system

has experienced relatively low levels of congestion in recent years, partly as a result of

significant amount of transmission investment.

The use of these financial instruments will be motivated by expected profits, particularly for

participants taking a speculative position (i.e., one without a corresponding physical position).

Market participants are likely to deploy their capital toward other uses if they believe that they

cannot earn a reasonable risk-adjusted return. In 2022, virtual supply transactions realized

their highest net profit per MWh of any of the past five years, while virtual demand was

unprofitable as it has been in four of the last five years.

219

Meanwhile, FTR holders realized a

modest profit in 2022, a year after making over $25 million in net profits.

Virtual transactions also perform an important market function by helping converge day-ahead

and real-time market prices. However, virtual transactions are not costless – they are subject to

highly variable uplift charges – and this cost can limit the ability of virtual transactions to

perform this important market function. This report finds that the NCPC charge rate increased

in 2022 to its highest level since 2018, in large part due to the overall increase in energy costs

and prices in 2022.

4.1 Virtual Transactions

This section presents our assessment of virtual transaction activity and performance.

Specifically, the first subsection (4.1.1) covers trading volumes, the second subsection (4.1.2)

covers profitability, including the impact of uplift charges, and the final subsection (4.1.3) looks

at the relationship between virtual transactions and price convergence.

219

Virtual demand is often used as a hedging tool for generators that are at risk of tripping in real time, while virtual supply

tends to be more of a speculative instrument.

2022 Annual Markets Report 136 ISO New England Inc.

ISO-NE PUBLIC

Key Takeaways

In general, the volume of cleared virtual transactions increased slightly in 2022. Average hourly

cleared virtual supply increased by 4% (from 594 MW to 649 MW per hour) and average cleared

virtual demand increased by 12% (from 372 MW to 418 MW per hour) year-over-year. The

average volumes of cleared virtual supply in 2022 were notably higher during the daytime hours

than other times of the day, reaching as high as 1,069 MW per hour. NCPC charges for virtual

transactions increased in 2022. From 2018 through 2021, this rate averaged about $0.56/MWh,

but increased to $0.88/MWh in 2022. This was the highest NCPC charge rate since 2018

($0.89/MWh). Despite this increased NCPC charge rate, virtual supply transactions made an

average of $1.99/MWh, the highest net profit for either transaction type in the past five years.

However, virtual demand lost an average of $1.51/MWh, the third highest losses observed over

the reporting period for any bid type. Some virtual participant profit came from external NCPC

credits that result from relieving congestion at the external interfaces.

4.1.1 Virtual Transaction Volume

In 2022, higher levels of virtual supply tended to be submitted and cleared during the middle

part of the day, while higher levels of demand tended to be submitted and cleared during the

morning and evening ramp periods. This can be seen in Figure 4-1, which shows the average

hourly volume of submitted and cleared virtual transactions by time of day in 2022. Virtual

supply is depicted as positive values, while virtual demand is depicted as negative values.

Figure 4-1: Average Hourly Submitted and Cleared Virtual Transaction Volumes by Time of Day, 2022

The average volumes of cleared virtual supply in 2022 were notably higher during the daytime

hours than other times of the day. Between hours ending 9 through 17, virtual supply averaged

about 900 MW per hour compared to 500 MW during the rest of the day. We have observed a

clear relationship between virtual supply and photovoltaic generation, particularly on high

solar output days. Most photovoltaic generation registers as settlement-only generation

(2,000)

(1,000)

1,000

2,000

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

Aerage Hourly Volume (MW)

Hour Ending

Cleared Virtual Supply Submitted Virtual Supply

Cleared Virtual Demand Submitted Virtual Demand

2022 Annual Markets Report 137 ISO New England Inc.

ISO-NE PUBLIC

(SOG).

220

Since SOGs cannot clear in the day-ahead market, virtual participants anticipate the

additional real-time generation and clear virtual supply in the day-ahead market in their

place.

221

These virtual supply offers effectively replace the price-taking SOGs that show up in the

real time.

Meanwhile, the average volume of cleared virtual demand tends to be slightly higher during the

morning and evening ramping periods, when loads are higher and prices tend to be higher in

the real-time.

4.1.2 Virtual Transaction Profitability

Virtual transactions profit from differences between day-ahead and real-time prices. However,

transaction costs in the form of NCPC charges can turn otherwise profitable virtual transactions

into unprofitable transactions on a net basis.

222

This limits the ability of virtual transactions to

converge prices between day-ahead and real-time prices, which is one of their intended market

functions.

223

This price-converging function is becoming increasingly important as low-marginal cost

intermittent generation enters the market and tends to produce more energy in the real-time

market rather than it clears in the day-ahead market. The IMM continues to recommend that

the ISO review the allocation of NCPC charges to virtual transactions to ensure the charges are

consistent with principles of cost causation.

224

Figure 4-2 illustrates the profitability of virtual transactions along with the impact of NCPC

charges on profitability. This figure displays the average annual gross and net profit of virtual

transactions since 2018 (left axis). The bars are categorized by year and bid type with virtual

demand shown in red and virtual supply shown in blue. The top of each bar represents gross

profit, the bottom represents net profit, and the length of the bar represents the per-MWh NCPC

charge. Additionally, the dashed black line shows the percentage of hours each year in which

virtual transactions were profitable on a gross basis (right axis).

225

220

By the end of 2022, s ettlement-only photovoltaic generation had an installed capacity of about 1,950 MWs.

221

The differences in the supply mix between the day-ahead and real-time energy markets are looked at in Section 3.2.2.

222

The ISO allocates the following NCPC charges to cleared virtual transactions: (1) Real-time Economic NCPC: all cleared

virtual transactions (supply and demand) incur a charge to contribute towards the payment of real-time economic NCPC

because they are considered real-time deviations; and (2) Day-ahead Economic NCPC: virtual demand bids are also

charged day-ahead economic NCPC based on their share of day-ahead load obligation. This charge is typically much

smaller because the total day-ahead economic NCPC is divided among a much larger quantity of energy.

Virtual transactions can also incur NCPC charges associated with congestion at the non-CTS (coordinated transaction

scheduling) external interfaces. Because these NCPC charges do not have a broad market impact or apply to virtual

transactions at most locations, they are not considered in much detail in this report.

223

The role of virtual transactions in price convergence is discussed in more detail in Section 4.1.3.

224

For more information on recommended market design changes, see Section 8

225

The line is flat for observations in the same year because the value is computed as the number of hours that all virtual

transactions together were profitable on a gross basis, as a percentage of total hours in the year.

2022 Annual Markets Report 138 ISO New England Inc.

ISO-NE PUBLIC

Figure 4-2: Average Annual Gross and Net Profits for Virtual Transactions

In 2022, virtual supply made an annual average gross profit ($2.84/MWh) while virtual demand

a gross loss (-$0.60/MWh).

Virtual supply profitability was the highest over the last five years (on a gross basis),

$0.77/MWh higher than 2021 ($2.07/MWh). Virtual supply tended to be profitable during the

middle of the operating day, when day-ahead prices were typically higher than real-time prices.

Between hours ending 9 and 17, participants made a total gross profit of $15.6 million in 2022,

which accounted for 97% of the virtual supply’s gross profit ($16.1 million) during the year.

The relationship between virtual supply and photovoltaic generation helped drive these higher

profits. Additionally, load-serving entities (LSEs) frequently cleared excess demand in the day-

ahead market on days with higher behind-the-meter photovoltaic generation during Spring

2022. This contributed to higher day-ahead prices and higher profits for virtual supply.

226

227

Virtual demand, conversely, incurred a gross loss of -$0.60/MWh on average in 2022. However,

this was a smaller loss than 2021 (-$1.29/MWh). In total, virtual demand, on a gross basis, lost

$2.2 million during 2022. Lower real-time prices led to participants losing a total of $5.2 million

during hours ending 9 to 17 on virtual demand, which drove overall losses. Virtual demand

made a profit of $3.0 million during the rest of the operating day.

Average NCPC charges for virtual transactions increased compared to 2021 (from $0.52/MWh

to $0.88/MWh). NCPC charges increased as total NCPC charges increased for the system.

228

2022, virtual supply stayed profitable after the netting of NCPC charges, making a net profit of

$1.99/MWh, on average. Virtual demand lost $1.51/MWh, on average, after accounting for

NCPC charges.

226

See the Internal Market Monitor’s Section 2.2 of the Spring 2022 Quarterly Markets Report for more information on

over-clearing in the day-ahead market.

227

The price premium in the day-ahead energy market is discussed in more detail in Section 3.1.3.

228

For more information on why NCPC increased in 2022, see Section 3.3.

40%

42%

44%

46%

48%

50%

52%

54%

56%

58%

60%

($5)

($4)

($3)

($2)

($1)

Demand

Supply

Demand

Supply

Demand

Supply

Demand

Supply

Demand

Supply

2018 2019 2020 2021 2022

Percentage of Hours with Positive Gross Profit (Line)

Virtual Profit in $/MWh (Bars)

Virtual Demand Profit (LA) Virtual Supply Profit (LA)

Hours Virtuals Made Gross Profit (RA)

Gross Profit

Net Profit

NCPC Charge

2022 Annual Markets Report 139 ISO New England Inc.

ISO-NE PUBLIC

Most Profitable Locations for Virtual Supply

Details of the top 10 most profitable locations for virtual supply in 2022, after accounting for

transaction costs and NCPC charges/credits (ranked by total net profit), are shown in Table 4-1

below.

Table 4-1: Top 10 Most Profitable Locations for Virtual Supply

Location

Type

Submitted

MWh

Cleared

MWh

Gross

Profit ($k)

Net Profit

($k)

Gross

Profit per

MWh

Net Profit

per MWh

# of

Participants

.I.SHOREHAM138 99

Ext Node

452,621

19,273

$(1)

$1,953

$(0.06)

$101.31

UN.BULL_HL 34.5BLHW

Gen Node

322,409

193,569

$1,890

$1,706

$9.76

$8.81

.H.INTERNAL_HUB

Hub

1,865,879

1,174,050

$2,412

$1,454

$2.05

$1.24

.Z.MAINE

Load Zone

602,049

452,628

$1,352

$956

$2.99

$2.11

.Z.SEMASS

Load Zone

856,181

657,390

$1,389

$836

$2.11

$1.27

.Z.NEMASSBOST

Load Zone

499,703

453,640

$1,194

$834

$2.63

$1.84

.Z.NEWHAMPSHIRE

Load Zone

396,047

311,580

$1,038

$767

$3.33

$2.46

UN.BULL_HL 34.5HANW

Gen Node

152,635

95,640

$818

$721

$8.55

$7.54

.Z.VERMONT

Load Zone

306,198

252,857

$813

$592

$3.21

$2.34

.Z.RHODEISLAND

Load Zone

258,489

218,771

$653

$463

$2.98

$2.12

The most profitable location for virtual supply was an external node, .I.SHOREHAM138 99,

where two participants received external credits for relieving congestion in the export

direction. This node is the external node for the Cross Sound Cable interface, one of the two

interfaces connecting New England to Long Island. Participants made nearly $2 million in

external credits at this interface, which led the largest net profit of any location, despite losing

money on a gross basis.

Seven of the top ten locations consisted of the Hub and six of the eight load zones.

229

High total

net profits at these locations were in line with the lower real-time prices at these locations,

especially during the middle of the operating day.

The other two locations are associated with wind power generation. Certain wind generators

are part of the set of resources known as do-not-exceed (DNE) dispatchable generators, or

DDGs. Wind generators often clear lower volumes in the day-ahead market, but produce more

real-time output at low or even negative real-time prices. Virtual supply participants fill this gap

by clearing virtual supply at prices more in line with real-time expectations, particularly on

windy days.

230

229

The other two load zones (.Z.WCMASS and .Z.CONNECTICUT) ranked 14

and 15

in total net profit, respectively.

230

These locations tend to be riskier as well, given the difficulty of forecasting wind generation. For example, if a

participant expects high wind output in the real-time market, they might clear virtual supply in the day-ahead market at a

low price, and expect to profit off negative real-time prices. However, if the wind generation does not meet day-ahead

expectations, these locations will likely be unconstrained, and the participant wo uld have to pay its day-ahead obligation

back at a higher real-time price.

2022 Annual Markets Report 140 ISO New England Inc.

ISO-NE PUBLIC

Most Profitable Locations for Virtual Demand

Details for the 10 most profitable locations for virtual demand in 2022, after accounting for

transaction charges and all relevant NCPC charges/credits (ranked by total net profit), are

shown in Table 4-2 below.

231

Table 4-2: Top 10 Most Profitable Locations for Virtual Demand

Location

Type

Submitted

MWh

Cleared

MWh

Gross

Profit ($k)

Net Profit

($k)

Gross

Profit per

MWh

Net Profit

per MWh

# of

Participants

.I.HQHIGATE120 2

Ext Node

275,246

46,789

$462

$484

$9.88

$10.33

.I.HQ_P1_P2345 5

Ext Node

332,710

81,360

$130

$364

$1.60

$4.47

UN.MONTVLLE69 MO10

Gen Node

7,448

7,438

$64

$58

$8.65

$7.76

AR.MADISON 34.5MADBD

ARD Node

1,298

1,236

$52

$49

$41.72

$39.64

LD.KINGSTON23

Load Node

2,397

1,238

$44

$42

$35.37

$33.86

LD.RUMFD_IP34.5

Load Node

2,540

2,485

$40

$37

$16.28

$14.79

UN.BINGHAM 34.5BNGW

Gen Node

8,553

3,157

$38

$34

$12.05

$10.76

LD.SALISBRY69

Load Node

10,605

8,204

$33

$28

$4.01

$3.36

UN.PARIS 34.5GRPW

Gen Node

48,145

9,566

$27

$20

$2.82

$2.12

LD.FREIGHT 13.8

Load Node

3,248

2,753

$23

$20

$8.38

$7.36

The most profitable locations consisted mostly of nodes with relatively lower total profits and

low trading activity during 2022 (compared to virtual supply). Like virtual supply, the most

profitable locations were at external nodes (.I.HQHIGATE120 2 and .I.HQ_P1_P2_2345 5),

associated with the Highgate and Phase II interfaces. Typically, transaction costs associated

with virtual transactions reduce net profits. However, participants made a larger net profit at

these external nodes as they received external NCPC credits for relieving congestion with

uneconomic bids at the external interface in the day-ahead market.

232

Virtual demand at the Highgate interface was particularly profitable on December 24, 2022

when generator trips and import reductions led to capacity shortage conditions and very high

231

For more information about the additional charges for virtual transactions, see Schedule 2 of the ISO Funding

Mechanism.

232

Virtual participants receive external NCPC credits for relieving congestion at the external interfaces. For example, if two

participants are willing to import 150 MWs (300 MWs total) over the Highgate interface, they may each submit fixed

import transactions, which will clear at any LMP. However, the interface has a total transfer capability (TTC) of 225 MWs,

so the interface cannot support all 300 MWs. Without virtual transactions, the two fixed import transactions would be pro-

rated to prevent the bids violating the TTC. However, if a participant cleared a low priced 75 MW virtual demand bid, this

would provide counter-flow across the interface and lower the net imports from 300 MW to 225 MW, the maximum

allowed over the interface. In this scenario, if the cleared virtual demand bid is priced below the system LMP, their

transaction is uneconomic, and require a make-whole payments. The two participants importing power in the day-ahead

market would pay the uplift to the participant relieving congestion with the virtual demand bid.

2022 Annual Markets Report 141 ISO New England Inc.

ISO-NE PUBLIC

real-time prices.

233

. Participants cleared an average of 174 MW of virtual demand, with a net

profit over $1 million on this day, and were unprofitable the rest of the year.

The other eight most profitable locations averaged less than 1.1 MWs per hour of cleared virtual

demand.

4.1.3 Virtual Transactions and Price Convergence

One of the primary benefits that virtual transactions can provide is to improve market

efficiency, by helping achieve the necessary real-time generation commitments at the lowest

cost. Market participants can, by pursuing profitable opportunities, use virtual transactions to

converge day-ahead commitments closer to real-time commitments. Improved price

convergence is reflective of improved commitment convergence.

234

Consequently, one might

expect to see higher levels of price convergence associated with higher levels of cleared virtual

transactions. The relationship between the volume of virtual transactions (measured on the left

y-axis [LA]) and the level of price convergence (measured on the right y-axis [RA]) is shown in

Figure 4-3 below. This figure presents two measures of price convergence:

235

1) The mean absolute difference (in $/MWh) between the day-ahead and real-time Hub

prices (blue line series).

2) The median absolute difference between day-ahead and real-time Hub prices as a

percentage of the day-ahead Hub LMP (gray line series).

Figure 4-3: Virtual Transaction Volumes and Price Convergence

233

See Section 3.2.7 for more information about the events on December 24, 2022.

234

Section 3.1.3 discusses price convergence in more depth.

235

For both of these metrics, the price difference is the absolute value of the day-ahead and real-time price difference. The

absolute value is used because we are interested in virtual transactions’ potential impact on price convergence, including

both positive and negative price differences. For the second metric, the price difference is divided by the day-ahead LMP

to help normalize for s ystematic differences between prices in different years. The median is used to reduce the i nfluence

of outliers.

$19.12 /MWh

14.4%

(20)

(15)

(10)

(5)

(4,000)

(3,000)

(2,000)

(1,000)

1,000

2,000

3,000

4,000

2018 2019 2020 2021 2022

$/MWh and %

Aerage Hourly Volume (MW)

Cleared Virtual Supply [LA] Offered Virtual Supply [LA]

Cleared Virtual Demand [LA] Bid Virtual Demand [LA]

Avg. Absolute LMP Diff. ($/MWh) [RA] Median LMP Diff. (% of DA LMP) [RA]

2022 Annual Markets Report 142 ISO New England Inc.

ISO-NE PUBLIC

The measures of price convergence provide a mixed picture about the level of convergence

between day-ahead and real-time prices in 2022 compared to prior years. The average absolute

price difference between the day-ahead and real-time Hub prices (blue line) was $19.12/MWh,

the highest difference over the reporting period. However, this increase aligns with higher

LMPs during 2022. When assessed as a percent of the day-ahead price (gray line), price was

comparable to 2021 (14.4% vs. 13.9%) and has generally improved compared to levels prior to

2021.

The figure also shows that, in general, the quantity of submitted virtual transactions increased

in 2022, after decreasing over the prior four years. In 2022, participants submitted an average

of 1,969 MWs of virtual transactions per hour, up 30% on 2021, with more activity across the

Hub and eight load zones. Submitted virtual transactions remained almost 30% lower than

submitted volumes in 2018.

236

Cleared transactions averaged 1,066 MW per hour in 2022, up

10% on 2021. Both cleared virtual supply and cleared virtual demand increased; virtual supply

by 9% and virtual demand by 12%.

4.2 Financial Transmission Rights

This section presents our assessment of financial transmission rights (FTRs) activity and

performance. The first subsection (4.2.1) covers FTR auction volumes, the second subsection

(4.2.2) covers FTR funding, and the final subsection (4.2.3) covers profitability at a market and

locational level. Given their outsized impact on FTR market outcomes, special attention is given

to FTR paths that source from .I.ROSETON 345 1 (“Roseton”), which is ISO-NE’s external node

for trading across the New York - New England (“NYNE”) interface.

237

Key Takeaways

FTRs volumes decreased by 8%, from 32,443 MW per hour in 2021 to 29,847 MW per hour in

2022, consistent with the general trend over the past few years. FTRs were fully funded in 2022,

meaning there was sufficient revenue collected through the energy markets’ congestion revenue

fund to pay FTR holders. After a very profitable year in 2021 ($25.9 million), FTR holders made a

collective profit of just $0.6 million in 2022. FTR activity associated with the NYNE interface was

one reason for this decreased profitability. Profit for FTRs sourcing from this interface decreased

by $16.7 million between 2021 ($9.8 million) and 2022 (-$6.8 million), partly from lower levels

of congestion between the two regions as a result of changing system dynamics.

4.2.1 FTR Volume

The volume of FTRs that participants hold depends on a number of factors, including

participants’ expectations of congestion in the day-ahead market. If participants expect more

congestion in the day-ahead market than in prior years, they may purchase more MWs of FTRs

to hedge against this congestion. Conversely, if participants expect less congestion in the day-

236

One participant contributed significantly to the decrease in submitted virtual transactions between 2018 and 2022. In

2018, this participant submitted an average of 941 MWs per hour, but submitted only 11 MW per hour in 2022.

237

The New York - New England (NYNE) interface is s ometimes referred to as the New York North interface, the New York

Northern AC interface, or the Roseton interface.

2022 Annual Markets Report 143 ISO New England Inc.

ISO-NE PUBLIC

ahead market, they may purchase fewer MWs of FTRs. Another important factor is the set of

transmission limits that the ISO uses in the auctions it conducts to award FTRs. The ISO

performs a market feasibility test in each FTR auction that ensures that the awarded set of FTRs

respects the transmission system’s physical and operational limits.

238

Essentially, these limits

restrict the MW volume of FTRs that can be purchased in FTR auctions.

