vvEPA
United States
Environmental Protection
Agency
Air and
Radiation
(62Q4J)
March 2023
"Post-IRA 2022
Reference Case"
Documentation for
EPA's Power Sector Modeling
Platform v6
Using the Integrated Planning
Model
Post-IRA 2022 Reference Case
vwc c_a>
wc.vwst
-------
Cover: EPA's Power Sector Modeling Platform v6 is used by the U.S. Environmental Protection Agency
as a platform to conduct various scenario and sensitivity analysis on the key drivers of the power sector
behavior and to project the impact of emissions policies on the electric power sector in the 48 contiguous
states and the District of Columbia in the lower continental U.S. Representation of the electric power
sector in Canada is also included for purposes of integrated projections. The map appearing on the cover
shows the 67 model regions used to characterize the operation of the U.S. electric power system in the
lower continental U.S. and 11 model regions in Canada. EPA's Power Sector Modeling Platform v6 using
the Integrated Planning Model (IPM®) was developed by EPA's Clean Air Markets Division with technical
support from ICF, Inc. The IPM is a product of ICF, Inc. and is used in support of its public and private
sector clients. IPM® is a registered trademark of ICF Resources, L.L.C.
-------
Documentation for
EPA's Power Sector Modeling Platform v6
Using the Integrated Planning Model
Post-IRA 2022 Reference Case
U.S. Environmental Protection Agency
Clean Air Markets Division
1200 Pennsylvania Avenue, NW (6204J)
Washington, D.C. 20460
(www.epa.gov/airmarkets)
March 2023
"Post-IRA 2022 Reference Case"
-------
Acknowledgment
This document was prepared by U.S. EPA's Clean Air Markets Division, Office of Air and Radiation. ICF
Incorporated, an operating company of ICF, provided technical support under EPA Contract
68HE0C18D0001.
i
-------
TABLE OF CONTENTS
Acknowledgment i
1. Introduction 1-1
1.1 Executive Summary 1-1
1.2 Review and Ongoing Improvement of the Integrated Planning Model 1-6
2. Modeling Framework 2-1
2.1 IPM Overview 2-1
2.1.1 Purpose and Capabilities 2-1
2.1.2 Applications 2-2
2.2 Model Structure and Formulation 2-2
2.2.1 Objective Function 2-3
2.2.2 Decision Variables 2-3
2.2.3 Constraints 2-4
2.3 Key Methodological Features of IPM 2-5
2.3.1 Model Plants 2-5
2.3.2 Model Run Years 2-6
2.3.3 Cost Accounting 2-7
2.3.4 Modeling Wholesale Electricity Markets 2-7
2.3.5 Load Duration Curves (LDCs) 2-7
2.3.6 Fuel Modeling 2-10
2.3.7 Transmission Modeling 2-10
2.3.8 Operating Reserves Modeling 2-10
2.3.9 Perfect Competition and Perfect Foresight 2-11
2.3.10 Scenario Analysis and Regulatory Modeling 2-11
2.4 Hardware and Programming Features 2-11
2.5 Model Inputs and Outputs 2-12
2.5.1 Data Parameters for Model Inputs 2-12
2.5.2 Model Outputs 2-13
3. Power System Operation Assumptions 3-1
3.1 Model Regions 3-1
3.2 Electric Load Modeling 3-2
3.2.1 Distributed Solar Photovoltaics 3-6
3.2.2 Demand Elasticity 3-7
3.2.3 Net Internal Demand (Peak Demand) 3-7
3.2.4 Regional Load Shapes 3-8
3.3 Transmission 3-8
3.3.1 Inter-regional Transmission Capability 3-8
3.3.2 Joint Transmission Capacity and Energy Limits 3-9
3.3.3 Transmission Link Wheeling Charge 3-10
3.3.4 Transmission Losses 3-10
3.3.5 New Transmission Builds 3-10
3.4 International Imports 3-12
3.5 Capacity, Generation, and Dispatch 3-12
3.5.1 Availability 3-12
3.5.2 Capacity Factor 3-13
-------
3.5.3 Turndown 3-15
3.6 Reserve Margins 3-15
3.7 Operating Reserves 3-16
3.7.1 Operating Reserve Requirements 3-17
3.7.2 Generation Characteristics 3-18
3.8 Power Plant Lifetimes 3-19
3.9 Heat Rates 3-19
3.10 Existing Legislations and Regulations Affecting Power Sector 3-20
3.10.1 Inflation Reduction Act 3-20
3.10.2 SO2 Regulations 3-20
3.10.3 NOx Regulations 3-21
3.10.4 Multi-Pollutant Environmental Regulations 3-24
3.10.5 CO2 Regulations 3-29
3.10.6 Non-Air Regulations Impacting EGUs 3-30
3.10.7 State-Specific Environmental Regulations 3-31
3.10.8 New Source Review (NSR) Settlements 3-32
3.10.9 Emission Assumptions for Potential (New) Units 3-32
3.10.10 Renewable Portfolio Standards and Clean Energy Standards 3-32
3.10.11 Canada CO2 and Renewable Regulations 3-34
3.11 Emissions Trading and Banking 3-35
3.11.1 Intertemporal Allowance Price Calculation 3-35
3.12 45Q - Credit for Carbon Dioxide Sequestration 3-38
Generating Resources 4-1
4.1 National Electric Energy Data System (NEEDS) 4-1
4.2 Existing Units 4-1
4.2.1 Population of Existing Units 4-1
4.2.2 Capacity 4-4
4.2.3 Plant Location 4-5
4.2.4 Online Year 4-5
4.2.5 Unit Configuration 4-6
4.2.6 Model Plant Aggregation 4-6
4.2.7 Cost and Performance Characteristics of Existing Units 4-9
4.2.8 Life Extension Costs for Existing Units 4-16
4.3 Planned-Committed Units 4-17
4.3.1 Population and Model Plant Aggregation 4-17
4.3.2 Capacity 4-17
4.3.3 State and Model Region 4-17
4.3.4 Online and Retirement Year 4-17
4.4 Potential Units 4-18
4.4.1 Methodology for Deriving the Cost and Performance Characteristics of Conventional
Potential Units 4-18
4.4.2 Cost and Performance for Potential Conventional Units 4-18
4.4.3 Short-Term Capital Cost Adder 4-19
4.4.4 Regional Cost Adjustment 4-19
4.4.5 Cost and Performance for Potential Renewable Generating and Non-Conventional
Technologies 4-25
4.5 Inflation Reduction Act Impacts on New Units 4-39
-------
4.6 Nuclear Units 4-42
4.6.1 Existing Nuclear Units 4-42
4.6.2 Potential Nuclear Units 4-43
5. Emission Control Technologies 5-1
5.1 Sulfur Dioxide Control Technologies - Scrubbers 5-1
5.1.1 Methodology for Obtaining SO2 Controls Costs 5-2
5.1.2 SO2 Controls for Units with Capacities from 25 MW to 100 MW (25 MW < capacity <
100 MW) 5-3
5.2 Nitrogen Oxides Control Technology 5-5
5.2.1 Combustion Controls 5-5
5.2.2 Post-combustion NOx Controls 5-5
5.2.3 Methodology for Obtaining SCR and SNCR Costs for Coal Steam Units 5-6
5.2.4 Methodology for Obtaining SCR Costs for Oil/Gas Steam Units 5-8
5.3 Biomass Co-firing 5-9
5.4 Mercury Control Technologies 5-9
5.4.1 Mercury Content of Fuels 5-9
5.4.2 Mercury Emission Modification Factors 5-10
5.4.3 Mercury Control Capabilities 5-13
5.4.4 Methodology for Obtaining ACI Control Costs 5-17
5.5 Hydrogen Chloride (HCI) Control Technologies 5-17
5.5.1 Chlorine Content of Fuels 5-17
5.5.2 HCI Removal Rate Assumptions for Existing and Potential Units 5-19
5.5.3 HCI Retrofit Emission Control Options 5-19
5.6 Fabric Filter (Baghouse) Cost Development 5-21
5.7 Coal-to-Gas Conversions 5-23
5.7.1 Boiler Modifications for Coal-To-Gas Conversions 5-23
5.7.2 Natural Gas Pipeline Requirements for Coal-To-Gas Conversions 5-24
5.8 Retrofit Assignments 5-25
6. CO2 Capture, Storage, and Transport 6-1
6.1 CO2 Capture 6-1
6.1.1 CO2 Capture for Potential EGUs 6-1
6.1.2 CO2 Capture for Existing EGUs with CCS retrofit 6-2
6.2 CO2 Storage 6-3
6.3 CO2 Transport 6-7
7. Coal 7-1
7.1 Coal Market Representation 7-1
7.1.1 Coal Supply Regions 7-2
7.1.2 Coal Demand Regions 7-3
7.1.3 Coal Quality Characteristics 7-3
7.1.4 Coal Emission Factors 7-4
7.1.5 Coal Grade Assignments 7-10
7.2 Coal Supply Curves 7-10
7.2.1 Nature of Supply Curves Developed for EPA Platform v6 7-10
7.2.2 Cost Components in the Supply Curves 7-11
7.2.3 Procedures Employed in Determining Mining Costs 7-12
iv
-------
7.2.4 Procedure Used in Determining Mine Productivity 7-13
7.2.5 Procedure to Determine Total Recoverable Reserves by Region and Type 7-13
7.2.6 New Mine Assumptions 7-14
7.2.7 Other Notable Procedures 7-14
7.2.8 Cumulative Supply Curve Development 7-16
7.2.9 EPA Platform v6 Assumptions and Outlooks for Major Supply Basins 7-18
7.3 Coal Transportation 7-19
7.3.1 Coal Transportation Matrix Overview 7-20
7.3.2 Overview of Rail Rates 7-22
7.3.3 Truck Rates 7-26
7.3.4 Barge and Lake Vessel Rates 7-26
7.3.5 Transportation Rates for Imported Coal 7-26
7.3.6 Other Transportation Costs 7-27
7.3.7 Long-Term Escalation of Transportation Rates 7-27
7.3.8 Market Drivers Moving Forward 7-29
7.3.9 Other Considerations 7-31
7.4 Coal Exports, Imports, and Non-Electric Sectors Demand 7-32
8. Natural Gas 8-1
8.1 GMM 8-3
8.2 Translating GMM Results to IPM Natural Gas Supply Curves 8-5
8.2.1 Supply Curves for EPA Platform v6 8-7
8.2.2 Basis 8-8
8.2.3 Delivered Price Adders 8-9
8.3 GMM Assumptions 8-9
8.3.1 GMM Resources Data and Reservoir Description 8-9
8.3.2 Oil Prices 8-12
8.3.3 Gas Production 8-12
8.3.4 Demand Assumptions 8-13
8.3.5 LNG Exports and Pipeline Exports to Mexico 8-14
9. Other Fuels and Fuel Emission Factor Assumptions 9-1
9.1 Fuel Oil 9-1
9.2 Biomass Fuel 9-2
9.3 Nuclear Fuel 9-3
9.4 Waste Fuels 9-3
9.5 Hydrogen Fuel 9-4
9.6 Fuel Emission Factors 9-4
10. Financial Assumptions 10-1
10.1 Introduction and Summary 10-1
10.2 Introduction to Risk 10-1
10.2.1 Deregulation - Market Structure Risks 10-2
10.3 Federal Income Tax Law Changes 10-3
10.4 Calculation of the Financial Discount Rate 10-4
10.4.1 Introduction to Discount Rate Calculations 10-4
10.4.2 Summary of Results 10-5
10.5 Discount Rate Components 10-7
v
-------
10.6 Market Structure: Utility-Merchant Financing Ratio 10-7
10.7 Capital Structure: Debt-Equity Share 10-7
10.7.1 Introduction and Shares for Utilities and IPPs 10-7
10.7.2 Utility and Merchant 10-8
10.7.3 Merchant by Technology 10-8
10.8 Cost of Debt 10-9
10.8.1 Merchant Cost of Debt 10-9
10.8.2 Utility Cost of Debt 10-10
10.9 Return on Equity (ROE) 10-10
10.9.1 Introduction and Beta 10-10
10.9.2 Risk-Free Rate and Equity Risk Premium 10-11
10.9.3 Beta 10-11
10.9.4 Equity Size Premium 10-12
10.9.5 Nominal ROEs 10-12
10.9.6 WACC/Discount Rate 10-13
10.10 Calculation of Capital Charge Rate 10-13
10.10.1 Introduction to Capital Charge Rate Calculations 10-13
10.10.2 Capital Charge Rate Components 10-14
vi
-------
LIST OF TABLES
Table 1-1 Key Updates and Specifications in the EPA Platform v6 Post-IRA 2022 Reference Case 1-2
Table 1-2 Plant Types in v6 1-4
Table 1-3 Emission Control Technologies in v6 1-5
Table 2-1 Model Run Year and Year Mapping in v6 2-6
Table 2-2 Load Duration Curves used in EPA Platform v6 Post-IRA 2022 Reference Case 2-13
Table 3-1 Mapping of NERC Regions and NEMS Regions with v6 Model Regions 3-3
Table 3-2 Electric Load Assumptions in v6 3-5
Table 3-3 Regional Electric Load Assumptions in v6 3-5
Table 3-4 National Non-Coincidental Net Internal Demand in v6 3-7
Table 3-5 Annual Joint Capacity and Energy Limits to Transmission Capabilities between Model Regions
in v6 3-9
Table 3-6 International Electricity Imports (billions kWh) in v6 3-12
Table 3-7 Availability Assumptions in v6 3-12
Table 3-8 Seasonal Hydro Capacity Factors (%) in v6 3-13
Table 3-9 Planning Reserve Margins in v6 3-16
Table 3-10 Operating Reserve Requirement Assumptions by Type in v6 3-17
Table 3-11 Operating Reserve Regions in v6 3-18
Table 3-12 Operating Reserve Contribution Assumptions by Technology in v6 3-18
Table 3-13 Lower and Upper Limits Applied to Heat Rate Data in v6 3-19
Table 3-14 State-of-the-Art Combustion Control Configurations by Boiler Type in v6 3-24
Table 3-15 Ozone-Season NOx Emission Caps (Tons) for Fossil Units greater than 25MW in v6 3-25
Table 3-16 G1 and G2 CSAPR Update State Budgets, Variability Limits, and Assurance Levels for
Ozone-Season NOx (Tons) - 2021 through 2054 3-26
Table 3-17 Revised CSAPR Update State Budgets, Variability Limits, and Assurance Levels for Ozone-
Season NOx for G3 states (tons) 3-26
Table 3-18 Renewable Portfolio Standards in v6 3-32
Table 3-19 State RPS Solar Carve-outs in v6 3-33
Table 3-20 Clean Energy Standards in v6 3-33
Table 3-21 Offshore Wind Mandates in v6 3-33
Table 3-22 Fossil Generation Limits (GWh) in v6 3-34
Table 3-23 Canada Renewable Electricity Requirements (%) in v6 3-35
Table 3-24 Trading and Banking Rules in v6 - Part 1 3-35
Table 3-25 CASPR Trading and Banking Rules in v6 - Part 2 3-36
Table 3-26 Emission and Removal Rate Assumptions for Potential (New) Units in v6 3-37
Table 3-27 Recalculated NOx Emission Rates for SCR Equipped Units Sharing Common Stacks with
Non-SCR Units in v6 3-38
Table 3-28 Regional Net Internal Demand in EPA Platform v6 Post-IRA 2022 Reference Case 3-39
Table 3-29 Annual Transmission Capabilities of U.S. Model Regions in EPA Platform v6 Post-IRA 2022
Reference Case 3-39
Table 3-30 Turndown Assumptions for Coal Steam Units in EPA Platform v6 Post-IRA 2022 Reference
Case 3-39
Table 3-31 State Power Sector Regulations included in EPA Platform v6 Post-IRA 2022 Reference
Case 3-39
Table 3-32 New Source Review (NSR) Settlements in EPA Platform v6 Post-IRA 2022 Reference
Case 3-39
vii
-------
Table 3-33 State Settlements in EPA Platform v6 Post-IRA 2022 Reference Case 3-39
Table 3-34 Citizen Settlements in EPA Platform v6 Post-IRA 2022 Reference Case 3-39
Table 3-35 Availability Assumptions in EPA Platform v6 Post-IRA 2022 Reference Case 3-39
Table 3-36 BART Regulations included in EPA Platform v6 Post-IRA 2022 Reference Case 3-39
Table 4-1 Data Sources for NEEDS v6 4-2
Table 4-2 Rules Used in Populating NEEDS v6 4-2
Table 4-3 Summary Population (through 2021) of Existing Units in NEEDS v6 4-4
Table 4-4 Hierarchy of Data Sources for Capacity in NEEDS v6 4-4
Table 4-5 Capacity-Parsing Algorithm for Steam Units in NEEDS v6 4-5
Table 4-6 Data Sources for Unit Configuration in NEEDS v6 4-6
Table 4-7 Aggregation Profile of Model Plants as Provided at Set up of v6 4-7
Table 4-8 VOM Assumptions in v6 4-10
Table 4-9 FOM Assumptions in v6 4-12
Table 4-10 Life Extension Cost Assumptions Used in v6 4-16
Table 4-11 Summary of Planned-Committed Units in NEEDS v6 4-17
Table 4-12 Performance and Unit Cost Assumptions for Potential (New) Capacity from Conventional
Technologies in v6 4-20
Table 4-13 Short-Term Capital Cost Adders for New Power Plants in v6 (2019$) 4-21
Table 4-14 Regional Cost Adjustment Factors for Conventional and Renewable Generating
Technologies in v6 4-22
Table 4-15 Performance and Unit Cost Assumptions for Potential (New) Renewable and Non-
Conventional Technologies in v6 4-24
Table 4-16 Offshore Fixed Regional Potential Wind Capacity (MW) by Wind Class, Resource Class, and
Cost Class in v6 4-25
Table 4-17 Offshore Floating Regional Potential Wind Capacity (MW) by Wind Class, Resource Class,
and Cost Class in v6 4-26
Table 4-18 Offshore Fixed Average Capacity Factor by Wind Class and Resource Class in v6 4-27
Table 4-19 Offshore Floating Average Capacity Factor by Wind Class and Resource Class in v6 4-28
Table 4-20 Onshore Reserve Margin Contribution by Wind Class in v6 4-29
Table 4-21 Offshore Fixed Reserve Margin Contribution by Wind Class in v6 4-29
Table 4-22 Offshore Floating Reserve Margin Contribution by Wind Class in v6 4-29
Table 4-23 Capital Cost Adder (2019$/kW) for New Offshore Fixed Wind Plants in v6 4-30
Table 4-24 Capital Cost Adder (2019$/kW) for New Offshore Floating Wind Plants in v6 4-30
Table 4-25 Example Calculations of Wind Generation, Reserve Margin Contribution, and Capital Cost
for Onshore Wind in WECC_CO for Wind Class 7, Resource Class 5, and Cost Class 1 4-31
Table 4-26 Solar Photovoltaic Reserve Margin Contribution by Resource Class in v6 4-32
Table 4-27 Regional Assumptions on Potential Geothermal Electric Capacity in v6 4-33
Table 4-28 Potential Geothermal Capacity and Cost Characteristics by Model Region in v6 4-33
Table 4-29 Potential Non-Powered Dam in v6 4-35
Table 4-30 Potential New Stream Development in v6 4-37
Table 4-31 Bounds and Reserve Margin Contribution for Potential (New) Battery Storage in v6 4-38
Table 4-32 Energy Community Tax Credit Increment for Solar and Wind Units 4-40
Table 4-33 Energy Storage Mandates in v6 4-41
Table 4-34 Planned-Committed Units by Model Region in NEEDS for EPA Platform v6 Post-IRA 2022
Reference Case 4-43
Table 4-35 Onshore Average Capacity Factor by Wind Class, Resource Class, and Vintage in EPA
Platform v6 Post-IRA 2022 Reference Case 4-43
Table 4-36 Onshore Regional Potential Wind Capacity (MW) by Wind Class, Resource Class, and Cost
Class in EPA Platform v6 Post-IRA 2022 Reference Case 4-43
-------
Table 4-37 Wind Generation Profiles in EPA Platform v6 Post-IRA 2022 Reference Case (kWh of
Generation per MW of Capacity) 4-43
Table 4-38 Capital Cost Adder (2019$/kW) for New Onshore Wind Plants by Resource and Cost Class
in EPA Platform v6 Post-IRA 2022 Reference Case 4-44
Table 4-39 Solar Photovoltaic Regional Potential Capacity (MW) by Resource and Cost Class in EPA
Platform v6 Post-IRA 2022 Reference Case 4-44
Table 4-40 Solar Thermal Regional Potential Capacity (MW) by Resource and Cost Class in EPA
Platform v6 Post-IRA 2022 Reference Case 4-44
Table 4-41 Solar Photovoltaic Generation Profiles in EPA Platform v6 Post-IRA 2022 Reference Case
(kWh of Generation per MW of Capacity) 4-44
Table 4-42 Solar Photovoltaic Regional Capital Cost Adder (2019$/kW) for Potential Units by Resource
and Cost Class in EPA Platform v6 Post-IRA 2022 Reference Case 4-44
Table 4-43 Solar Thermal Regional Capital Cost Adder (2019$/kW) for Potential Units by Resource and
Cost Class in EPA Platform v6 Post-IRA 2022 Reference Case 4-44
Table 4-44 Solar Photovoltaic Average Capacity Factor by Resource Class and Vintage in EPA Platform
v6 Post-IRA 2022 Reference Case 4-44
Table 4-45 Solar Thermal Capacity Factor by Resource Class and Season in EPA Platform v6
Post-IRA 2022 Reference Case 4-44
Table 4-46 Potential Electric Capacity from New Landfill Gas Units in EPA Platform v6 Post-IRA 2022
Reference Case (MW) 4-44
Table 4-47 Characteristics of Existing Nuclear Units in EPA Platform v6 Post-IRA 2022 Reference
Case 4-44
Table 5-1 Retrofit Emission Control Options in v6 5-1
Table 5-2 Retrofit SO2 Emission Control Performance Assumptions in v6 5-2
Table 5-3 Illustrative Scrubber Costs (2019$) for Representative Capacities and Heat Rates in v6 5-4
Table 5-4 Retrofit NOx Emission Control Performance Assumptions in v6 5-5
Table 5-5 Illustrative Post Combustion NOx Control Costs (2019$) for Coal Plants for Representative
Sizes and Heat Rates under the Assumptions in v6 5-7
Table 5-6 Post-Combustion NOx Controls Costs (2019$) for Oil/Gas Steam for Representative Sizes
and Heat Rates under the Assumptions in v6 5-8
Table 5-7 Coal Units with Biomass Co-firing Option in v6 5-9
Table 5-8 Mercury Concentration Assumptions for Non-Coal Fuels in v6 5-10
Table 5-9 Mercury Emission Modification Factors Used in v6 5-11
Table 5-10 Definition of Acronyms for Existing Controls 5-12
Table 5-11 Keyto BurnerType Designations in Table 5-9 5-13
Table 5-12 Assignment Scheme for Mercury Emissions Control Using Activated Carbon Injection
in v6 5-15
Table 5-13 Illustrative Activated Carbon Injection (ACI) Costs (2019$) for Representative Sizes and
Heat Rates under the Assumptions in v6 5-18
Table 5-14 HCI Removal Rate Assumptions for Potential (New) and Existing Units in v6 5-19
Table 5-15 Retrofit HCI and SO2 Emission Control Performance Assumptions in v6 5-20