Participants held fewer FTRs (by MWs) per hour, on average, in 2022 than in any other year

over the past ten years. The trend in declining FTR volumes can be seen in Figure 4-4, which

shows the average MW volume of FTRs that were in effect each hour by year (black diamonds)

for the past five years.

239

This figure also shows the average hourly MW volume of FTRs

purchased and sold by auction type (i.e., annual, prompt-month, or out-month) during each

year.

240

FTR purchases are depicted as positive values, while FTR sales are depicted as negative

values.

Figure 4-4: Average FTR MWs in Effect per Hour by Year

Market participants held an average of 29,847 MWs of FTRs per hour in 2022, representing an

8% decrease from the average amount of FTRs in effect in 2021 (32,443 MWs per hour). This

decrease was largely the result of reduced purchases in the annual auctions; these purchases in

2022 (15,241 MWs per hour) were down by 13% compared to 2021 (17,519 MWs per hour).

238

This test is performed in order to increase the likelihood of revenue adequacy, which means that there is sufficient

congestion revenue collected in the energy market and from FTR holders with negative target allocations to fully

compensate all FTR holders with positive target allocations. This is further discussed in Section 4.2.2.

239

The averages are hourly-weighted MW volumes. This weighting accounts for the fact that there are more off-peak

hours than on-peak hours in a year. The volume of FTRs in effect each year represents the hourly-weighted average MW

volume of FTRs purchased less the hourly-weighted average MW volume of FTRs sold.

240

An annual auction refers to an auction where participants purchase (or sell) FTRs whose term is one calendar year,

while both prompt-month and out-month auctions refer to auctions where participants purchase (or sell) FTRs whose term

is one month. Prompt-month refers to the monthly auctions for FTRs that are in effect for the month immediately after

when the auction takes place, while out-month refers the monthly auctions for FTRs that are in effect for any other month

remaining in the calendar year (excluding the prompt month). Out-month auctions did not exist until the Balance of

Planning Period (“BoPP) project was implemented in September 2019.

(5,000)

5,000

10,000

15,000

20,000

25,000

30,000

35,000

40,000

2018 2019 2020 2021 2022

Average MW per Hour

Annual Purchased MW Annual Sold MW

Prompt-month Purchased MW Prompt-month Sold MW

Out-month Purchased MW Out-month Sold MW

Average MW in Effect

2022 Annual Markets Report 144 ISO New England Inc.

ISO-NE PUBLIC

Average prompt-month FTR purchases rose by 1% in 2022 (11,773 MWs per hour) compared

to 2021 (11,668 MWs per hour), while average out-month FTR purchases decreased by 12% in

2022 (2,987 MWs per hour) compared to 2021 (3,412 MWs per hour). In general, FTR holders

sell very few FTRs each year (just 153 MWs in 2022), as can be seen below the horizontal axis

in Figure 4-4.

4.2.2 FTR Funding

FTR funding refers to the ability to pay FTR holders the full value of their positive target

allocations. Positive target allocations arise when the congestion component at the sink location

(point of delivery) of an FTR path is larger the congestion component at the source location

(point of injection). When there is sufficient revenue to pay all the positive target allocations,

FTRs are said to be fully funded. Fully funding FTRs is an important aspect of a well-functioning

FTR market, because it gives market participants confidence that they will receive the full value

of their FTRs.

FTRs were fully funded in 2022 and have been in each of the last five years, as can be seen in

Figure 4-5 below. The graph captures the inflows (positive values) into the Congestion Revenue

Fund (“CRF”) – mainly day-ahead congestion revenue – and the outflows from the fund – mainly

positive target allocations.

241

The balance in the CRF at the end of each year is shown by the

blue line.

242

Figure 4-5: FTR Funding and Congestion Revenue Fund Components by Year

241

The CRF is used to pay FTR holders with positive target allocations. This fund collects money from three s ources: (1)

day-ahead congestion revenue, (2) real-time congestion revenue, and (3) the holders of FTRs with negative target

allocations. See Section III.5 of ISO-NE Market Rule 1 for more information about transmission congestion revenue and FTR

funding.

242

The CRF balance is defined here as the



day-ahead congestion revenue + real-time congestion revenue + abs(negative

target allocations) – positive target allocations].

100%

($80)

($60)

($40)

($20)

$20

$40

$60

$80

$100

2018 2019 2020 2021 2022

Congestion Revenue Fund Components

($ millions)

Day-Ahead Congestion Revenue Negative Allocations

Positive Allocations Real-time Congestion Revenue

% - Percent of Positive Allocations Paid Congestion Revenue Fund Year-End

2022 Annual Markets Report 145 ISO New England Inc.

ISO-NE PUBLIC

Positive target allocations in 2022 ($45.2 million) declined by $10.0 million from their 2021

value ($55.2 million). This was consistent with the combination of lower FTR volumes and

lower levels of transmission congestion. Day-ahead congestion revenue also decreased in 2022

($48.9 million) from its 2021 value ($51.1 million), although the decline was more modest.

Real-time congestion revenue added an additional $2.1 million to the CRF in 2022, marking the

first time in the reporting period that this value was positive. Negative target allocations in

2022 ($4.9 million) dropped from their 2021 value ($12.1 million). The CRF year-end balance at

the end of 2022 was $10.7 million; this surplus was distributed proportionately to entities that

paid congestion costs during the year.

243

4.2.3 FTR Profitability

Profit in the FTR market is measured as the sum of the positive target allocations and the

revenue from FTR sales, minus the negative target allocations and the cost of FTR purchases. In

a competitive FTR market, one would not expect to see excessive (risk-adjusted) profits or

losses sustained over numerous years. Prolonged periods of high profitability would likely spur

the entry of new participants (or at least an increase in FTR bid prices among existing

participants), raising the cost to purchase FTRs and reducing FTR profitability.

244

Analysis of

FTR profitability takes on additional importance in the context of participation costs (in

economic terms, a “barrier to entry”) that exist in this marketplace in the form of financial

assurance requirements.

245

As a group, FTR holders were profitable in 2022. Figure 4-6 below shows total profit (purple

line) as well as each of the different profit components. In this figure, FTR sales revenue and

positive target allocations are shown as positive values (as they increase FTR profitability),

while FTR purchase costs and negative target allocations are shown as negative values (as they

reduce FTR profitability). Further, this figure classifies purchase costs and sales revenues by

auction type (i.e., annual, prompt-month, or out-month).

243

In 2022, the participants that received this money included generator owners, participants that engaged in virtual and

external transactions, and load-serving entities, among others. See Section III.5.2.6 of Market Rule 1 for more information

about the distribution of excess congestion revenue.

244

Conversely, prolonged periods of large losses might motivate existing participants to exit the market (or at least

decrease their FTR bid prices), lowering the cost to purchase FTRs and increasing FTR profitability.

245

See Section VI of the ISO-NE Financial Assurance Policy for the financial assurance requirement provisions placed on FTR

transactions. https://www.iso-ne.com/static-assets/documents/2017/09/sect_i_ex_ia.pdf

2022 Annual Markets Report 146 ISO New England Inc.

ISO-NE PUBLIC

Figure 4-6: FTR Costs, Revenues, and Profits

In 2022, the total profit from FTRs was $0.6 million. This represents a substantial decrease from

2021 when total FTR profit was $25.9 million, but a modest increase from 2020 when total FTR

profit was -$0.8 million. An increase in FTR purchase costs was one of the primary drivers of

reduced profits. Participants spent $22.7 million more (or 128%) to procure FTRs in 2022

($40.4 million) than they did in 2021 ($17.7 million); spending increased by over $11 million in

both annual and monthly auctions. Another contributing factor was the reduction in positive

target allocations, which was generally reflective of lower levels of congestion on the system.

Most Profitable FTR Paths

Analyzing FTR outcomes at the path level provides insight into paths that were more congested

in the day-ahead energy market than reflected in auction clearing prices. Table 4-3 below

provides details of the 10 most profitable FTR paths in 2022. The purchase amount indicates

the total dollars spent in FTR auctions, while the sale amount shows the amount participants

earned from sales in the FTR auctions.

Table 4-3: Top 10 Most Profitable FTR Paths in 2022

Source Location

Sink Location

Purchase

Amount

($k)

Sale

Amount

($k)

Positive

Target

Allocations

($k)

Negative

Target

Allocations

($k)

Profit

($k)

UN.BULL_HL 34.5WEVW

LD.HARRINGT34.5

$(227)

$654

$428

.I.SALBRYNB345 1

.H.INTERNAL_HUB

$(151)

$637

$(59)

$427

UN.POWERSVL115 GNRT

.H.INTERNAL_HUB

$(3,182)

$3,630

$(51)

$397

LD.IRASBURG46

LD.STJHNSBY34.5

$(201)

$546

$(1)

$345

LD.S_NAUGTK13.8

LD.BEACN_FL13.8

$349

$(6)

$343

UN.HIGHGATE46 SHEL

LD.LYNDONVL34.5

$(76)

$353

$(1)

$278

LD.DEBLOIS 34.5

LD.TUNK 34.5

$(58)

$327

$276

($80)

($60)

($40)

($20)

$20

$40

$60

$80

2018 2019 2020 2021 2022

FTR Profit

($ millions)

Annual Auction Purchase Cost Annual Auction Sale Revenue

Monthly Auction Purchase Cost Monthly Auction Sale Revenue

BoPP Auction Purchase Cost BoPP Auction Sale Revenue

Negative Allocations Positive Allocations

FTR Profit

2022 Annual Markets Report 147 ISO New England Inc.

ISO-NE PUBLIC

Source Location

Sink Location

Purchase

Amount

($k)

Sale

Amount

($k)

Positive

Target

Allocations

($k)

Negative

Target

Allocations

($k)

Profit

($k)

UN.PONTOOK 34.5PONT

LD.LOSTNATN34.5

$(189)

$440

$251

UN.BULL_HL 34.5HANW

LD.HARRINGT34.5

$(106)

$352

$246

LD.DEBLOIS 34.5

LD.BOGGY_BK44

$(57)

$282

$234

All 10 of the most profitable FTR paths included locations that experienced negative congestion

pricing. Participants can hedge this type of congestion by procuring FTRs that source from

within the area experiencing negative congestion pricing and that sink in a location outside that

constrained area (the Hub is often used by participants) as paths of this type would produce

positive target allocations. Additionally, with the exception of one path, every path was a

prevailing flow FTR path.

246

This means they were paths defined in the direction that congestion

was expected to occur based on FTR auction clearing prices (i.e., participants had to pay to

acquire these paths, which can be seen by the negative value for each FTR path in the purchase

amount column).

Congestion on the New York – New England Interface

Changes in profitability of FTRs that source from Roseton, ISO-NE’s external node for trading

across the NYNE interface, can contribute significantly to overall FTR market outcomes. This is

because the NYNE interface tends to be one of the most frequently binding transmission

constraints in the day-ahead market.

247

Typically, participants purchase FTRs that source from

Roseton and sink somewhere within the ISO-NE system, as Roseton tends to experience

negative congestion pricing in the day-ahead market. To provide some perspective, the

purchase costs for FTRs sourcing from Roseton represented 52% of all the FTR auction

purchase costs in 2022, while the positive target allocations for FTRs sourcing from Roseton

represented 31% of all positive target allocations.

FTRs sourcing from Roseton were unprofitable in 2022, meaning that the value of congestion in

the day-ahead market was less than the cost anticipated in FTR auction clearing prices. Total

annual profits (purple line) are shown in Figure 4-7, which also shows the purchase costs, sale

revenues, and positive and negative target allocations for FTRs that sourced from Roseton.

246

The exception is the path from LD.S_NAUGTK13.8 to LD.BEACN_FL13.8.

247

See Section 3.2.9 for more information about the most frequently binding constraints in 2022.

2022 Annual Markets Report 148 ISO New England Inc.

ISO-NE PUBLIC

Figure 4-7: FTR Profits and Costs for FTRs Sourcing from Roseton

Profitability decreased by $16.7 million between 2021 ($9.8 million) and 2022 (-$6.8 million).

This decrease in profitability was largely the result of a sizeable jump in purchase costs

associated with this group of FTRs; participants paid $13.1 million more to acquire FTRs

sourcing from Roseton in 2022 ($20.9 million) than they did in 2021 ($7.7 million). At the same

time, the holders of these FTRs received $3.2 million less in positive target allocations in 2022

($14.2 million) than they did in 2021 ($17.4 million). One reason for this decrease was that the

NYNE interface was significantly less constrained in the day-ahead market in 2021 (when it

bound in 15.5% of hours) compared to 2022 (when it bound in 5.9% of hours) as average flows

of power from New York into New England decreased year over year.

248

See Section 5 for more information about the why net interchange decreased over the NYNE interface.

($25)

($20)

($15)

($10)

($5)

$10

$15

$20

$25

2018 2019 2020 2021 2022

FTR Profit

($ millions)

Auction Sale Revenue Auction Purchase Cost Negative Allocations

Positive Allocations FTR Profit

2022 Annual Markets Report 149 ISO New England Inc.

ISO-NE PUBLIC

Section 5

External Transactions

This section examines trends in external transactions in the day-ahead and real-time energy

markets. The first section (5.1) provides an overview of external transactions across all external

interfaces, while the second section (0) looks specifically at the performance of Coordinated

Transaction Scheduling (CTS) with New York.

External transactions are energy market transactions that allow market participants to transfer

power between New England and its neighboring control areas, and represent an important

part of the overall supply and demand picture.

249

Transferring power between different control

areas can help reduce total production costs across control areas by allowing power to flow

from lower priced to higher priced control areas, and provide reliability benefits to the

interconnected systems.

In 2022, net interchange (imports minus exports) was at its lowest level over the five-year

period. Net interchange averaged 1,914 MW per hour, down from 2,144 MW per hour 2021.

Over 90% of total net interchange came from Canadian interfaces. At New York North, net

interchange decreased by 189 MW per hour compared to 2021 due to increased congestion in

New York. This contributed to higher prices at this interface in New York compared to New

England. At New Brunswick, a nuclear generator outage lasted from April to July, resulting in

less net interchange over that period. Outside of New York North and New Brunswick, volumes

of net interchange remained in line with prior years. In 2022, external transactions received

more uplift than in each of the prior four years. In the day-ahead market, external transactions

received $3.1 million in uplift, most of which ($2.7 million) was paid to virtual transactions for

relieving congestion at external interfaces. In the real-time market, external transactions

received $1.2 million in uplift as make-whole payments for external transactions scheduled out-

of-merit.

The higher prices in New York compared to New England led to New England being a net

importer of power over the Roseton interface just 51% of the time, a decrease from 69% of the

time in 2021. Coordinated transaction scheduling (CTS) at the interface moved power from the

lower-priced region to the higher-priced in just 57% of hours, which was in line with prior

years. The average absolute price difference between New England and New York was about

$25/MWh, 96% higher than in 2021, but consistent with the overall increase in energy prices in

New York and New England. The average absolute price spread forecast error also increased

with higher energy prices, more than doubling from 2021 up to almost $24/MWh. ISO forecast

error continues to dampen the positive impacts of CTS by producing loss-making schedules and

risk for participants at the Roseton interface. Possibly in response to forecast-error-driven

inefficiencies, many CTS participants take on day-ahead positions and offer price-insensitive

real-time bids and offers. In 2022, the quantity of both imports and exports offering at prices

less than -$50/MWh increased from 2021.

249

A control area, or balancing authority area, is an area comprising a collection of generation, transmission and

load within metered boundaries for which a responsible entity (defined by NERC to be a balancing authority) integrates

resource plans for that area, maintains the area’s load-resource balance, and supports the area’s

interconnection frequency in real time.

2022 Annual Markets Report 150 ISO New England Inc.

ISO-NE PUBLIC

5.1 External Transactions

This section reviews outcomes, trends and drivers of import and export (external) transactions.

Specifically, we assess overall flows between New England and its neighboring control areas

and provide a breakdown across the six interfaces (subsection 5.1.1). We also assess the

pattern of fixed bidding versus price-sensitive bidding (5.1.2), and the drivers of uplift (NCPC)

payments to external transactions (5.1.3).

Key Takeaways

In 2022, net interchange (or net imports) averaged 1,914 MWs per hour, an 11% (or 231 MWs)

decrease compared to 2021 and the lowest average net interchange over the five year period.

Average net interchange fell due to decreased net interchange across the New York North and

New Brunswick interfaces.

At New York North, average net interchange decreased by 189 MWs per hour compared to 2021

(401 MWs vs. 590 MWs). Increased congestion in New York led to higher New York prices and

increased exports over the interface. At New Brunswick, average net interchange decreased by

68 MWs (228 MWs vs. 296 MWs) compared to 2021 due to the extended outage of a nuclear

generator in New Brunswick.

External transactions received more uplift payments than in 2021 despite higher energy prices in

2021. In the day-ahead market, virtual transactions received a majority of the uplift payments for

relieving congestion at external interfaces. In the real-time market, uplift payments increased

due to increased forecast error at non-CTS interfaces leading to increased clearing of out-of-merit

external transactions.

5.1.1 External Transaction Volumes

The average hourly system-wide net interchange from the day-ahead and real-time markets are

shown in the line series of Figure 5-1 below. The bar series chart the hourly average imported

volume (positive values) and exported volume (negative values). The real-time import and

export volumes are shown as the incremental additions to the amounts cleared in the day-

ahead market.

2022 Annual Markets Report 151 ISO New England Inc.

ISO-NE PUBLIC

Figure 5-1: Hourly Average Day-Ahead and Real-Time Pool Net Interchange

Real-time net imports averaged 1,914 MW each hour, the lowest level of net interchange over

the five-year period, and down by 11% (or 231 MW) on 2021. The continued decline in net

interchange from 2020 was due to an increase in export transactions. Total exports were up by

22% (or 225 MW) compared to 2021, with the greatest changes at the New York North

interface.

The close proximity of the day-ahead net interchange (orange) and real-time net interchange

(red) line series highlights that day-ahead market outcomes across the external nodes do, on

average, align well with real-time scheduled flows (historically within 2%).

250

However, average

real-time net interchange was less than day-ahead net interchange by 8% (or 177 MW) in 2022

(i.e., less power was imported in real-time than planned for in the day-ahead market) due to the

increase in real-time export bids at New York North.

A breakdown of flows across the Canadian Interfaces

Figure 5-2 below shows the annual hourly average real-time net interchange volumes (red line)

as well as the gross import (positive values) and export (negative values) volumes at each of the

three Canadian interconnections. The average hourly real-time total transfer capability (TTC)

ratings for each interface in the import and export directions are plotted using the black dashed

lines.

250

Virtual transactions cleared at external interfaces in the day-ahead market are included in the day-ahead net

interchange value. In the day-ahead energy market, virtual supply and demand are treated similarly to imports or exports.

2022 Annual Markets Report 152 ISO New England Inc.

ISO-NE PUBLIC

Figure 5-2: Real-Time Net Interchange at Canadian Interfaces

New England continues to import significant volumes of power from Canada, averaging 1,781

MW per hour in 2022, or 93% of total net imports. To put this in perspective, net imports from

Canada represented 13% of real-time load in 2022. Net imports declined by 4% (or 79 MW)

from 2021.

Average flows across Phase II and Highgate have been relatively steady over the past five years;

combined flows in 2022 were just 11 MW per hour less than 2021. Reduced flows at the New

Brunswick interface explain the majority of the decrease (68 MW). In 2022, a 660-MW New

Brunswick nuclear generator experienced an extended outage, which resulted in decreased

imports and increased exports over the interface.

A breakdown of flows across the New York Interfaces

Figure 5-3 below shows the annual hourly average real-time net interchange volumes (red line)

as well as the gross import and export volumes at each of the three New York interconnections.

The average hourly real-time total transfer capability (TTC) ratings for each interface in the

import and export directions are plotted using black dashed lines.

251

The total transfer capability (TTC) rating is the MW amount of power that can be reliably transferred from one system

to the other over the transmission interface.

(1,200)

(800)

(400)

400

800

1,200

1,600

2018 2019 2020 2021 2022 2018 2019 2020 2021 2022 2018 2019 2020 2021 2022

Phase II New Brunswick Highgate

Hourly average MW

Imports Exports Net interchange TTC

2022 Annual Markets Report 153 ISO New England Inc.

ISO-NE PUBLIC

Figure 5-3: Real-Time Net Interchange at New York Interfaces

On a net basis, New England imports power over the New York North interface and exports

power over both the Cross Sound and Northport-Norwalk interfaces. Combining flows at all

three interfaces, ISO-NE averaged just 133 MW per hour, the lowest volume of the five-year

period.

At New York North, net interchange (red line) decreased for the second consecutive year, down

32% (or 189 MW per hour) compared to 2021 and down 57% (or 533 MW) compared to the

five-year high value (2020). While cleared import transactions were comparable to 2021, real-

time exports increased by 31% (or 209 MW). The increase in exports was due to increased

congestion and higher prices in Eastern New York. The retirement of a nuclear power plant

(870-MW Indian Point 3 unit), and increased planned transmission outages during 2022 were

major drivers of higher New York prices compared to prices in New England.

252

253

At Cross-Sound Cable, net interchange continued in the export direction in 2022 (212 MW) and

was comparable to 2021.

254

Exports remained well above levels from 2018 – 2020 (121 MW).

New York’s prices ($95.39/MWh) were higher than New England ($83.89/MWh) at the

interface, on average, which led to the continued high volumes of export transactions compared

to 2018 – 2020.

252

Quarterly Report on the New York ISO Electricity Markets Third Quarter of 2021. See

https://www.potomaceconomics.com/wp-content/uploads/2021/12/NYISO-Quarterly-Report_2021Q3__11-29-2021.pdf

253

For more on congestion in New York, see Potomac Economics Quarterly Report on New York ISO Electricity Markets

Third Quarter of 2022: https://www.potomaceconomics.com/wp-content/uploads/2022/12/NYISO-Quarterly-

Report_2022Q3__11-21-2022.pdf

254

Imports at the Cross Sound Cable averaged less than 0.1 MW since 2019 and have averaged less than 0.9 MWs every

year over the reporting period.

(1,200)

(800)

(400)

400

800

1,200

1,600

2018 2019 2020 2021 2022 2018 2019 2020 2021 2022 2018 2019 2020 2021 2022

New York North Cross Sound Northport-Norwalk

Hourly Average MW

Imports Exports Net Interchange TTC

2022 Annual Markets Report 154 ISO New England Inc.

ISO-NE PUBLIC

At Northport-Norwalk net exports were down slightly on 2021 (55 MW per hour in 2022 versus

84 MW in 2022) due to a decrease in fixed exports submitted by participants (77 MW vs 95 MW

per hour).

255

5.1.2 External Transaction Offer Composition

In Section 3 (see 3.2.2 and 3.2.5) we assessed overall market supply and demand-side

participation in terms of the extent to which supply offers or demand bids are price-taking

(fixed) versus price-making (priced). This can have important implications for price formation

in the energy markets. In this section, we present a similar analysis focusing on external

transactions, which participate on both the supply and demand sides of the market (i.e., supply

offers for imports and demand bids for exports).

At the Canadian interfaces, the majority of transactions continue to be fixed, but levels of fixed

transactions decreased compared to 2021. By contrast, at the New York interfaces transactions

are mostly priced, and the split between priced and fixed remained steady compared to 2021.

Canadian Interfaces

The composition of transactions that cleared at the Canadian interfaces in the day-ahead and

real-time markets by fixed and priced is shown in Figure 5-4 below. Volumes are average MW

per hour.

Figure 5-4: Transaction Types by Market and Direction at Canadian Interfaces (Average MW per hour)

Imports at the Canadian interfaces continue to be predominantly fixed, but levels of fixed

transactions decreased compared to levels in 2021 in both the day-ahead and real-time

markets. In 2022, priced transactions accounted for 29% of volumes in the day-ahead and 25%

of volumes in real-time market. On average, flows are higher in the real-time market due to

additional real-time fixed transactions. In particular, participants submitted more priced

transactions (and conversely fewer fixed transactions) over the Phase II interface, leading to the

255

An 11 MW increase in imports over Northport-Norwalk also contributed to the decrease in net exports.

1,667

1,336

1,242

1,664

1,363

327

544

529

184

489

1,859

1,539

1,358

1,767

1,418

275

487

559

150

463

(500)

500

1,000

1,500

2,000

2,500

DA RT DA RT DA RT DA RT DA RT

Imports (+)/Exports(

Fixed Priced

2018

2019 2020

2021 2022

2022 Annual Markets Report 155 ISO New England Inc.

ISO-NE PUBLIC

difference in system-level transactions. At Phase II, participants cleared an average 407 MW of

priced transactions in 2022 compared to 58 MW on average in 2021.

On the exports side, volumes increased, but remained low at an average of just 100 MW. Both

the decrease in imports and increase in exports occurred at the New Brunswick interface from

April to July when a nuclear generator was on an extended outage.

New York Interfaces

The composition of transactions that cleared at the New York interfaces in the day-ahead and

real-time markets by fixed and priced is shown in Figure 5-5.

256

Figure 5-5: Transaction Types by Market and Direction at New York Interfaces (Average MW per hour)

Most day-ahead cleared import transactions at the New York interfaces are fixed (65% in 2022)

while most exports are priced (96%). However, the share of total priced day-ahead external

transactions (both imports and exports) increased slightly in 2022, to about 60%, due to more

priced exports. In fact, export transactions have increased significantly in the past two years. A

driver of higher exports was higher prices at the NYN interface, particularly due to constraints

associated with the retirement of Indian Point 3 and increased transmission work that caused

congestion at Central-East interface.

257

A higher percentage of real-time imports (77%) are priced due to the bidding mechanics of

Coordinated Transaction Scheduling (CTS) at the New York North interface. The majority of

total cleared import transaction volumes come from New York North. Under CTS, all real-time

transactions are evaluated based on price, although participants may offer prices as low as -

256

Volumes not listed in the figure all averaged less than 100 MW.

257

For more on congestion in New York, see Potomac Economics Quarterly Report on New York ISO Electricity Markets

Third Quarter of 2022: https://www.potomaceconomics.com/wp-content/uploads/2022/12/NYISO-Quarterly-

Report_2022Q3__11-21-2022.pdf

441

699

815

639

668

447

323

308

382

362

(354)

(268)

(227)

(580)

(707)

168

276

303

967

1,281

1,265

996

999

(442)

(536)

(563)

(892)

(1,091)

(1,500)

(1,000)

(500)

500

1,000

1,500

2,000

DA RT DA RT DA RT DA RT DA RT

Imports (+)/Exports(

Fixed Priced

2018

2019

2020

2021

2022

2022 Annual Markets Report 156 ISO New England Inc.

ISO-NE PUBLIC

$1,000/MWh, which effectively schedules the transaction as fixed. For example, about 637 MW

of imports and 390 MW of export transactions were willing to clear in each hour at a greater

than $50/MWh loss.

258

A majority of the fixed import transactions also occurred at New York

North. The higher volume of fixed import transactions over the past two years mostly reflects

the growth in wheeled transactions through New York into New England.

259

5.1.3 External Transaction Uplift (NCPC) Payments

External transactions are eligible to receive uplift (or NCPC) payments when revenues are not

sufficient to recover their costs. These payments often occur when external transactions clear

on an ISO price forecast but are unable to recover as-offered costs through energy market

transactions. External transactions (or virtual transactions placed at external nodes) can also

receive uplift for relieving congestion at non-CTS external interfaces since congestion is not

captured in the LMP. These payments occur when a transaction that is out-of-the-money at the

system price clears in the direction counter to the constraint (e.g., an export or virtual demand

bid when the interface is import-constrained) allowing a counter-party to clear in excess of the

interface limit.

260

These otherwise uneconomic transactions require uplift in absence of

congestion pricing.

The annual uplift credit totals at all external nodes in both the day-ahead and real-time markets

are presented in Table 5-1 below.

Table 5-1: NCPC Credits at External Nodes

Year

Day-ahead credits

($million)

Real-time credits

($million)

2018

$0.30

$2.73

2019

$0.02

$1.02

2020

$0.00

$1.39

2021

$1.04

$0.53

2022

$3.13

$1.17

Typically, total uplift paid at external nodes is very small compared with other types of uplift

(see Section 3.3). In the day-ahead market, these payments typically occur when there is a

surplus of fixed external transactions in excess of the TTC and a virtual or priced external

258

Participant CTS bidding behavior is discussed below in Section 5.2.3.

259

A wheeled transaction is when power is flowed from one system to another over a third party’s transmission lines. For

example, a participant might use these transactions to flow power from PJM through New York and into New England.

260

For example, consider an interface with an import TTC of 100MW and an LMP of $100. If there are 200MW of imports

offered at $0, only 100MW can clear (due to the TTC), unless there is a transaction to offset the re maining 100MW of

excess imports. If a 100MW export is offered at $50, it can provide counterflow. The $50 exports are willing to purchase

the $0 imports. However, because congestion is not captured in the LMP, the energy settlement for the export will result

in a loss; the exports are only willing to pay $50, and the LMP is $100. Therefore, these 100MW of exports must be paid

$50/MWh to make them whole, for a total NCPC payment of $5,000. The NCPC charges are only levied to the participants

importing over the interface.

2022 Annual Markets Report 157 ISO New England Inc.

ISO-NE PUBLIC

transaction provides “counter-flow.”

261

In 2022, most of the day-ahead uplift was due to virtual

transactions relieving congestion by providing counter-flow for surplus fixed transactions at

external interfaces.

In 2022, day-ahead uplift credits totaled $3.1 million, significantly higher than prior years.

Virtual transactions received nearly $2.7 million in uplift credits and external transactions

received about $0.5 million. Most of the uplift credits ($2 million) accrued at the Cross Sounds

Cable interface for congestion relief in the export direction, with the remainder mainly at New

Brunswick and Highgate for relieving congestion in the import direction.

Total real-time uplift credits during 2022 were higher than in 2021. This was because (1) more

external transactions were bid as priced in 2022, making more transactions eligible to receive

NCPC compared to 2021, and (2) more transactions (at non-CTS interfaces

262

) were in-rate

based on forecast prices, but were out-of-rate based actual prices used in settlement . Accurate

price forecasting of LMPs helps reduce NCPC paid to external transactions. In 2022, forecast

error at the non-CTS interfaces increased, with actual LMPs much higher than forecasted LMPs.

Real-time LMPs at all interfaces averaged nearly $7.00/MWh higher than the forecasted LMPs,

which was up from $4.46/MWh in 2021.

261

Since congestion is not priced at the interface, the participant relieving congestion with virtual supply receives uplift if

the bid was cleared out of economic merit order. Often, this occurs when there is a large decrease in an interface TTC and

participants have not adjusted their fixed bidding behavior.

262

At the CTS interface, out-of-rate transactions are not entitled to NCPC, but also do not incur NCPC charges.

2022 Annual Markets Report 158 ISO New England Inc.

ISO-NE PUBLIC

5.2 Coordinated Transaction Scheduling

In this section, we provide an updated assessment of how coordinated transaction scheduling

(CTS) is functioning. Specifically, we review CTS performance metrics against its high-level

primary goals (subsection 5.2.1), including the level of inefficient scheduling (5.2.2), and CTS

impacts on efficient bidding opportunities (5.2.3).

Key Takeaways

In 2022, New York energy prices exceeded New England prices for the first time in the last five

years. As a result, New England was a net importer of power over the Roseton interface 51% of

the time, a decrease from 69% of the time in 2021.

The Roseton interface did not move power from the lower-to-higher priced region significantly

more often (about 57% of the time) than previous years. The average absolute price difference

between New England and New York was about $25/MWh, 96% higher than in 2021, but

consistent with the overall increase in energy prices in New York and New England. The average

absolute price spread forecast error also increased with higher energy prices, more than

doubling from 2021 up to almost $24/MWh. Rather than submitting price-sensitive spread bids,

participants continued a trend of clearing in the day-ahead market and offering price-insensitive

transactions in real time. In 2022, the quantity of both imports and exports offering at prices less

than -$50/MWh increased from 2021.

ISO forecast error continues to dampen the positive impacts of CTS by producing loss-making

schedules and risk for participants at the Roseton interface. Possibly in response to forecast-

error-driven inefficiencies, many CTS participants take on day-ahead positions and offer price-

insensitive real-time bids and offers.

Since price forecast error is unlikely to be completely eliminated, minimizing the impact of price

forecast error through changes to CTS mechanics or settlement may better incentivize

participants to offer at cost. We will continue to evaluate these issues and encourage ISO-NE to

review its price forecasting tools and explore opportunities to improve forecast accuracy.

5.2.1 CTS Performance

In this section, we analyze CTS performance against two measures of efficiency: the flow of

power from the lower-to higher-cost region and price convergence between regions.

A summary of CTS power flows between the two control areas is shown in Table 5-2 below. The

percentage of time power flowed into each control area is shown in the Net Flow columns.

263

The percentage of time the flow was directionally correct (i.e., power flowed from lower- to

263

Fixed wheeling transactions at the NYN interface are ignored in all of the analyses contained in this s ection. These

transactions are not cleared in the CTS process. On average, in 2022 there were 292 MW of fixed -wheeling transactions

net importing over the NYN interface in each interval.

2022 Annual Markets Report 159 ISO New England Inc.

ISO-NE PUBLIC

high-cost region, based on the forecasted or actual prices) is shown in the Correct Flow

columns.

264

Table 5-2: Summary of CTS Flow Outcomes

Net Flow (% of intervals), to:

Correct Flow (% of intervals),

based on:

Year

ISO-NE

NYISO

Forecast Price

Spread

Actual Price

Spread

2018

77%

23%

61%

63%

2019

91%

49%

58%

2020

95%

40%

55%

2021

69%

31%

52%

56%

2022

51%

49%

59%

57%

In 2022, power flowed into New England from New York 51% of the time, the lowest frequency

in the five-year period. In 2022, the decrease in importing-intervals was driven by an increase

in price-insensitive export bids. Although ISO-NE exported over the NYN interface more often in

2022, the impact on correct flows was modest. In 2022, power flowed in the correct (economic)

direction, based on the actual price spread between New York and New England 57% of the

time – a similar percentage to the prior three years.

Economic flow direction metric

Ideally, power would flow from the control area with lower energy prices to the control area

with higher prices until prices converge or a physical limit is hit (the TTC or the ramp limit). A

breakdown of the intervals in which each of these constraints were binding and the absolute

price difference between the control areas in each interval over the last five years is illustrated

in Figure 5-6 below. The overall price difference between control areas is shown by the height

of the area, in descending order of price difference. The following is a description of how to read

the graph.

 The green areas on the chart represent the best possible outcomes; when power flowed

in the correct (economic) direction and the TTC or ramp constraint were binding.

 The yellow area shows intervals where power flowed in the correct (economic)

direction without binding constraints.

 The orange area shows intervals where power flowed in the wrong (uneconomic)

direction without binding constraints.

 The red areas on the right show the less desirable outcomes: when power flowed in the

wrong (uneconomic) direction and the TTC or ramp constraint were binding. In other

words, when CTS was diverging prices as much as the constraints allowed.

264

The prices used in this subsection are proxy prices that represent the marginal cost of energy on each side of the NYN

interface. The NYISO pricing node is “N.E._GEN_SANDY PD” (Sandy Pond) and the ISO-NE node is “.I.ROSETON 345 1”

(Roseton). Congestion pricing is removed from external prices to ensure we are better-capturing the marginal cost of

energy in each control area at the border. When the ramp or flow limit binds, the prices at the interface reflect the bids

and offers that set price based on the forecast, and not necessarily the marginal cost of energy in each control area.

2022 Annual Markets Report 160 ISO New England Inc.

ISO-NE PUBLIC

Figure 5-6: CTS Outcome Summary

Overall, there was a similar percentage of intervals when power flowed in the correct direction

(dark green and yellow areas) or the interface was ramp-constrained but adjusting in the

correct direction (light green areas) in 2021 (56%) and 2022 (57%). However, in 2022, CTS

performance continued to decline (since 2018) by flowing in the correct direction up to the TTC

or ramp limit (the green areas) only 13% of the time, 4% less often than in 2021, and

continuing a downward trend from 2018.

265

The reduction in this performance metric is driven

by an increase in price-insensitive export transactions. Due to the volume of price-insensitive

imports and exports offered over the NYN interface, CTS has limited ability to adjust to price

differences between New England and New York. In past years, CTS has produced schedules

with the most efficient outcomes (ramp constrained in the correct direction or flowing at the

flow limit in the correct direction) almost exclusively in the import direction. In 2020 and 2021,

New England was importing over 95% of the time when a physical constraint (TTC or ramp)

was binding. In 2022, CTS was more balanced in the direction of these outcomes; New England

was importing only 58% of the time when a constraint was binding. Even though CTS schedules

were correctly binding in the export direction more often, the total percentage of the time CTS

constraints were binding in the correct direction decreased.

265

In some cases, CTS could be converging prices as efficiently as possible without a ramp constraint or flow constraint

binding, if prices are equal on either side of the interface. Although there were only 2.5 cumu lative hours in which the

price difference was $0/MWh, 7% of the total time in 2022 there was no constraint binding and the price difference was

under $2/MWh. This also represents a decrease from 2021, when no constraint was binding and the price difference was

under $2/MWh for 21% of the time.

2022 Annual Markets Report 161 ISO New England Inc.

ISO-NE PUBLIC

Price convergence metric

As a primary consideration for the implementation of CTS, it is important to evaluate its impact

on price convergence between the markets. The height of each bar in Figure 5-6 above

represents the absolute price difference in each interval. Less area under each curve represents

better price convergence. In 2022, the average absolute price difference (the average of the

lines in each year) between NE and NY was $25.06, 96% higher than in 2021. This is visually

apparent because there are much larger areas under each curve, shown in the graph above, in

2022. However, overall NE LMPs were 93% higher in 2022 (not shown). Therefore, as a

percentage of average LMPs, the absolute price difference in 2021 and 2022 were similar;

average absolute price differences were about 30% of average NE LMPs in each year.

Tie line capacity utilization metric

When prices differ between control areas and neither the TTC or ramp constraint is binding,

CTS is not converging prices optimally, and there is underutilized capacity on the transmission

ties. Figure 5-7 below shows the available, but unused, capacity up to the physical constraint

(TTC or ramp) in the low- to high-price direction.

266

This surplus capacity could be used to help

converge prices further. This analysis is bucketed by price difference because at lower price

difference levels, we would not expect CTS to utilize as much of the capacity.

267

The percentages

on the left side of the bar show the percentage of time each price bin occurred during the year.

266

This figure does not account for s chedule cuts. In some circumstances, after schedules are generated for an interval,

schedules are cut and actual flows do not match pricing outcomes. The ramp limit in this chart i s calculated from the

previous interval’s scheduled net interchange, rather than actual net interchange after cuts. In cases where there are

schedule cuts, the available capacity in this chart may not represent the available capacity used in CTS clearing.

267

For instance, if the price difference is less than $1/MWh, the interface may have already reached the optimal flow to

converge prices, and clearing one more MW could result in a larger price difference. However, this scenario is less likely as

the price difference gets larger.

2022 Annual Markets Report 162 ISO New England Inc.

ISO-NE PUBLIC

Figure 5-7: CTS Unused Capacity, 2022

Even when the price difference between regions is high, on average, CTS did not fully utilize the

available capacity to help converge prices. In scenarios where price differences were greater

than $50/MWh, there was about 200 MW of average unused interface capacity. Since

implementation, CTS has consistently left NYN interface capacity underutilized. Figure 5-7

indicates that CTS is not effectively utilizing the available capacity to improve price

convergence.

5.2.2 CTS and ISO Scheduling Efficiency

CTS participants are not compensated when energy-price incentives are misaligned with CTS

transaction schedules. In other words, transactions are scheduled based on forecast prices, but

settled based on actual prices. Therefore, participants bear the risk resulting from scheduling

inefficiencies due to price forecast error.

A summary of forecast versus actual prices, as well as the average and absolute forecasting

errors, is provided in Table 5-3 below. Similar to above analyses, unless otherwise noted (i.e., in

the Spread, with Cong. column), NYN proxy prices net of congestion are shown to better capture

the marginal cost of energy on each side of the interface.

268

Proxy prices do not include external congestion. The average forecast error and average absolute forecast error will not

change if congestion is added to both the forecast and actual LMPs.

26%

20%

24%

12%

0 50 100 150 200 250 300

Less than $1

Between $1 and $5

Between $5 and $10

Between $10 and $25

Between $25 and $50

Between $50 and $100

Over $100

Available Interface Capacity to Converge Prices (MW)

Absolute NE

-NY Price Difference ($/MWh)

2022 Annual Markets Report 163 ISO New England Inc.

ISO-NE PUBLIC

Table 5-3: Forecast Error in CTS Solution

Actual LMP ($/MWh)

Forecast LMP ($/MWh)

Average Forecast Error

($/MWh)

Average Absolute Forecast

Error ($/MWh)

ISO-NE

NYISO

Spread

Spread,

with

Cong.

ISO-NE

NYISO

Spread

ISO-NE

NYISO

Spread

ISO-NE

NYISO

Spread

2018

$42.53

$37.52

$5.01

($1.51)

$41.45

$35.71

$5.74

($1.07)

($1.81)

$0.73

$8.09

$8.42

$13.46

2019

$29.66

$26.47

$3.19

($0.72)

$28.64

$26.83

$1.81

($1.02)

$0.36

($1.38)

$4.69

$5.07

$7.96

2020

$22.45

$20.46

$1.99

($0.67)

$21.72

$21.33

$0.39

($0.73)

$0.87

($1.60)

$3.76

$4.04

$6.34

2021

$43.03

$41.07

$1.96

($1.65)

$41.32

$41.66

($0.34)

($1.71)

$0.60

($2.31)

$5.50

$7.93

$10.78

2022

$82.94

$86.39

($3.45)

($4.63)

$79.80

$84.57

($4.77)

($3.14)

($1.82)

($1.32)

$11.23

$18.51

$23.87

Price forecasting continues to be a challenge for the ISOs. The Average Absolute Forecast Error

columns show the amount that the forecast differed from the actual prices (ignoring

directionality). In 2022, average absolute forecast error increased to $23.87/MWh, more than

double the 2021 error ($10.78/MWh), indicating that the CTS forecasts became less accurate.