Table 5-16 Illustrative Dry Sorbent Injection (DSI) Costs (2019$) for Representative Sizes and Heat
Rates in v6 5-22
Table 5-17 Illustrative Particulate Controls Costs (2019$) for Representative Sizes and Heat Rates
in v6 5-22
Table 5-18 Cost and Performance Assumptions for Coal-to-Gas Retrofits in v6 5-23
Table 5-19 First Stage Retrofit Assignment Scheme in v6 5-26
Table 5-20 Second and Third Stage Retrofit Assignment Schemes in v6 5-27
Table 5-21 Cost of Building Pipelines to Coal Plants in EPA Platform v6 Post-IRA 2022 Reference
Case 5-28
ix
-------
Table 6-1 Cost and Performance Assumptions for Potential USC and NGCC with and without Carbon
Capture in v6 6-1
Table 6-2 Performance and Unit Cost (2019 $) Assumptions for Carbon Capture in v6 6-2
Table 6-3 Lower-48 CO2 Sequestration Capacity by Region (Gigatonnes) in v6 6-6
Table 6-4 CO2 Storage Cost Curves in EPA Platform v6 Post-IRA 2022 Reference Case 6-8
Table 6-5 C02 Transportation Matrix in EPA Platform v6 Post-IRA 2022 Reference Case 6-8
Table 7-1 Coal Supply Regions in EPA Platform v6 7-2
Table 7-2 Coal Rank Heat Content Ranges 7-4
Table 7-3 Coal Grade SO2 Content Ranges 7-4
Table 7-4 Coal Quality Characteristics by Supply Region and Coal Grade in v6 7-5
Table 7-5 Coal Clustering by Coal Grade - SO2 Emission Factors (Ibs/MMBtu) 7-7
Table 7-6 Coal Clustering by Coal Grade - Mercury Emission Factors (Ibs/TBtu) 7-7
Table 7-7 Coal Clustering by Coal Grade - Ash Emission Factors (Ibs/MMBtu) 7-8
Table 7-8 Coal Clustering by Coal Grade - HCI Emission Factors (Ibs/MMBtu) 7-8
Table 7-9 Coal Clustering by Coal Grade - CO2 Emission Factors (Ibs/MMBtu) 7-9
Table 7-10 Example of Coal Assignments Made in v6 7-10
Table 7-11 Basin-Level Groupings Used in Preparing v6 Coal Supply Curves 7-10
Table 7-12 Rail Competition Definitions 7-23
Table 7-13 Assumed Eastern Rail Rates for 2020 (2019 mills/ton-mile) 7-24
Table 7-14 Assumed Midwestern Rail Rates for 2020 (2019 mills/ton-mile) 7-24
Table 7-15 Assumed Non-PRB Western Rail Rates for 2020 (2019 mills/ton-mile) 7-25
Table 7-16 Assumed PRB Western Rail Rates for 2020 (2019 mills/ton-mile) 7-25
Table 7-17 Assumed Truck Rates for 2020 7-26
Table 7-18 Assumed Barge Rates for 2020 7-26
Table 7-19 Assumed Other Transportation Rates for 2020 7-27
Table 7-20 EIA AEO Diesel Fuel Forecast, 2020-2050 7-30
Table 7-21 Summary of Expected Escalation for Coal Transportation Rates, 2020-2050 7-31
Table 7-22 Coal Exports in v6 (Million Short Tons) 7-32
Table 7-23 Residential, Commercial, and Industrial Demand in v6 (Million Short Tons) 7-32
Table 7-24 Coal Import Limits in v6 (Million Short Tons) 7-33
Table 7-25 Coal Transportation Matrix in EPA Platform v6 Post-IRA 2022 Reference Case 7-35
Table 7-26 Coal Supply Curves in EPA Platform v6 Post-IRA 2022 Reference Case 7-35
Table 8-1 Supply/Demand Balance and Henry Hub Price for a GMM Run Underlying the Natural Gas
Supply Curves in v6 8-5
Table 8-2 Delivered Price Adders 8-9
Table 8-3 Refiners' Acquisition Cost of Crude (RACC) 8-12
Table 8-4 United States and Canada Projected Dry Gas Production by Source (Bcfd) 8-13
Table 8-5 GMM United States and Canada Gas Demand Projection (Bcfd) 8-14
Table 8-6 LNG Export Volumes and Capacity (Bcfd) 8-15
Table 8-7 U.S. Pipeline Exports to Mexico (Bcfd) 8-15
Table 8-8 EIA Style Gas Report for EPA Platform v6 Post-IRA 2022 Reference Case 8-15
Table 8-9 Natural Gas Basis for EPA Platform v6 Post-IRA 2022 Reference Case 8-15
Table 8-10 Natural Gas Supply Curves for EPA Platform v6 Post-IRA 2022 Reference Case 8-15
Table 9-1 Fuel Oil Prices by NEMS Region in v6 9-1
Table 9-2 Waste Fuels in v6 9-3
Table 9-3 Fuel Emission Factor Assumptions in v6 9-4
Table 9-4 Biomass Supply Curves for EPA Platform v6 Post-IRA 2022 Reference Case 9-4
Table 10-1 Summary Tax Changes 10-4
x
-------
Table 10-2 Financial Assumptions for Utility and Merchant Cases 10-5
Table 10-3 Weighted Average Cost of Capital in v6 10-6
Table 10-4 Share of Annual Thermal Capacity Additions by Market 10-7
Table 10-5 Capital Structure Assumptions in v6 10-9
Table 10-6 Nominal Debt Rates in v6 10-9
Table 10-7 Utilities Used to Calculate Cost of Debt 10-10
Table 10-8 Estimated Annual Levered Beta forS15ELUT Utility Index Based on Daily Returns 10-11
Table 10-9 Real Capital Charge Rate - Blended (%) in v6 10-13
Table 10-10 Real Capital Charge Rate - IPP (%) 10-13
Table 10-11 Real Capital Charge Rate - Utility (%) 10-14
Table 10-12 Book Life, Debt Life, and Depreciation Schedules in v6 10-14
xi
-------
LIST OF FIGURES
Figure 1-1 Modeling and Data Structures in EPA Platform v6 1-6
Figure 2-1 Hypothetical Chronological Hourly Load Curve and Seasonal Load Duration Curve for
Summer Season 2-8
Figure 2-2 Stylized Depiction of a Six Segment Load Duration Curve Dispatch Modeling 2-9
Figure 2-3 Stylized Dispatch Order in Illustrative Load Segments 2-10
Figure 3-1 EPA Platform v6 Model Regions 3-2
Figure 3-2 Modeling Process for Obtaining Projected NOx Emission Rates 3-22
Figure 3-3 How One of the Four NOx Modes Is Ultimately Selected for a Unit 3-23
Figure 4-1 Derivation of Plant Fixed O&M Data 4-12
Figure 5-1 Process for Lateral Cost Estimation 5-24
Figure 7-1 Map of the Coal Supply Regions in v6 7-3
Figure 7-2 Coal Mine Productivity (2000-2019) 7-15
Figure 7-3 Average Annual Cost Growth Assumptions by Region (2021-2050) 7-15
Figure 7-4 Maximum Annual Coal Production Capacity per Year (Million Short Tons) 7-16
Figure 7-5 Illustration of Preliminary Step in Developing a Cumulative Coal Supply Curve 7-17
Figure 7-6 Illustration of Final Step in Developing a Cumulative Coal Supply Curve 7-17
Figure 7-7 Example Coal Supply Curve in Stepped Format 7-18
Figure 7-8 Calculation of Multi-Mode Transportation Costs (Example) 7-21
Figure 7-9 Rail Cost Indices Performance (1Q2016-1Q2020) 7-28
Figure 7-10 Long-Run Marginal Cost Breakdown by Transportation Mode 7-29
Figure 8-1 GMM Gas Quantity and Price Response 8-2
Figure 8-2 IPM/GMM Interaction 8-2
Figure 8-3 Geographic Coverage of GMM 8-3
Figure 8-4 Demand Region Definition 8-6
Figure 8-5 Supply Region Definition 8-7
Figure 8-6 Supply Curves for 2028, 2030, 2035, 2040, 2045, and 2050 8-8
xii
-------
LIST OF ATTACHMENTS
Attachment 3-1 Incremental Demand Accounting for on-the-books EPA OTAQ GHG Final Rule (not
reflected in AEO2021) in EPA Platform v6 Post-IRA 2022 Reference Case 3-
Attachment 3-2 NOx Rate Development in EPA Platform v6 Post-IRA 2022 Reference Case 3-
Attachment 4-1 Nuclear Power Plant Life Extension Cost Development Methodology in EPA Platform
v6 Post-IRA 2022 Reference Case 4-
Attachment 5-1 Wet FGD Cost Methodology 5-
Attachment 5-2 SDA FGD Cost Methodology 5-
Attachment 5-3 SCR Cost Methodology for Coal-Fired Boilers 5-
Attachment 5-4 SCR Cost Methodology for Oil-Gas-Fired Boilers 5-
Attachment 5-5 SNCR Cost Methodology for Coal-Fired Boilers 5-
Attachment 5-6 SNCR Cost Methodology for Oil-Gas-Fired Boilers 5-
Attachment 5-7 DSI Cost Methodology 5-
Attachment 5-8 Hg Cost Methodology 5-
Attachment 5-9 PM Cost Methodology 5-
Attachment 5-10 Combustion Turbine NOx Control Technology Memo 5-
Attachment 6-1 CO2 Reduction Retrofit Cost Development Methodology 6
Attachment 7-1 Mining Cost Estimation Methodology and Assumptions 7-
xiii
-------
1.
Introduction
1.1 Executive Summary
This document describes the nature, structure, and capabilities of the Integrated Planning Model (IPM)
and the assumptions underlying the EPA's Power Sector Modeling Platform version 6 Post-IRA 2022
Reference Case (EPA Platform v6) that was developed by the U.S. Environmental Protection Agency
(EPA) with technical support from ICF, Inc. IPM is a multi-regional, dynamic, and deterministic linear
programming model of the U.S. electric power sector. The model provides forecasts of least cost
capacity expansion, electricity dispatch, and emission control strategies, while meeting energy demand,
environmental, transmission, dispatch, and reliability constraints. IPM can be used to evaluate the cost
and emissions impacts of proposed policies to limit emissions of sulfur dioxide (SO2), nitrogen oxides
(NOx), carbon dioxide (CO2), mercury (Hg), and hydrogen chloride (HCI) from the electric power sector.
This introduction chapter summarizes the key modeling capabilities and major data elements that are
described in greater detail in the subsequent chapters.
EPA Platform v6 incorporates important structural improvements and data updates with respect to the
previous version (v5). A new version number (moving from v5 to v6) indicates a substantial change to the
architecture. For example, the EPA Platform v6 has significantly more detailed representation of the load
segments and seasons. Further, this updated version of EPA Platform v6 uses demand projections from
the Energy Information Agency's (EIA) Annual Energy Outlook (AEO) 2021.
EPA Platform v6 documentation includes assumptions and data values that were used to produce the
Post-IRA 2022 Reference Case. For subsequent runs that examine various alternative futures, we include
separate documentation that makes clear where any assumptions or data values differ from the Post-IRA
2022 Reference Case conditions shown in this core documentation.
When policy analysis is conducted using EPA Platform v6, relevant assumptions and documentation will
be provided elsewhere accordingly.
EPA Platform v6 is a projection of electricity sector activity that considers only those Federal and state air
emission laws and regulations, and legislations whose provisions were either in effect or enacted as
documented in Section 3.10. Section 3.10 contains a detailed discussion of the environmental
regulations included in EPA Platform v6 Post-IRA 2022 Reference Case, which are summarized below.
• Inflation Reduction Act of 2022
• Proposed Good Neighbor Plan, a federal regulatory measure affecting EGU emissions from 25 states
to address transport under the 2015 National Ambient Air Quality Standards (NAAQS) for ozone.
• The Revised Cross-State Air Pollution Rule (CSAPR) Update, a federal regulatory measure affecting
EGU emissions from 12 states to address transport under the 2008 National Ambient Air Quality
Standards (NAAQS) for ozone.
• The Standards of Performance for Greenhouse Gas Emissions from New, Modified, and
Reconstructed Stationary Sources: Electric Utility Generating Units1 through rate limits.
• The Mercury and Air Toxics Rule (MATS),2 which was finalized in 2011. MATS establishes National
Emissions Standards for Hazardous Air Pollutants (NESHAP) for the "electric utility steam generating
unit" source category.
1 80 FR 64510
2 82 FR 16736
1-1
-------
• Current and existing state regulations. A summary of these state regulations can be found in Table
3-31.
• Current and existing Renewable Portfolio Standards and Clean Energy Standards (see Section
3.10.10)
• EPA Platform v6 reflects the latest actions EPA has taken to implement the Regional Haze
Regulations and Guidelines for Best Available Retrofit Technology (BART) Determinations Final
Rule3. The regulation requires states to submit revised State Implementation Plans (SIPs) that include
(1) goals for improving visibility in Class I areas on the 20% worst days and allowing no degradation
on the 20% best days and (2) assessments and plans for achieving Best Available Retrofit
Technology (BART) emission targets for sources placed in operation between 1962 and 1977. Since
2010, EPA has approved SIPs or, in a few cases, put in place regional haze Federal Implementation
Plans for several states. The BART limits approved in these plans (as of summer 2020) that will be in
place for EGUs are represented in the EPA Platform v6 (see Table 3-36).
• EPA Platform v6 reflects California AB 32 CO2 allowance price projections and the Regional
Greenhouse Gas Initiative (RGGI) rule (see Section 3.10.5).
• EPA Platform v6 also includes three non-air federal rules affecting EGUs: National Pollutant
Discharge Elimination System-Final Regulations to Establish Requirements for Cooling Water Intake
Structures at Existing Facilities and Amend Requirements at Phase I Facilities, Hazardous, and Solid
Waste Management System; Disposal of Coal Combustion Residuals from Electric Utilities; and the
Effluent Limitation Guidelines and Standards for the Steam Electric Power Generating Point Source
Category. (See Section 3.10.6)
• EPA Platform v6 reflects renewable portfolio standards and air emission regulations affecting EGUs
in Canada.
Table 1-1 lists key updates included in EPA Platform v6 Post-IRA 2022 Reference Case incremental to
the previous release of EPA Platform v6 Summer 2021 Reference Case with the corresponding data
sources. The updates are listed in the order in which they appear in the documentation.
Table 1-1 Key Updates and Specifications in the EPA Platform v6 Post-IRA 2022 Reference Case
Description
For More
Information
Modeling Framework
Modeling time horizon out to 2059 with seven model run years. These include 2028,
2030, 2035, 2040, 2045, 2050, and 2055.
Table 2-1
Power System Operation
Power system operations are updated based on recent data from EIA, NERC, and
FERC.
Chapter 3
The electricity demand projection is based on AEO 2021. Also added are the
incremental demand provided by EPA's Office of Transportation and Air Quality
(OTAQ) on the book rules4 that are not captured in the AEO 2021 demand
projections.
Section 3.2
The SPP reserve margins are updated to 15%.
Section 3.6
Inventory of state emission regulations is updated.
Section 3.10
IRA Provisions, GNP Proposal, CSAPR, MATS, and BART are reflected.
Section 3.10.4,
Section 4.5
All CCR/ELG cost adders are removed for the list of units with CCR-based
retirements (with a retirement date no later than 12/31/2028).
Section 3.10.6
3 70 FR 39104
4 https://vwwy.epa.qov/requlations-emissions-vehicles-and-enqines/final-rule-revise-existinq-national-qhq-emissions
1-2
-------
Description
For More
Information
Inventory of RPS and CES standards are updated.
Table 3-18, Table
3-20
Generating Resources
NEEDS planned units, retirements, and emission control configurations are updated
based on 2018 EIA Form 860, October 2021 EIA Form 860M, December 2021 EIA
Table 4-1
Form 860M, AEO 2020, AMPD 2019, and recent lists of deactivations from PJM,
MISO, and ERCOT.
The default NOx rates of selected NEEDS units with NOx rates are updated using
Section 3.10.3
NOx emissions data from NEI 2018 and state-level inventories.
Minimum capacity factor requirements of 10% are applied to existing coal steam
Section 3.5.2
units in regions without capacity markets.
Nuclear units have a prespecified life and are no longer endogenously retired.
Section 4.6.1
Cost and performance characteristics for potential (new) units are updated based on
Table 4-12 and Table
AEO 2021 and NREL ATB 2021.
4-15
Wind and solar technologies have revised cost and resource base estimates based
on NREL ATB 2021.
Section 4.4.5
Energy storage options of both 4-hour and 8-hour durations are based on NREL
ATB 2021. Also, a new capacity credit methodology is implemented for energy
Section 4.4.5
storage units that respond to the level of penetration of energy storage in a region.
Tax credit extensions from the Inflation Reduction Act of 2022 are implemented for
Section 4.4.5
wind, solar, hydro, geothermal, landfill gas, energy storage, biomass, and 45Q.
Emission Control Technologies
Cost and performance assumptions for SCR controls are updated
Section 5.2.3
Pipeline lateral costs for coal-to-gas-retrofits are updated
Section 5.2.3
Carbon Capture, Transport, and Storage
45Q is modeled in the 2030 and 2035 run years.
Section 3.12
Cost and performance assumptions for CCS controls are updated. Capital cost
reductions are implemented overtime for CCS retrofits
Section 6.1.2
CO2 transportation cost adders reflect a transport cost algorithm that is updated
based on a single, separate pipeline being used for each power plant all the way
Section 6.3
from the source to the sink.
Natural Gas
Natural gas assumptions as of the end of 2021 are modeled through annual gas
supply curves and IPM region-level seasonal basis differentials.
Chapter 8
Other Fuels
A hydrogen fuel price of 7.4 $/MMBtu is assumed.
Chapter 9
Financial assumptions
Cost adder for new non-peaking fossil units associated with future CO2 emissions is
no longer applied.
Section 10.7.3
1-3
-------
Table 1-2 lists the types of plants included in the EPA Platform v6.
Table 1-2 Plant Types in v6
Conventional Technologies
Coal Steam
Oil/Gas Steam
Combustion Turbine
Combined-Cycle Combustion Turbine
Integrated Gasification Combined-Cycle (IGCC) Coal
Ultra-Supercritical Coal with and without Carbon Capture
Fluidized Bed Combustion
Nuclear
Renewables and Non-Conventional Technologies
Hydropower
Pumped Storage
Energy Storage
Biomass
Onshore Wind
Offshore Wind
Fuel Cells
Distributed Solar Photovoltaics
Solar Photovoltaics
Solar Thermal
Geothermal
Landfill Gas
Other1
Note:
1 Included are fossil and non-fossil waste plants.
1-4
-------
Table 1-3 lists the emission control technologies available for meeting emission limits in EPA Platform v6.
Table 1-3 Emission Control Technologies in v6
Sulfur Dioxide (SO2)
Limestone Forced Oxidation (LSFO)
Lime Spray Dryer (LSD)
Nitrogen Oxides (NOx)
Combustion controls
Selective catalytic reduction (SCR)
Selective non-catalytic reduction (SNCR)
Mercury (Hg)
Combinations of SO2, NOx, and particulate control technologies
Activated Carbon Injection
Hydrogen Chloride (HCI)
Dry Sorbent Injection (with milled Trona)
Carbon Dioxide (CO2)
Heat rate improvement
Coal-to-gas
Carbon Capture and Sequestration
Notes:
Fuel switching between coal types is also a compliance option
for reducing emissions in EPA Platform v6.
Figure 1-1 provides a schematic of the components of the modeling and data structure used for EPA
Platform v6. The document contains separate chapters devoted to all the key components shown in
Figure 1-1. Chapter 2 provides an overview of IPM's modeling framework (also referred to as the IPM
Engine), highlighting the mathematical structure, notable features of the model, programming elements,
and model inputs and outputs. The remaining chapters are devoted to different aspects of EPA Platform
v6. Chapter 3 covers the operating characteristics of the power system. Chapter 4 explores the
characterization of electric generation resources. Emission control technologies and carbon capture,
transport, and storage are discussed in chapters 5 and 6. The next three chapters discuss the
representation of and assumptions for fuels. Coal is covered in chapter 7, natural gas in chapter 8, and
other fuels (i.e., fuel oil, biomass, nuclear fuel, and waste fuels) in chapter 9 (along with fuel emission
factors). Finally, chapter 10 summarizes the financial assumptions.