The Average Forecast Error columns take direction into account. Since 2018, actual ISO-NE

prices have been higher than the forecast. In 2022, NYISO actual prices were also, on average,

higher than the forecast. Since the 2022 errors were in the same direction, they helped to

diminish the impact of each individual ISO’s forecast error; the forecast error in the spread (-

$1.32/MWh) was smaller than each ISO’s individual forecast error (-$3.14 and -$1.82/MWh).

The Actual LMP column shows that in 2022, NYISO energy prices were higher than ISO-NE for

the first time over the past five years.

269

Due to the NY premium, 2022 was the first year in the

study period in which the price spread with congestion, shown in the Actual LMP, Spread with

Cong. column, was directionally similar to the price spread without congestion. In 2018-2021,

interface constraints, coupled with bid and offer behavior from CTS participants, drove a sub-

optimal outcome where, on average, bids and offers converging prices (i.e., moving energy from

the lower- to higher-cost region) were loss generating. This occurs when participants make bids

and offers that are less than zero (i.e., willing to clear at a loss).

270

For instance, consider a scenario in which there are 2,000 MW of imports offered at a -

$10/MWh spread price, the CTS flow limit is 1,400 MW, and the New England price is

forecasted to be $20/MWh higher than New York. CTS will schedule 1,400 W of the 2,000 MW

of imports (up to the flow limit) because the offered spread bids are priced lower than the

forecasted spread price. A -$10/MWh spread bid will set the congestion price at the interface,

because the interface is constrained—in other words, based on the $20/MWh price spread

more of the -$10/MWh imports would like to clear, but cannot due to the flow limit. The CTS

congestion price will be -$30/MWh, because if one more MW could clear over the interface

because the flow limit was relaxed 1 MW, the system would benefit $30/MWh (the $20/MWh

269

The day-ahead NY price premium at the NYN interface was $4.63/MWh, identical to the real-time NY price premium

with congestion.

270

Bid prices can be positive, negative, or zero. A positive bid price indicates the participant is willing to move power when

the price in the destination market exceeds the price in the source market by at least the bid price (i.e., buy low and sell

high). A negative bid price indicates a willingness to trade power when the energy price is higher at the source than at the

destination, by as much as the negative bid price (i.e., to buy high and sell low).

2022 Annual Markets Report 164 ISO New England Inc.

ISO-NE PUBLIC

less expensive power from New York could flow into New England plus the $10/MWh the

participant is willing to pay to flow the import). If the actual price spread is $20/MWh, the same

as the forecast, the CTS spread price with congestion will be -$10/MWh ($20/MWh actual price

spread minus $30/MWh congestion price). This outcome highlights a scenario in which CTS is

flowing from the lower- to higher-cost region (New England prices are $20/MWh higher

without congestion pricing), but the import offers are loss making (the settled CTS price spread

is -$10/MWh with congestion pricing).

The risk of uneconomic scheduling due to forecast error could be one driver of negative real-

time CTS bid and offer prices. One strategy for mitigating risk is to lock in (hedge) the day-

ahead position by submitting low-priced real-time transactions to minimize day-ahead/real-

time deviations.

Table 5-4 shows the average bid prices of 2022 bids and offers cleared in each price bin, by

direction, along with forecasted and settled LMPs.

271

When forecasted LMP differences are

greater than average bid prices, bids were scheduled and are shown in green. All forecasted

LMP differences are greater than average bid prices in this table because if forecasted LMP

differences are less than the bid price, the bids are not scheduled. When settled LMP differences

are greater than average bid prices, bids were profitable and are shown in green. Due to

forecast error, some settled LMP differences are less than bid prices and are shown in red. This

indicates that, if participants were offering at cost, these bids were loss making due to forecast

error in 2022.

Table 5-4: Profit Scenarios, 2022

Direction

Bid Price Bin

Average Bid

Price

Forecasted LMP

Difference

Settled LMP

Difference

Hourly

Scheduled MW

Export

Less than -$25

(585.85)

5.09

4.48

413.9

-$25 and -$5

(16.61)

5.14

3.60

102.3

-$5 and $0

(1.17)

6.63

3.50

101.9

$0 and $5

2.52

9.38

4.47

200.7

$5 and $25

9.31

15.97

7.26

39.0

Greater than $25

68.60

151.44

85.89

0.7

Import

Less than -$25

(755.52)

(7.02)

(6.54)

727.2

-$25 and -$5

(17.76)

(5.19)

(6.30)

242.0

-$5 and $0

(2.53)

1.23

(0.76)

7.6

$0 and $5

2.76

10.66

(9.26)

9.9

$5 and $25

9.76

24.38

(2.70)

2.1

Greater than $25

48.56

68.42

(11.50)

0.2

Table 5-4 above shows how forecast error has impacted CTS participants in each bid and offer

price bin. Settled LMP differences were negative for all import bins in 2022, indicating that,

271

To better compare LMPs with bid prices, forecasted and settled LMPs in this table are weighted by scheduled MWs,

reflect external congestion, and show price differences in the direction of bids. Imports clear when NE – NY > bid price and

exports clear when NY – NE > bid price; the NE-NY LMP difference is included for imports and the NY-NE LMP difference is

included for exports.

2022 Annual Markets Report 165 ISO New England Inc.

ISO-NE PUBLIC

although the offers were economic based on the offer price, they were moving power counter-

intuitively, from the higher- to lower-cost region. Due to price forecast error, import offers

priced above $0/MWh, as well as export transactions priced between $5 and $25/MWh, were,

on average, loss generating due to forecast error (i.e., scheduled transactions with positive costs

were uneconomic once prices materialized). Faced with these risks, participants may prefer

offering in the day-ahead market (where there is no forecast error) and minimizing real-time

deviations with price-insensitive bids in real time. The Hourly Scheduled MW column highlights

these dynamics, with about half of export transactions and about three-quarters of import

transactions offered at prices less than -$25/MWh. Consequently, the NYN interface has limited

price-sensitive bids and offers needed to adjust to regional price differences.

5.2.3 CTS and Participant Bidding Opportunities

Forecast error drives inefficient scheduling that reduces efficient bidding opportunities. In

other words, if the price forecast error was zero, then participants should be willing to offer at

de minimus price (say $0.01/MWh) in the import and export direction and flow to the higher-

price market. However, from the table in the prior section, we see that in many cases a

$0.01/MWh import would be scheduled to flow from the lower- to high-price market but, due

to forecast error, would end up flowing at a loss because the actual spread can materialize in

the opposite direction from the forecasted spread. A rational bidder would create a bidding

strategy to account for forecast error inefficiencies. In this subsection, we take a closer look at

actual participant bidding behavior.

Average CTS transaction curves, by year, are shown in Figure 5-8 below. Import offers for 2022

are shown in the first graph (gold curves) followed by 2022 export bids (red curves). Lastly,

imports and exports are aggregated to produce a net supply curve (orange curves). The import

and export curves show the average volume of energy willing to clear at each New England -

New York price spread. The aggregate supply curve shows the net flow that would be produced

if all of the economic import and export transactions were to clear.

For example, in 2022, at a price spread of $0/MWh (i.e., NE price is equal to the NY price), about

1,075 MW of imports would have cleared, 725 MW of exports would have cleared, and the net

flow of CTS transactions would have been 350 MW, on average. The typical import TTCs, less

the average number of wheeling transactions, are shown in dotted lines.

272

The net imports

cannot clear above these lines, and when the price difference is forecasted to be greater than

the intersection (about $500/MWh New England – New York when the TTC is 1,200MW) a CTS

bid will set the congestion prices at NYN.

273

272

The export TTC (plus wheeling transactions) is not shown because the average net imports curve does not cross the

limit.

273

Only one of these TTCs will be active at a time. Both are shown to visualize the difference in flows and prices when

either is binding.

2022 Annual Markets Report 166 ISO New England Inc.

ISO-NE PUBLIC

Figure 5-8: Price Sensitivity of Offered CTS Transactions

Figure 5-8 highlights a few key takeaways about participant bidding behavior at a high level. On

average:

1) There were many price-insensitive imports willing to clear in each interval.

 About 637 MW of imports were willing to clear in each hour at a greater than

$50/MWh loss, an increase from 576 MW in 2021.

2) There were also many more exports willing to clear at a loss.

 About 390 MW per hour were willing to clear at a greater than $50/MWh loss,

compared to only 218 MW in 2021.

3) CTS participants, as a whole, were net importers even when New England prices were

lower.

 Participants would import when prices were up to $7/MWh lower in New

England, and would import about 352 MW at a price difference of $0/MWh.

4) The TCC in the import direction bound at a much higher average price premium.

 In 2021, a relatively small New England price premium – about $11/MWh – was

enough to bind the TTC when the TTC was 1,400 MW. In 2022, the additional

exports willing to provide counter-flow increased the price premium at which

the TTC bound to $1,000/MWh.

The large quantity of price-insensitive bids reflect participants with day-ahead positions. 50%

of exports and 71% of imports scheduled in real-time were hedged with a day-ahead position.

While this is a reasonable business strategy to mitigate forecast error risk (or based on

contractual positions), this bidding behavior inhibits CTS from adjusting to changes in price

between New York and New England.

274

This section highlights CTS mechanics that degrade participant incentives. In reality, market participants may offer in

ways that appear uneconomic for many reasons, including capturing environmental credits or long-term contracts.

500

1,000

1,500

2,000

-$50 -$40 -$30 -$20 -$10 $0 $10 $20 $30 $40 $50

Average Offered

Exports (MW)

Export Demand Curve

500

1,000

1,500

2,000

-$50 -$40 -$30 -$20 -$10 $0 $10 $20 $30 $40 $50

Average Offered

Imports (MW)

Import Supply Curve

-1,000

-500

500

1,000

1,500

2,000

-$50 -$40 -$30 -$20 -$10 $0 $10 $20 $30 $40 $50

Average Net Offered Imports

- Exports

(MW)

New England minus New York Price Difference ($/MWh)

2018 2019

2020 2021

2022 1,400 TTC Minus Fixed Imports

1,600 TTC Minus Fixed Imports

New York Price Higher

New England Price Higher

Willing to Net

Export

Net Supply Curve

(Imports - Exports)

2022 Annual Markets Report 167 ISO New England Inc.

ISO-NE PUBLIC

Since price forecast error is unlikely to be completely eliminated, minimizing the impact of

price forecast error through changes to CTS mechanics or settlement may better incentivize

participants to offer at cost. We will continue to evaluate these issues and encourage ISO-NE to

review its price forecasting tools and explore opportunities to improve forecast accuracy.

2022 Annual Markets Report 168 ISO New England Inc.

ISO-NE PUBLIC

Section 6

Forward Capacity Market

This section reviews key trends and performance of New England’s forward capacity market

(FCM). The seventeenth forward capacity auction (FCA) is covered in Section 6.1. The auction

cleared with a surplus of capacity at a payment rate well below the estimated cost of new entry.

The outcomes of FCA 17 were closely aligned with the prevailing trend of surplus capacity and

low clearing rates.

Section 6.2 reviews FCM trends over the past seven years. The surplus conditions can be

attributed to low amounts of retirements and steady additions of new generation, notably from

renewable or policy-sponsored projects (wind, solar, and battery storage). New additions in

New England have transitioned away from gas-fired generation and demand response and

toward mostly renewable projects that will come online over the next three years.

A capacity scarcity condition under the Pay for Performance (PfP) rules was triggered in

December 2022 for the first time since 2018. Import and nuclear resources performed above

expectations, as measured by their financial obligation, while gas-fired and dual-fuel (gas/oil)

resources performed below expectations.

6.1 Review of Seventeenth Forward Capacity Auction (FCA)

This section provides a closer review of FCA 17, the most recent primary auction held in March

2023.

275

Key Takeaways

Qualified capacity participating in FCA 17 exceeded the Net Installed Capacity Requirement by

over 7 GW (37,386 MW qualified compared to 30,305 MW). About 6 GW left during the auction,

comprised of 2,236 MW of existing resources, 3,106 MW of import resources, and 674 MW of

new resources. Cleared capacity totaled 31,370 MW, leaving a surplus of about 1 GW (1,065

MW) at the auction clearing price of $2.59/kW-month.

Over 770 MW of new resources cleared, with battery storage projects (400 MW) and solar

projects (124 MW) representing the largest entrant types. The vast majority of capacity that left

FCA 17 de-listed for one year under the Dynamic De-list Bid Threshold of $2.59/kW-month. The

largest groups of generators that dynamically de-listed were gas-fired (800 MW), oil-fired (778

MW), and coal-fired (438 MW) generators.

FCA 17 was structurally competitive with limited ability and incentive to exercise market power.

We did not observe bidding behavior that was consistent with the exercise of market power.

Based on the pre-auction mitigations, excess capacity during the auction, and liquidity of

dynamic de-list bids, it is our opinion that a competitive process drove the results of the auction.

275

Further detail on the auction is contained in the IMM’s Winter 2023 quarterly markets report. See https://www.iso-

ne.com/markets-operations/market-monitoring-mitigation/internal-monitor.

2022 Annual Markets Report 169 ISO New England Inc.

ISO-NE PUBLIC

6.1.1 Qualified and Cleared Capacity

The qualified and cleared capacity in FCA 17 compared to Net ICR (blue bars) is illustrated in

Figure 6-1 below. Total qualified and cleared capacity are broken down across three

dimensions: capacity type, capacity zone and resource type.

Figure 6-1: Qualified and Cleared Capacity in FCA 17

In FCA 17, qualified capacity exceeded Net ICR by 7,081 MW, or 23%. New qualified capacity

totaled 1,447 MW, down by just over 200 MW from the FCA 16 value (1,696 MW). New battery

storage projects held the largest portion of new qualified capacity, totaling over 876 MW. The

ISO reintroduced the renewable technology resource (RTR) exemption in FCA 17.

276

Total new

qualified capacity under the RTR exemption reached the maximum auction allowance of 300

MW.

As excess supply declined during the auction rounds, the surplus (above Net ICR) fell from

7,081 MW to 1,065 MW at the end of the auction. The capacity that dropped out of the auction

(6,016 MW) comprised of existing resources that de-listed and new supply resources that

exited the market at prices at or greater than the associated zonal clearing price. The first

orange “Total Cleared” bar (capacity type) illustrates that existing capacity accounted for 98%

of cleared capacity, with just 2% (773 MW) of capacity obligations held by new resources.

Two capacity zones were modelled in addition to Rest-of-Pool: the export-constrained zone of

Northern New England (NNE), and the nested export-constrained zone of Maine. The qualified

and cleared values are illustrated in the second orange bars (by Capacity Zone). There was no

price separation between capacity zones in FCA 17.

276

More information on the RTR exemption for FCA 17 can be found in the FERC filing accepting tariff revisions centered

on the removal of the minimum offer price rule (MOPR).

Net ICR, 30,305

Surplus, 7,081

Exisiting, 35,939

New, 1,447

RoP, 28,156

NNE, 5,520

Maine, 3,710

Gas, 14,324

Oil, 5,731

Import, 3,673

Nuclear, 3,352

Hydro, 3,188

Passive, 2,358

5,000

10,000

15,000

20,000

25,000

30,000

35,000

40,000

45,000

Total Cleared Capacity Type Capacity Zone Resource Type

Total Qualified: 37,386 MW

Active DR, Coal, Solar, Wind,

Other, Battery Storage: 4,760

Active DR,

Net ICR, 30,305

Surplus, 1,065

Exisiting, 30,597

New, 773

RoP, 23,804

NNE, 4,416

Maine, 3,150

Gas, 13,508

Oil, 4,941

Nuclear, 3,352

Hydro, 3,188

Passive, 2,317

5,000

10,000

15,000

20,000

25,000

30,000

35,000

40,000

45,000

Total Cleared Capacity Type Capacity Zone Resource Type

Total Cleared: 31,370 MW

Active DR, Coal, Solar, Wind,

Other, Battery Storage, Import:

4,065

2022 Annual Markets Report 170 ISO New England Inc.

ISO-NE PUBLIC

6.1.2 Results and Competitiveness

In addition to the amount of qualified capacity eligible to participate in the auction, several

other factors contribute to auction outcomes. These factors, including the ISO-provided auction

parameters as well as participant behavior, are summarized in Figure 6-2 below.

On the demand side, the demand curve, Net CONE, and Net ICR are shown in black. FCA 17 was

the fourth auction with a demand curve that relied solely on the Marginal Reliability Impact

(MRI) methodology.

277

On the supply side, the qualified and cleared capacities are shown (solid and dashed red lines,

respectively). The clearing price of $2.59/kW-month is shown at the intersection of the cleared

MW and the demand curve, just below the dynamic de-list bid threshold (DDBT) price of

$2.59/kW-month. Lastly, the blue, green, purple, and orange markers represent the end-of-

round prices, and the corresponding dots depict excess end-of-round supply.

278

Figure 6-2: System-wide FCA 17 Demand Curve, Prices, and Quantities

The auction closed in the fourth round for all capacity zones and interfaces except for slight

price separation at the New Brunswick interface. The fourth round opened with 3,053 MW of

excess capacity above demand at the system level (purple dot) and a starting price equal to the

DDBT price, allowing existing resources to exit the market through dynamic de-list bids.

279

277

The MRI methodology estimates how an incremental change in capacity impacts system reliability at various capacity

levels. Prior to FCA 14, a transitional approach was taken, with the demand curve reflecting a hybrid of the previous linear

demand curve and the new convex-shaped MRI curve. The transition period began with FCA 11 and could last for up to

three FCAs, unless certain conditions relating to Net ICR growth are met, pursuant to Section III.13.2.2.1 of the Tariff.

278

The colored dots and lines move from cooler colors at higher prices and capacity, to warmer colors at lower prices and

less capacity.

279

Excess system capacity only includes import capacity up to the capacity transfer limit.

28,000 30,000 32,000 34,000 36,000 38,000 40,000 42,000

Price ($/kW

-month)

Qualfied Capacity: 37,386 MW

Clearing Values

(31,370, $2.59)

Round 1 Excess MW and Price:

(4,547, $8.59)

Round 2 Excess MW and Price:

(4,036, $5.59)

Round 3 Excess MW and Price:

(3,053, $2.59)

DDBT:

$2.59/kW-mo

Net CONE:

$7.36/kW-mo

NICR:

30,305 MW

2022 Annual Markets Report 171 ISO New England Inc.

ISO-NE PUBLIC

All 3,053 MW excess supply exited the auction in the final round at the round starting price of

$2.59/kW-month. A fully-rationable dynamic de-list bid set the final clearing price of $2.59/kW-

month for all capacity zones and interfaces (except New Brunswick). The market clearing

engine could not clear all new supply offers and de-list bids attempting to leave the auction at

$2.589/kW-month without falling below system-wide demand.

280

FCA 17 Competitiveness

The IMM conducts a pivotal supplier test prior to the auction based on portfolios with existing

capacity relative to the system and local requirements (see Section 2.4.1).

281

However, capacity

conditions change as the auction proceeds (new resources leave, existing capacity de-lists, the

demanded quantity changes) and a supplier that was not pivotal at the start of the auction

(when the IMM made the pivotal status determination) may become pivotal during the auction

as the surplus falls.

282

There were no pivotal suppliers at the start of the auction. As the auction entered its fourth

round, with excess capacity of 3 GW, there was one pivotal supplier with a portfolio slightly

above the system-wide excess capacity. This indicates that the possibility of exercising seller-

side market power was theoretically possible, yet the pivotal supplier’s de-list bids comprised

of less than 14% of all submitted de-list bids in the final auction rounds. Given the surplus

capacity conditions associated with prices below the dynamic de-list bid threshold and the

significant amount of de-list bids submitted by non-pivotal suppliers, it is unlikely that the

pivotal participant could exercise seller-side market power.

In summary, since the auction was structurally competitive there was limited ability and

incentive to exercise market power. Further, we did not observe bidding behavior that was

consistent with the exercise of market power. Based on the pre-auction mitigations, excess

capacity during the auction, and liquidity of dynamic de-list bids, it is our opinion that a

competitive process drove the results of the auction.

6.1.3 Results of the Substitution Auction (CASPR)

In FCA 17, the substitution auction did not proceed, as there were no active demand bids. The

two retiring resources that qualified for the substitution auction did not receive CSOs in the

primary auction, a prerequisite for submitting a demand bid in the substitution auction.

280

Due to minimum rationing limits, the necessary CSO to meet system-wide demand could not be allocated evenly among

all resources attempting to exit the auction. Instead, the market clearing engine rationed one resource’s de-list bid to

retain the necessary amount of CSO for system-wide s upply to meet system-wide demand.

281

The initial pivotal supplier calculation is limited to pre-auction calculations and applies only the mitigation of static de-

list bids.

282

In fact, suppliers that have been deemed pivotal prior to the auction may not be pivotal at the start of the auction (if

the quantity demanded along the sloped demand curve is greater than Net ICR or Local Sourcing Requirement in import-

constrained capacity zones).

2022 Annual Markets Report 172 ISO New England Inc.

ISO-NE PUBLIC

6.2 Forward Capacity Market Outcomes

This section reviews the overall trends in prices and volumes in the FCM; specifically we cover

the FCM credits and charges, trends in the resource mix, and finally, activity in secondary

markets for trading capacity.

Key Takeaways

Capacity market outcomes remained relatively constant year-over-year, driven by an unchanged

FCA 16 and 17 Rest-of-Pool clearing price of $2.59/kW-month. Since FCA 10, capacity prices have

decreased 63% ($4.44/kW-month) and Net ICR has decreased 11% (3,846 MW), resulting in a

record low $0.9 billion in projected capacity payments for capacity commitment period (CCP)

2026-2027 (FCA 17).

New entrants to New England have varied over the past seven auctions, with gas and energy

efficiency resources making up a majority of new capacity additions from FCA 10-13. Since FCA

14, renewable technologies, typically sponsored-policy resources, have made up 56% of all new

generator additions. Major retirements have comprised of nuclear, oil, and gas resources, and an

additional 1,100 MW of new gas resources have been terminated since FCA 10, as they failed to

meet critical development milestones.

December 2022 saw the first Pay-for-Performance event since 2018. The December 2022 event

lasted 17 five-minute intervals (1 hour 25 minutes) and resulted in the transfer of $35.9 million

from under-performing resources to over-performing resources. Import and nuclear resources

received the most over-performance payments, while gas resources were charged the most for

under-performance. A significant amount of gas generation was out of economic merit order in

the day-ahead market due to very high natural gas costs, and in real time could not respond to

high prices during the emerging event due to start time constraints.

Market Rule Changes, Clearing Prices, and Payments

FCM prices are determined by the interaction of capacity supply and demand. When capacity is

in relatively short supply, we expect higher prices. When supply is more abundant, we expect

the opposite. In addition to evolving supply and demand conditions, the capacity market design

has also undergone a number of significant market rule changes in recent years.

A brief history of the FCM: Market design changes

The trend in Rest-of-Pool FCA clearing prices against the backdrop of some of the major design

changes over the reporting period is shown in Figure 6-3 over an eight year period, including

3½ years of actual delivery and 4½ years of future delivery periods. The impact of market rule

changes on auction outcomes is varied, in terms of both level and direction, and the graph is

intended to provide useful context rather than infer a particular relationship between

individual changes and clearing prices.

2022 Annual Markets Report 173 ISO New England Inc.

ISO-NE PUBLIC

Figure 6-3: FCA Clearing Prices in the Context of Market Rule Changes

The minimum offer price rule (MOPR) was implemented in FCA 8, intended to mitigate the

impact of below-cost offers (in practice mostly subsidized offers) on clearing prices. From FCA

9, the Pay-for-Performance (PFP) market rules were implemented and placed a financial

obligation on capacity resources to perform during capacity scarcity events. Combined, the

MOPR and PFP rules encouraged a greater degree of priced participation in the auctions, with

more new and existing resources submitting offers (as opposed to price-taking).

The sloped demand curve was introduced in FCA 9 with the objective of improving price

formation and reducing price volatility, by reflecting the willingness of load to procure a given

amount of capacity (reliability) at various price levels.

283

When the auction clears at a surplus

(over NICR), a sloped demand curve results in a clearing price below Net CONE, which occurred

in all auctions shown above (FCA 10-17). Following a three-year transition period, as the

system demand curve shifted to a non-linear form in FCA 11, the value of excess capacity

decreased, and clearing prices continued to drop.

In FCA 13, two rule changes were implemented with implications for price formation. First, the

ISO retained Mystic 8 and 9 (~1,400 MW) for their fuel security attributes under a cost-of-

service agreement, rather than those resources retiring (which will happen in May 2024). The

Mystic resources were treated as price takers in the FCA, with a resulting downward impact on

prices.

Second, CASPR was introduced as a secondary means to allow sponsored policy resources,

which may not clear due to MOPR, to obtain a CSO from retiring resources. The design objective

283

A linear sloped system demand curve was implemented for FCA 9, but the zonal demand curves remained vertical. In

FCA 10 linear sloped demand curves were used at both the system and zonal levels. More recently, for FCA 11 both sloped

and non - linear demand curves (except for a portion of the system curve) were implemented based on the MRI

methodology.

10 11 12 13 14 15 16 17

(2019-2020) (2020-2021) (2021-2022) (2022-2023) (2023-2024) (2024-2025) (2025-2026) (2026-2027)

RoS Clearing Price ($/kW

-mo)

FCA (Years)

FCA Clearing Price

Minimum Offer Price Review (from CCP 8, to be eliminated from FCA 19)

Pay-for-Performance (from CCP 9)

Mystic Fuel Security RMR

Substitution Auction (CASPR)

Nested Capacity Zones

Sloped Demand Curves - MRI

2022 Annual Markets Report 174 ISO New England Inc.

ISO-NE PUBLIC

was to facilitate the orderly exit and entry of existing and new resources through a secondary

(substitution) auction, while mitigating the price-suppressing impact on the primary auction.

Rest-of-Pool and Zonal FCA Clearing Prices

The changes in new and existing capacity clearing prices for each FCA are illustrated in Figure

6-4 below. The different colored lines represent the price paid to resources in each modeled

capacity zone.

Figure 6-4: Forward Capacity Auction Clearing Prices

Prior to FCA 10, higher capacity prices sent a signal to market participants that load was willing

to pay for more capacity that would improve system reliability.

284

Once new generation

responded to price signals and entered the FCM in greater volumes, clearing prices fell steadily,

beginning at $7.03/kW-month in FCA 10 and continuing to drop in FCA 11.

In FCAs 12 through 14, a downward trend in New England capacity requirements reduced

clearing prices below the dynamic de-list bid threshold (DDBT) – where de-list bids are not

subject to an IMM cost review. In each auction, the closing round started at the DDBT price.

Dynamic de-list bids from existing resources set the system-wide clearing price in FCAs 12-13.

In FCA 15, the Rest-of-Pool (RoP) clearing price increased for the first time in six auctions, as did

price separation among the capacity zones. Significant decreases in capacity in the South

Eastern New England (SENE) capacity zone, led by the retirement of Mystic 8 and 9, meant that

the zone cleared one round earlier and at a higher price than the RoP price. The RoP and

Northern New England capacity zones cleared in the fifth round of the auction, two rounds past

the DDBT threshold of $4.30/kW-month. Due to large amounts of cleared capacity in the

export-constrained Northern New England capacity zone, the zonal clearing rate cleared

slightly below the RoP benchmark.

284

Clearing prices for FCA 8 and 9 were $15.00/kW-month and $9.551/kW-month, respectively.

$2.48

$3.98

$7.03

$5.30

$4.63

$3.80

$2.00

$2.61

$2.59

10 11 12 13 14 15 16 17

(2019-2020) (2020-2021) (2021-2022) (2022-2023) (2023-2024) (2024-2025) (2025-2026) (2026-2027)

$/kW-month

NNE SENE RoP Dynamic Delist Bid Threshold

2022 Annual Markets Report 175 ISO New England Inc.

ISO-NE PUBLIC

In FCA 16, the RoP clearing price was comparable to FCA 15. Price separation also occurred in

FCA 16, but to a much lesser degree (SENE cleared $0.04/kW-mo higher). All zones cleared

immediately after the dynamic de-list bid threshold was reached in the fourth round.

In FCA 17, the clearing price remained constant year-over-year. Price separation only occurred

at the New Brunswick interface, which cleared 177 MW of import capacity at a price of

$2.55/kW-month (not shown). Despite over 3,000 MW of excess supply entering the fourth

round of the FCA, the auction cleared at that round’s starting price (equal to the DDBT).

Payments by Commitment Period

FCM payments are primarily a function of clearing prices and capacity (volume) in each FCA.

Total payments for capacity commitment periods (CCPs) 10-17 are shown in Figure 6-5 below,

along with the key determinants: Rest-of-Pool clearing prices and cleared capacity.

285

The gray

boxes in the bottom graph represent the amount of surplus capacity above the Net ICR that

cleared in each FCA.

Figure 6-5: FCM Payments and FCA Volumes by Commitment Period

Capacity prices have generally trended downwards over the eights auctions, consistent with a

surplus of cleared system capacity, and a decrease in the capacity required (NICR) to meet the 1

day-in-10 year loss of load reliability standard.

Payments peaked with high clearing prices for FCA 10 and fell significantly over the remaining

four auctions as both cleared capacity and prices declined; total payments are projected to

285

The blue bars represent gross FCM payments by commitment period. Payments for CCPs 13-17 are projected payments

based on FCA outcomes, as those periods have not yet been settled. Payments for incomplete periods, CCP 13 through

CCP 17, have been estimated as:    × 1,000 ×   × 12 for each resource.

2022 Annual Markets Report 176 ISO New England Inc.

ISO-NE PUBLIC

decrease by 68% from CCP 10 to 14. Lower clearing prices were consistent with lower demand

(NICR) and system capacity surplus exceeding 1 GW.

In FCA 15, capacity prices and payments rebounded slightly due to an increase in Net ICR and

the retirements of major supply resources (Mystic 8 and 9). Higher zonal and RoP clearing

prices in FCA 15 increased projected payments to $1.3 billion. In FCAs 16 and 17, projected

payments declined with lower clearing prices and capacity; total payments for CCP 16 and 17

are projected to be $1.0 billion and $0.9 billion, respectively – among the lowest in the FCM’s

history.

Charges – or reductions in base FCM payments – are grouped together and have been negligible

in the context of annual payments (green bar is not visible). Failure-to-Cover charges have

affected a small portion of FCM payments since implementation in 2019, with resources only

charged a maximum of about $1.0 million to date. Pay-for Performance (PFP) are essentially a

transfer from under-performing capacity resources (charges) to over-performing resources

(credits) and are expected to net to zero. PFP outcomes are covered in Section 6.2.2.

6.2.1 Capacity Resource Mix

This section discusses trends and major changes in cleared capacity since FCA 10, including a

review of major retirements and new entry driving changes to capacity supply. There are three

categories of capacity resources that can participate in the FCM: generation, demand response

and import resources. The share of these categories in the context of total capacity (gray box),

with generation broken down by fuel type ,and demand response by passive (primarily energy

efficiency) and active (price-responsive) resources is illustrated in Figure 6-6.

Figure 6-6: Capacity Mix by Fuel Type

Since FCA 10, total cleared capacity has trended down and has followed the downward trend in

Net ICR. Despite the decrease in cleared capacity, the auctions have maintained surplus cleared

capacity of at least 1 GW (~4% on average) above NICR in each of the past eight auctions.

2022 Annual Markets Report 177 ISO New England Inc.

ISO-NE PUBLIC

Gas-fired generation capacity has also trended downwards, driven by large generator

retirements and terminations.

286

Material reductions include the termination of Burrillville

Energy Center (485 MW in FCA 12)

287

, retirement of Mystic 8 and 9 (1,413 MW in FCA 15), and

termination of Killingly Energy Center (632 MW in FCA 16)

288

, together accounting for over 2.5

GW.

Other notable changes were to passive demand resources, solar generation, and battery storage

projects. Between FCAs 9 and 13, capacity from passive demand sharply increased, reaching a

peak of 3,355 MW, in line with significant utility investment supported by state policy goals to

increase energy efficiency. Since FCA 13, however, passive demand capacity has decreased

down to 2,317 MW as existing capacity has amortized over time.

289

Solar capacity has jumped from 20 MW in FCA 9 to 710 MW in FCA 17. More efficient solar

technology has reduced project costs and the renewable technology resource exemption helped

solar projects enter the capacity market.

290

Battery storage projects are a significant new

entrant to New England’s capacity market. FCA 17 added 400 MW of new battery storage

capacity to the existing 700 MW. Since being introduced to the market in FCA 13, battery

storage projects have increased their share of capacity from 5 MW to 1,100 MW. The total share

of cleared capacity for solar and battery storage resources reached 6% in FCA 17. Wind capacity

comprises just 1% of the FCM (~360 MW), but has increased by 227 MW since FCA 10.

Retirement of Capacity Resources

Resources can submit retirement bids in the FCA to fully exit both the capacity and the energy

markets. Major retired generating resources, defined here as capacity exceeding 50 MW, from

FCA 10 through FCA 17 are shown in Table 6-1 below.

286

Terminated resources clear capacity in previous FCAs but have their capacity revoked from FCAs following termination.

287

In September 2018, ISO-NE filed to terminate the 485 MW CSO of the Burrillville Energy Center, which was accepted by

the Commission. Per the filing, the project sponsor had not made sufficient progress to achieve Clear River Unit 1’s critical

path schedule milestones. With the insufficient progress, the commercial operation date for Clear River Unit 1 was more

than two years beyond June 1, 2019, which is the start of the Capacity Commitment Period in which the resource first

obtained a CSO.

288

In November 2021, ISO-NE filed to terminate the 632 MW CSO of Killingly Energy Center, which was accepted and

upheld by the Commission. Per the filing, the project sponsor had not made s ufficient progress to achieve Killingly Energy

Center’s critical path schedule milestones. With the insufficient progress, the co mmercial operation date for Killingly

Energy Center was more than two years beyond June 1, 2022, which is the start of the Capacity Commitment Period in

which the resource first obtained a CSO.

289

Energy efficiency assets are modelled with a Measure Life, or expected length of performance. The qualified capacity of

a passive demand resource, most commonly an aggregation of energy efficiency assets, will decrease over time as

individual assets reach their Measure Life.

290

The renewable technology resource (RTR) exemption allowed a s et MW of state-sponsored renewable resources into

the FCM without being subjected to buyer-side mitigation rules. In almost all cases, buyer-side mitigation rules denied

state-sponsored resources entry into the primary FCA.

2022 Annual Markets Report 178 ISO New England Inc.

ISO-NE PUBLIC

Table 6-1: Generating Resource Retirements over 50 MW

FCA #

(Commitment Period)

Resource Name

Fuel Type

Capacity Zone

FCA MW

FCA 10 (2019/20)

Pilgrim Nuclear

Nuclear

SEMA

677

FCA 12 (2021/22)

Bridgeport Harbor 3

Oil

Connecticut

383

FCA 13 (2022/23)

Mystic 7

Oil

NEMA/Boston

575

FCA 14 (2023/24)

Yarmouth 1

Oil

Maine

FCA 14 (2023/24)

Yarmouth 2

Oil

Maine

FCA 14 Total (resources > 50 MW)

101 MW

FCA 15 (2024/25)

Mystic 9

Gas

NEMA/Boston

710

FCA 15 (2024/25)

Mystic 8

Gas

NEMA/Boston

703

FCA 15 (2024/25)

West Springfield 3

Gas

WCMA

FCA 15 (2024/25)

CDECCA

Gas

Connecticut

FCA 15 Total (resources > 50 MW)

1,560 MW

FCA 16 (2025/26)

Potter 2 CC

Gas

SEMA

Total Major Retirements since FCA 10

3,368 MW

Energy policy and market dynamics have been cited as reasons leading to increased retirement

pressure on nuclear, coal- and oil-fired generators. Increasing emissions prices and

environmental regulations have led to increased production and capital costs. Many of the

retiring resources are older resources that may require environmental upgrades or major

overhauls.

FCAs 15 and 16 saw the first large retirements of older natural gas-fired generators as the

market shifted toward more efficient gas technology and brought in large volumes of renewable

projects. There were no retirements over 50 MW in FCA 17 despite continued low capacity

prices.

New Entry of Generation Capacity Resources

There has been a significant amount of new entry generation capacity over the past eight

auctions, totaling about 5.4 GW (averaging over 680 MW per auction). New generation capacity

by fuel type since FCA 10 is shown in Figure 6-7. The shaded red bars indicate gas capacity that

was ultimately terminated due to not realizing commercial operation.

2022 Annual Markets Report 179 ISO New England Inc.

ISO-NE PUBLIC

Figure 6-7: New Generation Capacity by Fuel Type

Natural gas-fired resources comprised the majority of new additions between FCA 10 and FCA

13, totaling ~2,530 MW. However, over 1,100 MW of the new additions were ultimately

terminated as the Burrillville Energy Center and Killingly Energy Center failed to meet key

development milestones. FCA 10 saw the largest amount of new delivered generation entry,

with an additional 960 MW of new natural gas-fired capacity including Bridgeport Harbor 5

(484 MW) and Canal 3 (333 MW). In FCA 14, no new, large gas-fired resources cleared in the

auction. Instead, an increase in state-sponsored solar resources and new wind resources were

the primary sources of new cleared generation.

In FCA 15, new gas-fired capacity entered the market again, driven by a 334 MW repowering of

the existing Ocean State Power facility. Battery storage projects also cleared a significant

amount (596 MW) of new capacity for the first time in an FCA.

In the past two auctions, new entry has largely comprised state sponsored policy resources. In

FCA 16, battery storage projects continued to enter (but only 100 MW), while the largest

increase came from new solar projects (208 MW).

In FCA 17, battery storage projects made up the largest share of new capacity (400 MW). Solar

and wind cleared 120 MW and 100 MW of new capacity, respectively. A decreased minimum

offer price for battery storage projects and the reintroduction of the renewable technology

resource exemption facilitated the entry of new renewable capacity.

Exit of Import Capacity Resources

From FCA 10-16, the volume of cleared import capacity has fluctuated from between 1,000 MW

to 1,500 MW across the four external ties in the FCM, with most capacity at the New York AC

lines (50%) and Phase II (30%). However, cleared import capacity was down significantly in the

most recent auction, FCA 17, at just 567 MW comprising 390 MW at New York and 177 MW at

New Brunswick. This section reviews changes in cleared import capacity.

$10

$12

$14

$16

200

400

600

800

1,000

1,200

1,400

1,600

10 11 12 13 14 15 16 17

(2019-2020) (2020-2021) (2021-2022) (2022-2023) (2023-2024) (2024-2025) (2025-2026) (2026-2027)

FCA (Years)

Gas Hydro Wind Solar Battery Storage Nuclear

Clearing Price (right axis)

Burrillville Energy Center (485 MW)

Killingly Energy Center (632 MW)

1,000

2,000

3,000

4,000

5,000

6,000

Total

2022 Annual Markets Report 180 ISO New England Inc.

ISO-NE PUBLIC

Import resources using existing tie line capacity are deemed “new capacity” in the market rules.

However, for the purpose of this review, and indeed for a market power mitigation perspective,

import resources are similar to existing capacity resources. We therefore view these resources

as essentially de-listing capacity when they exit the market. Cleared CSO MW and de-list MW

totals from import capacity resources from FCA 10–17 are shown in Figure 6-8 below. The

import limit represents the maximum amount of import capacity that can clear over New

England’s external interfaces in each FCA.

Figure 6-8: FCA Import De-list MW

From FCA 10 – 16, import resources de-listed an average of 2,549 MW, or 65% of their qualified

capacity. A key factor driving import resource behavior is the expected price they can receive in

their neighboring control areas (“the next best alternative”). For example, if there is a higher

likelihood that New York capacity prices will be higher, one would expect less capacity to clear

across New York North. In FCA 17, we observed the largest percentage of import capacity de-

listing (85%). Despite an import limit of 1,600 MW, import resources cleared the lowest

amount of CSO over the reporting period at just 567 MW.

6.2.2 Pay-for-Performance Outcomes

One Pay-for-Performance (PfP) event occurred in 2022, the first since September 2018 and only

the second event since the rules were introduced in June 2018. On December 24, from 4:40 PM

to 6:00 PM (17 five-minute intervals), system-wide energy supply failed to meet customer load

plus the 30-minute reserve requirement, triggering a capacity scarcity condition. That day,

Winter Storm Elliot brought in lower-than-expected temperatures and influenced sudden,

unexpected generator outages. The supply-side disruptions pushed New England’s grid into its

first PfP event since 2018.

During the scarcity condition, only 19,995 MW of the contracted 30,114 CSO MW provided

energy and reserves to New England. The average load and reserve requirement for the scarcity

condition was 20,283 MW, resulting in a balancing ratio of 67% - meaning that capacity

resources were financially obligated to provide two-thirds of their CSO position in the form of

energy or reserves. Under-performance resulted in $35.9 million in charges, which were

1,450

1,235

1,217

1,188

1,059

1,487

1,504

567

1,570

2,119

2,784

3,772

3,388

2,145

2,061

3,106

3,020

3,355

4,001

4,960

4,447

3,632

3,565

3,673

1,463

1,750

1,680

1,700

1,719

1,965

1,870

1,600

1,000

2,000

3,000

4,000

5,000

6,000

10 11 12 13 14 15 16 17

(2019-2020) (2020-2021) (2021-2022) (2022-2023) (2023-2024) (2024-2025) (2025-2026) (2026-2027)

FCA (years)

Cleared CSO Total Delist Import Limit

2022 Annual Markets Report 181 ISO New England Inc.

ISO-NE PUBLIC

transferred to over-performing resources. Resource performance by PfP charges and credit are

aggregated by fuel type, and ranked, in Figure 6-9.