1-5
-------
Figure 1-1 Modeling and Data Structures in EPA Platform v6
Parsing Outputs
I ndiv idual Boiler Lev el Data
Retail Price Projection
Notes:
* information on existing and planned electric generating units (EGUs) is contained in the National Electrical
Energy Data System (NEEDS) data base maintained for EPA bylCF, Inc. Planned EGUs are those which
were under construction or had obtained financing at the time EPA's Platform v6 was finalized.
Outputs for Air Quality
Modeling
Criteria Air Pollutants
Non-criteria Air Pollutants
Toxics Air Pollutants
Point Source Locators
**IPM Engine is the model structure described in Chapter 2
1.2 Review and Ongoing Improvement of the Integrated Planning Model
A customized, fully documented version of the data assumptions underlying IPM has been developed and
used by EPA to help inform power plant air regulatory and legislative efforts for more than 25 years,
following the enactment of the Clean Air Act Amendments of 1990. The model has been tailored to meet
the unique environmental considerations important to EPA, while also fully capturing the detailed and
complex economic and electric dispatch dynamics of power plants across the country. EPA's goal is to
explain and document the agency's use of the model in a transparent and publicly accessible manner,
while also providing for concurrent channels for improving the model's assumptions and representation by
soliciting constructive feedback to improve the model. This includes making all inputs and assumptions to
the model, as well as output files from the model, publicly available on EPA's website (and, when applied
to inform a rulemaking, in the relevant publicly accessible regulatory docket).
EPA's use of IPM depends upon a variety of environmental, policy, and regulatory considerations. EPA's
version of the model input assumptions has undergone significant updates and architectural
improvements every 2-4 years to best reflect the evolving dynamics of the power sector, and smaller
ongoing updates (1-2 times a year) to reflect changes in fleet composition (retirements, new capacity
1-6
-------
builds, and installed retrofits). Currently, EPA's implementation of IPM is in its sixth major version, not
including Coal and Electric Utility Model (CEUM), the model used by EPA before its use of IPM.
Federal Regulatory efforts:
EPA has used IPM for many regulatory efforts affecting the power sector, including:
• The NOx SIP Call, the Clean Air Interstate Rule (2004-2006), the Clean Air Visibility Rule, the
Clean Air Mercury Rule (2005), the Cross-State Air Pollution Rule and Updates (2010-2023), the
Mercury and Air Toxics Rule (2012), the Clean Power Plan (2015), Affordable Clean Energy Rule
(2019) and various Ozone, PM NAAQS, and regional haze regulatory efforts.
National Legislative efforts:
EPA has used IPM to support legislative efforts that affect the power sector, including:
• The Clear Skies Act (2002-2005), the Clean Air Planning Act (2002-2005), the Clean Power Act
(2002-2005), the Climate Stewardship and Innovation Act (2007), the Low Carbon Economy Act
(2007-2008), the Lieberman-Warner Climate Security Act (2007-2008), and the American Clean
Energy and Security Act (2008-2009).
Notable Versions and Updates/Improvements/Enhancements:
EPA Base Case using IPM - 1996
• Designed for projections covering the US with 4 run years
• Disaggregated the US into 17 IPM model regions
• Modeled coal and gas markets through coal and gas supply curves
EPA Base Case using IPM - 1998
• Updated unit inventory of power plants
• Increased the number of IPM model regions covering the US from 17 to 21
• Disaggregated New York into 4 IPM model regions
• Increased the number of run years from 4 to 6
EPA Base Case 2000 using IPM Version 2.1 (2000-2003)
• Updated unit inventory of power plants
• Increased the number of IPM model regions covering the US from 21 to 26
• Increased the modeling time horizon to 2030
• Increased the overall number of emission control technology options modeled
• Incorporated Activated Carbon Injection (ACI) retrofit options for mercury control modeling
• Expanded coal supply representation
EPA Base Case 2004 using IPM Version 2.1.9 (2004)
• Updated unit inventory of power plants
• Improved the characterization of SO2 and NOx emissions
• Revised coal choice assumptions for individual coal units
• Updated natural gas supply curves, incorporating recommendations from the natural gas peer
review
1-7
-------
EPA Base Case 2006 using IPM Version 3 (2005-2009)
• Updated unit inventory of power plants
• Improved environmental pollution control retrofit assumptions
• Increased the number of IPM model regions covering the US from 26 to 32 to enhance regional
representation
• Increased the number of load segments from 5 to 6 to enhance electric load representation
• Updated natural gas supply curves based on ICF's North American Natural Gas Systems
Analysis (NANGAS) model
• Updated coal supply curves
• Enhanced electric transmission capabilities and imports/exports
• Enhanced power plant representation detail
EPA Base Case using IPM Version 4.10 (2010-2013)
• Updated unit inventory of power plants
• Integrated Canada into the modeling framework
• Incorporated HCI emissions and Dry Sorbent Injection retrofit options
• Improved resolution of carbon capture and storage modeling by including regional storage
representation and transportation network
• Updated coal supply modeling with significantly more resolution of coal mine data
• Incorporated natural gas resource model for North America to reflect emerging shale resource
• Enhanced power plant representation detail to support toxic air pollutant emissions and controls
EPA Base Case using IPM Version 5 (2014-2017)
• Updated unit inventory of power plants
• Doubled the number of IPM model regions from 36 to 64
• Revised environmental pollution control retrofit assumptions for conventional pollutants and toxic
emissions
• Incorporated additional technology options for new power plants
• Overhauled coal supply assumptions, with even further resolution to reflect mine-by-mine
geography and coal characteristics
• Improved coal transportation network by modeling each individual coal plant as its own coal
demand region
• Updated gas modeling assumptions to reflect natural gas shale supply/trends and pipeline
capacity expansion
EPA Base Case using IPM Version 6 (2017-2023)
• Continuously updated unit inventory of power plants
• Revised environmental pollution control retrofit assumptions for conventional pollutants and toxic
emissions
• Increased the number of seasons from 2 to 3 and the number of load segments for each season
from 6 to 24
• Aggregated hours in load segments based on predefined time of day categories.
• Inputs for generation profiles for wind and solar technologies at an hourly level.
• Implemented capacity credit assumptions for wind, solar, and energy storage units that
deteriorate with an increase in their penetration.
• Performed a comprehensive update of coal and natural gas supply and transportation
assumptions.
• Updated generation technology costs
• Enabled functionality to model endogenous transmission builds
• Implemented capability to model operating reserves
1-8
-------
• Revised the model time horizon to 2028-2059
• Implemented the impact of Inflation Reduction Act of 2022
Background on EPA Base Case using IPM Review:
Peer Reviews:
EPA conducts periodic peer review of the EPA Base Case application of IPM. The reviews have included
separate expert panels on the model itself and on EPA's key modeling input assumptions. For example,
separate panels of independent experts have been convened to review the EPA Base Case application of
IPM's coal supply and transportation assumptions, natural gas assumptions, and model formulation.
EPA IPM v6 Reference Case Peer Review
In September 2019, EPA commissioned a peer review of EPA's v6 Reference Case. An independent
contractor facilitated a formal peer review process in compliance with EPA's Peer Review Handbook
(U.S. EPA, 2006). A panel of peer reviewers with extensive expertise in energy policy, power sector
modeling and economics reviewed the EPA Version 6 Reference Case and provided feedback in the form
of a report.5 The peer reviewers evaluated the adequacy of the framework, assumptions, and supporting
data used in the EPA Version 6 Reference Case using IPM, and they suggested potential improvements.
Overall, the panel found much to commend EPA; stating that the modeling platform:
lends itself well to EPA analyses of air policy focused on the power sector
includes significant detail related to electricity supply and demand
includes data-rich representation both across different geographic areas and across time
provides a reasonable representation of power sector operations, generating technologies,
emissions performance and controls, and markets for fuels used by the power sector
is well suited to assess the costs and emissions impacts
documentation is well written, clearly organized, and detailed in its presentation of most model
characteristics
EPA has posted a response document to this Peer Review Report detailing the latest improvements in
capabilities and documentation, and potential future improvements.
EPA Base Case v5.13 Data Assumption Review
In 2015, an independent peer review panel provided expert feedback on whether the analytical
framework, assumptions, and applications of data in IPM were sufficient for the EPA's needs in estimating
the economic and emissions impacts associated with the power sector. The panel identified several
strengths associated with the model and underlying data and assumptions. For example, the report
stated that EPA's platform exceeds other model capabilities in providing a relevant feedback mechanism
between the electric power model and key fuel inputs that drive simulation results.6
Other strengths the panel identified include:
• The detail with which pollution control technology options and costs are represented
• The level of detail at which federal Clean Air Act (CAA) regulations are represented
• The ability of the model to allow for the detailed representation of a variety of potential changes in
energy and environmental policies, including important features of market-based programs
• The accuracy of the emissions control costs and their relationship to retirement decisions
5 https://www.epa.gov/power-sector-modeling/ipm-peer-reviews
6 https://www.epa.gov/power-sector-modeling/ipm-peer-reviews
1-9
-------
• The expansion of model regions from 32 to 64, which allows the model to better represent current
power market operations and existing transmission bottlenecks even within regional transmission
organization (RTO) regions
• Continuous updates of the representation of domestic coal and natural gas market conditions
The peer review panel has also provided several areas for investigation and additional recommendations
for the EPA's consideration, including:
• Improved documentation of the input assumptions
• Changes to certain cost functions and financial assumptions
• Consideration of certain improvements to the Base Case architecture (additional seasonal
representation, representation of electric demand, transmission considerations, and renewable
energy representation among others)
The EPA Platform v6 using IPM addresses many of the recommendations (seasons, renewable energy
representation, regional representation, etc.). The peer review has also led to additional work at EPA to
further understand and better represent some of the emerging issues in the power sector. EPA intends to
add more capabilities and continue to refine the modeling platform to reflect these comments and adopt
those changes at an appropriate time after further research and testing of the model.
Coal Market Assumptions Review
In 2003, a group of experts in the field of cost, quality, reserves, and availability of coal were selected as
peer reviewers to assess whether the choice, use, and interpretation of data and methodology employed
in the derivation of the IPM coal supply curves was appropriate and analytically sound. The peer
reviewers were charged with:
• Evaluating the appropriateness of the overall methodology used to develop the new coal supply
curves
• Assessing the adequacy of the individual components employed in building the coal supply
curves in terms of both the approach and data used
• Assessing the technical soundness of the resulting coal supply curves for each coal type and
supply region in terms of the cost/quantity relationship and the characteristics associated with the
coal (e.g., sulfur, heat, and mercury content)
• Assessing the appropriateness of the use of this set of supply curves for use in production cost
models in general (of which IPM is a particular example)
The review process produced useful and specific recommendations for improvements and updates to the
coal supply information represented in IPM, which were subsequently incorporated into the model.
Gas Market Assumptions Review
In 2003, a peer review of the natural gas supply assumptions implemented in EPA Base Case using IPM
v.2.1.6 (2003) was performed. The peer reviewers were charged with evaluating the following:
• The appropriateness of the representation of all the key natural gas market fundamentals in
NANGAS
• The reasonableness of the natural gas supply curves, non-electricity demand assumptions and
transportation adders
• The reasonableness of the iteration process between NANGAS and IPM
The review commended the comprehensiveness of the approach used to generate the gas supply curves
implemented in the EPA Base Case. The review further identified assumptions that could be revised in
generating a new set of natural gas supply curves, as well as nonelectric-sector gas demand curves, for
the next update of the EPA Base Case.
1-10
-------
IPM Formulation Review
Conducted in 2008, this peer review focused on IPM's core mathematical formulation. The objective of
the review was to obtain expert feedback on the adequacy of the formulation in representing the
economic and operational behavior of the power sector over a modeling time horizon of 20-50 years.
The panel identified several strengths of IPM, including:
• The model's ability to compute optimal capacity that combined short-term dispatch decisions with
long-term investment decisions
• The model's integration of relevant markets, including the electric power, fuel, and environmental
markets, into a single modeling framework
• And the model's ability to represent a very detailed level of data regarding the emissions
modeling capability
The peer review panel also provided several areas for investigation and recommendations for the EPA's
consideration. These peer reviews led to changes, enhancements, and updates to the IPM framework to
better represent the power sector and related markets (i.e., fossil fuels).
Regulatory Review:
The formal rulemaking process provides an opportunity for expert review and comment by key
stakeholders. Formal comments as part of a rulemaking are reviewed and evaluated, and changes and
updates are made to IPM where appropriate. Stakeholders to EPA regulatory efforts are a diverse group,
including regulated entities and impacted industries, fuel supply companies, states, environmental
organizations, developers of other models of the U.S. electricity sector, and others. The feedback
provides a highly detailed review of input assumptions, model representation, and model results.
Other Uses and Reviews:
• IPM has been used by many regional organizations for regulatory support, including the Regional
Greenhouse Gas Initiative (RGGI), the Western Regional Air Partnership (WRAP), and the Ozone
Transport Assessment Group (OTAG). IPM has also been used by other Federal agencies (e.g.,
FERC, USDA), environmental groups, and many electric utilities.
• The Science Advisory Board reviewed EPA's application of IPM as part of the CAAA Section 812
prospective study 1997-1999.
• The President's Council of Economic Advisors (2002-2003) performed head-to-head comparison
of IPM and ElA's NEMS system for use in multi-pollutant control analysis.
• IPM has been used in several comparative model exercises sponsored by Stanford University's
Energy Modeling Forum and other organizations.
EPA Platform v6 using IPM represents a major iteration of EPA's application of IPM, with notable
structural and platform improvements and enhancements, as well as universal updates to reflect the most
current set of data and assumptions, coupled with continuous routine input data and assumption updates.
1-11
-------
2. Modeling Framework
ICF developed the Integrated Planning Model (IPM) to support analysis of the electric power sector. The
EPA, in addition to other state air regulatory agencies, utilities, and public and private sector entities, has
used IPM extensively for various air regulatory analyses, market studies, strategy planning, and economic
impact assessments.
IPM is a long-term capacity expansion and production-costing model of the electric power sector. Its
mathematical formulation is based on a Linear Programming (LP) structure. The structure provides for
several advantages, one of which is the guarantee of a globally optimal solution. Fast and efficient
commercial solvers exist to solve LP models. The solved dual variables (also known as shadow prices)
of each constraint modeled in IPM inform EPA rulemaking or policy analysis process in regard to the
marginal cost pricing of energy, capacity, fuels, and emission allowances. Also, reasonable solution
times for an LP model allow EPA to gain insights by modeling a large number of scenarios in a relatively
short period of time.
The first section of this chapter provides a brief overview of the model's purpose, capabilities, and
applications. The following sections are devoted to describing the IPM's model structure and formulation
(2.2), key methodological characteristics (2.3), and programming features (2.4), including its handling of
model inputs and outputs. Readers may find some overlap between sections. For example, transmission
decision variables and constraints are covered in the discussion of model structure and formulation in
section 2.2, and transmission modeling is covered as a key methodological feature in section 2.3.7. The
different perspectives of each section are designed to provide readers with information that is
complementary rather than repetitive.
2.1 IPM Overview
IPM is a well-established model of the electric power sector designed to help government and industry
analyze a wide range of issues related to this sector. The model represents economic activities in key
components of energy markets - fuel markets, emission markets, and electricity markets. Since the
model captures the linkages in electricity markets, it is well suited for developing integrated analyses of
the impacts of alternative regulatory policies on the power sector. In the past, applications of IPM have
included capacity planning, environmental policy analysis and compliance planning, wholesale price
forecasting, and power plant asset valuation.
2.1.1 Purpose and Capabilities
IPM is a dynamic linear programming model that generates optimal decisions under the assumption of
perfect foresight. It determines the least-cost method of meeting energy and peak demand requirements
over a specified period. In its solution, the model considers a number of key operating or regulatory
constraints that are placed on the power, emissions, and fuel markets. The constraints include, but are
not limited to, emission limits, transmission capabilities, renewable generation requirements, and fuel
market constraints. The model is designed to accommodate complex treatment of emission regulations
involving trading, banking, and special provisions affecting emission allowances (e.g., bonus allowances
and progressive flow control), as well as traditional command-and-control emission policies.
IPM represents power markets through model regions that are geographical entities with distinct
operational characteristics. The model regions are largely consistent with the North American Electric
Reliability Council (NERC) assessment regions, and with the organizational structures of the Regional
Transmission Organizations (RTOs), and the Independent System Operators (ISOs) that handle dispatch
on most of the U.S. grid. IPM represents the least-cost arrangement of electricity supply (capacity and
generation) within each model region to meet assumed future load (electricity demand) while constrained
by a transmission network of bulk transfer limitations on interregional power flows. All utility-owned
existing electric generating units, including renewable resources, as well as independent power producers
and cogeneration facilities selling electricity to the grid, are modeled.
2-1
-------
IPM provides a detailed representation of new and existing resource options. These include fossil,
nuclear, renewable, storage, and non-conventional options. Fossil options include coal steam, oil/gas
steam, combined cycles, and simple cycle combustion turbines. Renewable options include wind, landfill
gas, geothermal, solar thermal, solar photovoltaic, and biomass. Storage options include pump storage
and battery storage. Non-conventional options include fuel cell.
IPM can incorporate a detailed representation of fuel markets and can endogenously forecast fuel prices
for coal, natural gas, and biomass by balancing fuel demand and supply for electric generation. The
model also includes detailed fuel quality parameters to estimate emissions from electric generation.
IPM provides estimates of air emission changes, regional wholesale energy and capacity prices,
incremental electric power system costs, changes in fuel use, and capacity and dispatch projections.
2.1.2 Applications
IPM's structure, formulation, and set-up make it adaptable and flexible. The necessary level of data,
modeling capabilities exercised, and computational requirements can be tailored to the strategies and
policy options being analyzed. This adaptability has made IPM suitable for a variety of applications.
These include:
Air Regulatory Assessment: Since IPM contains extensive air regulatory modeling features, state and
federal air regulatory agencies have used the model extensively in support of air regulatory assessment.
Integrated Resource Planning: IPM can be used to perform least-cost planning studies that
simultaneously optimize demand-side options (load management and efficiency), renewable options and
traditional supply-side options.
Strategic Planning: IPM can be used to assess the costs and risks associated with alternative utility and
consumer resource planning strategies as characterized by the portfolio of options included in the input
database.
Options Assessment: IPM allows industry and regulatory planners to screen alternative resource options
and option combinations based upon their relative costs and contributions to meeting customer demands.
Cost and Price Estimation: IPM produces realistic estimates of energy prices, capacity prices, fuel prices,
and allowance prices. Industry and regulatory agencies have used these cost reports for due diligence,
planning, litigation, and economic impact assessment.
2.2 Model Structure and Formulation
IPM employs a linear programming structure that is particularly well-suited for analysis of the electric
sector to help decision makers plan system capacity and model the dispatch of electricity from individual
units or plants. The model consists of three key structural components:
• A linear objective function
• A series of decision variables
• A set of linear constraints
• The sections below describe the objective function, key decision variables, and constraints
included in IPM for EPA Platform v6.
2-2
-------
2.2.1 Objective Function
IPM's objective function is to minimize the total, discounted net present value of the costs of meeting
demand, power operation constraints, and environmental regulations over the entire planning horizon.
The objective function represents the summation of all the costs incurred by the electricity sector on a net
present value basis. These costs, which the linear programming formulation attempts to minimize,
include the cost of new plant and pollution control construction, fixed and variable operating and
maintenance costs, and fuel costs. Many of these cost components are captured in the objective function
by multiplying the decision variables by a cost coefficient. Cost escalation factors are used in the
objective function to reflect changes in cost overtime. The applicable discount rates are applied to derive
the net present value for the entire planning horizon from the costs obtained for all years in the planning
horizon.
2.2.2 Decision Variables
Decision variables represent the values for which the IPM model is solving, given the cost-minimizing
objective function described in Section 2.2.1 and the set of electric system constraints detailed in Section
2.2.3. The model determines values for these decision variables that represent the optimal least-cost
solution for meeting the assumed constraints. Key decision variables represented in IPM are described in
detail below.
Generation Dispatch Decision Variables: IPM includes decision variables representing the generation
from each model power plant.7 For each model plant, a separate generation decision variable is defined
for each possible combination of fuel, season, model run year, and segment of the seasonal load duration
curve applicable to the model plant. (See Section 2.3.5 below for a discussion of load duration curves.)
In the objective function, each plant's generation decision variable is multiplied by the relevant heat rate
and fuel price (differentiated by the appropriate step of the fuel supply curve) to obtain a fuel cost. It is
also multiplied by the applicable variable operation and maintenance (VOM) cost rate to obtain the VOM
cost for the plant.
Capacity Decision Variables: IPM includes decision variables representing the capacity of each existing
model plant and capacity additions associated with potential (new) units in each model run year. In the
objective function, the decision variables representing existing capacity and capacity additions are
multiplied by the relevant fixed operation and maintenance (FOM) cost rates to obtain the total FOM cost
for a plant. The capacity addition decision variables are also multiplied by the investment cost and capital
charge rates to obtain the capital cost associated with the capacity addition.
Operating Reserve Decision Variables: IPM includes decision variables representing the contribution of
each model plant to meet operating reserve requirements. While a model plant can contribute to both
energy and operating reserve requirements, the total contribution is limited by the total capacity of the
model plant.
Transmission Decision Variables: IPM includes decision variables representing the electricity
transmission along each transmission link between model regions in each run year. In the objective
function, these variables are multiplied by variable transmission cost rates to obtain the total cost of
transmission across each link.