291

Figure 6-9: Total PfP Payments by Resource Type (December 2022)

Import resources were the best-performing resource type, with over $6.5 million (74%) of their

total $8.81 million in payments going to import transactions without an associated CSO.

Following import resources, nuclear resources received the second highest net payment; these

generators operate as baseload resources at their full CSO MW during all hours and typically

over-perform relative to their CSO. Both import resources and nuclear resources were two of

the three best performing resource groups during the 2018 PfP event as well.

Gas- and gas/oil-fired resources were the worst performers. While cold weather led to forced

outages for some of the gas fleet, the leading cause of gas resource under-performance was high

gas prices. Spot natural gas prices climbed to over $30/MMBtu, resulting in gas generation

pushed out of economic merit order and, instead, relatively cheaper oil generation was

scheduled in the day-ahead energy market.

292

Once system conditions tightened in real time and

system-wide LMPs increased significantly, operating limitations prevented many gas-fired

resources from coming online in time to contribute to meeting the load and reserve

requirement.

6.2.3 Secondary Forward Capacity Market Results

After each FCA and up to just before the commitment period, resources can adjust their CSOs

through a variety of reconfiguration auctions in the secondary market. The average annual

volume by secondary market products (stacked bars corresponding to the left axis) and volume

291

In this figure, “Dual” refers to dual-fuel (gas/oil) assets and “CC” refers to combined-cycle assets.

292

More information on fuel prices in Q4 2022 can be found in Section 1.2.2.

$8.81

$5.60

$2.46

$1.49

$1.34

$1.29

$1.07

$0.45

$0.06

$0.03

$0.02

-$0.04

-$0.47

-$1.02

-$5.89

-$15.19

$4.08

$10.92

$7.02

$5.63

$9.15

$0.94

$18.42

$3.04

$0.18

$0.01

$2.01

$0.09

$16.56

$1.70

$22.64

$36.46

-$20

-$10

$10

$20

$30

$40

Payments (Millions)

Resource Type

Base FCM Supply Credits (Total: $141.1m)

PfP Credits/Charges (Total: $35.9m)

PfP Winners

PfP Losers

2022 Annual Markets Report 182 ISO New England Inc.

ISO-NE PUBLIC

as a percentage of cleared volume in the corresponding FCA (red line corresponding to the right

axis) are shown in Figure 6-10 below.

293

Figure 6-10: Traded Volumes in Reconfiguration Auctions

Historically, traded volumes in the secondary markets have been much lower than in the

primary auctions. Since CCP 9, the trade volumes as a percentage of FCA volumes have

increased steadily, ranging from 6% in CCP 10 to over 10% in CCP 13. The majority of

secondary trading occurs during annual reconfiguration auctions and monthly reconfiguration

auctions. CCP 13, beginning in June 2022, saw the largest portion of secondary trading in annual

reconfiguration auctions (dark green), reaching almost 2,700 MW.

293

Volumes are shown as average annual weighted values. A monthly product gets a weight of 1/12th; an annual product

a weight of 1, etc.

10%

12%

1,000

2,000

3,000

4,000

5,000

6,000

9 10 11 12 13

(2018-2019) (2019-2020) (2020-2021) (2021-2022) (2022-2023)

Percent of FCA Cleared MW

Commitment Period (Years)

Annual Bilateral (LA) Monthly Bilateral (LA)

Annual Reconfiguration (LA) Monthly Reconfiguration (LA)

Percent of FCA Volume (RA)

2022 Annual Markets Report 183 ISO New England Inc.

ISO-NE PUBLIC

Section 7

Ancillary Services

This section reviews the performance of ancillary services in ISO New England’s forward and

real-time markets. While there are six main types of ancillary services (listed below), this

section focuses on real-time operating reserves, forward reserves, and regulation services.

The six main types of ancillary services are:

 Real-time operating reserves represent additional generating capacity that is available to

respond to unexpected contingencies (such as the unexpected loss of a generator or

transmission line) during operation of the real-time energy market (Section 7.1).

 Forward reserves represent the procurement of offline operating reserves in advance of the

delivery period; that is, the ability to start and ramp quickly in the event of system

contingencies (Section 7.2).

 Regulation service is provided by resources that alter their energy output over very short

time intervals (minute-to-minute) to balance supply and demand and maintain system

frequency levels in the real-time energy market (Section 7.3).

 The Winter Reliability Program was implemented by the ISO from 2013 to 2018 to remedy

fuel supply issues that threatened reliability. The program paid market participants to

purchase sufficient fuel inventories (oil or LNG) or provide additional demand response

during the winter months. The program ended after winter 2018.

294

 Voltage support helps the ISO maintain an acceptable range of voltage on the transmission

system, and is necessary for the reliable flow of electricity. The ISO regulates voltage

through reactive power dispatch, and the generators that provide this service receive

voltage support payments.

295

 Blackstart service is provided by generators that are able to start quickly without outside

electrical supply. The ISO selects and compensates strategically located generators for

providing blackstart service. This service is necessary to facilitate power system restoration

in the event of a partial or complete system shutdown.

Ancillary service costs by submarket are shown in Figure 7-1 below.

296

The gray boxes above

each bar show the total ancillary service cost for each year.

294

A similar out of market program known as known as Interim Compensation Treatment (ICT) is scheduled to be

implemented for Winters 2023/24 and 2024/25.

295

Trans mission customers who use regional network service or through-or-out service incur voltage support charges. If

the ISO commits a resource for voltage support in the energy market and it does not recover its effective offer, the

resource is eligible for NCPC. The ISO Tariff contains detailed rules regarding compensation for voltage support. See

Schedule 2 of Section II: Open Access Transmission Tariff (the OATT), available at: https://www.iso-ne.com/static-

assets/documents/regulatory/tariff/sect_2/oatt/sect_ii.pdf

296

The Voltage Service category includes payments for capacity costs, lost opportunity costs, costs of energy consumed,

and costs of energy produced.

2022 Annual Markets Report 184 ISO New England Inc.

ISO-NE PUBLIC

Figure 7-1: Ancillary Service Costs by Product

Overall, ancillary costs in 2022 were the highest over the last five years, totaling $177 million,

an increase of over 60%, or $68 million, on 2021 ($109 million). The largest increase was in

forward reserve costs, which rose by $44.2 million due to increased supply offer pricing by

participants with resources capable of providing ten-minute non-spinning reserves. Regulation

and net real-time reserve costs rose by $13.0 and $11.8 million, respectively, primarily due to

higher energy prices. Blackstart and voltage service costs were similar to 2021 costs. The only

year with Winter Reliability Program payments was 2018, since the program expired in March

2018.

7.1 Real-Time Operating Reserves

The following section reviews real-time operating reserve products and outcomes. The first

subsection (7.1.1) summarizes the level of real-time reserve payments and explores the impact

that fast-start pricing (FSP).

297

The second subsection (7.1.2) explores the frequency and

magnitude of real-time reserve prices, including the frequency of reserve constraint penalty

pricing.

297

Fast-start pricing (FSP) was implemented in March 2017 to improve price formation and performance incentives in the

real-time energy market. See Section 5.5 of our Summer 2017 Quarterly Markets report for detail on fast-start pricing:

https://www.iso-ne.com/static-assets/documents/2017/12/2017-summer-quarterly-markets-report.pdf

$30

$10

$11

$23

$41

$37

$23

$19

$63

$30

$24

$20

$25

$38

$16

$23

$21

$19

$14

$21

$29

$35

$154

$114

$103

$109

$177

100

150

200

2018 2019 2020 2021 2022

$ millions

Net Real-Time Reserve Net Forward Reserve Regulation

Winter Reliability Program Voltage Service Black Start

2022 Annual Markets Report 185 ISO New England Inc.

ISO-NE PUBLIC

Key Takeaways

During 2022, extreme weather events in summer and winter led to higher total reserve

payments and prices, particularly for ten and thirty-minute non-spinning reserves (TMNSR and

TMOR). Offline reserve payments totaled $8.4 million, compared to $3.7 million in 2021. Nearly

70% of the non-spinning reserve payments occurred during the extreme weather events in

summer and winter.

While the frequency of spinning reserves (TMSR) declined in 2022 due to lower requirements,

payments increased due to higher energy prices; non-zero TMSR prices increased by 88%.

For the first time since 2018, the reserve constraint penalty factors (RCPFs) for non-spinning

reserve products bound and there was a deficit of reserves. This occurred due to supply issues

on December 24, and led to capacity scarcity conditions during 17 five-minute intervals (almost

1.5 hours).

7.1.1 Real-Time Operating Reserve Payments

The ISO maintains system-wide and local real-time operating reserves to respond to

unexpected losses. There are three products for the system: ten-minute spinning reserves

(TMSR), ten-minute non-spinning reserves (TMNSR), and thirty-minute operating reserves

(TMOR). Local reserve zones only have TMOR.

Real-time operating reserve payment totals may change significantly on a percentage basis

from year to year because of changes in operating reserve requirements, fuel costs, and system

conditions. However, total payments are relatively small compared to overall energy market

and capacity market payments.

The payments presented in Figure 7-2 below are a measure of the value of real-time reserves.

The height of each bar represents the payments by system and local reserve products, with each

reflecting the product of total real-time reserve designations and market clearing prices. The

black diamond shows total net real-time reserve payments.

298

The diamond will be lower than the height of the bars when real-time payments are removed to ensure resources paid

in the forward reserve market are not paid again in the real-time reserve market.

2022 Annual Markets Report 186 ISO New England Inc.

ISO-NE PUBLIC

Figure 7-2: Real-Time Reserve Payments

Total gross real-time reserve payments in 2022 totaled $27.3 million, approximately 0.2% of

total wholesale market costs. Reserve payments doubled (up by $13.6 million) compared to

2021 ($13.7 million). Real-time payments were reduced by $4.5 million, for resources holding a

forward reserve market obligation resulting in net reserve payments of $22.7 million (or 108%

higher than in 2021).

299

An increase in extreme weather events in summer and winter led to more non-zero non-

spinning reserve frequency and higher re-dispatch costs in the co-optimization process,

resulting in higher reserve clearing prices. Specifically, during twelve days in summer and one

day in winter, gross reserve payments were $15.0 million, or 55% of total payments.

300

Due to increased reserve prices, spinning reserve (TMSR) payments increased by almost 90%

(by $8.9 million) compared to 2021 ($10.0 million). Payments increased despite a lower

frequency (down by 12%) of non-zero TMSR pricing intervals.

301

Payments for non-spinning

reserves increased due to the tight system conditions during the system events in summer and

winter. TMNSR ($4.6 million) and TMOR ($3.8 million) payments increased by 62% and 329%

year-over-year, respectively; 70% of the payments were accrued during the system event days

in July, August, and December.

Impact of Fast-Start Pricing on Operating Reserve Payments

An objective of the fast-start pricing (FSP) rules is to better reflect the short-term operating cost

of fast-start generators in real-time market pricing.

302

On average, FSP has increased the price of

energy.

Consequently, the opportunity costs to provide reserves has increased as well, resulting

299

Section 2.6.1 discusses FRM payments. For reference, net FRM payments in 2022 were roughly $63.2 million.

300

See Section 3.2.7 for more information on these system events.

301

Section 7.1 explains the decline in TMSR pricing frequency.

302

The impact of fast-starting pricing on real-time energy prices is discussed in Section 3.1.2.

$10

$15

$20

$25

$30

$35

$40

2018 2019 2020 2021 2022

$ millions

TMSR TMNSR TMOR

NEMA/Boston TMOR CT TMOR SWCT TMOR

Net RT Reserve Payments

2022 Annual Markets Report 187 ISO New England Inc.

ISO-NE PUBLIC

in higher reserve prices. Figure 7-3 below shows the impact of fast-start pricing on real-time

reserve payments over the past five years.

303

Figure 7-3: Impact of Fast-Start on Reserve Payments

Without fast-start pricing, real-time reserve payments would have been approximately $9.0

million in 2022, compared to the actual amount of $26.9 million.

304

Fast-start pricing has had a

significant impact on real-time reserve payments, increasing payments by over 60% ($57

million) between 2018 and 2022. A detailed assessment of the impact of fast-start pricing in

provided in Section 3.1.2 of this report.

7.1.2 Real-Time Operating Reserve Prices and Frequency

Average reserve prices are a function of two factors: frequency and magnitude. Frequency

represents the number of intervals with non-zero reserve pricing. Magnitude is the average

real-time reserve price for all non-zero five-minute pricing intervals. Figure 7-4 below shows

both the frequency (left panel) and magnitude (right panel) of non-zero reserve prices by

reserve product.

303

We approximate the impact of fast-start pricing by comparing prices from the dispatch and pricing software solutions.

The dispatch solution acts as a proxy for pricing outcomes prior to fast-start pricing rules.

304

The total payments ($26.9 million) are an approximation using the dispatch and pricing software solutions. These values

differ from the settlement payments ($27.3 million) discussed above due processes in-between the pricing run and

settlement outcome.

$17.4

$4.2

$3.0

$3.2

$9.0

$15.4

$5.9

$7.8

$10.5

$17.9

$32.8

$10.1

$10.8

$13.7

$26.9

$10

$15

$20

$25

$30

$35

2018 2019 2020 2021 2022

$ millions

Without Fast-Start Pricing With Fast-Start Pricing

Total RT reserve

payments

FSP impact

2022 Annual Markets Report 188 ISO New England Inc.

ISO-NE PUBLIC

Figure 7-4: Frequency and Average of Non-Zero Reserve Prices

In 2022, TMSR pricing was non-zero for 13% of all intervals, down from 15% in 2021 (left-

panel). Therefore, in 87% of intervals no resource had an opportunity cost to provide reserves

(i.e., prices were $0/MWh). From 2021 to 2022, the average TMSR margin increased due to

lower TMSR requirements during the second half of 2022, which was discussed in Section 1.2.1.

TMSR prices averaged $19.09/MWh in non-zero pricing intervals, an increase from

$10.15/MWh in 2021 (right panel). The higher non-zero TMSR price was driven by higher fuel

prices, energy prices, and tight system conditions.

The frequency of the non-spinning products (TMNSR and TMOR) pricing was below 0.6% in all

intervals for 2021 and 2022. There were only 28 hours of TMNSR and 21 hours of TMOR in

2022, compared to 24 and 7 hours, respectively in 2021. The system events in 2022 had a

positive impact on reserve prices, and a muted impact on reserve frequency. Similar to average

TMSR prices, average TMNSR and TMOR prices increased year-over-year. Additionally, in 2022

there were more instances when reserve costs were capped at their Reserve Constraint Penalty

Factors (RCPFs). In 2022, there was no local TMOR pricing.

We can evaluate the impact of reserve frequency and magnitude by looking at real-time reserve

prices for all pricing intervals in Figure 7-5 below.

$50

$100

$150

$200

$250

$300

$350

2018 2019 2020 2021 2022

$/MWh

Price

10%

15%

20%

25%

2018 2019 2020 2021 2022

% of hours

Frequency

NEMA/Boston, CT, and SWCT

TMOR underneath TMOR

2022 Annual Markets Report 189 ISO New England Inc.

ISO-NE PUBLIC

Figure 7-5: Average Real-Time Reserve Prices for all Pricing Intervals

The increase in average prices across all hours for all reserve products drove the 108% year-

over-year increase in reserve payments. Average TMSR prices during all pricing intervals (i.e.,

zero- and non-zero pricing intervals) increased by 66%, from $1.55/MWh in 2021 to

$2.57/MWh in 2022. The increase in the magnitude of non-zero prices (88%) was partially

offset by the decrease in the frequency (12%). In addition, as non-spinning reserve prices are

included in the price of TMSR, the increase in TMSR prices reflects increases in both TMOR and

TMNSR prices.

Reserve Constraint Penalty Factors (RCPFs)

RCPFs for reserve products are triggered due to either a shortage of available capacity to meet

the reserve requirements or re-dispatch costs that exceed RCPF values. The percentage of five-

minute intervals during which the RCPFs were triggered for each reserve constraint are shown

in Figure 7-6 below.

$0.00

$0.50

$1.00

$1.50

$2.00

$2.50

$3.00

$3.50

2018 2019 2020 2021 2022

$/MWh

TMSR TMNSR SWCT TMOR CT TMOR NEMA/Boston TMOR TMOR

NEMA/Boston, CT, and SWCT

TMOR underneath TMOR

2022 Annual Markets Report 190 ISO New England Inc.

ISO-NE PUBLIC

Figure 7-6: Reserve Constraint Penalty Factor Activation Frequency

Overall, the low level of reserve scarcity is consistent with high reserve margins in recent years.

In 2022, the RCPF for TMSR bound 0.6% of total intervals, or roughly two days of the year. The

TMSR RCPF bound most frequently due to its relatively low RCPF value ($50/MWh). The

replacement TMOR RCPF, which has a lower penalty factor than TMOR ($250/MWh vs

$1,000/MWh) bound in 43 five-minute intervals (0.02% of total intervals or ~ 3.5 hours).

During the December 24 system event, which was caused by unexpected outages and reduced

imports that led to tight system conditions, there were 17 intervals (85 minutes) where the

minimum TMOR RCPF ($1,000/MWh) bound and one interval where the TMNSR RCPF

($1,500/MWh) bound. There were 17 capacity scarcity condition intervals under the capacity

markets pay-for-performance rules, which is discussed further in Section 6.2.

0.00%

0.20%

0.40%

0.60%

0.80%

1.00%

TMSR TMNSR TMOR Replacement

TMOR

NEMA/Boston

TMOR

SWCT TMOR CT TMOR

% of RCPF activation for all intervals

2018 2019 2020 2021 2022

2022 Annual Markets Report 191 ISO New England Inc.

ISO-NE PUBLIC

7.2 Forward Reserves

This section assesses outcomes in the forward reserve market (FRM), which procures non-

spinning reserve products in advance of summer and winter seasons. Specifically, in this section

we review the trends in auction demand quantities over the past five years (7.2.1), auction

results (0), and the resulting levels of forward reserve payments, noting the impact that

penalties have had on these payments (7.2.3).

Key Takeaways

Over the last few years, the quantities of reserve capacity required in the FRM have been

reasonably stable at the system level. Additionally, in the eight most recent auctions

(Summer 2019 through Winter 2022-23), there has been no need for a local requirement in

any local reserve zone due to transmission improvements.

FRM auction prices for TMOR have generally been below $2,000/MW-month, and in some

auctions have been below $1,000/MW-month over the review period. While TMNSR auction

prices frequently have been below $2,000/MW-month, the most recent auctions (Summer

2022 and Winter 2022-23) have seen increasing price levels. The Summer 2022 auction, in

particular, had TMNSR prices of $7,386/MW-month, reflecting a significant increase in

participant offer prices. The IMM reviewed participant offers in the Summer 2022 auction

and did not find evidence of the exercise of market power.

FRM payments significantly exceed real-time operating reserve payments ($63 million vs.

$27 million in 2022), reflecting significantly different structures between these forward and

spot markets. While FRM payments have declined significantly from 2018 to 2021, increased

auction clearing prices for TMNSR in 2022 resulted in a large increase (233%) in FRM

payments.

7.2.1 Market Requirements

The FRM auction is intended to ensure that there are adequate reserves to meet ten-minute

non-spinning reserve (TMNSR) and thirty-minute operating reserve (TMOR) requirements.

305

The FRM requirements for the New England control area are based on the forecast of the first

(TMNSR) and second (TMOR) largest contingency supply losses for each procurement period.

The system-wide requirements from Summer 2018 through Winter 2022-23 are shown in

Figure 7-7 below.

305

Reserve requirements represent the minimum quantity of unloaded capacity needed to meet the system’s first and

second contingencies.

2022 Annual Markets Report 192 ISO New England Inc.

ISO-NE PUBLIC

Figure 7-7: Forward Reserve Market System-wide Requirements

Over the past ten auctions, the TMNSR purchase amount has represented the expected single

contingency of the HQ Phase II Interconnection. The TMOR purchase amount has represented

the expected single second contingency of either Mystic 8/9 or Seabrook.

306

Therefore, the

requirements have been relatively consistent around 1,300-1,600 MW for TMNSR and around

800 MW for TMOR. The reasonably small fluctuations in seasonal requirements reflect seasonal

variation in expected capabilities for Phase II and Mystic 8/9 (or Seabrook), and relatively

stable expectations for non-spinning reserve needs (affecting TMNSR), generator performance

when called upon for system contingencies (affecting TMNSR), and replacement reserve needs

(affecting TMOR).