Emission Allowance Decision Variables: For emission policies where allowance trading applies, IPM
includes decision variables representing the total number of emission allowances for a given model run
year that are bought and sold in that or subsequent run years. In the objective function, these year-
differentiated allowance decision variables are multiplied by the market price for allowances prevailing in
7 Model plants are aggregate representations of real-life electric generating units. They are used by IPM to model the
electric power sector. For a discussion of model plants in EPA Platform v6, see Section 4.2.6.
2-3
-------
each run year. This formulation allows IPM to capture the inter-temporal trading and banking of
allowances.
Fuel Decision Variables: For each type of fuel and each model run year, IPM defines decision variables
representing the quantity of fuel delivered from each fuel supply region to model plants in each demand
region. Coal decision variables are further differentiated according to coal rank (bituminous, sub-
bituminous, and lignite), sulfur grade, chlorine content and mercury content. These fuel quality decision
variables do not appear in the IPM objective function, but in constraints which define the types of fuel that
each model plant is eligible to use and the supply regions that are eligible to provide fuel to each specific
model plant.
2.2.3 Constraints
Model constraints are implemented in IPM to accurately reflect the characteristics of, and the conditions
faced by, the electric sector. Among the key constraints included in EPA Platform v6 are:
Reserve Margin Constraints: Regional reserve margin constraints capture system reliability requirements
by defining a minimum margin of reserve capacity (in megawatts) per year beyond the total capacity
needed to meet future peak demand that must remain in service to that region. These reserve capacity
constraints are derived from reserve margin targets that are assumed for each region based on
information from NERC, RTOs, or ISOs. If existing plus planned capacity is not sufficient to satisfy the
annual regional reserve margin requirement, the model will build the required level of new capacity.
Section 3.6 further discusses reserve margin assumptions.
Operating Reserve Constraints: These constraints specify the operating reserve requirements by product
type and region that need to be met by the power system.
Demand Constraints: The model categorizes regional annual electricity demand into seasonal load
curves which are used to form winter (December 1 - February 28), winter shoulder (March 1 - April 30
and October 1 - November 30), and summer (May 1 - September 30) load duration curves (LDC). The
seasonal load segments, when taken together, represent all the hourly electricity load levels that must be
satisfied in a particular region, season, and model run year. As such, the LDC defines the minimum
amount of generation required to meet the region's electricity demand during the specific season. These
requirements are specified by demand constraints.
Capacity Factor Constraints: These constraints specify how much electricity each plant can generate,
given its capacity and seasonal availability.
Turn Down Constraints: The model uses turn down constraints to account for the cycling capabilities of
generation resources, i.e., whether they can be shut down at night or on weekends, must operate at all
times, or must operate at least at some minimum capacity level. The constraints ensure that the model
reflects the distinct operating characteristics of peaking, cycling, and base-load units.
Emissions Constraints: IPM can endogenously consider an array of emissions constraints for SO2, NOx,
HCI, mercury, and CO2. Emission constraints can be implemented on a plant-by-plant, regional, or
system-wide basis. The constraints can be defined in terms of a total tonnage cap (e.g., tons of SO2) or a
maximum emission rate (e.g., Ib/MMBtu of NOx). The scope, timing, and definition of the emission
constraints depend on the required analysis.
Transmission Constraints: IPM can simultaneously model any number of regions linked by transmission
lines. The constraints define either a maximum capacity on each link or a maximum level of transmission
on two or more links (i.e., joint limits) to different regions.
2-4
-------
Fuel Supply Constraints: These constraints define the types of fuel that each model plant is eligible to
use and the supply regions that are eligible to provide fuel to each specific model plant. A separate
constraint is defined for each model plant.
2.3 Key Methodological Features of IPM
IPM is a flexible modeling tool for obtaining short- and long-term projections of production activity in the
electric generation sector. The projections obtained using IPM are not statements of what will happen.
Rather, they are estimates of what might happen given the assumptions and methodologies used.
Chapters 3 to 10 contain detailed discussions of the cost and performance assumptions specific to EPA
Platform v6. The present section provides an overview of the essential methodological and structural
features of IPM that extend beyond the assumptions that are specific to EPA Platform v6.
2.3.1 Model Plants
Model plants are a central structural component that IPM uses: (1) to represent aggregations of existing
generating units, (2) to represent retrofit and retirement options that are available to existing generating
units, and (3) to represent potential (new) generating units that the model can build.
Existing Units: Theoretically, there is no predefined limit on the number of generating units that can be
included in IPM. However, to keep model size and solution time within acceptable limits, EPA utilizes
model plants to represent aggregations of actual individual generating units. The aggregation algorithm
groups units with similar characteristics for representation by model plants with a combined capacity and
weighted-average characteristics that are representative of all the units comprising the model plant.
Model plants are defined to maximize the accuracy of the model's cost and emissions estimates by
capturing variations in key features of those units that are critical in the EPA Platform v6 and anticipated
policy case runs. For EPA Platform v6, EPA employed an aggregation algorithm, which allowed 23,290
actual existing electric generating units to be represented by 3,633 model plants. Section 4.2.6 describes
the aggregation procedure.
Retrofit and Retirement Options: IPM also utilizes model plants to represent the retrofit and retirement
options that are available to existing generating units. EPA Platform v6 provides existing model plants
with a wide range of options for retrofitting with emission control equipment as well as with an option to
retire. (See Chapter 5 for a detailed discussion of the options that are included.) Model plants that
represent potential (new) generation resources are not given the option to take on a retrofit or to retire.
The options available to each model plant are pre-defined at the model set-up. The retrofit and retirement
options are themselves represented in IPM by model plants, which, if actuated during a model run, take
on all or a portion of the capacity initially assigned to a model plant, which represents existing generating
units.8 In setting up IPM, parent-child-grandchild relationships are pre-defined between each existing
model plant (parent) and the specific retrofit and retirement model plants (children and grandchildren) that
may replace the parent model plant during the course of a model run. The child and grandchild model
plants are inactive unless the model finds it economical to engage one of the options provided, e.g.,
retrofit with particular emission controls or retire.
Theoretically, there are no limits on the number of successive retrofit and retirement options that can be
associated with each existing model plant. However, model size and computational considerations
dictate that the number of successive retrofits be limited. In EPA Platform v6, a maximum of three stages
of retrofit options are provided. For example, an existing model plant may retrofit with an activated
carbon injection (ACI) for mercury control in one model run year (stage 1), with a selective catalytic
8 IPM has a linear programming structure whose decision variables can assume any value within the specified
bounds subject to the constraints. Therefore, IPM can generate solutions where model plants retrofit or retire a
portion of the model plants capacity. IPM's standard model plant outputs explicitly present these partial investment
decisions.
2-5
-------
reduction (SCR) for NOx control in the same or subsequent run year (stage 2), and with a carbon capture
and sequestration (CCS) for CO2 control in the same or subsequent run year (stage 3). However, if it
exercises this succession of retrofit options, no further retrofit or retirement options are possible beyond
the third stage.
Potential (New) Units: IPM also uses model plants to represent new generation capacity that may be built
during a model run. All the model plants representing new capacity are pre-defined at set-up. They are
differentiated by type of technology, regional location, and years available. When it is economically
advantageous to do so (or otherwise required by reserve margin constraints to maintain electric
reliability), IPM builds one or more of these predefined model plants by raising its generation capacity
from zero during a model run. In determining whether it is economically advantageous to build new
plants, IPM considers cost differentials between technologies, expected technology cost improvements
(by differentiating costs based on a plant's vintage, i.e., build year), and regional variations in capital costs
that are expected to occur overtime.
Parsing: Since EPA Platform v6 results are presented at the model plant level, EPA has developed a
post-processor, a parsing tool, designed to translate results at the model plant level into generating unit-
specific results. The parsing tool produces unit-specific emissions, fuel use, emission control retrofit, and
capacity projections based on model plant results. Another post-processing activity involves deriving
inputs for air quality modeling from IPM outputs. This entails using emission factors to derive the levels of
pollutants needed in EPA's air quality models from emissions and other parameters generated by IPM. It
also involves using decision rules to assign point source locators to these emissions. (See Figure 1-1 for
a graphical representation of the relationship of the post-processing tools to the overall IPM structure.)
2.3.2 Model Run Years
Another important structural feature of IPM is the use of model run years to represent the full planning
horizon being modeled. Although IPM can represent an individual year in an analysis time horizon,
mapping each year in the planning horizon into a representative model run year enables IPM to perform
multiple year analyses while keeping the model size manageable. IPM considers the costs in all years in
the planning horizon while reporting results only for model run years. (See Section 2.3.3 below for further
details.)
The analysis time horizon for EPA Platform v6 extends from 2028 through 2059. The seven years
designated as model run years and the mapping of calendar years to the model run years is shown in
Table 2-1.
Table 2-1 Model Run Year and Year Mapping in v6
Run Year
Years Represented
2028
2028
2030
2029 -2031
2035
2032 -2037
2040
2038 - 2042
2045
2043 - 2047
2050
2048 - 2052
2055
2053-2059
Often models like IPM include a final model run year that is not used in the analysis of results. This
technique reduces the likelihood that modeling results in the last represented year will be skewed due to
the modeling artifact of having to specify an end point in the planning horizon, whereas, in reality,
economic decision-making will continue to take information into account from years beyond the model's
time horizon. This should be considered when assessing model projections from the last output year.
2-6
-------
2.3.3 Cost Accounting
As noted, IPM is a dynamic linear programming model that solves for the least cost investment and
electricity dispatch strategy for meeting electricity demand subject to resource availability and other
operating and environmental constraints. The cost components that IPM considers in deriving an optimal
solution include the costs of investing in new capacity options, the cost of installing and operating
pollution control technology, fuel costs, and the operation and maintenance costs associated with unit
operations. Several cost accounting assumptions are built into IPM's objective function that ensures a
technically sound and unbiased treatment of the cost of all investment options offered in the model.
These features include:
• All costs in IPM's single multi-year objective function are discounted to a base year. Since the
model solves for all run years simultaneously, discounting to a common base year ensures that
IPM properly captures complex inter-temporal cost relationships.
• Capital costs in IPM's objective function are represented as the net present value of levelized
stream of annual capital outlays, not as a one-time total investment cost. The payment period
used in calculating the levelized annual outlays never extends beyond the model's planning
horizon: it is either the book life of the investment or the years remaining in the planning horizon,
whichever is shorter. This approach avoids presenting artificially higher capital costs for
investment decisions taken closer to the model's time horizon boundary simply because some of
that cost would typically be serviced in years beyond the model's view. This treatment of capital
costs ensures both realism and consistency in accounting for the full cost of each of the
investment options in the model.
• The cost components informing IPM's objective function represent the composite cost over all
years in the planning horizon rather than just the cost in the individual model run years. The
approach permits the model to capture more accurately the escalation of the cost components
overtime.
2.3.4 Modeling Wholesale Electricity Markets
IPM is also designed to simulate electricity production activity in a manner that would minimize production
costs, as is the intended outcome in wholesale electricity markets. For this purpose, although not
designed to capture retail distribution costs, the model captures transmission costs and losses between
IPM model regions. However, the model implicitly includes distribution losses since net energy for load,9
rather than delivered sales,10 is used to represent electricity demand in the model. Further, the
production costs calculated by IPM are the wholesale production costs. In reporting costs, the model
does not include embedded costs, such as carrying charges of existing units, which may ultimately be
part of the retail cost incurred by end-use consumers.
2.3.5 Load Duration Curves (LDCs)
IPM uses LDCs to provide realism to the dispatching of electric generating units. Unlike a chronological
electric load curve, which is simply an hourly record of electricity demand, the LDCs are created by
rearranging the hourly chronological electric load data from the highest to lowest (MW) value. To
aggregate such load detail into a format enabling this scale of power sector modeling, EPA Platform v6
uses a 24-step piecewise linear representation of the LDC.
IPM can include any number of user-defined seasons. A season can consist of a single month or several
months. EPA Platform v6 contains three seasons: summer (May through September), winter (December
9 Net energy for load is the electrical energy requirements of an electrical system, defined as system net generation,
plus energy received from others, less energy delivered to others through interchange. It includes distribution losses
10 Delivered sales is the electrical energy delivered under a sales agreement. It does not include distribution losses.
2-7
-------
through February), and a winter shoulder season (October, November, March, and April). The summer
season corresponds to the ozone season for modeling seasonal NOx policies. The remaining seven
months are split into a three-month winter season and a four-month winter shoulder season to better
capture winter peak and seasonality in wind and solar hourly generation profiles. Separate summer,
winter, and winter shoulder season LDCs are created for each of IPM's model regions. Figure 2-1 below
presents side-by-side graphs of a hypothetical chronological hourly load curve and a corresponding load
duration curve for a summer season.
The use of seasonal LDCs rather than annual LDCs allows IPM to capture seasonal differences in the
level and patterns of customer demand for electricity. For example, in most regions air conditioner cycling
only impacts customer demand patterns during the summer season. The use of seasonal LDCs also
allows IPM to capture seasonal variations in the generation resources available to respond to the
customer demand depicted in an LDC. For example, power exchanges between utility systems may be
seasonal in nature. Some air regulations affecting power plants are also seasonal in nature. This can
impact the type of generation resources that are dispatched during a particular season. Further, because
of maintenance scheduling for individual generating units, the capacity and utilization for these supply
resources also vary between seasons.
Figure 2-1 Hypothetical Chronological Hourly Load Curve and Seasonal Load Duration Curve for
Summer Season
Chronological Hourly Load Curve
Seasonal Load Duration Curve
MW
Hours in Season 3572 Hours in Season
In EPA Platform v6, regional forecasts of peak and total electricity demand from AEO 2020 and hourly
load curves from FERC Form 714 and ISO/RTOs11 are used to derive seasonal load duration curves for
each IPM run year in each IPM region. The results of this process are individualized seasonal LDCs that
capture the unique hourly electricity demand profile of each region. The LDCs change overtime to reflect
projected changes in load factors because of future variations in electricity consumption patterns.12
3672
Within IPM, LDCs are represented by a discrete number of load segments, or generation blocks, as
illustrated in Figure 2-2 for a six-load segment LDC. EPA Platform v6 uses 24 load segments in its
seasonal LDCs.
Figure 2-2 illustrates and the following text describes the 24-segment LDCs. Length of time and system
demand are the two parameters, which define each segment of the load duration curve. The load
segment represents the amount of time (along the x-axis) and the capacity that the electric dispatch mix
must be producing (represented along the y-axis) to meet system load. The hours in the LDC are initially
11 The 2018 load curves are used for all IPM model regions. For further details, see Section 3.2.4.
12 For further details regarding the source of the load factors used in EPA Platform v6, see Section 3.2.3.
2-8
-------
clustered into six groups. Group 1 incorporates 1% of all hours in the season with the highest load.
Groups 2 to 6 have 4%, 10%, 30%, 30%, and 25% of the hours with progressively lower levels of
demand. Each of these 6 groups of hours are further separated into four time of day categories to result
in a possible maximum of 24 load segments. This approach better accounts for the impact of solar
generation during periods of high demand. The fourtime-of-day categories are 8PM - 6AM, 6AM - 9AM,
9AM - 5PM, and 5PM - 8PM. Plants are dispatched to meet load based on economic considerations
and operating constraints. The most cost-effective plants are assigned to meet load in all 24 segments of
the load duration curve. Section 2.3.6 discusses dispatch modeling in more detail.
Table 2-2 contains data of the 2028 seasonal LDCs in each of the 67 model regions in the lower
continental U.S.
Figure 2-2 Stylized Depiction of a Six Segment Load Duration Curve Dispatch Modeling
In IPM, the dispatching of electricity is based on the variable cost of generation. In the absence of any
operating constraints, units with the lowest variable cost generate first. The marginal generating unit, i.e.,
the generating unit that generates the last unit of electricity, sets the energy price. Physical operating
constraints also influence the dispatch order. For example, IPM uses turndown constraints to prevent
base load units from cycling, i.e., switching on and off. Turndown constraints often override the dispatch
order that would result based purely on the variable cost of generation. Variable costs in combination
with turndown constraints enable IPM to dispatch generation resources in a realistic fashion.
Figure 2-3 depicts a stylized dispatch order based on the variable cost of generation. Two hypothetical
load segments are subdivided according to the type of generation resources available to respond to the
load requirements represented in the segments. The generation resources with the lowest operating cost
(i.e., hydro and nuclear) respond first to the demand represented in the LDC and are accordingly at the
bottom of dispatch stack." They are dispatched for the maximum possible number of hours represented
in the LDC because of their low operating costs. Generation resources with the highest operating cost
(i.e., peaking turbines) are at the top of the dispatch stack," since they are dispatched last and for the
minimum possible number of hours. In the load segment with a non-dispatchable generating resource
(i.e., solar or wind), the conventional generation resources are dispatched to the residual load level,
where residual load is defined as the difference between the total load and the load met by the non-
dispatchable resource.
2-9
-------
Figure 2-3 Stylized Dispatch Order in Illustrative Load Segments
MW
Without Intermittent With Intermittent
Capacity Capacity
Note: Figure 2-3 does not include all plant types modeled in EPA Platform v6. Intermittent renewable technologies
such as wind and solar are considered non-dispatchable and are assigned a specific hourly generation profile.
2.3.6 Fuel Modeling
IPM can model the full range of fuels used for electric generation. The cost, supply, and (if applicable)
quality of each fuel included in the model are defined during model set-up. Fuel price and supply are
represented in one of two approaches: (1) through a set of supply curves (coal, natural gas, and biomass)
or (2) through an exogenous price stream (fuel oil and nuclear fuel). With the first approach, the model
endogenously determines the price for the fuel by balancing supply and demand. IPM uses fuel quality
information (e.g., the sulfur, chlorine, or mercury content of different types of coal from different supply
regions) to determine the emissions resulting from combustion of the fuel.
EPA Platform v6 includes coal, natural gas, fuel oil, nuclear fuel, biomass, and fossil and non-fossil waste
as fuels for electric generation. Chapters 7 to 9 examine the specific assumptions for these fuels.
2.3.7 Transmission Modeling
IPM includes a detailed representation of existing transmission capabilities between model regions. The
maximum transmission capabilities between regions are specified by transmission constraints. Additions
to transmission lines are represented by decision variables defined for each eligible link and model run
year. In IPM's objective function, the decision variables representing transmission additions are
multiplied by new transmission line investment cost and capital charge rates to obtain the capital cost
associated with the transmission addition. Section 3.3 describes the specific transmission assumptions.
2.3.8 Operating Reserves Modeling
Operating reserves are part of a set of services referred to as essential reliability services required to
maintain the reliability and stability of the electric grid.13 Although definitions vary by market and region,
13 Essential reliability services have also often been referred to as ancillary services.
2-10
-------
the main services required to ensure reliable grid operation in the U.S. include operating reserves,
voltage support, and black start capability. Operating reserves consist of several services and products,
including frequency responsive reserves, regulating reserves, contingency reserves, and ramping
reserves. The grid operates across timescales ranging from milliseconds to years. Because supply and
demand must always be balanced, services must be provided to ensure stability across all timescales.
Energy and capacity services ensure that there is sufficient supply to meet demand over a specified
period, with a reserve margin in the event of an outage of a generating unit. Operating reserves ensure
that there are sufficient resources with the characteristics required to always balance supply and demand.
IPM has the capability to model operating reserve services at a regional level and can account for the
impact of solar and wind technologies on operating reserves requirements. Section 3.7 describes the
specific operating reserve assumptions.
2.3.9 Perfect Competition and Perfect Foresight
IPM assumes perfect competition and perfect foresight. Perfect competition means that IPM models
production activity in wholesale electric markets on the premise that these markets subscribe to all
assumptions of a perfectly competitive market. The model does not explicitly capture any market
imperfections such as market power, transaction costs, informational asymmetry, or uncertainty.
However, if desired, appropriately designed sensitivity analyses or redefined model parameters can be
used to gauge the impact of market imperfections on the wholesale electric markets.
Perfect foresight implies that agents precisely know the nature and timing of conditions in future years
that affect the ultimate costs of decisions along the way. For example, under IPM there is complete
foreknowledge of future electricity demand, fuel supplies, and other variables (including regulatory
requirements) that are subject to uncertainty and limited foresight. Models like IPM frequently assume
perfect foresight to establish a decision-making framework that can estimate cost-minimizing courses of
action given the best-guess expectations of these future variables that can be constructed at the time the
projections are made.
2.3.10 Scenario Analysis and Regulatory Modeling
IPM offers detailed and flexible modeling features that enables scenario analysis involving different
outlooks of key drivers of the power sector and environmental regulations. In particular, treatment of
environmental regulations is endogenous in IPM. By providing a comprehensive representation of
compliance options, IPM enables environmental decisions to be made within the model based on least
cost considerations, rather than exogenously imposing environmental choices on model results. For
example, unlike other models that enter allowance prices as an exogenous input during model set-up,
IPM obtains allowance prices as an output of the endogenous optimization process of finding the least
cost compliance options in response to air regulations. (In linear programming terminology, they are the
shadow prices of the respective emission constraints — a standard output from solving a linear
programming problem.) IPM can capture a wide variety of regulatory program designs including
emissions trading policies, command-and-control policies, and renewable portfolio standards.
Representation of emissions trading policies can include allowance banking, trading, borrowing, bonus
allowance mechanisms, and progressive flow controls. Air regulations can be tailored to specific
geographical regions and can be restricted to specific seasons. Many of these regulatory modeling
capabilities are deployed in EPA Platform v6.
2.4 Hardware and Programming Features
IPM produces model files in standard mathematical programming system (MPS) format. The model runs
on most PC-platforms. Hardware requirements are dependent on the size of a particular model run. For
example, with almost 11.6 million decision variables and 2.8 million constraints, EPA Platform v6 is run on
a 64-bit Windows Server 2019 Standard platform with Intel® Xeon® Gold 6240R Processor, 35.75MB
Cache, 2.40 GHz (2 processor)/24Core and 512 GB of RAM. Due to the size of the EPA Platform v6,
2-11
-------
FICO Xpress Optimization Suite 8.8.0 (a 64-bit, commercial-grade solver with capability of optimizing
mixed integer(MIP), linear and non-linear problems using multi-threaded parallel processing) is used.
Two data processors, a front-end and the post-processing tool, support the model. The front-end creates
the necessary inputs that IPM uses. The post-processing tool maps IPM model-plant level outputs to
individual electric generating units (a process referred to as parsing- see Section 2.3.1) and creates input
files in flat-file format as required by EPA's air quality models.
In preparation for a model run, IPM requires an extensive set of input parameters. The input parameters
are discussed in Section 2.5.1. Results from a model run are presented in a series of detailed reports.
The reports are described in Section 2.5.2.
2.5 Model Inputs and Outputs
2.5.1 Data Parameters for Model Inputs
IPM requires input parameters that characterize the U.S. electric power system, economic outlook, fuel
supply and air regulatory framework. Chapters 3-10 contain detailed discussions of the values assigned
to these parameters in EPA Platform v6. The present section lists the key input parameters required by
IPM:
Electric System
Existing Generation Resources
• Plant Capacity
• Heat Rate
• Fuels Used
• Emission Limits and Emission Rates for NOx, SO2, HCI, CO2, and mercury
• Existing Pollution Control Equipment and Retrofit Options
• Availability
• Fixed and Variable Operation & Maintenance Costs
• Minimum Generation Requirements (Turn Down Constraints)
• Generation Profiles for Non-Dispatchable Resources
New Generation Resources
• Cost and Operating Characteristics
• Resource Limits and Generation Profiles
• Limitations on Availability
Other System Requirements
• Regional Specification
• Inter-regional Transmission Capabilities
• Reserve Margin Requirements for Reliability
• System Specific Generation Requirements
Economic Outlook
Electricity Demand
• Firm Regional Electricity Demand
• Load Curves
2-12
-------
Financial Outlook
• Capital Charge Rates
• Discount Rate
Fuel Supply
Fuel Supply Curves for Coal, Gas, and Biomass
• Fuel Price
• Fuel Quality
• Transportation Costs for Coal, Natural Gas, and Biomass
Regulatory Outlook
Air Regulations for NOx, SO2, HCI, CO2, and Mercury
• Other Air Regulations
• Non-air Regulations (affecting electric generating unit operations)
2.5.2 Model Outputs
IPM produces a variety of output reports. These range from detailed reports, which describe the results
for each model plant and run year, to summary reports, which present results for regional and national
aggregates. Individual topic areas can be included or excluded at the user's discretion. Standard IPM
reports cover the following topics:
• Generation mix
• Capacity mix
• Capacity additions and retirements
• Capacity and energy prices
• Power production costs (capital, fixed and variable operation & maintenance costs, and fuel
costs)
• Fuel consumption
• Fuel supply and demand
• Fuel prices for coal, natural gas, and biomass
• Emissions (NOx, SO2, HCI, CO2, and mercury)
• Emission allowance prices
List of tables that are uploaded directly to the web:
Table 2-2 Load Duration Curves used in EPA Platform v6 Post-IRA 2022 Reference Case
2-13
-------
3. Power System Operation Assumptions
This chapter describes the assumptions pertaining to the North American electric power system as
represented in the EPA Platform v6 Post-IRA 2022 Reference Case (EPA Platform v6).
3.1 Model Regions
EPA Platform v6 models the power sector in the contiguous United States, and 10 Canadian provinces
(with Newfoundland and Labrador represented as two regions on the electricity network even though
politically they constitute a single province14) as an integrated network.15
There are 67 IPM model regions covering the contiguous United States.16 The IPM model regions are
largely consistent with the regional configuration presented in the NERC Long-Term Reliability
Assessments.17 IPM model regions reflect the administrative structure of regional transmission
organizations (RTOs) and independent system operators (ISOs). Further disaggregation allows a more
accurate characterization of the operation of the United States power markets by providing the ability to
represent transmission bottlenecks across RTOs and ISOs, as well as key transmission limits within
them. Other items of note in the IPM regional definition include:
• The NERC assessment regions of MISO, PJM, and SPP cover the areas of the corresponding
RTOs and are designed to better represent transmission limits and dispatch in each area. In
IPM, model regions are designed to represent planning areas within each RTO and/or areas with
internal transmission limits. Accordingly, MISO area is disaggregated into 14 IPM regions. PJM
assessment area is disaggregated into 9 IPM regions, and SPP is disaggregated into 5 IPM
regions.
• New York is disaggregated into 8 IPM regions, to better represent flows around New York City
and Long Island, and to better represent flows across New York State from Canada and other
United States regions. The NERC assessment region SERC is divided into Kentucky, TVA, AECI,
the Southeast, and the Carolinas. New England is disaggregated into CT, ME, and rest of New
England regions. ERCOT is also disaggregated into 3 IPM regions. IPM retains the NERC
assessment areas within the overall WECC regions, and further disaggregates these areas using
sub-regions from the WECC Power Supply Assessment. In total, WECC is disaggregated into 16
IPM regions.
Figure 3-1 contains a map showing the EPA Platform v6 model regions.
Table 3-1 defines the abbreviated region names appearing on the map and gives a crosswalk between
the IPM model regions, the NERC assessment regions, and regions used in the Energy Information
Administration's (ElA's) National Energy Model System (NEMS) that is the basis for ElA's Annual Energy
Outlook (AEO) reports.
14 This results in a total of 11 Canadian model regions being represented in EPA Platform v6.
15 Because United States and the Canadian power markets are being modeled in an integrated manner, IPM can
model the transfer of power in between the two countries endogenously. This transfer of power is limited by the
available transmission capacity in between the two countries. Hence, it is possible for the model to build capacity in
one country to meet demand in the other country when economic and is operationally feasible.
16 The 67 U.S. IPM model regions include 64 power market regions and 3 power switching regions.
17 IPM regions also generally conform to the boundaries of the National Energy Modeling System (NEMS) model to
provide for a more accurate translation of demand projections taken from the Annual Energy Outlook (AEO).
3-1
-------
3.2 Electric Load Modeling
Net energy for load and net internal demand are inputs to IPM that together are used to represent the
grid-demand for electricity. Net energy for load is the projected annual electricity grid-demand, prior to
accounting for intra-regional transmission and distribution losses. Net internal demand (peak demand) is
the maximum hourly demand within a given year after removing interruptible demand. Table 3-2 shows
the electricity demand assumptions (expressed as net energy for load) used in EPA Platform v6. It is
based on the net energy for load in AEO 2021 Reference Case.18 Also added the incremental demand
from EPA's OTAQ's on the book rules that are not captured in the AEO 2021 demand projections.19
Incremental demand was calculated by running MOVES for the full country to calculate total energy
consumption for all Zero Emission Vehicles (ZEVs) by EPA's OTAQ.
Figure 3-1 EPA Platform v6 Model Regions
18 The electricity demand in EPA Platform v6 for the U.S. lower 48 states and the District of Columbia is obtained for
each IPM model region by disaggregating the Total Net Energy for Load projected for the corresponding NEMS
Electric Market Module region as reported in the Electricity and Renewable Fuel Tables 54.1-54,25 at
https://www.eia.gov/outlooks/archive/aeo21/tables_ref.php.
19 Incremental demand accounting for the on-the-books EPA OTAQ GHG https://www.epa.qov/requlations-
emissions-vehicles-and-enaines/firial-rule-revise-existina-national-ahg-emissions) final rule that has not been
reflected in AEO 2021 is documented in Attachment 3-1.
3-2
-------
For purposes of documentation, Table 3-2 and Table 3-3 present the net energy for load on a national-
and regional-level, respectively. EPA Platform v6 models net energy for load in each of the 67 U.S. IPM
regions in the following steps:
• The net energy for load in each of the 25 NEMS electricity regions is taken from the AEO 2021
Reference Case.
• NERC balancing areas are assigned to both IPM regions and NEMS regions to determine the share
of the NEMS net energy for load in each NEMS region that falls into each IPM region. These shares
are calculated in the following steps.
• Map the NERC Balancing Authorities/ Planning Areas in the United States to the 67 IPM regions.
• Map the Balancing Authorities/ Planning Areas in the United States to the 25 NEMS regions.
• Using the 2016 hourly load data from FERC Form 714, ISOs, and RTOs, calculate the
proportional share of load in the 25 NEMS regions that share a geography with the 67 IPM
regions.
• Using the calculated load shares for each NEMS region that falls into each IPM region, calculate
the total net energy for load for each IPM region from the NEMS regional load in the AEO 2021
Reference Case.
Table 3-1 Mapping of NERC Regions and NEMS Regions with v6 Model Regions
NERC Assessment
AEO 2021 NEMS
Region
Region
Model Region
Model Region Description
TRE (1)
ERC_REST
ERCOT_Rest
TRE (1)
ERC_GWAY
ERCOT_Tenaska Gateway Generating Station
ERCOT
TRE (1)
ERC_FRNT
ERCOT_Tenaska Frontier Generating Station
TRE (1)
ERC_WEST
ERCOT_West
TRE (1)
ERC_PHDL
ERCOT_Panhandle
FRCC
FRCC (2)
FRCC
FRCC
MAPP
MISW (3), SPPN (19)
MIS_MAPP
MISO_MT, SD, ND
MISC (4)
MIS IL
MISO_lllinois
MISC (4)
MISJNKY
MISOJndiana (including parts of Kentucky)
MISW (3)
MISJA
MISOJowa
MISW (3)
MIS_MIDA
MISOJowa-MidAmerican
MISE (5)
MIS_LMI
MISO_Lower Michigan
MISC (4)
MIS_MO
MISO_Missouri
MISO
MISW (3)
MIS_WUMS
MISO_Wisconsin- Upper Michigan (WUMS)
MISW (3)
MIS_MNWI
MISO_Minnesota and Western Wisconsin
MISS (6)
MIS_WOTA
MISO_WOTAB (including Western)
MISS (6)
MIS_AMSO
MISO_Amite South (including DSG)
MISS (6)
MIS_AR
MISO_Arkansas
MISS (6)
MIS_MS
MISO_Mississippi
MISS (6)
MIS_LA
MISO_Louisiana
ISNE (7)
NENG_CT
ISONE_Connecticut
ISO-NE
ISNE (7)
NENGREST
ISONE_MA, VT, NH, Rl (Rest of ISO New
England)
ISNE (7)
NENG_ME
ISONE_Maine
NYUP (9)
NY_Z_C&E
NY_Zone C&E
NYUP (9)
NY_Z_F
NY_Zone F (Capital)
NYUP (9)
NY_Z_G-I
NY_Zone G-l (Downstate NY)
NYISO
NYCW (8)
NY_Z_J
NY_Zone J (NYC)
NYCW (8)
NY_Z_K
NY_Zone K (LI)
NYUP (9)
NY_Z_A
NY_Zone A (West)
NYUP (9)
NY_Z_B
NY_Zone B (Genesee)
3-3
-------
NERC Assessment
AEO 2021 NEMS
Region
Region
Model Region
Model Region Description
NYUP (9)
NY Z D
NY_Zone D (North)
PJME (10)
PJM WMAC
PJM Western MAAC
PJME (10)
PJM EMAC
PJM EMAAC
PJME (10)
PJM SMAC
PJM SWMAAC
PJMW (11)
PJM West
PJM West
PJM
PJMW (11)
PJM AP
PJM AP
PJMC (12)
PJM COMD
PJM ComEd
PJMW (11)
PJM ATS I
PJM ATS I
PJMD (13)
PJM Dom
PJM Dominion
PJME (10)
PJM PENE
PJM PENELEC
SERC-E
SRCA (14)
S VACA
SERC VACAR
SRCE (16)
S C KY
SERC_Central_Kentucky
SERC-N
MISC (4), SPPS (17)
S D AECI
SERC Delta AECI
SRCE (16)
S C TVA
SERC Central TVA
SERC-SE
SRSE (15)
S SOU
SERC Southeastern
SPPN (19)
SPP NEBR
SPP Nebraska
SPPC (18)
SPP N
SPP North- (Kansas, Missouri)
SPP
SPPS (17)
SPP_KIAM
SPP_Kiamichi Energy Facility
SPPS (17)
SPP WEST
SPP West (Oklahoma, Arkansas, Louisiana)
SPPS (17)
SPP_SPS
SPP SPS (Texas Panhandle)
SPPN (19)
SPP WAUE
SPP WAUE
California/Mexico
(CA/MX)
CANO (21)
WEC_CALN
WECC Northern California (not including
BANC)
CASO (22)
CASO (22)
WEC_LADW
WEC SDGE
WECC_LADWP
WECC_San Diego Gas and Electric
CASO (22)
WECC SCE
WECC Southern California Edison
NWPP (23)
WECC MT
WECC Montana
CANO (21)
WEC BANC
WECC BANC
BASN (25)
WECC ID
WECC Idaho
Northwest Power Pool
BASN (25)
WECC NNV
WECC Northern Nevada
(NWPP)
BASN (25), SRSG
(20)
WECC_SNV
WECC_Southern Nevada
BASN (25)
WECC UT
WECC Utah
NWPP (23)
WECC PNW
WECC Pacific Northwest
Rocky Mountain Reserve
Group (RMRG)
RMRG (24)
BASN (25), RMRG
(24)
WECC_CO
WECC_WY
WECC_Colorado
WECC_Wyoming
Southwest Reserve
Sharing Group (SRSG)
SRSG (20)
WECC AZ
WECC Arizona
SRSG (20)
SRSG (20)
WECC_NM
WECC I ID
WECC_New Mexico
WECCJmperial Irrigation District (IID)
CN AB
Canada Alberta
CN BC
Canada British Columbia
CN MB
Canada Manitoba
CN NB
Canada New Brunswick
CN NF
Canada New Foundland
Canada
CN_NL
CN_PE
CN_NS
CN_ON
CN_PQ
CN_SK
Canada_Labrador
Canada_Prince Edward island
Canada_Nova Scotia
Canada_Ontario
Canada_Quebec
Canada_Saskatchewan
3-4
-------
Table 3-2 Electric Load Assumptions in v6
Year
Net Energy for Load (Billions of kWh)
2028
4,405
2030
4,508
2035
4,796
2040
5,102
2045
5,428
2050
5,778
2055
6,134
Table 3-3 Regional Electric Load Assumptions in v6
IPM Region
Net Energy for Load (Billions of kWh)
2028
2030
2035
2040
2045
2050
2055
ERC FRNT
0
0
0
0
0
0
0
ERC GWAY
0
0
0
0
0
0
0
ERC PHDL
0
0
0
0
0
0
0
ERC REST
388
397
423
451
481
512
545
ERC WEST
34
34
37
39
42
45
48
FRCC
264
271
290
311
332
356
380
MIS AMSO
37
37
40
42
45
47
50
MIS AR
43
44
47
50
53
56
59
MIS IA
23
23
25
26
27
29
30
MIS IL
53
54
57
59
62
66
69
MIS INKY
102
104
110
116
122
128
134
MIS LA
56
58
61
65
69
73
77
MIS LMI
108
110
116
122
128
135
142
MIS MAPP
9
9
10
10
11
11
12
MIS MIDA
29
29
31
33
34
36
38
MIS MNWI
95
97
102
107
113
118
124
MIS MO
42
43
46
49
51
54
57
MIS MS
27
27
29
31
33
35
38
MIS WOTA
38
39
41
43
46
49
52
MIS WUMS
70
71
75
79
83
88
92
NENG CT
32
33
35
37
40
42
45
NENG ME
12
13
14
15
16
17
18
NENGREST
88
90
96
102
109
116
123
NY Z A
16
16
17
18
19
20
21
NY Z B
10
10
11
11
12
13
13
NY Z C&E
24
24
26
27
29
30
32
NY Z D
4
4
5
5
5
5
6
NY Z F
12
13
13
14
15
16
17
NY Z G-l
19
19
21
22
23
24
26
NY Z J
56
56
58
61
64
68
72
NY Z K
23
23
24
26
27
29
31
PJM AP
52
53
56
59
62
66
69
PJM ATS I
73
74
79
83
88
93
98
PJM_COMD
101
103
109
114
120
127
133
PJM Dom
114
116
123
131
140
150
160
PJM EMAC
148
152
161
170
180
192
203
PJM PENE
19
19
20
22
23
24
26
PJM SMAC
67
68
72
76
80
85
90
PJM West
213
217
229
240
253
266
279
PJM WMAC
59
60
63
67
70
75
79
S C KY
35
36
38
40
42
44
46
S C TVA
178
181
191
200
211
221
231
S D AECI
19
19
20
21
22
23
24
S SOU
259
265
282
299
317
336
355
S VACA
241
246
262
279
297
316
336
SPP KIAM
0
0
0
0
0
0
0
3-5
-------
IPM Region
Net Energy for Load (Billions of kWh)
2028
2030
2035
2040
2045
2050
2055
SPP N
80
82
87
92
98
103
109
SPP NEBR
31
32
34
35
37
39
41
SPP SPS
36
37
39
41
44
47
50
SPP WAUE
25
26
27
28
29
31
32
SPP WEST
109
112
120
128
137
146
155
WEC BANC
15
16
17
18
19
21
23
WEC CALN
120
123
132
143
154
168
181
WEC LADW
31
33
37
41
45
48
52
WEC SDGE
22
22
24
26
29
31
34
WECC AZ
103
107
115
125
135
146
158
WECC CO
73
76
82
88
96
104
112
WECC ID
26
27
30
32
35
38
41
WECC 11D
5
5
5
6
6
6
7
WECC MT
14
14
15
16
18
19
21
WECC NM
24
25
28
31
34
37
40
WECC NNV
15
15
17
18
19
21
23
WECC PNW
182
186
198
214
230
249
268
WECC SCE
110
113
121
131
142
153
166
WECC SNV
29
30
33
35
38
41
45
WECC UT
41
42
46
50
54
59
64
WECC WY
25
26
28
30
32
35
38
3.2.1 Distributed Solar Photovoltaics
Distributed solar photovoltaic (DPV) generation constitutes a significant and growing source of new
electricity generation in the United States. As a result, DPV generation has become increasingly pertinent
from an integrated resource planning perspective because it has the potential to significantly impact the
shapes of the residual load curves that are available for the grid-connected generation sources to meet.
The DPV implementation in EPA Platform v6 seeks to reflect this impact to the load shape by directly
representing the magnitude and timing of the electricity demand projected to be satisfied by distributed
solar PV as part of the total net energy for load.
Electricity Demand Assumptions: Electricity demand assumptions are represented by the total net energy
for load from the AEO 2021 Reference Case. To account for DPV generation, the AEO 2021 Reference
Case projections of end-use solar photovoltaic generation are added to AEO 2021 Reference Case
projections of net energy for load.
Unit-Level Data Assumptions: Non-dispatchable DPV model plants at the IPM region and state level are
implemented in IPM to capture the impact of the DPV generation on the shapes of the residual load
curves available for the grid-connected generation sources to meet. Their generation patterns are
governed by assumed DPV generation profiles provided by NREL.
The capacity and capacity factors of DPV model plants are calculated as follows. First, the AEO 2021
Reference Case end-use solar photovoltaic generation and capacity data that are available at the NEMS
region level are apportioned to IPM region level, using the methodology for mapping the electricity
demand projections from NEMS regions to IPM regions. Then, the IPM region level data are further
apportioned to the state level, using state shares of regional energy sales as reported by the 2016 EIA
Form 861. The data are next used to derive IPM region and state level capacity factor data. Finally, the
resulting IPM region and state level capacity data are hardwired to the DPV model plants, while the
capacity factor data are implemented by appropriately scaling the NREL's IPM region and state level DPV
hourly generation profiles. For this analysis, NREL's DPV hourly generation profiles for the highest
resource class in each of the IPM region and state categories were scaled by multiplying the hourly
3-6
-------
generation values with the ratio between the AEO 2021 Reference Case capacity factor and the capacity
factor underlying the NREL's hourly generation profiles.
3.2.2 Demand Elasticity
EPA Platform v6 has the capability to endogenously adjust electricity demand based on changes to with
the price of power. However, this capability is exercised only for sensitivity analyses where different price
elasticities of demand are specified for purposes of comparative analysis. The default assumption is that
the electricity demand shown in Table 3-2, which was derived from EIA modeling that already considered
price elasticity of demand, is static as IPM solves for least-cost electricity supply. The approach
maintains a consistent expectation of future load between the EPA Platform and the corresponding EIA
Annual Energy Outlook reference case (e.g., between EPA Platform v6 and the AEO 2021 Reference
Case).
3.2.3 Net Internal Demand (Peak Demand)
EPA Platform v6 has separate regional winter, winter shoulder, and summer peak demand values, as
derived from each region's seasonal load duration curve (found in Table 2-2). Peak projections for the
2028-2029 period were estimated based on NERC ES&D 2019 load factors20, and the estimated energy
demand projections shown in Table 3-3. For post 2029 years when NERC ES&D 2019 load factors were
not available, the NERC ES&D 2019 load factors for 2029 were projected forward using growth factors
embedded in the AEO 2021 Reference Case load factor projections.
Table 3-4 illustrates the national sum of each region's seasonal peak demand, and Table 3-28 presents
each region's seasonal peak demand. Because each region's seasonal peak demand need not occur at
the same time, the national peak demand is defined as non-coincidental (i.e., national peak demand is a
summation of each region's peak demand at whatever point in time that region's peak occurs across the
given time period).
Table 3-4 National Non-Coincidental Net Internal Demand in v6
Year
Peak Demand (GW)
Winter
Winter Shoulder
Summer
2028
719
662
806
2030
737
679
827
2035
790
727
886
2040
849
780
954
2045
915
838
1,030
2050
986
902
1,114
2055
1,046
957
1,182
Notes:
This data is an aggregation of the model-region-specific peak demand loads.