Some zones are constrained in terms of how much power they can import from other zones and

so, instead of having a single reserve requirement for each reserve product the entire system,

the ISO identifies requirements at a zonal level and at the system level. The aggregate reserve

requirements for the past ten auctions for the import-constrained reserve zones of Connecticut,

NEMA/Boston, and Southwest Connecticut are shown in Figure 7-8 below. The local 30-minute

operating reserve requirement can be met through 10- or 30-minute reserve supply offers in

each local reserve zone.

306

As noted in the ISO’s assumptions memoranda for the individual FRM auctions, the FRM system requirements also may

be biased up or down and, in the case of TMOR, include a replacement reserve adjustment. See: https://www.iso-

ne.com/markets-operations/markets/reserves/?document-type=Forward Reserve Market Assumptions

200

400

600

800

1,000

1,200

1,400

1,600

1,800

TMNSR TMOR

2022 Annual Markets Report 193 ISO New England Inc.

ISO-NE PUBLIC

Figure 7-8: Aggregate Local Forward Reserve (TMOR) Requirements

The summer and winter procurement periods have experienced a significant reduction in local

FRM requirements. Only NEMA/Boston and Southwest Connecticut needed a local reserve

requirement in the 2018 auctions. Since then, external reserve support (ERS) has been

sufficient to alleviate the need for local reserve requirements. This results from a considerable

increase in ERS for Connecticut and Southwest Connecticut due mainly to transmission

upgrades. Similarly, transmission upgrades in NEMA/Boston have increased ERS for that area,

resulting in no need for a local requirement in the last four summer and winter periods. The

ERS in NEMA/Boston has typically exceeded the local second contingency by more than 1,000

MW in these auctions.

307

External reserve support (ERS) refers to the ability of a local reserve zone to obtain operating reserves from other

reserve zones. The ERS reflects the amount of available transfer capability on the transmission interface for the local

reserve zone. See Manual M-36, Forward Reserve and Real-Time Reserve, Section 2.2.

100

150

200

250

300

350

Connecticut Southwest Connecticut NEMA Boston

Summer Winter

2022 Annual Markets Report 194 ISO New England Inc.

ISO-NE PUBLIC

7.2.2 Auction Results

This section covers FRM auction pricing outcomes from the Summer 2018 auction through the

Winter 2022-23 auction. The TMNSR and TMOR clearing prices by reserve zone for each

auction are shown in Figure 7-9 below.

308

Figure 7-9: Forward Reserve Prices by FRM Procurement Period

System-wide TMOR pricing was either stable or declining in the Summer and Winter auctions,

respectively, throughout the review period. This decline is consistent with lower auction offer

prices by participants.

309

TMNSR prices, however, increased in 2022 for both the Summer and

Winter auctions; the Summer 2022 auction prices increased dramatically, rising from

$1,150/MW-month in Summer 2021 to $7,386/MW-month in Summer 2022. TMNSR prices

also increased in the Winter 2022-23 auction relative to the Winter 2021-22 auction prices,

rising from $740/MW-month to $2,500/MW-month. The TMNSR prices in 2022 reflect

significantly higher offer prices. When reviewing participants’ auction offers, the IMM has not

found evidence of the exercise of market power.

In NEMA/Boston, limited forward reserve supply resulted in high auction clearing prices for the

Summer 2018 and Winter 2018/19 auctions. In fact, there was a shortfall of supply in this zone

in the Winter 2018-19 auction, which resulted in the clearing prices for this zone being set at

the offer cap. However, a local reserve requirement for NEMA/Boston has not been needed for

308

Forward reserve auction clearing prices are calculated for each reserve product in each reserve zone. The requirements

for the Connecticut reserve zone can be fulfilled by reserve offers for the Southwest Connecticut reserve zone. When

supply offers for forward reserve are not adequate to meet a requirement, the clearing price for that product is set to the

offer price cap. When enough supply is offered under the price cap to meet the requirement in a particular zone, the

auction clearing price for that product is set equal to the price of the marginal supply offer.

309

In general, a number of factors can affect TMNSR and TMOR clearing prices, including: (1) offer prices for TMNSR and

TMOR, (2) the ability to substitute lower-priced TMNSR supply for TMOR supply when there is low-priced TMNSR supply in

excess of the TMNSR requirement, and (3) cleared high-priced TMOR supply needed for local requirements that reduces

the amount of TMOR supply needed to meet the rest-of-system requirement, among other factors.

1,000

2,000

3,000

4,000

5,000

6,000

7,000

8,000

9,000

10,000

$/MW

-Month

Systemwide TMNSR Systemwide TMOR SWCT TMNSR SWCT TMOR

CT TMNSR CT TMOR NEMABOST TMNSR NEMABOST TMOR

2022 Annual Markets Report 195 ISO New England Inc.

ISO-NE PUBLIC

the eight most recent auctions (occurring in 2019 to 2022), as external reserve support has

supplanted that need.

7.2.3 FRM Payments

FRM payments are based on auction obligations, auction clearing prices, and the actual delivery

of the obligation in the real-time energy market.

310

Participants obtain FRM obligations by

participating in forward reserve auctions or by obtaining an obligation from another

participant that has an auction-based obligation.

311

312

Participants must convert their

obligations into the physical delivery of operating reserve capacity by assigning obligations to

resources for the real-time energy market. When a participant’s assignments to resources are

less than the participant’s FRM obligation, it results in failure-to-reserve penalties. Additionally,

participants can incur failure-to-activate penalties when a resource that has been assigned an

FRM obligation fails to provide energy (when called upon by the ISO).

Annual FRM payment and penalty data over the past five years are provided in Figure 7-10

below. The chart indicates the annual auction-based payments as positive stacked bar values

and penalties as negative stacked bar values; the line graph indicates annual payments net of

total penalties.

313

Figure 7-10: FRM Payments and Penalties by Year

310

See Manual M-36, Forward Reserve and Real-Time Reserve, Section 6.

311

Hourly FRM obligations may be transferred by participants on a daily basis up to two days after the delivery period.

These transfers take place through “internal bilateral transactions” that allow the ISO to determine whether the holder of

the obligation delivered the physical capacity needed to back the obligation in the real-time energy market. See ISO

Manual M-36, Forward Reserve and Real-Time Reserve, Section 3.1.2.

312

FRM auction obligations are specific to participants and are not specific to resources.

313

“FTR” refers to failure-to-reserve and “FTA” refers to failure-to-activate.

-10

2018 2019 2020 2021 2022

Payments and Penalties ($ millions)

TMNSR Credit TMOR Credit TMNSR FTR Penalties

TMOR FTR Penalties TMNSR FTA Penalties TMOR FTA Penalties

Net Payment

2022 Annual Markets Report 196 ISO New England Inc.

ISO-NE PUBLIC

Net forward reserve payments increased significantly in 2022 due to the much higher TMNSR

auction clearing prices discussed previously. Prior to 2022, net forward reserve payments

declined from 2018 through 2021. The decline primarily reflects a reduction in TMOR clearing

prices in 2020 and 2021.

314

Penalties have been low relative to gross payments and have been stable in the 2% to 8% range

of total payments over the period. These penalties have been predominately for failing to

reserve (97%). Since failure-to-reserve penalties result in forfeiture of auction-based payments

for unassigned obligations, the payments are directly influenced by FRM clearing prices. Total

penalties declined prior to 2022 (in part because FRM clearing prices declined), but increased

by 118% to $1.9 million in 2022 (when FRM clearing prices significantly increased). Total

penalties in 2022 represented just 3% of gross FRM payments.

314

Summer TMOR auction prices fell by 53% (or $999/MW-month) in 2020 (relative to summer 2019), and fell by

additional 33% in the Summer 2021 auction. A 32% decline in TMOR auction prices for Winter 2020-21 (compared to

Winter 2019-20) added to the reduced payments in 2020 and 2021.

2022 Annual Markets Report 197 ISO New England Inc.

ISO-NE PUBLIC

7.3 Regulation

This section examines the participation, outcomes, and competitiveness of the regulation

market in 2022. Specifically, we review regulation clearing prices (7.3.1), payments (7.3.2) and

the amount of regulation capability needed by the ISO (7.3.3).

Key Takeaways

The regulation market produces two clearing prices, for capacity and service. Clearing prices for

capacity, which makes up about 90% of the overall market value, increased from $19.23/MWh

in 2021 to $30.96/MWh in 2022 (a 61% change), reflecting a rise in energy market opportunity

costs. Regulation service prices increased modestly from $0.21/mile in 2021 to $0.27/mile in

2022.

Regulation payments increased by 52% in 2022, primarily reflecting the increase in capacity

prices; 2021 payments were $25.3 million compared to $38.4 million in 2022.

Regulation requirements in 2022 were relatively steady compared to 2021, needing just over 90

MW of regulation capacity per hour, on average. The resource mix of committed regulation

capacity has changed significantly over the review period. In 2018, alternative regulation

resources (such as batteries) accounted for 10% of committed capacity; by 2020, such resources

accounted for 25%, and by 2022, these resources accounted for 50%.

7.3.1 Regulation Prices

Regulation Clearing Prices (RCP) are calculated in real-time and are based on the regulation

offer of the highest-priced generator providing the service. There are two types of regulation

clearing prices: “service” and “capacity.”

315

Clearing prices for the past five years are shown in

Table 7-1 below.

316

315

The s ervice price represents the direct cost of providing the regulation service (also known as regulation “mileage”).

Mileage represents the up and down movement of generators providing regulation and is measured as the absolute MW

variation in output per hour. These direct costs may include increased operating and maintenance costs, as well as

incremental fuel costs resulting from the generator operating less efficiently when providing regulation service. The

capacity price may represent several types of costs, including: (1) the expected value of lost energy market opportunities

when providing regulation service,

315

(2) the value of intertemporal opportunities that would be lost from providing

regulation, (3) elements of fixed costs such as incremental maintenance to ensure a generator’s continued performance

when providing regulation, and (4) fuel market or other risks associated with providing regulation.

316

The prices in the table are simple average prices for each year.

2022 Annual Markets Report 198 ISO New England Inc.

ISO-NE PUBLIC

Table 7-1: Regulation Prices

Year

Regulation Service Clearing Price

($/Mile)

Regulation Capacity Clearing Price

($/MW per Hour)

Min

Avg

Max

Min

Avg

Max

2018

0.00

0.25

10.00

0.00

28.30

2,331.55

2019

0.00

0.28

10.00

0.75

21.96

258.67

2020

0.00

0.21

10.00

0.40

16.12

396.08

2021

0.00

0.21

10.00

0.00

19.23

699.11

2022

0.00

0.27

10.00

0.00

30.96

1,068.09

Regulation capacity prices increased by 61% in 2022, reflecting an increase in the “opportunity

cost” and “incremental cost saving” components of regulation capacity pricing. The opportunity

cost component of the regulation price indicates the expected value of foregone energy market

opportunities, when providing regulation. The increase in opportunity costs is consistent with a

significant increase in real-time energy market LMPs, which rose by 89% in 2022 compared to

2021. The increase in incremental cost savings is affected by other regulation offer components,

and reflects the cost difference between the marginal offer and the next most expensive offer.

Regulation service prices also increased compared to 2021. In 2022, the average service price

was $0.27/mile, up $0.06/mile compared to the prior year. Mileage payments represent a small

share of overall regulation payments (13% or $5.1 million in 2022).

Between 2018 and 2021, regulation capacity prices declined and then increased (decreasing by

19% in 2019, decreasing by 22% in 2020 and increasing by 19% in 2021). In each instance, the

variation in capacity prices was strongly influenced by variation in the “opportunity cost”

component of regulation capacity pricing; regulation opportunity costs reflected declining real-

time energy market LMPs between 2018 and 2020 and increasing energy market LMPs in 2021.

7.3.2 Regulation Payments

Compensation to generators providing regulation includes a regulation capacity payment, a

service payment, a make-whole payment, and an operating reserve adjustment.

317

Annual

regulation payments over the past five years are shown in Figure 7-11 below.

318

The reserve payment deduction is shown as a negative value in the exhibit; the positive values

represent total payments (prior to reserve payment deductions) for the regulation capacity and

service (mileage) provided by regulation resources during the period. The make-whole

317

The operating reserve adjustment represents a deduction to regulation payments. Under certain circumstances, part of

a regulation resource’s regulating range may overlap with the resource’s operating reserve range. Since generators do not

actually provide operating reserves within the regulating range, reserve compensation needs to be deducted from the

resource’s market compensation. The reserve payment deduction represents the MW quantity overlap of the regulating

range and operating reserve range, multiplied by the operating reserve price.

318

In the figure, capacity payments include regulation uplift payments. Regulation uplift is provided when opportunity cost

estimates included in regulation capacity prices are insufficient to cover actual energy market opportunity costs incurred

by regulation resources.

2022 Annual Markets Report 199 ISO New England Inc.

ISO-NE PUBLIC

payment is included in capacity payment totals, since it represents an uplift payment when the

capacity payments do not fully compensate resources for energy market opportunity costs.

Figure 7-11: Regulation Payments

Payments to regulation resources totaled $38.4 million in 2022, 52% more than the $25.3

million in 2021 (these totals exclude the reserve payment adjustment). The increase in 2022

payments resulted primarily from a 59% increase in capacity payments; this increase in

capacity payments is consistent with the above-noted increase in capacity prices (61%). A

modest increase in regulation service prices and payments ($0.8 million increase in service

payments) in 2022 also contributed to the overall increase in regulation payments.

In the earlier years (2018-2021), capacity payments constituted 81% to 92% of overall

regulation payments. Year-to-year variation in regulation payments, consequently, resulted

primarily from variation in capacity prices and associated payments. As noted in the preceding

section, capacity prices are significantly influenced by energy market opportunity costs and

tend to follow increases and decreases in real-time energy market LMPs.

7.3.3 Regulation Requirements, Resource Mix, and Performance

The regulation requirement in New England varies throughout the day and is typically highest in

the morning and the late evening. The higher regulation requirement during these hours is the

result of greater load variability (load ramping up in the morning and down in the evening). The

average hourly regulation requirement by hour of day for 2022 is shown in Figure 7-12 below.

2022 Annual Markets Report 200 ISO New England Inc.

ISO-NE PUBLIC

Figure 7-12: Average Hourly Regulation Requirement, 2022

The average hourly regulation requirement of 91.2 MW in 2022, a negligible change (0.5 MW)

from the 2021 requirement (90.7 MW). In each hour, the ISO commits resources to meet the

requirement. Two different types of resources can provide regulation in the ISO’s regulation

market: traditional generators and alternative regulation resources (ATRRs). Almost all of the

ATRRs are battery resources that may function solely as regulation resources or may operate as

a combination of energy market services: consumption (battery charging), generation (battery

discharging), and regulation.

The regulation market resource mix for 2018 to 2022 is shown in Figure 7-13.

Figure 7-13: Regulation Resource Mix

100

120

140

160

180

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

Hour of Day

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

2018 2019 2020 2021 2022

% of Committed Regulation Capacity

ATRR Gen

2022 Annual Markets Report 201 ISO New England Inc.

ISO-NE PUBLIC

The resource mix of committed regulation capacity has changed significantly. In 2018, ATRRs

(blue shading) provided approximately 10% of committed regulation capacity; by 2022, ATRRs

provided 50%. This change follows continuing increases in the installed capacity of battery

resources in the ISO’s energy markets. Regulation capacity available from ATRRs has increased

from 19 MW to 127 MW over the period. The change in resource mix also suggests that battery

resources are lower-cost regulation resources (i.e., have lower-cost regulation market offers),

as these ATRRs have increasingly displaced traditional generators in the merit order for

regulation market commitment.

Finally, regulation performance is measured relative to a NERC standard. With the ISO’s

implementation of NERC BAL-001-2 standards in 2016, the ISO uses violations of Balancing

Authority ACE Limits (BAAL) to measure performance. Violations result from exceeding ACE

limits for more than 30 consecutive minutes; in 2022, there were no BAAL violations.

2022 Annual Markets Report 202 ISO New England Inc.

ISO-NE PUBLIC

Section 8

Market Design or Rule Changes

This section provides an overview of the major market design and rule changes that were

recently implemented or are being assessed or planned for future years. Table 8-1 below lists

the design changes summarized in this section.

319

Table 8-1: Market Design or Rule Changes

Major Design or Rule Changes Recently Implemented

Major Design or Rule Changes in Development or

Implementation for Future Years

Transmission Cost Allocation to Network Customers with

Behind-the-Meter Generation

Incorporate Solar into Do-Not-Exceed Dispatch

Competitive Transmission Solicitation Enhancements

Inventoried Energy Program

Extended-Term Transmission Planning Tariff Changes

FERC Order 2222, Distributed Energy Resources

Competitive Capacity Markets without a Minimum Offer

Price Rule (MOPR)

Resource Capacity Accreditation (RCA) in the Forward

Capacity Market

Day-Ahead Ancillary Services Initiative

8.1 Major Design Changes Recently Implemented

The following subsections provide an overview of changes recently implemented.

8.1.1 Transmission Cost Allocation to Network Customers with Behind-the-Meter Generation

FERC accepted Participating Transmission Owners (PTO) proposal in 2022

In our Spring 2020 Quarterly Markets Report (QMR), we conducted an analysis of transmission

cost allocation issues with respect to the treatment of behind-the-meter (BTM) generation

during monthly peak demand hours.

320

We expressed our concern about potential widespread

non-compliance with the tariff requirement to reconstitute peak load by adding back BTM

generation. Further, the IMM recognized that the transmission cost allocation rules were

established over 20 years ago and should be re-evaluated.

In 2020 and 2021, the IMM engaged with stakeholders in the review of a Participating

Transmission Owners (PTO) proposal developed in response to our analysis.

321

In July 2021, the

319

An overview of key ISO projects is also available on the ISO website at https://www.iso-ne.com/committees/key-

projects

320

IMM, Spring 2020 Quarterly Markets Report (Augus t 17, 2020 – Revision 1), https://www.iso-ne.com/static-

assets/documents/2020/07/2020-spring-quarterly-markets-report.pdf

321

IMM, IMM Feedback on the Participating Transmission Owners’ (PTOs) Transmission Cost Allocation Proposal (January

20, 2021), https://www.iso-ne.com/static-assets/documents/2021/01/a03_tc_2021_06_imm_feedback_ptoac.docx

2022 Annual Markets Report 203 ISO New England Inc.

ISO-NE PUBLIC

Participating Transmission Owners Administrative Committee (PTO AC), joined by the ISO,

322

filed a proposal to modify the monthly Regional Network Load (RNL) calculation to exclude

BTM generation. In filed comments, the IMM described why the PTO proposal was deficient

and should be rejected.

323

In 2022, FERC issued an order accepting the PTO’s proposal effective

September 1, 2021.

324

8.1.2 Competitive Transmission Solicitation Enhancements

FERC accepted ISO’s lessons learned tariff changes in 2022

FERC Order No. 1000 required the ISO, along with other ISOs/RTOs across the US, to change

aspects of their regional and interregional transmission planning and cost-allocation processes.

As part of its compliance with this order, the ISO created a Request for Proposal (RFP) process

to solicit competitive proposals for certain transmission upgrades, such as non-time sensitive

(more than three year’s out) transmission needs in the region.

From December 2019 to July 2020, the ISO conducted its first RFP under Order 1000 to address

necessary transmission upgrades to maintain reliability in the Boston area due to the

retirement of the Mystic generating station.

325

Following the RFP, the ISO and stakeholders held

“lessons learned” discussions, and in December 2021, the ISO and NEPOOL jointly proposed

tariff changes to improve the competitive transmission planning process.

326

In 2022, FERC

issued an order accepting the ISO’s filing.

327

8.1.3 Extended-Term Transmission Planning Tariff Changes

FERC accepted first phase tariff revisions in 2022

In December 2021, the ISO and NEPOOL jointly filed proposed tariff changes to allow the ISO to

perform extended-term (beyond 10 years) system planning analyses requested by the New

England states on a recurring basis. The proposal was in response to a request by the states to

“implement a state-led, proactive scenario-based planning process for long-term analysis of

322

The ISO joined the filing in its capacity as the administrator of the ISO-NE Tariff and to facilitate the proposed revisions

in eTariff but took no position on the PTO Proposal. Filing Letter, p. 1, footnote 4 (“the ISO does not take a position on th e

proposed revisions”).

323

IMM, Comments of the Internal Market Monitor on the Proposal to Exclude Behind -the-Meter Generation from

Transmission Cost Allocation, Docket No. ER21-2337-000 (July 22, 2021), https://www.iso-ne.com/static-

assets/documents/2021/07/imm_comments_on_pto_proposal.pdf

324

FERC, Letter Order Accepting Tariff Revisions, Docket No. ER21-2337-002 (February 22, 2022), https://www.iso-

ne.com/static-assets/documents/2022/02/er21-2337-002_order_accept_monthly_regional_load_calculation.pdf

325

ISO New England ISO Newswire, “ISO-NE makes selection in first Order 1000 transmission RFP,”(July 24, 2020),

https://isonewswire.com/2020/07/24/iso-ne-makes-selection-in-first-order-1000-transmission-rfp/

326

ISO New England Inc. and New England Power Pool, Transmission Planning Improvements, Docket No. ER-22-733-000

(December 28, 2021), https ://www.iso-ne.com/static-

assets/documents/2021/12/transmission_planning_improvements.pdf

327

FERC, Order Accepting Tariff Revisions, Docket No. ER22-733-000 (February 25, 2022), https://www.iso-ne.com/static-

assets/documents/2022/02/er22-733-000_2_25_22_order_accepting_transmission_planning_improvements.pdf

2022 Annual Markets Report 204 ISO New England Inc.

ISO-NE PUBLIC

state mandates and policies as a routine planning practice.”