20 Load factors can be calculated at the NERC assessment region level based on the NERC ES&D 2019 projections
of net energy for load and net internal demand. All IPM regions that map to a particular NERC assessment region
are assigned the same load factors. In instances where sub regional level load factor details could be estimated in
selected ISO/RTO zones, those load factors were assigned to the associated IPM region.
3-7
-------
3.2.4 Regional Load Shapes
EPA Platform v6 uses the year 2018 as the "normal weather year"21 for all IPM regions. The 2018
chronological hourly load data were assembled by aggregating individual utility load curves taken from
Federal Energy Regulatory Commission Form 714 data and individual ISOs and RTOs.
3.3 Transmission
The contiguous United States and Canada can be represented by several power markets that are
interconnected by a transmission grid. This section details the assumptions about the transfer
capabilities and costs used to represent this transmission grid in EPA Platform v6.
3.3.1 Inter-regional Transmission Capability
Table 3-2922 shows the firm and non-firm Total Transfer Capabilities (TTCs) between model regions.
TTC is a metric that represents the capability of the power system to import or export power reliably from
one region to another. The purpose of TTC analysis is to identify the sub-markets created by
commercially significant constraints. Firm TTCs, also called Capacity TTCs, specify the maximum power
that can be transferred reliably, even after the contingency loss of a single transmission system element
such as a transmission line or a transformer (a condition referred to as N-1, or "N minus one"). Firm
TTCs provide a high level of reliability and are used for capacity transfers. Non-firm TTCs, also called
Energy TTCs, represent the maximum power that can be transferred reliably when all facilities are under
normal operation (a condition referred to as N-0, or "N minus zero"). Non-firm TTCs specify the sum of
the maximum firm transfer capability between sub-regions and incremental curtailable non-firm transfer
capability. Non-firm TTCs are used for energy transfers since they provide a lower level of reliability than
Firm TTCs, and transactions using Non-firm TTCs can be curtailed under emergency or contingency
conditions.
The amount of energy and capacity transferred on a given transmission link is modeled on a seasonal
basis for all run years in the EPA Platform v6. All the modeled transmission links have the same TTCs for
all seasons. The maximum values for firm and non-firm TTCs were obtained from public sources such as
market reports and regional transmission plans, wherever available. Where public sources were not
available, the maximum values for firm and non-firm TTCs are based on ICF's expert view. ICF analyzes
the operation of the grid under normal and contingency conditions, using industry-standard methods, and
calculates the transfer capabilities between regions. To calculate the transfer capabilities, ICF uses
standard power flow data developed by the market operators, transmission providers, or utilities, as
appropriate.
Furthermore, each transmission link between model regions shown in Table 3-29 represents a one-
directional flow of power on that link. This means that the maximum amount of flow of power possible
from region A to region B may be more or less than the maximum amount of flow of power possible from
region B to region A, due to the physical nature of electron flow across the grid.
21 The term "normal weather year" refers to a representative year whose weather is closest to the long-term (e.g., 30
year) average weather. The selection of a "normal weather year" can be made, for example, by comparing the
cumulative annual heating degree days (HDDs) and cooling degree days (CDDs) in a candidate year to the long-term
average. For any individual day, heating degree days indicate how far the average temperature fell below 65
degrees F; cooling degree days indicate how far the temperature averaged above 65 degrees F. Cumulative annual
heating and cooling degree days are the sum of all the HDDs and CDDs, respectively, in a given year.
22 In the column headers in Table 3-29, the term "Energy TTC (MW)" is equivalent to non-firm TTCs and the term
"Capacity TTC (MW)" is equivalent to firm TTCs.
3-8
-------
3.3.2 Joint Transmission Capacity and Energy Limits
Table 3-5 shows the annual joint limits to the transmission capabilities between model regions, which are
identical for the firm (capacity) and non-firm (energy) transfers. The joint limits were obtained from public
sources where available or based on ICF's expert view. A joint limit represents the maximum
simultaneous firm or non-firm power transfer capability of a group of interfaces. It restricts the amount of
firm or non-firm transfers between one model region (or group of model regions) and a different group of
model regions. For example, the New England market is connected to the New York market by four
transmission links:
• NENG_CT to NY_Z_G-I: 600 MW
• NENGREST to NY_Z_F: 800 MW
• NENGREST to NY_Z_D: 0 MW
• NENG_CT to NY_Z_K: 734 MW
Without any simultaneous transfer limits, the total transfer capability from New England to New York
would be 2,134 MW. However, current system conditions and reliability requirements limit the total
simultaneous transfers from New England to New York to 1,730 MW, as shown in Table 3-5. IPM uses
joint limits to ensure that this and similar reliability limits are not violated. Therefore, each individual link
can be utilized to its limit as long as the total flow on all links does not exceed the joint limit.
Table 3-5 Annual Joint Capacity and Energy Limits to Transmission Capabilities between Model
Regions in v6
Region Connection
Transmission Path
Capacity TTC
(MW)
Energy TTC
(MW)
NY Zone G-l (Downstate NY) & NY Zone J (NYC) to
NY Zone K (LI)
NY Z G-l to NY Z K
NY Z J to NY Z K
1,528
NY Zone K(LI) to NY Zones G-l (Downstate NY) & NY Zone
J (NYC)
NY Z K to NY Z G-l
NY Z K to NY Z J
104
ISO NEto NY I SO
NENG CT to NY Z G-l
NENGREST to NY Z F
NENG CT to NY Z K
NENGREST to NY Z D
1,730
NY I SO to ISO NE
NY Z G-l to NENG CT
NY Z F to NENGREST
NY Z Kto NENG CT
NY Z D to NENGREST
1,730
PJM West & PJM_PENELEC & PJM_AP to PJM_ATSI
PJM West to PJM ATS I
PJM PENE to PJM ATSI
PJM APto PJM ATSI
9,925
PJM_ATSI to PJM West & PJM_PENELEC & PJM_AP
PJM ATSI to PJM West
PJM ATSI to PJM PENE
PJM ATSI to PJM AP
9,925
PJM_West & PJM_Dominion to SERC VACAR
PJM West to S VACA
PJM Dom to S VACA
2,208
3,424
SERC VACAR to PJM_West & PJM_Dominion
S VACA to PJM West
S VACA to PJM Dom
2,208
3,424
MIS_MAPP & SPP_WAUE to MIS_MNWI
MIS MAPPtoMIS MNWI
SPP WAUEto
MIS MNWI
3,000
5,000
MIS_MNWI to MIS_MAPP & SPP_WAUE
MIS MNWI to MIS MAPP
MIS MNWI to
SPP WAUE
3,000
5,000
SERC_Central_TVA & SERC_Central_Kentucky to PJM West
S C TVAto PJM West
S C KY to PJM West
3,000
4,500
PJM West to SERC_Central_TVA & SERC_Central_Kentucky
PJM West to S C TVA
PJM West to S C KY
3,000
4,500
MISJNKY to PJM_COMD & PJM_West
MIS INKY to PJM COMD
MIS INKY to PJM West
4,586
6,509
PJM_COMD & PJM_West to MIS_ INKY
PJM COMD to MIS INKY
PJM West to MIS INKY
5,998
8,242
3-9
-------
Region Connection
Transmission Path
Capacity TTC
(MW)
Energy TTC
(MW)
NY_Z_J & NY_Z_G-I to PJM_EMAC
NY Z J to PJM EMAC
NY Z G-l to PJM EMAC
1,975
PJM_EMAC to NY_Z_J & NY_Z_G-I
PJM EMAC to NY Z J
PJM EMAC to NY Z G-l
2,975
NY_Z_C&E & NY_Z_A to PJM_PENELEC
NY Z C&Eto
PJM PENE
NY Z A to PJM PENE
1,050
PJM_PENELEC to NY_Z_C&E & NY_Z_A
PJM PENE to
NY Z C&E
PJM PENE to NY Z A
1,365
PJM_SMAC & PJM_WMAC to PJM_EMAC
PJM SMACto
PJM EMAC
PJM WMACto
PJM EMAC
9,752
PJM AP, PJM DOM, PJM EMAC & PJM WMACto
PJM_SMAC
PJM APto PJM SMAC
PJM DOM to PJM SMAC
PJM EMAC to
PJM SMAC
PJM WMACto
PJM SMAC
9,158
PJM AP, PJM ATSI&PJM DOM to PJM PENELEC,
PJM_SMAC & PJM_WMAC
PJM APto PJM PENE
PJM APto PJM SMAC
PJM APto PJM WMAC
PJM ATSI to PJM PENE
PJM DOM to PJM SMAC
2,252
6,500
CN_AB to CN_BC & WECC_MT
CN ABtoWECC MT
CN AB to CN BC
1,000
CN_BC & WECC_MT to CN_AB
WECC_MT to CN_AB
CN_BC to CN_AB
1,110
3.3.3 Transmission Link Wheeling Charge
The transmission link wheeling charge is the cost of transferring electric power from one region to
another. The EPA Platform v6 has no wheeling charges within individual IPM regions and no charges
between IPM regions that fall within the same RTO. The wheeling charges, expressed in 2019 mills/kWh,
are shown in Table 3-29 in the column labeled "Transmission Tariff."
3.3.4 Transmission Losses
The EPA Platform v6 assumes a 2.8 percent inter-regional transmission loss of energy transferred in the
Western interconnection and a 2.4 percent inter-regional transmission loss of energy transferred in
Eastern Interconnection and ERCOT. These factors are based on average loss factors calculated from
standard power flow data developed by the transmission providers.
3.3.5 New Transmission Builds
EPA Platform v6 includes new endogenous transmission build options starting in 2028.23 An important
dynamic driving this change is the increased deployment of new renewable generation capacity that is at
a significant distance from the load centers driving its deployment. Consequently, the inability to deploy
additional transmission capacity endogenously may be unduly limiting the economic potential of new
renewable capacity. More generally, enabling transmission capacity expansion allows IPM to co-optimize
23 New transmission options in EPA Platform v6 are built simultaneously in both directions as transmission lines when
built can allow bidirectional flows.
3-10
-------
generation and transmission builds and solve for the optimal mix of generation and transmission additions
to meet capacity and energy needs.
For these transmission build options, representative costs were derived from NREL's Jobs and Economic
Development Impact (JEDI) model. Inputs to the JEDI model included the likely voltage rating, a
representative length of line between each region, and the type of terrain expected to be traversed. The
approach included:
• Determination of likely voltage rating. The cost of transmission lines varies with voltage rating.
Higher voltage ratings typically have higher costs per unit length. To minimize maintenance,
inventory, and other costs, it is likely that a new transmission line in an area will be rated at a
voltage similar to transmission lines already existing in the area. Further, it is likely that an
interregional line would be rated at or close to the highest voltage rating of the area's backbone
transmission system due to economies of scale. ICF reviewed the backbone transmission
system in each of the model regions to determine the likely voltage rating that would be used for
new transmission lines. For example, the backbone transmission system in the Northeast (New
York and the New England states) is rated 345 kV. While the systems also have underlying 230
kV and lower voltage transmission lines, it is likely that new inter-regional transmission lines
would be rated 345 kV. In most of the southeastern U.S. states the backbone voltage is 500 kV;
therefore, we assume that a line between Florida and Southern Company, for example, would
likely be rated 500 kV.
• Estimation of representative line lengths. The cost of transmission lines also varies with the
length of line. The length of a particular line will depend on several factors, including the
location of existing interconnecting substations, existing rights-of-way, area of need within the
zone, and other factors. The length cannot be determined in advance without knowing the
specific application. For this analysis EPA made a simplifying assumption that lines would be
built between the geographic centers of the regions. In instances where the transmission line
lengths that are calculated using the centroid approach are longer than a typical maximum for
the assumed line voltage, the typical maximum24 length was used to estimate the unit cost of the
line.
• Assessment of terrain. Transmission line costs also vary with terrain. For example, a line
traversing a mountainous region would have a higher capital cost than a line in a flat, rural area.
Terrain classifications in the JEDI model include "Desert/Remote", "Mountainous", and "Flat With
Access". The model also allows for specification of population densities, including "In Town",
"Near Town", and "Rural". Terrain classifications and population densities were assigned that
best represented the area that lines between the regions would likely traverse. For example, the
terrain traversed by a line between New York City and Long Island was classified as Flat With
Access and the population density was specified as In Town, while a line between Nebraska and
the Oklahoma-Missouri area was classified as Flat With Access and Rural.
Together, this information was used to determine the total cost of a new transmission line between each
pair of contiguous IPM regions. ICF then calculated a unit cost in $/kW for each transmission link using
estimates of the power (MW) ratings for each transmission line. The bidirectional unit costs for new
transmission lines are shown in Table 3-29.
24 The typical maximum line lengths by voltage class were estimated based on a review of projects that were under
construction or complete in 2015-2018 EIA Form 411 datasets. The EIA Form 411 data was supplemented with
information from the year 2016 EEI report Transmission Projects: At a Glance that describes major high voltage
projects proposed by investor-owned utilities.
3-11
-------
3.4 International Imports
The United States electric power system is connected with the transmission grids in Canada and Mexico
and the three countries actively trade in electricity. The Canadian power market is endogenously
modeled in EPA Platform v6, but Mexico is not. International electric trading between the United States
and Mexico is represented by an assumption of net imports based on information from AEO 2021
Reference Case. Table 3-6 summarizes the assumptions on net imports into the United States from
Mexico.
Table 3-6 International Electricity Imports (billions kWh) in v6
2028
2030
2035
2040
2045
2050
2055
Net Imports from Mexico
2.93
2.93
2.93
2.65
2.65
2.65
2.65
Note 1: Source: AEO 2020 Reference Case
Note 2: Imports & exports transactions from Canada are
endogenously modeled in IPM.
3.5 Capacity, Generation, and Dispatch
While the capacity of existing units is an exogenous input into IPM, the dispatch of those units is an
endogenous decision. The capacity of existing generating units included in EPA Platform v6 can be
found in the National Electrical Energy Data System (NEEDS v6), a database which provides IPM with
information on all currently operating and planned-committed electric generating units. NEEDS v6 is
discussed in Chapter 4.
A unit's generation over a time period is defined by its dispatch pattern. IPM determines the optimal
economic dispatch profile given the operating and physical constraints imposed on the unit. In EPA
Platform v6, unit-specific operational and physical constraints are represented through availability,
capacity factor, and turndown constraints.
3.5.1 Availability
Power plant availability is the percentage of time that a generating unit is available to provide electricity to
the grid. Availability takes into account both scheduled maintenance and forced outages; it is formally
defined as the ratio of a unit's available hours adjusted for the derating of capacity (due to partial outages)
to the total number of hours in a year when the unit was in an active state. For most types of units in IPM,
availability parameters are used to specify an upper bound on generation to meet demand. Table 3-7
summarizes the availability assumptions used in EPA Platform v6, which are based on data from NERC
Generating Availability Data System (GADS) 2014-2018 and AEO 2020 Reference Case. NERC GADS
summarizes the availability data by plant type and size class. Unit-level availability assignments in EPA
Platform v6 are made based on the unit's plant type and size as presented in NEEDS v6. Table 3-35
shows the availability assumptions for all generating units in EPA Platform v6.
Table 3-7 Availability Assumptions in v6
Plant Type
Annual Availability
(%)
Biomass
83
Coal Steam
73-84
Combined Cycle
85
Combustion Turbine
85-91
Energy Storage
96
Fossil Waste
90
Fuel Cell
87
Geothermal
87
3-12
-------
Plant Type
Annual Availability
(%)
Hydro
76-83
IGCC
77-84
Landfill Gas
90
Municipal Solid Waste
90
Non-Fossil Waste
90
Nuclear
68-99
Oil/Gas Steam
68-84
Offshore Wind
95
Onshore Wind
95
Pumped Storage
82
Solar PV
90
Solar Thermal
90
Notes:
Ranges in unit level availabilities are based on varying plant sizes.
In the EPA Platform v6, separate seasonal (winter, winter shoulder, and summer) availabilities are
defined. For the fossil and nuclear unit types shown in Table 3-35, seasonal availabilities differ only in
that no planned maintenance is assumed to be conducted during the on-peak - summer (June, July, and
August) months for summer peaking regions and on-peak - winter (December, January, and February)
months for winter peaking regions. Characterizing the availability of hydro, solar, and wind technologies
is more complicated due to the seasonal and locational variations of the resources. The procedures used
to represent seasonal variations in hydro are presented in Section 3.5.2 and of wind and solar in Section
4.4.5.
3.5.2 Capacity Factor
For non-dispatchable technologies - such as run-of-river hydro, wind, and solar - IPM uses generation
profiles, not availabilities, to define the upper bound on the generation obtainable from the unit. The
capacity factors that result from the implementation of generation profiles are the percentage of the
maximum possible power generated by the unit. The seasonal capacity factor assumptions for hydro
facilities contained in Table 3-8 were derived from EIA Form 923 data for the 2009-2018 period. A
discussion of capacity factors and generation profiles for wind and solar technologies is contained in
Section 4.4.5 and Table 4-18, Table 4-19, Table 4-35, Table 4-44, and Table 4-45.
Table 3-8 Seasonal Hydro Capacity Factors (%) in v6
Model
Winter Capacity
Winter Shoulder
Summer Capacity
Annual Capacity
Region
Factor
Capacity Factor
Factor
Factor
ERC_REST
11%
12%
14%
12%
FRCC
51%
45%
38%
44%
MIS_AR
44%
43%
47%
45%
MIS IA
40%
47%
55%
49%
MIS IL
57%
63%
63%
61%
MIS INKY
47%
47%
61%
53%
MIS_LA
56%
63%
64%
62%
MIS LMI
57%
68%
48%
57%
MIS_MAPP
72%
72%
79%
75%
MIS MIDA
19%
22%
23%
22%
MIS_MNWI
47%
54%
58%
54%
MIS MO
37%
43%
50%
45%
MIS_WOTA
22%
22%
20%
21%
MIS_WUMS
56%
66%
59%
60%
NENG CT
41%
43%
36%
40%
NENG_ME
61%
58%
53%
57%
NENGREST
40%
44%
34%
39%
3-13
-------
Model
Winter Capacity
Winter Shoulder
Summer Capacity
Annual Capacity
Region
Factor
Capacity Factor
Factor
Factor
NY_Z_A
72%
69%
66%
68%
NY_Z_B
46%
45%
43%
45%
NY Z C&E
52%
52%
52%
52%
NY_Z_D
85%
77%
77%
79%
NY Z F
54%
53%
50%
52%
NY_Z_G-I
30%
30%
29%
29%
PJM AP
49%
48%
41%
45%
PJM_ATSI
19%
21%
24%
22%
PJM_COMD
38%
42%
47%
43%
PJM_Dom
24%
20%
17%
20%
PJM EMAC
43%
42%
29%
37%
PJM_PENE
53%
55%
43%
50%
PJM West
33%
31%
30%
31%
PJM_WMAC
43%
44%
31%
38%
S_C_KY
31%
27%
25%
27%
S C TVA
54%
41%
35%
42%
S_D_AECI
16%
18%
19%
18%
S SOU
30%
24%
18%
23%
S_VACA
28%
22%
19%
23%
SPP N
14%
16%
18%
16%
SPP_NEBR
35%
40%
47%
42%
SPP WAUE
36%
40%
48%
42%
SPP_WEST
24%
24%
29%
26%
WEC BANC
21%
23%
31%
26%
WEC_CALN
23%
27%
41%
32%
WEC LADW
14%
16%
24%
19%
WEC_SDGE
25%
29%
46%
35%
WECC AZ
27%
28%
31%
29%
WECC_CO
30%
24%
33%
29%
WECCJD
35%
36%
47%
40%
WECC IID
29%
34%
54%
41%
WECC_MT
38%
39%
50%
43%
WECC NM
20%
21%
27%
23%
WECC_NNV
42%
53%
60%
53%
WECC PNW
46%
42%
45%
44%
WECC_SCE
22%
28%
48%
35%
WECC SNV
19%
24%
26%
24%
WECC_UT
33%
35%
43%
38%
WECC WY
19%
25%
54%
36%
Note: Annual capacity factor is provided for information purposes only. It is not used for modeling purposes.
Capacity factor limits are used to define the upper bound on generation obtainable from nuclear units
because nuclear units will typically dispatch to their availability, and, consequently, capacity factor and
availability limits are equivalent. The capacity factors (and, consequently, the availabilities) of existing
nuclear units in EPA Platform v6 vary from region to region and overtime. Further discussion of the
nuclear capacity factor assumptions in EPA Platform v6 is contained in Section 4.6.
Minimum capacity factor requirements of 10% are applied to existing coal steam units in regions without
capacity markets in EPA Platform v6. NYISO, ISONE, PJM, and MISO are assumed to have capacity
markets. In EPA Platform v6, oil/gas steam units are assigned minimum capacity factors under certain
conditions. These minimum capacity factor constraints reflect stakeholder comments that if left
unconstrained, IPM does not project as much operation from oil/gas steam units as has occurred
historically. This dynamic is often the result of local transmission constraints, unit-specific grid reliability
requirements, or other drivers that are not captured in EPA's modeling. EPA examined its modeling
treatment of these units and introduced minimum capacity factor constraints to better reflect the real-world
3-14
-------
behavior of these units. The approach is designed to balance the continued operation of these units in
the near-term with allowing economic forces to influence decision-making over the modeling time horizon.