328

In February 2022, FERC accepted

the proposed tariff changes.

329

8.2 Major Design or Rule Changes in Development or Implementation for Future Years

The following market design or rule changes are either (i) currently being assessed or are in the

design phase or (ii) have been completed and the planned implementation date is in the future.

8.2.1 Incorporate Solar into Do-Not-Exceed Dispatch

Planned implementation for December 2023

In November 2022, the ISO submitted to FERC proposed tariff changes to incorporate front-of-

meter solar resources into the Do-Not-Exceed (DNE) dispatch rules. Existing DNE dispatch rules

allow for inclusion of certain intermittent resources—wind and run-of-river hydro

generators—into the ISO’s real-time economic dispatch and avoids the need for manual

dispatch of these resources. With DNE dispatch, the ISO sends to each DNE resource, a DNE

dispatch point, which is the maximum output level that the generator must not exceed.

In January 2023, FERC issued a letter order accepting the ISO’s proposal to incorporate solar

into DNE dispatch with an effective date of December 5, 2023.

330

As of this effective date, the

existing DNE rules for wind and run-of-river hydro will extend to front-of-meter solar

resources.

8.2.2 Inventoried Energy Program

Planned implementation for Winters 2023/24 and 2024/25

The Inventoried Energy Program (IEP) provides an interim solution to compensate and incent

inventoried energy during winter seasons of 2023/24 and 2024/25. The IEP allows resources

to sell inventoried energy that will be held for certain trigger conditions.

331

Participating

resources can sell this inventoried energy at either a forward settlement rate for the winter

season or a spot rate. If a resource sells inventoried energy forward, it must either (i) maintain

this amount of inventoried energy during each trigger condition or (ii) buy out of any shortfall

at the spot rate, for each trigger condition.

In 2022, following a FERC order that was in response to a decision from the U.S. Court of

Appeals for the D.C. Circuit, the ISO made a compliance filing to make nuclear, coal, biomass,

328

NESCOE, Report to the Governors (June 2021), https://nescoe.com/wp-

content/uploads/2021/06/Advancing_Vision_Report_6-29-21.pdf

329

FERC, Letter Order Accepting ISO New England Inc's et al 12/27/2021 Filing of Proposed Tariff Revisions to Attachment

K of its Open Access Transmission Tariff etc., Docket No. ER22-727-000 (February 25, 2022), https://www.iso-

ne.com/static-assets/documents/2022/02/er22-727-000_2_25_22_ltr_order_accepting_longer-term_planning.pdf

330

FERC, Letter order accepting ISO New England Inc.'s November 30, 2022 filing of revisions to section III.1.11.3(e) of

Market Rule 1 of its Transmission, Markets and Services Tariff etc. under ER23-517., Docket No. ER23-517-000 (January 19,

2023), https://www.iso-ne.com/static-assets/documents/2023/01/er23-517-000_1-19-

23_ltr_order_accept_incorporate_solar_sources_dne.pdf

331

A trigger condition occurs when the average of the daily high and low temperature measured at Bradley International

Airport in Windsor Locks, CT is 17 or lower.

2022 Annual Markets Report 205 ISO New England Inc.

ISO-NE PUBLIC

and hydroelectric generators ineligible for the program.

332

In 2023, the ISO has filed further

proposed changes to the IEP program that include an update the settlement rate calculation to

better reflect evolving global LNG market conditions.

8.2.3 FERC Order 2222, Distributed Energy Resources

FERC issued order in 2023 on the ISO’s compliance filing

On September 17, 2020, FERC issued Order 2222, which found that existing ISO/RTO market

rules were unjust and unreasonable, because they contained barriers to the participation of

distributed energy resources aggregations (DERAs).

333

The purpose of Order 2222 is to remove

these barriers and allow DERAs to provide all services that they are technically capable of

providing. Specifically, the order outlined 11 directives for ISOs/RTOs to follow, including

allowing participation of DERAs, allowing DER aggregators to register DERAs under one or

more participation models

334

, and establishing a minimum size requirement for DERAs of no

more than 100 kW.

During 2020 and 2021, the ISO worked with stakeholders to develop the tariff revisions

necessary to come into compliance with Order 2222. The ISO’s proposed tariff changes were

brought through the complete stakeholder process. At its January 2022 meeting, the NEPOOL

Participant’s Committee voted to support the proposal (71.10% in favor).

On February 2, 2022, the ISO, joined by NEPOOL and the PTO AC, filed a compliance proposal

for Order 2222.

335

The proposal creates two new participation models for the energy and

ancillary services market (called Demand Response DERA and Settlement Only DERA) and

modifies existing models to accommodate the physical and operational characteristics of

DERAs. The proposal includes many other changes to comply with the order, including

introducing a new participation model for the FCM (called a Distributed Energy Capacity

Resource), setting a minimum size of 100 kW for DERAs, specifying locational requirements,

and changing existing metering and telemetry rules.

The ISO requested two effective dates: 1) November 1, 2022 for FCM-related revisions, which

would be in time for the FCA 18 qualification process, and 2) November 1, 2026 for changes

related to the energy and ancillary services market.

332

In December 2022, the ISO and NEPOOL also jointly filed tariff changes related to incorporating financial assurance and

billing policy changes related to the IEP, which were accepted by FERC in early 2023. See FERC, letter order accepting ISO

New England Inc.'s 12/22/2022 filing of revisions to its Transmission, Markets, and Services Tariff to revise the Financial

Assurance Policy etc. under ER23-705, Docket No. ER23-705-000 (February 14, 2023), https://www.iso-ne.com/static-

assets/documents/2023/02/er23-705-000.pdf

333

DERAs are aggregations of small-scale power generation or storage technologies, such as electric storage, intermittent

generation, distributed generation, demand response, energy efficiency, thermal storage, and electric vehicles and their

charging equipment. FERC , “FERC Order No. 2222: Fact Sheet,” webpage (last updated September 28, 2020),

https://www.ferc.gov/media/ferc-order-no-2222-fact-sheet

334

A “participation model” refers to rules created for a specific type of resource that has unique physical and operational

characteristics (see Order 2222, footnote 7 on p. 5). For example, a generator is a type of participation model in ISO-NE.

335

ISO New England Inc., Revisions to ISO New England Inc. Transmission, Markets and Services Tariff to Allow for the

Participation of Distributed Energy Resource Aggregations in New England Markets, FERC filing, Docket No. ER22-983-000

(February 2, 2022), https://www.iso-ne.com/static-assets/documents/2022/02/order_no_2222_filing.pdf

2022 Annual Markets Report 206 ISO New England Inc.

ISO-NE PUBLIC

In March 2023, FERC issued an order accepting in part and rejecting in part the ISO’s

compliance filing, subject to further compliance filings. In its order, the FERC found that the

ISO’s compliance filing satisfied six of the eleven aforementioned directives in Order 2222 but

only partially satisfied the remaining five directives, which include issues related to storage

participation models, information and data requirements, metering and telemetry

requirements, and other issues.

336

8.2.4 Competitive Capacity Markets without a Minimum Offer Price Rule (MOPR)

FERC accepted MOPR elimination filing in 2022

In May 2022, FERC accepted the ISO’s and NEPOOL’s joint proposed tariff revisions to remove

the MOPR.

337

These approved changes eliminate the core components of the MOPR (i.e., offer

review trigger prices) as well as the substitution auction effective for Forward Capacity Auction

(FCA) 19 (capacity commitment period June 2028–May 2029). There is a two-year transition

period (FCAs 17 and 18) where the MOPR remains in effect; during the transition period, the

Renewable Technology Resource (RTR) exemption has been reinstated, which will allow a

greater number of sponsored policy resources to enter the market.

338

Beginning in FCA 19, the MOPR will be replaced with a revised buyer-side market power

framework. Under this new approach, resources that fall into certain categories are not subject

to buyer-side market power review; these include resources with capacity of 5 MW or less,

passive demand-response resources, those with no load-side interest (i.e., “competitive

entrants”), and certain federally or state-sponsored resources. Resources that do not fall into

any of these categories will be subject to a buyer-side market power review by the IMM.

339

8.2.5 Resource Capacity Accreditation (RCA) in the Forward Capacity Market (FCM)

ISO currently reviewing the projected filing date to FERC

In 2022, the ISO began the stakeholder process for the resource capacity accreditation (RCA)

project. This initiative aims to assess and implement methodologies of accrediting resources in

the FCM that will better reflect their contributions to resource adequacy. Under the proposed

RCA reforms, the ISO would accredit resources based on their Marginal Reliability Impact

(MRI), which reflects each resource’s incremental contribution to resource adequacy. As part of

this effort, the ISO is also pursuing changes to improve accreditation during winter by modeling

gas limitations and accounting for fuel storage capability and contracting arrangements when

accrediting gas-fired resources.

336

For a s ummary of the FERC’s order on compliance filing, see https://www.iso-ne.com/static-

assets/documents/2023/03/a05_2023_03_07-09_order_2222_nepool_counsel_memo.pdf.

337

FERC, Order Accepting Tariff Revisions, Docket No. ER22-1528-000 (May 27, 2022), https://www.iso-ne.com/static-

assets/documents/2022/05/er22-1528-000_5-27-2022_order_accept_mopr_removal.pdf

338

The RTR exemption would be 300 MW in FCA 17 and 400 MWs in FCA 18 (less substitution auction MWs in FCA 17).

During the transition period, the substitution auction test price would also be eliminated.

339

This review consists of a conduct test, wherein a resource fails the conduct test if its requested offer is below the

resource’s offer floor price. Those that fail the conduct test have the opportunity to avoid mitigation if they can sufficiently

demonstrate that they would be unlikely to realize a material, net financial b enefit from lowering FCA prices.

2022 Annual Markets Report 207 ISO New England Inc.

ISO-NE PUBLIC

8.2.6 Day-Ahead Ancillary Services Initiative

ISO plans to file with FERC by end of 2023

The Day-Ahead Ancillary Services Initiative (DASI) sets out to procure and transparently price

specific ancillary services needed for system reliability. Currently, there is no day-ahead

reserves market in ISO-NE. The ISO is proposing to create four new reserve products that would

be co-optimized with energy and priced in the day-ahead market.

340

These day-ahead reserve

products are not forward sales of reserves that settle against real-time reserve prices; instead,

they have a unique call-option settlement structure, where the settlement is based on a pre-

determined strike price and the actual real-time Hub LMP.

The ISO plans to submit proposed tariff revisions for DASI to FERC by the end of 2023.

8.3 Additional Notable Studies

The following subsection provides an overview of additional notable studies that are not part of

any planned market design or implementation work.

8.3.1 New England's Future Grid Initiative

Reports published in 2022

In 2022, the ISO published two major reports for the Future Grid Initiative, which is a

stakeholder-led effort that seeks to help the region prepare for and support New England’s

transition to a future grid.

341

The initiative has two parallel tracks: the Future Grid Reliability

Study and Pathways to the Future Grid.

Future Grid Reliability Study (FGRS)

The purpose of the FGRS is “to assess and discuss the future state of the regional power system

in light of current state energy and environmental laws.”

342

In particular, the two phases of the

FGRS together will assess if existing markets will be sufficient to attract and retain resources

needed for reliability and identify potential operational and reliability challenges that will need

to be addressed.

In July 2022, the ISO published a report for Phase 1 of the FGRS that evaluated grid reliability

issues under several decarbonization scenarios. In the upcoming Phase 2 of the FGRS, the ISO

plans to determine which types of clean energy resources can be “revenue sufficient” and meet

resource adequacy needs and identify any potential reliability gaps.

Pathways to the Future Grid

340

Three of these products mirror the three real-time reserve products: TMSR, TMNSR, and TMOR; the fourth product is

called Energy Imbalance Reserve, which is procured to essentially help fill any potential “energy gap” betw een forecasted

load and cleared day-ahead physical supply.

341

See p. 12 of the NEPOOL 2021 Annual Report https ://nepool.com/wp-

content/uploads/2021/12/Annual_Report_2021.pdf

342

See pp. 1-2 of NEPOOL Future Grid Reliability Study, Study Framework for Phase 1 Economic Study Request (March 12,

2021) https://nepool.com/wp-content/uploads/2021/03/FG_20200331_a04_framework_document_redlined.docx

2022 Annual Markets Report 208 ISO New England Inc.

ISO-NE PUBLIC

The Pathways to the Future Grid seeks to explore and evaluate market frameworks to support

the region’s clean energy transition. In assisting with this project, the ISO and the consulting

firm Analysis Group worked with stakeholders to evaluate four different frameworks for

decarbonizing the New England power sector:

 Status quo: continued use of long-term contracts as the primary tool to meet

decarbonization objectives.

 Forward Clean Energy Market (FCEM): introduction of clean energy credits, which

represent the clean energy attributes of generation and are procured in a forward

market.

 Net carbon pricing: implementation of a price on carbon, where suppliers are charged

based on their carbon emissions and these charges are rebated to load.

 Hybrid: a combination of the FCEM and net carbon pricing frameworks. Only new

resources are eligible for the FCEM and the carbon price level is set to provide revenue

adequacy for existing clean resources.

The Analysis Group modeled each of these frameworks to help compare and assess these

different approaches to decarbonizing. The modeling assumes that New England states meet a

regional target for the power sector of 80% CO

emissions reduction by 2040 (relative to 1990

levels) and quantifies how market and economic outcomes (e.g., LMPs, social costs) differ under

each framework.

In April 2022, the final Pathways Study report was published. Following this report, the ISO is

working with states and stakeholders to gather feedback on the report.

2022 Annual Markets Report 209 ISO New England Inc.

ISO-NE PUBLIC

Acronyms and Abbreviations

Description

°F

degrees Fahrenheit

alternating current

ACE

area control error

ADCR

Active Demand Capacity Resources

AMR

Annual Markets Report

ARA

annual reconfiguration auction

ARD

asset-related demand

ART

Annual Reconfiguration Transaction

ancillary service

BAA

balancing authority area

BAAL

Balancing Area ACE Limits

BAL-001-2

NERC’s Real Power Balancing Control Performance Standard

BAL-003

NERC’s Frequency Response and Frequency Bias Setting

Standard

bbl

barrel (unit of oil)

Bcf

billion cubic feet

BTM

behind-the-meter

Btu

British thermal unit

market concentration of the four largest competitors

CASPR

Competitive Auctions with Sponsored Policy Resources

combined cycle (generator)

CCP

capacity commitment period

CDD

cooling degree day

CMR

Code of Massachusetts Regulations

carbon dioxide

CONE

cost of new entry

CPS 2

NERC Control Performance Standard 2

CSC

Cross Sound Cable

CSO

capacity supply obligation

State of Connecticut, Connecticut load zone, Connecticut

reserve zone

combustion turbine

CTL

capacity transfer limit

CTS

Coordinated Transaction Scheduling

DAGO

day-ahead generation obligation

DALO

day-ahead load obligation

DARD

dispatchable asset related demand

DDBT

dynamic de-list bid threshold

2022 Annual Markets Report 210 ISO New England Inc.

ISO-NE PUBLIC

Acronyms and Abbreviations

Description

DDG

do-not-exceed dispatchable generators

DDT

dynamic de-list threshold

Dec

decrement (virtual demand)

DFC

dual fuel commissioning

distributed generation

DLOC

dispatch lost opportunity costs NCPC

DNE

do not exceed

DOE

US Department of Energy

demand response

EGEL

Electricity Generator Emissions Limits (program)

EIA

US Energy Information Administration (of DOE)

EMM

External Market Monitor

EMOC

Energy Market Opportunity Cost

EMOF

Energy Market Offer Flexibility

EPA

Environmental Protection Agency

ERS

external reserve support

ETU

Elective Transmission Upgrade

FCA

Forward Capacity Auction

FCM

Forward Capacity Market

FERC

Federal Energy Regulatory Commission

FRM

Forward Reserve Market

FSP

Fast-Start Pricing

FTR

Financial Transmission Right

gas turbine

GHG

greenhouse gas

gigawatt

GW-month

gigawatt-month

GWh

gigawatt-hour

GWSA

Global Warming Solutions Act

HDD

heating degree day

hour ending

Hydro-Québec

HQICCS

Hydro-Québec Installed Capacity Credit

IBT

internal bilateral transaction

ICE

Intercontinental Exchange, Inc.

ICR

Installed Capacity Requirement

ICT

Interim Compensation Treatment

IMAPP

Integrating Markets and Public Policy

IMM

Internal Market Monitor

2022 Annual Markets Report 211 ISO New England Inc.

ISO-NE PUBLIC

Acronyms and Abbreviations

Description

Inc

increment (virtual supply)

ISO

Independent System Operator,

ISO New England

ISO tariff

ISO New England Transmission, Markets, and Services Tariff

kilowatt

kWh

kilowatt-hour

kW-month

kilowatt-month

kW/yr

kilowatt per year

symbol for the competitiveness level of the LMP

left axis

LCC

Local Control Center

LEG

limited-energy generator

LMP

locational marginal price

LNG

liquefied natural gas

LOC

lost opportunity cost

LOLE

loss- of-load expectation

LS/ERI

Lower SEMA/Eastern RI Import interface

LSE

load-serving entity

LSCPR

local second-contingency-protection resource

LSR

local sourcing requirement

M-36

ISO New England Manual for Forward Reserve

State of Massachusetts

MAPE

mean absolute percent error

MassDEP

Massachusetts Department of Environmental Protection

MCL

maximum capacity limit

MDE

manual dispatch energy

State of Maine and Maine load zone

M/LCC 2

Master/Local Control Center Procedure

No. 2, Abnormal Conditions Alert

MMBtu

million British thermal units

MOPR

Minimum Offer Price Rule

MRA

monthly reconfiguration auction

MRI

marginal reliability impact

megawatt

MWh

megawatt-hour

N-1

first contingency

N-1-1

second contingency

NCPC

Net Commitment-Period Compensation

NEL

net energy for load

NEMA

Northeast Massachusetts, Boston load zone

2022 Annual Markets Report 212 ISO New England Inc.

ISO-NE PUBLIC

Acronyms and Abbreviations

Description

NEMA/Boston

Northeast Massachusetts/Boston

local reserve zone

NEPOOL

New England Power Pool

NERC

North American Electric Reliability Corporation

State of New Hampshire,

New Hampshire load zone

NHME

New Hampshire-Maine Import interface

NICR

net Installed Capacity Requirement

NNE

northern New England

No.

Number

NPCC

Northeast Power Coordinating Council

State of New York

NYNE

New York-New England interface

NYISO

New York Independent System Operator

OATT

Open Access Transmission Tariff

OP 4

ISO Operating Procedure No. 4

OP 7

ISO Operating Procedure No. 7

OP 8

ISO Operating Procedure No. 8

ORTP

offer-review trigger price

PER

peak energy rent

PFP

pay-for-performance

PJM

PJM Interconnection, L.L.C.

pnode

pricing node

PPR

pay-for-performance penalty rate

PRD

price-responsive demand

PROBE

Portfolio Ownership and Bid Evaluation

PST

pivotal supplier test

PTO

Participating Transmission Owners

PURA

Public Utilities Regulatory Authority

photovoltaic

quarter

reconfiguration auction

right axis

RAA

reserve adequacy assessment

RCA

Reliability Coordinator Area

RCP

regulation clearing price

RCPF

Reserve Constraint Penalty Factor

RFP

Requests for Proposals

RGGI

Regional Greenhouse Gas Initiative

2022 Annual Markets Report 213 ISO New England Inc.

ISO-NE PUBLIC

Acronyms and Abbreviations

Description

State of Rhode Island, Rhode Island

load zone

RMCP

reserve market clearing price

RNL

regional network load

RNS

regional network service

RoP

rest of pool

RoS

rest of system

RRP OC

rapid-response pricing opportunity costs NCPC

RSI

Residual Supply Index

RTDR

real-time demand response

RTLO

real-time load obligation

RTO

Regional Transmission Organization

RTR

renewable technology resource

SCR

special-constraint resource

SEMA

Southeast Massachusetts load zone

SENE

southeastern New England

SMD

Standard Market Design

SWCT

Southwest Connecticut

THI

Temperature-Humidity Index

TMNSR

10-minute non-spinning reserve

TMOR

30-minute operating reserve

TMSR

10-minute spinning reserve

TPRD

transitional price-responsive demand

TTC

total transfer capability

UDS

unit dispatch system

United States

UTC

up-to-congestion

State of Vermont and Vermont load zone

WCMA

Western/Central Massachusetts

WRP

Winter Reliability Program

WTI

West Texas Intermediate