As a result, the minimum capacity factor limitations are relaxed overtime (and are terminated even earlier
if the capacity in question reaches 60 years of age). Historical operational data indicate that oil/gas
steam units with high-capacity factors have maintained a high level of generation over many years. To
reflect persistent operation of these units, minimum capacity factors for higher capacity factor units are
phased out more slowly than those constraints for lower capacity factor units. The steps in assigning
these capacity constraints are as follows:
i) For each oil/gas steam unit, calculate an annual capacity factor over a ten-year baseline (2009-
2018).
ii) Identify the minimum capacity factor over this baseline period for each unit.
iii) Terminate the constraints in the earlier of (a) the run-year in which the unit reaches 60 years of
age, or (b) based on the assigned minimum capacity factor and the model year indicated in the
following schedule:
• For model year 2023, remove minimum constraint from units with capacity factor < 5%
• For model year 2025, remove minimum constraint from units with capacity factor < 10%
• For model year 2028, remove minimum constraint from units with capacity factor < 15%
• For model year 2030, remove minimum constraint from units with capacity factor < 20%.
3.5.3 Turndown
Turndown assumptions in EPA Platform v6 are used to prevent coal and oil/gas steam units from
operating as peaking units, which would be inconsistent with their operational capabilities and assigned
costs. The turndown constraints in EPA Platform v6 require coal steam and oil/gas steam units to
dispatch no less than a fixed percentage of the unit capacity in the 23 base and mid-load segments of the
load duration curve in order to dispatch 100% of the unit in the peak load segments of the LDC. Oil/gas
steam units are required to dispatch no less than 25% of the unit capacity in the 23 base- and mid-load
segments of the LDC in order to dispatch 100% of the unit capacity in the peak load segment of the LDC.
Operating under the fixed percentage of base- and mid-load segments does not preclude the unit from
operating during peak hours, it merely reduces the share of peak hours in which it can operate. The unit
level turndown percentages for coal units were estimated based on a review of hourly Air Markets
Program Data (AMPD) data and are shown in Table 3-30.
3.6 Reserve Margins
A reserve margin is a measure of the system's generating capability above the amount required to meet
the net internal demand (peak load) requirement. It is defined as the difference between total dependable
capacity and annual system peak load divided by annual system peak load. The reserve margin capacity
contribution for variable renewable units is described in Section 4.4.5; the reserve margin capacity
contribution for other units is the capacity in the NEEDS for existing units or the capacity build by IPM for
new units. In practice, each NERC region has a reserve margin requirement, or comparable reliability
standard, which is designed to encourage electric suppliers in the region to build beyond their peak
requirements to ensure the reliability of the electric generation system within the region.
In IPM, reserve margins are used to represent the reliability standards that are in effect in each NERC
region. Individual reserve margins for each NERC region are derived from reliability standards in NERC's
electric reliability reports. The IPM regional reserve margins are imposed throughout the entire time
horizon. EPA Platform v6 reserve margin assumptions are shown in Table 3-9.
3-15
-------
Table 3-9 Planning Reserve Margins in v6
Model Region
Reserve Margin
CN_AB
10.2%
CN BC
10.2%
CN_MB
12.0%
CN NB
20.0%
CN_NF
20.0%
CN NL
20.0%
CN_NS
20.0%
CN ON
24.7%
CN_PE
20.0%
CN PQ
12.8%
CN_SK
11.0%
ERC FRNT
13.8%
ERC_GWAY
13.8%
ERC_PHDL
13.8%
ERC REST
13.8%
ERC_WEST
13.8%
FRCC
18.5%
MIS_AR
16.8%
MIS MS
16.8%
MISJA
16.8%
MIS IL
16.8%
MISJNKY
16.8%
MIS LA
16.8%
MIS_LMI
16.8%
MIS MAPP
16.8%
MIS_MIDA
16.8%
MIS MNWI
16.8%
MIS_MO
16.8%
MIS_AMSO
16.8%
MIS WOTA
16.8%
MIS_WUMS
16.8%
NENG CT
17.8%
NENG_ME
17.8%
NENGREST
17.8%
NY_Z_A
15.0%
NY Z B
15.0%
NY_Z_C&E
15.0%
NY Z D
15.0%
NY Z F
15.0%
Model Region
Reserve Margin
NY_Z_G-I
15.0%
NY Z J
15.0%
NY_Z_K
15.0%
PJM AP
15.7%
PJM_ATSI
15.7%
PJM_COMD
15.7%
PJM_Dom
15.7%
PJM EMAC
15.7%
PJM_PENE
15.7%
PJM SMAC
15.7%
PJM_West
15.7%
PJM WMAC
15.7%
S_C_KY
15.0%
S_C_TVA
15.0%
S D AECI
15.0%
S_SOU
15.0%
S VACA
15.0%
SPP_KIAM
15.0%
SPP N
15.0%
SPP_NEBR
15.0%
SPP SPS
15.0%
SPP_WAUE
15.0%
SPP WEST
15.0%
WEC_BANC
15.9%
WEC CALN
13.8%
WEC_LADW
13.8%
WEC SDGE
13.8%
WECC_AZ
11.0%
WECC_CO
12.5%
WECC ID
15.9%
WECC_IID
11.0%
WECC MT
15.9%
WECC_NM
11.0%
WECC NNV
15.9%
WECC_PNW
15.9%
WECC SCE
13.8%
WECC_SNV
15.9%
WECC UT
15.9%
WECC WY
12.5%
3.7 Operating Reserves
EPA Base Case v6 models operating reserve requirements in IPM to ensure that an appropriate mix of
supply resources will be included that is consistent with maintaining reliability standards, especially in
later years as new capacity deploys more rapidly. Operating reserves are typically deployed in order of
the response speed, from fast to slow. In general, the categories of reserves include:25
• Frequency-Responsive Reserves. This is the fastest response. It has traditionally been provided
through automatic action of synchronous generators that react to slow down and arrest frequency
25 Denholm, Paul, Yinong Sun, and Trieu Mai. 2019. An Introduction to Grid Services: Concepts, Technical
Requirements, and Provision from Wind. Golden, CO: National Renewable Energy Laboratory. NREL/TP-6A20-
72578. https://vwwy.nrel.gov/docs/fV19osti/72578.pdf.
3-16
-------
deviations as a result of the inertia of the machines or their governor action (also referred to as
primary frequency response or PFR). As a result of the increase in renewable integration and loss
of generators that provide inertial response, other products are emerging to provide frequency
response on a very fast (sub-minute) timescale.
• Regulating Reserves. This is the rapid response by generators to balance supply and demand to
maintain system frequency. Regulation reserve can address the random fluctuations in load that
create imbalances in supply and demand.
• Contingency Reserves. These reserves are deployed to cover the unplanned loss of power plants
or transmission lines. Contingency reserves generally include spinning, non-spinning, and
supplemental reserves. Spinning reserves respond quickly and are then supplemented or
replaced with non-spinning and supplemental reserves that are usually less costly.
• Ramping Reserves. This is used to address slower variations or events that occur over a longer
period, such as variable generation forecast errors. Ramping reserves, also known as load-
following or flexibility reserves, are an emerging product that is becoming more important with the
increasing penetration of variable generation sources such as wind and solar.
The operating reserves products currently procured in United States electricity markets include regulating
reserves, contingency reserve, and ramping reserves. FERC Order No. 842 requires that new generation
resources that participate in the electricity markets provide some form of frequency-responsive reserve to
support the reliability of the grid, but the Order does not mandate explicit compensation for the product.
EPA's implementation of operating reserve requirements is consistent with the products offered in the
electricity markets. The operating reserves modeled explicitly in EPA Platform v6 are regulating reserves,
contingency reserves, and ramping reserves. The plant types that can provide these reserves are listed in
Table 3-12. Based on current regulations, new generation resources that are built in the EPA Platform v6
are assumed to have the capability to provide frequency-responsive reserves. It is reasonable to expect
that sufficient frequency-responsive reserves will be available to support grid reliability in IPM analyses
even if the requirement is not modeled explicitly.
3.7.1 Operating Reserve Requirements
Operating reserve requirements typically depend on the load and load forecast error. As variable
renewable generation increase, it is likely that the operating reserve requirements will increase due to the
variability of the renewable resources.26 27 Table 3-10 shows operating reserve assumptions, which are
based on the National Renewable Energy Laboratory (NREL) report, Operating Reserves in Long-term
Planning Models.28 The long-term requirements include components that depend on the penetration of
wind and solar resources to address the expected increase in variability as more variable resources enter
the market.
Table 3-10 Operating Reserve Requirement Assumptions by Type in v6
Product
Operating Reserve
Load Requirement
Operating Reserve
Requirement for Wind
Operating Reserve
Requirement for Solar
Operating Reserve
Timescale
Spinning
3% of load
-
-
10 minutes
Regulation
1% of load
0.5% of wind capacity
0.3% of solar PV capacity
5 minutes
Flexibility
-
10% of wind capacity
4% of solar PV capacity
60 minutes
26 Western Wind and Solar Integration Study (WWSIS) Phase 1, National Renewable Energy Laboratory (GE
Energy), May 2010
27 Analysis of Wind Generation Impact on ERCOT Ancillary Services Requirements, Electric Reliability Council of
Texas (GE Energy), March 2008
28 Cole, W. et al., Operating Reserves in Long-term Planning Models (NREL), June 2018
3-17
-------
The operating reserve requirements when modeled in IPM have a significant impact on model size. To
counter this effect, EPA made two simplifying assumptions. First, the spinning reserve, regulation, and
flexibility requirements are combined into a single product. Second, these constraints may be
implemented only in the later years when renewable penetration and operating reserve requirements are
highest; this representation of operating reserve requirements can be activated or deactivated by run year
for any scenario analyzed using IPM. The operating reserve requirements in v6 are applied at the 17
regional groups summarized in Table 3-11.
Table 3-11 Operating Reserve Regions in v6
Operating Reserve
Region
v6 Model Region
ERCOT
ERC PHDL, ERC REST, and ERC WEST
FRCC
FRCC
ISO-NE
NENG CT, NENGRESTand NENG ME
MISO East
MIS WUMS, MIS Ml DA, MIS IA, MIS IL, MIS LMI, MIS INKY and MIS MO
MISO South
MIS MS, MIS AR, MIS AMSO, MIS WOTA and MIS LA
MISO West
MIS MAPPandMIS MNWI
NYISO
NY Z A, NY Z B, NY Z C&E, NY Z D, NY Z F, NY Z G-l, NY Z J and NY Z K
PJM East
PJM PENE, PJM EMAC, PJM WMAC and PJM SMAC
PJM West
PJM West, PJM AP, PJM COMD, PJM Dom and PJM ATSI
SERC-E
S VACA
SERC-N
S C TVA and S C KY
SERC-SE
S SOU
SPP
SPP WAUE, SPP SPS, SPP WEST, SPP NEBR, SPP N and S D AECI
WECC-CAMX
WEC SDGE, WECC SCE, WEC CALN and WEC LADW
WECC-NWPP
WECC MT, WECC ID, WECC PNW, WECC NNV, WECC UT, WECC SNV and
WEC BANC
3.7.2 Generation Characteristics
The ability of a generator to provide operating reserves varies with the technology type. The more flexible
a unit (i.e., faster ramp rate), the higher its operating reserve capability. Table 3-12 shows the assumed
operating reserve capabilities for different generation technologies and are based on the NREL's report,
Operating Reserves in Long-term Planning Models. For example, gas combustion turbines and
combined cycles have faster ramp rates than coal plants; therefore, the gas plants can provide more
operating reserves per unit capacity than coal plants. EPA also assumed that capacity meeting energy
needs cannot provide operating reserves at the same time. For example, if 75% of a generator's capacity
is serving the energy market, only 25% will be available to be offered into the operating reserve market.
Table 3-12 summarizes the ramp rates of power plant technologies. Since EPA Platform v6 is
incorporating a single composite operating reserves product, the maximum operating reserve
contributions are based on the 10-minute spinning reserve requirement.
Table 3-12 Operating Reserve Contribution Assumptions by Technology in v6
Technology
Assumed Ramp Rate (%/minute)
Maximum Operating Reserve Contribution (%)
Combustion Turbine
8
80
Combined Cycle
5
50
Coal Steam
4
40
Geothermal
4
40
CSP with Storage
10
100
Biomass
4
40
Oil/Gas Steam
4
40
Hydro
100
100
Energy Storage
100
100
Generation resources that are not fast starting cannot provide operating reserves unless they are already
operating. To provide operating reserves, the plant must also be dispatching into the energy market.
3-18
-------
3.8 Power Plant Lifetimes
EPA Platform v6 does not include any pre-specified assumptions about power plant lifetimes (i.e., the
duration of service allowed) except for nuclear units. All conventional fossil units (coal, oil/gas steam,
combustion turbines, and combined cycle), nuclear, and biomass units can be retired during a model run
if their retention is deemed uneconomic.
Nuclear Retirement: EPA Platform v6 does not assume that commercial nuclear reactors will be retired
upon license expiration. EPA Platform v6 incorporates life extension costs to enable these operating life
extensions. (See Sections 4.2.8 and 4.6). For unit specific retirement years, see NEEDS.
3.9 Heat Rates
Heat rates, expressed in British thermal units (Btus) per kilowatt-hour (kWh), are a measure of an electric
generating unit's (EGU's) efficiency. As in previous versions of NEEDS, it is assumed in NEEDS v6 that,
with the exception of deploying the heat rate improvement option described below, heat rates of existing
EGUs remain constant overtime. This assumption reflects two offsetting factors:
i) Plant efficiencies tend to degrade over time, and
ii) Increased maintenance and component replacement costs act to maintain, or improve, an EGU's
generating efficiency.
The heat rates for the model plants in EPA Platform v6 are based on values from the AEO 2020
Reference Case and are informed by fuel use and net generation data reported on Form EIA-923. These
values were screened and adjusted using a procedure developed by EPA (as described below) to ensure
that the heat rates used in EPA Platform v6 are within the engineering capabilities of the various EGU
types.
The result of an earlier EPA engineering analysis, the upper and lower heat rate limits shown in Table
3-13 were applied to coal steam, oil/gas steam, combined cycle, combustion turbine, and internal
combustion engines. If the reported heat rate for such a unit was below the applicable lower limit or
above the upper limit, the upper or lower limit was substituted for the reported value.
Table 3-13 Lower and Upper Limits Applied to Heat Rate Data in v6
Plant Type
Heat Rate (Btu/kWh)
Lower Limit Upper Limit
Coal Steam
8
300
14
500
Oil/Gas Steam
8
300
14
500
Combined Cycle - Natural Gas
5
500
15
000
Combined Cycle - Oil
6
000
15
000
Combustion Turbine - Natural Gas - 80 MW and above
8
700
18
700
Combustion Turbine - Natural Gas < 80 MW
8
700
36
800
Combustion Turbine - Oil and Oil/Gas - 80 MW and above
6
000
25
000
Combustion Turbine - Oil and Oil/Gas < 80 MW
6
000
36
800
IC Engine - Natural Gas
8
700
18
000
IC Engine - Oil and Oil/Gas - 5 MW and above
8
700
20
500
IC Engine - Oil and Oil/Gas < 5 MW
8
700
42
000
3-19
-------
3.10 Existing Legislations and Regulations Affecting Power Sector
This section describes the existing federal, regional, and state SO2, NOx, mercury, HCI and CO2
emissions regulations and legislations that are represented in EPA Platform v6. EPA Platform v6 also
includes three non-air federal rules affecting EGUs: Cooling Water Intakes (316(b)) Rule, Coal
Combustion Residuals from Electric Utilities (CCR), and the Effluent Limitations and Guidelines Rule.
The first four subsections discuss national and regional regulations. The next five subsections describe
state level environmental regulations, a variety of legal settlements, emission assumptions for potential
units, renewable portfolio standards, and Canadian regulations forCC>2 and renewables.
3.10.1 Inflation Reduction Act
The Inflation Reduction Act (IRA) contains a number of tax credit provisions that affect power sector
operations. The Clean Electricity Investment and Production Tax Credits (provisions 48E and 45Y of the
IRA) are described in more detail in Section 4.5. The credit for Carbon Capture and Sequestration
(provision 45Q) is described in Section 3.12. The impacts from the Zero-Emission Nuclear Power
Production Credit (provision 45U) are reflected through modifying nuclear retirement limits, as described
in Section 4.6.1. The Credit for the Production of Clean Hydrogen (provision 45V) is reflected through the
inclusion of an exogenous delivered price of hydrogen fuel, see Section 9.5. The Advanced
Manufacturing Production Tax Credit (45X) was reflected through adjustments to the short-term capital
cost added for renewable technologies, see Section 4.4.3.
3.10.2 SO2 Regulations
Unit-level Regulatory SO2 Emission Rates and Coal Assignments: Before discussing the national and
regional regulations affecting SO2, it is important to note that unit-level SO2 permit rates including SO2
regulations arising out of State Implementation Plan (SIP) requirements, which are not only state-specific
but also county-specific, are captured at model set-up in the coal choices given to coal fired existing units
in EPA Platform v6. Since SO2 emissions are dependent on the sulfur content of the fuel used, the SO2
permit rates are used in IPM to define fuel capabilities.
For instance, a unit with a SO2 permit rate of 3.0 Ibs/MMBtu would be provided only with those
combinations of fuel choices and SO2 emission control options that would allow the unit to achieve an out-
of-stack rate of 3.0 Ibs/MMBtu or less. If the unit finds it economical, it may elect to burn a fuel that would
achieve a lower SO2 rate than its specified permit limit. In EPA Platform v6, there are six different sulfur
grades of bituminous coal, four different grades of subbituminous coal, four different grades of lignite, and
one sulfur grade of residual fuel oil. There are two different SO2 scrubber options and one DSI option for
coal units. Further discussion of fuel types and sulfur content is contained in Chapter 7. Further
discussion of SO2 control technologies is contained in Chapter 5.
National and Regional SO2 Regulations: The national program affecting SO2 emissions in EPA Platform
v6 is the Acid Rain Program established under Title IV of the Clean Air Act Amendments (CAAA) of 1990,
which set a goal of reducing annual SO2 emissions by 10 million tons below 1980 levels. The program,
which became operational in 2000, affects all SO2 emitting electric generating units greater than 25 MW.
The program provides trading and banking of allowances overtime across all affected electric generation
sources.
The annual SO2 caps over the modeling time horizon in EPA Platform v6 reflect the provisions in Title IV.
For allowance trading programs like the Acid Rain Program that allow banking of unused allowances over
time, we usually estimate an allowance bank that is assumed to be available by the first year of the
modeling horizon (which is 2028 in EPA Platform v6). However, the Acid Rain Program has
demonstrated a substantial oversupply of allowances that continues to grow overtime, and we anticipate
projecting that the program's emission caps will not bind the model's determination of SO2 emissions
regardless of any level of initial allowance bank assumed. Therefore, EPA Platform v6 does not assume
any Title IV SO2 allowance bank amount for the year of 2028 (notwithstanding that a large allowance
bank will exist in that year in practice), because such an assumption would have no material impact on
3-20
-------
projections given the nonbinding nature of that program. Calculating the available 2028 allowances
involved deducting allowance surrenders due to NSR settlements and state regulations from the 2028
SO2 cap of 8.95 million tons. The surrenders totaled 977 thousand tons in allowances, leaving 7.973
million of 2021 allowances remaining. Specifics of the allowance surrender requirements understate
regulations and NSR settlements can be found in Table 3-31 and Table 3-32.
EPA Platform v6 also includes a representation of the Western Regional Air Partnership (WRAP)
Program, a regional initiative involving New Mexico, Utah, and Wyoming directed toward addressing
visibility issues in the Grand Canyon and affecting SO2 emissions starting in 2018. The WRAP
specifications for SO2 are presented in Table 3-24.
3.10.3 NOx Regulations
Much like SO2 regulations, existing NOx regulations are represented in EPA Platform v6 through a
combination of system level NOx programs and generation unit-level NOx limits. In EPA Platform v6, the
NOx SIP Call trading program, Cross State Air Pollution Rule (CSAPR), the CSAPR Update, and the
Revised CSAPR Update Rule are represented. Table 3-24 shows the specification for the entire
modeling time horizon.
By assigning unit-specific NOx rates based on 2019 data, EPA Platform v6 is implicitly representing Title
IV unit-specific rate limits and Clean Air Act Reasonably Available Control Technology (RACT)
requirements for controlling NOx emissions from electric generating units in ozone non-attainment areas
or in the Ozone Transport Region (OTR).29 Unlike SO2 emission rates, NOx rates are calculated off
historical data and reflect the fuel mix for that particular year at the unit. NEEDS represents up to four
scenario NOx rates based on historical data to capture seasonal and existing control variability. These
rates are constant and do not change independent of the fuel mix assumed in the model. If the unit
undertakes a post-combustion control retrofit or a coal-to-gas retrofit, then these rates would change in
the model projections.
NOx Emission Rates
Future emission projections for NOx are a product of a unit's utilization (heat input) and emission rate
(Ibs/MMBtu). A unit's NOx emission rate can vary significantly depending on the NOx reduction
requirements to which it is subject. For example, a unit may have a post-combustion control installed
(i.e., SCR or SNCR), but only operate it during the time of the year in which it is subject to NOx reduction
requirements (e.g., the unit only operates its post-combustion control during the ozone season).
Therefore, its ozone-season NOx emission rate would be lower than its non-ozone-season NOx emission
rate. Because the same individual unit can have such large variation in its emission rate, the model
needs a suite of emission rate modes from which it can select the value most appropriate to the
conditions in any given model scenario. The different emission rates reflect the different operational
conditions a unit may experience regarding upgrades to its combustion controls and operation of its
existing post-combustion controls. Four modes of operation are developed for each unit, with each mode
carrying a potentially different NOx emission rate for that unit under those operational conditions.
The emission rates assigned to each mode are derived from historical data (where available) and
presented in NEEDS v6. When the model is run, IPM selects one of these four modes through a decision
process depicted in Figure 3-3 below. The four modes address whether units upgrade combustion
controls and/or operate existing post-combustion controls; the modes themselves do not address what
happens to the unit's NOx rate if it is projected to add a new post-combustion NOx control. If a unit is
projected to add a new post-combustion control, then after the model selects the appropriate input mode
it adjusts that mode's emission rate downwards to reflect the retrofit of SCR or SNCR; the adjusted rate
will reflect the greater of a percentage removal from the mode's emission rate or an emission rate floor.
29 The OTR consists of the following states: Maine, New Hampshire, Vermont, Massachusetts, Rhode Island,
Connecticut, New York, New Jersey, Pennsylvania, Delaware, Maryland, District of Columbia, and northern Virginia.
3-21
-------
The full process for determining the NOx rate of units in EPA Platform v6 model projections is summarized
in Figure 3-2.
Figure 3-2 Modeling Process for Obtaining Projected NOx Emission Rates
Historical NOx
Emission Rate Data
(e.g., 2019)
*
NEEDS
Assignment of emission rates
(derived from historic data) to
each of four NOx modes. Modes
reflect different potential
operational conditions at a unit.
Model Projections
Assignment of NOx emission rate
based on one of four NEEDS modes
rates with potential adjustment if the
unit is projected to add post-
combustion retrofit control
technology.
NOx Emission Rates in NEEDS v6 Database
The NOx rates were derived, wherever possible, directly from actual monitored NOx emission rate data
reported to EPA under the Acid Rain and Cross-State Air Pollution Rule in 2019.30 The emission rates
themselves reflect the impact of applicable NOx regulations.31 For coal-fired units, NOx rates were used in
combination with empirical assessments of NOx combustion control performance to prepare a set of four
possible starting NOx rates to assign to a unit, depending on the specific NOx reduction requirements
affecting that unit in a model run.
The reason for having a framework of four potential NOx rate modes applicable to each unit in NEEDS is
to enable the model to select from a range of NOx rates possible at a unit, given its configuration of NOx
combustion controls and its assumed operation of existing post-combustion controls. There are up to four
basic operating states for a given unit that significantly impact its NOx rate, and thus there are four NOx
rate modes.
Mode 1 and mode 2 reflect a unit's emission rates with its existing configuration of combustion and post-
combustion (i.e., SCR or SNCR) controls.
• For a unit with an existing post-combustion control, mode 1 reflects the existing post-combustion
control not operating and mode 2 the existing post-combustion control operating. However:
o If a unit has operated its post-combustion control year-round during the most recent of
2019, 2017, 2016, 2015, 2014, 2011, 2009, or 2007 years then mode 1 = mode 2, which
reflects that the control will likely continue to operate year-round (and thus a "not run"
emission rate option is not needed as justified by historical data).
o If a unit has not operated its post-combustion control during the most recent of 2019,
2017, 2016, 2015, 2014, 2011, 2009, or 2007 years, mode 1 will be based on this data
and mode 2 will be calculated using the method described under Question 3 in
Attachment 3-2.
30 By assigning unit-specific NOx rates based on 2019 data, EPA Platform v6 is implicitly representing Title IV unit-
specific rate limits and Clean Air Act Reasonably Available Control Technology (RACT) requirements for controlling
NOx emissions from electric generating units in ozone non-attainment areas or in the Ozone Transport Region (OTR).
Unlike SO2 emission rates, NOx emission rates are assumed not to vary with coal type but are dependent on the
combustion properties of the generating unit. Under the EPA Platform v6, the NOx emission rate of a unit can only
change if the unit is retrofitted with NOx post-combustion control equipment or if it is assumed to install state-of-the-art
NOx combustion controls. In instances where a coal steam unit converts to natural gas, the NOx rate is assumed to
reduce by 50%.
31 Because 2019 NOx rates reflect CSAPR, we no longer apply any incremental CSAPR related NOx rate adjustments
exogenously for CSAPR affected units in EPA Platform v6.
3-22
-------
o If a unit has operated its post-combustion control seasonally in recent years (i.e., either
only in the summer or winter, but not both), mode 1 will be based on historic data from
when the control was not operating, and mode 2 will be based on historic data from when
the SCR was operating.
• For a unit without an existing post-combustion control, mode 1 = mode 2, which reflects the unit's
historic NOx rates from a recent year.
Mode 3 and mode 4 emission rates parallel modes 1 and 2 emission rates but are modified to reflect
installation of state-of-the-art combustion controls on a unit if it does not already have them.
• For units that already have state-of-the-art combustion controls: mode 3 = mode 1 and mode 4 =
mode 2.
Emission rates derived for each unit operating under each of these four modes are presented in NEEDS
v6. Note that not every unit has a different emission rate for each mode, because certain units cannot in
practice change their NOx rates to conform to all potential operational states described above.
Figure 3-3 How One of the Four NOx Modes Is Ultimately Selected for a Unit
Did the source operate a
post-combustion control in
2019?
State-of-the-art combustion controls (SOA combustion controls')
The definition of state-of-the-art varies depending on the unit type and configuration indicates the
incremental combustion controls that are required to achieve a state-of-the-art combustion control
configuration for each unit. For instance, if a wall-fired, dry bottom boiler (highlighted below) currently has
LNB but no overfire air (OFA), the state-of-the-art rate calculated for such a unit would assume a NOx
emission rate reflective of overfire air being added at the unit. As described in the attachment of this
chapter, the state-of-the-art combustion controls reflected in the modes are only assigned to a unit if it is
subject to a new (post-2019) NOx reduction requirement (i.e., a NOx reduction requirement that did not
apply to the unit during its 2019 operation that forms the historic basis for deriving NOx rates for units in
EPA Platform v6). Existing reduction requirements as of 2019 under which units have already made
combustion control decisions would not trigger the assignment of the state-of-the-art modes that reflect
additional combustion controls.
3-23
-------
Table 3-14 State-of-the-Art Combustion Control Configurations by Boiler Type in v6
Boiler Type
Existing NOx
Incremental Combustion Control
Necessary to Achieve State-of-the-Art
Combustion Control
Tangential Firing
Does not Include LNC1 and LNC2
Includes LNC1, but not LNC2
Includes LNC2, but not LNC3
Includes LNC1 and LNC2 or LNC3
LNC3
CONVERSION FROM LNC1 TO LNC3
CONVERSION FROM LNC2 TO LNC3
Wall Firing, Dry Bottom
Does not Include LNB and OFA
Includes LNB, but not OFA
Includes OFA, but not LNB
Includes both LNB and OFA
LNB + OFA
OFA
LNB
Note:
LNB = Low NOx Burner Technology, LNC1 = Low NOx coal-and air nozzles with close-coupled overfire air, LNC2 =
Low NOx Coal-and-Air Nozzles with Separated Overfire Air, LNC3 = Low NOx Coal-and-Air Nozzles with Close-
Coupled and Separated Overfire Air, OFA = Overfire Air.
The emission rates for each generating unit under each mode are included in the NEEDS v6 database,
described in Chapter 4. Attachment 3-2 gives further information on the procedures employed to derive
the four NOx mode rates.
Because of the complexity of the fleet and the completeness/incompleteness of historic data, there are
instances where the derivation of a unit's modeled NOx emission rate is more detailed than the
description provided above. For a more complete step-by-step description of the decision rules used to
develop the NOx rates, see Attachment 3-2.
3.10.4 Multi-Pollutant Environmental Regulations
Proposed GNP
On February 28, 2022, EPA proposed the Good Neighbor Plan (GNP) for the 2015 ozone National
Ambient Air Quality Standards (NAAQS). Starting in the 2023, 25 states will be subject to ozone season
NOx budgets consistent with Table 3-15. The programs' assurance provisions, which restrict the
maximum amount of exceedance of an individual state's emissions budget in each year through the use
of banked or traded allowances to 21% of the state's budget, are also implemented. The starting
allowance bank in 2023 is 22,319 tons, which is equal to the number of banked allowances at the start of
the GNP after old CSAPR Update / RCU allowances were converted. This is equal to the sum of the
states' 10.5% variability limits. In run year 2025, coal facilities greater than 100 MW lacking SCR controls
and certain oil/gas steam facilities greater than 100 MWthat lack existing SCR controls located in 23 of
these states must meet daily emission rate limits, effectively forcing affected units to install new SCR
controls, find other means of compliance, or retire.
3-24
-------
Table 3-15 Ozone-Season NOx Emission Caps (Tons) for Fossil Units greater than 25MW in v6
State
2023
2025 onwards
Alabama
7,702
7,577
Arkansas
10,756
4,786
Delaware
465
525
Illinois
8,958
7,354
Indiana
14,613
10,252
Kentucky
13,310
8,926
Louisiana
11,182
4,678
Maryland
1,435
1,534
Michigan
12,865
8,540
Minnesota
4,713
3,054
Mississippi
6,079
2,341
Missouri
13,965
9,698
Nevada
2,759
1,465
New Jersey
1,279
1,279
New York
4,657
3,978
Ohio
9,966
10,222
Oklahoma
12,424
5,203
Pennsylvania
10,721
8,259
Tennessee
5,123
4,850
Texas
47,932
26,897
Utah
18,127
3,169
Virginia
3,730
3,096
West Virginia
16,100
12,897
Wisconsin
7,214
4,201
Wyoming
11,130
5,441
Regional Cap
212,564
132,418
CSAPR
EPA Platform v6 includes the Cross-State Air Pollution Rule (CSAPR) Rule, CSAPR Update Rule, and
the Revised CSAPR Update Rule federal regulatory measures affecting 23 states to address transport
under the 1997, 2006, and 2008 National Ambient Air Quality Standards (NAAQS) for fine particle
pollution and ozone. CSAPR requires fossil-fired EGUs greater than 25 MW in a total of 22 states to
reduce annual SO2 emissions, annual NOx emissions, and/or ozone season NOx emissions to assist in
attaining the 1997 ozone and fine particle and 2006 fine particle National Ambient Air Quality Standards
(NAAQS). The CSAPR Phase 2 combined annual emissions budgets are 1,372,631 tons SO2 for CSAPR
SO2 Group 1 ;32 597,579 tons SO2 for CSAPR SO2 Group 2;33 and 1,069,256 tons for annual NOx.34 As
the budgets are significantly above current emission levels, i.e., they are not binding, the EPA did not
include a starting bank of allowances for these programs for simplicity.
The original Phase 2 combined ozone season NOx emissions budget was 0.59 million tons. However,
several of the state budgets were remanded. As the CSAPR Update Rule addresses the D.C. Circuit's
remand, the budgets for these states were updated to reflect those promulgated in the CSAPR Update
Rule. The programs' assurance provisions, which restrict the maximum amount of exceedance of an
individual state's emissions budget in a given year through the use of banked or traded allowances to
32 Illinois, Indiana, Iowa, Kentucky, Maryland, Michigan, Missouri, New Jersey, New York, North Carolina, Ohio,
Pennsylvania, Tennessee, Virginia, West Virginia, and Wisconsin.
33 Alabama, Georgia, Kansas, Minnesota, Nebraska, and South Carolina.
34 Alabama, Georgia, Illinois, Indiana, Iowa, Kansas, Kentucky, Maryland, Michigan, Minnesota, Missouri, Nebraska,
New Jersey, New York, North Carolina, Ohio, Pennsylvania, South Carolina, Tennessee, Virginia, West Virginia, and
Wisconsin.
3-25
-------
18% or 21% of the state's budget are also included. For more information on CSAPR, go to
https://www.epa.gov/csapr/overview-cross-state-air-pollution-rule-csapr.
The state budgets for Ozone Season NOx for the CSAPR Update Rule (that were not further adjusted in
the Revised CSAPR Update Rule) are shown in Table 3-16. Additionally, Georgia was modeled as a
separate region, with Georgia units unable to trade allowances with units in other states and received its
CSAPR Phase 2 budget and assurance level, as shown in Table 3-16. This is because Georgia, unlike
the other states covered by the CSAPR Update Rule, did not significantly contribute to a downwind
nonattainment or maintenance receptor for the 2008 NAAQS. Further, Georgia did not have a remanded
Ozone Season NOx budget related to a D.C. Circuit Court decision on the original Cross-State Air
Pollution Rule.
The programs' assurance provisions, which restrict the maximum amount of exceedance of an individual
state's emissions budget in each year through the use of banked or traded allowances to 21% of the
state's budget, are also implemented. This is equal to one-and-a-half times the sum of the states' 21%
variability limits. For more information on CSAPR, go to httpsi//www.epa.qov/csapr. For more
information on the CSAPR Update, go to https://www.epa.gov/airmarkets/final-cross-state-air-pollution-
rule-update.
Table 3-16 G1 and G2 CSAPR Update State Budgets, Variability Limits, and Assurance Levels for
Ozone-Season NOx(Tons) - 2021 through 2054
State
Budget
Variability Limit
Assurance Level
Alabama
13,211
2,774
15,985
Arkansas
9,210
1,934
11,144
Iowa
11,272
2,367
13,639
Kansas
8,027
1,686
9,713
Missouri
15,780
3,314
19,094
Mississippi
6,315
1,326
7,641
Oklahoma
11,641
2,445
14,086
Tennessee
7,736
1,625
9,361
Texas
52,301
10,983
63,284
Wisconsin
7,915
1,662
9,577
Georgia Budget, Variability Limit, and Assurance Level for Ozone-Season NOx
Georgia | 24,041 | 5,049 | 29,090
On March 15, 2021, EPA finalized the Revised Cross-State Air Pollution Rule Update for the 2008 ozone
National Ambient Air Quality Standards (NAAQS) to address the D.C. Circuit's remand of the CSAPR
Update Rule. Starting in the 2021, 12 of the 22 states covered in the CSAPR Update Rule will revise
ozone season NOx budgets consistent with Table 3-17. The programs' assurance provisions, which
restrict the maximum amount of exceedance of an individual state's emissions budget in each year
through the use of banked or traded allowances to 21% of the state's budget, are also implemented. The
starting allowance bank in 2023 is 22,488 tons, which is equal to the number of banked allowances at the
start of the Revised CSAPR Update program after old CSAPR Update allowances were converted. This
is equal to the sum of the states' 21% variability limits.
Table 3-17 Revised CSAPR Update State Budgets, Variability Limits, and Assurance Levels for
Ozone-Season NOxfor G3 states (tons)
State
Budget (tons)
Variability Limit (tons)
Assurance Level (tons)
2021
Illinois
9,102
1,911
11,013
Indiana
13,051
2,741
15,792
Kentucky
15,300
3,213
18,513
Louisiana
14,818
3,112
17,930
Maryland
1,499
315
1,814
Michigan
12,727
2,673
15,400
3-26
-------
State
Budget (tons)
Variability Limit (tons)
Assurance Level (tons)
New Jersey
1,253
263
1,516
New York
3,416
717
4,133
Ohio
9,690
2,035
11,725
Pennsylvania
8,379
1,760
10,139
Virginia
4,516
948
5,464
West Virginia
13,334
2,800
16,134
2022
Illinois
9,102
1,911
11,013
Indiana
12,582
2,642
15,224
Kentucky
14,051
2,951
17,002
Louisiana
14,818
3,112
17,930
Maryland
1,266
266
1,532
Michigan
12,290
2,581
14,871
New Jersey
1,253
263
1,516
New York
3,416
717
4,133
Ohio
9,773
2,052
11,825
Pennsylvania
8,373
1,758
10,131
Virginia
3,897
818
4,715
West Virginia
12,884
2,706
15,590
2023
Illinois
8,179
1,718
9,897
Indiana
12,553
2,636
15,189
Kentucky
14,051
2,951
17,002
Louisiana
14,818
3,112
17,930
Maryland
1,266
266
1,532
Michigan
9,975
2,095
12,070
New Jersey
1,253
263
1,516
New York
3,421
718
4,139
Ohio
9,773
2,052
11,825
Pennsylvania
8,373
1,758
10,131
Virginia
3,980
836
4,816
West Virginia
12,884
2,706
15,590
2024 -2059
Illinois
8,059
1,692
9,751
Indiana
9,564
2,008
11,572
Kentucky
14,051
2,951
17,002
Louisiana
14,818
3,112
17,930
Maryland
1,348
283
1,631
Michigan
9,786
2,055
11,841
New Jersey
1,253
263
1,516
New York
3,403
715
4,118
Ohio
9,773
2,052
11,825
Pennsylvania
8,373
1,758
10,131
Virginia
3,663
769
4,432
West Virginia
12,884
2,706
15,590
MATS
Finalized in 2011, the Mercury and Air Toxics Rule (MATS) establishes National Emissions Standards for
Hazardous Air Pollutants (NESHAPS) for the "electric utility steam generating unit" source category,
which includes those units that combust coal or oil for the purpose of generating electricity for sale and
3-27
-------
distribution through the electric grid to the public. EPA Platform v6 applies the input-based (Ibs/MMBtu)
MATS control requirements for mercury and hydrogen chloride to covered units.
EPA Platform v6 assumes that all active coal-fired generating units with a capacity greater than 25 MW
have complied with the MATS filterable PM requirements through the operation of either electrostatic
precipitator (ESP) or fabric filter (FF) particulate controls. No additional PM controls beyond those in
NEEDS v6 are modeled in EPA Platform v6.
EPA Platform v6 does not model the alternative SO2 standard offered under MATS for units to
demonstrate compliance with the rule's HCI control requirements. Coal steam units with access to lignite
in the modeling are required to meet the "existing coal-fired unit low Btu virgin coal" standard. For more
information on MATS, go to httpi//www.epa.qov/mats/.
Regional Haze
The Clean Air Act establishes a national goal for returning visibility to natural conditions through the
"prevention of any future, and the remedying of any existing impairment of visibility in Class I areas [156
national parks and wilderness areas], where impairment results from manmade air pollution." On July 1,
1999, EPA established a comprehensive visibility protection program with the issuance of the regional
haze rule (64 FR 35714). The rule implements the requirements of section 169B of the CAAA and
requires states to submit State Implementation Plans (SIPs) establishing goals and long-term strategies
for reducing emissions of air pollutants (including SO2 and NOx) that cause or contribute to visibility
impairment. The requirement to submit a regional haze SIP applies to all 50 states, the District of
Columbia, and the Virgin Islands. Among the components of a long-term strategy is the requirement for
states to establish emission limits for visibility-impairing pollutants emitted by certain source types
(including EGUs) that were placed in operation between 1962 and 1977. These emission limits are to
reflect Best Available Retrofit Technology (BART). States may perform individual point source BART
determinations, or meet the requirements of the rule with an approved BART alternative. An alternative
regional SO2 cap for EGUs under Section 309 of the regional haze rule is available to certain western
states whose emission sources affect Class 1 areas on the Colorado Plateau.
Since 2010, EPA has approved regional haze State Implementation Plans (SIPs) or, in a few cases, put
in place regional haze Federal Implementation Plans for several states. The BART limits approved in
these plans (as of January 2021) that will be in place for EGUs are represented in EPA Platform v6 as
follows.
• Source-specific NOx or SO2 BART emission limits, minimum SO2 removal efficiency requirements for
FGDs, limits on sulfur content in fuel oil, constraints on fuel type (e.g., natural gas only or prohibition
of certain fuels such as petroleum coke), or commitments to retire units are applied to the relevant
EGUs.
• EGUs in states that rely on CSAPR trading programs to satisfy BART must meet the requirements of
CSAPR.
• EGUs in states that rely on state power plant rules to satisfy BART must meet the emission limits
imposed by those state rules.
• For the three western states (New Mexico, Wyoming, and Utah) with approved Section 309 SIPs for
SO2 BART, emission constraints were not applied as current and projected emissions are well under
the regional SO2 cap.
Table 3-36 lists the NOx and SO2 limits applied to specific EGUs, and other implementations applied in
IPM. For more information on the Regional Haze Rule, go to https://www.epa.gov/visibilttv.
On June 28, 2021, EPA filed a status update with the United States Court of Appeals for the District of
Columbia Circuit noting that "the agency is convening a proceeding for reconsideration" of the August
3-28
-------
2020 rule known as the "Texas Regional Haze BART and Interstate Visibility Transport FIP." Any
changes from the that effort will be incorporated into EPA modeling when finalized.
3.10.5 CO2 Regulations
The Regional Greenhouse Gas Initiative (RGGI) is a CO2 cap and trade program affecting fossil fired
electric power plants 25 MW or larger in Connecticut, Delaware, Maine, Maryland, Massachusetts, New
Hampshire, New Jersey, New York, Rhode Island, Vermont, and Virginia. Table 3-24 shows the
specifications for RGGI that are implemented in EPA Platform v6. If/when other states join RGGI and
finalize/implement regulations, EPA will adjust its representation accordingly.
As part of California's Assembly Bill 32 (AB32), the Global Warming Solutions Act, a multi-sector GHG
cap-and-trade program was established that establishes long-term economy-wide emission
targets, starting in 2013 for electric utilities and large industrial facilities, with distributors of transportation,
natural gas, and other fuels joining the capped sectors in 2015. In addition to in-state sources, the cap-
and-trade program also covers the emissions associated with qualifying, out-of-state EGUs that sell
power into California. Due to the inherent complexity in modeling a multi-sector cap-and-trade program
where the participation of out-of-state EGUs is determined based on endogenous behavior (i.e., IPM
determines whether qualifying out-of-state EGUs are projected to sell power into California), EPA has
developed a simplified methodology to model California's economy-wide cap-and-trade program as
follows.
• Adopt the AB32 cap-and-trade allowance price from ElA's AE02020 Reference Case, which fully
represents the non-power sectors. All qualifying fossil-fired EGUs in California are subject to this
price signal, which is applied through the end of the modeled time horizon since the underlying
legislation requires those emission levels to be maintained.
• Assume the marginal CO2 emission rate for each IPM region that exports power to California to be
0.428 MT/MWh.
• For each IPM region that exports power to California, convert the $/ton CO2 allowance price
projection into a mills/kWh transmission wheeling charge using the marginal emission rate from the
previous step. The additional wheeling charge for qualifying out-of-state EGUs is equal to the
allowance price imposed on affected in-state EGUs. Applying the charge to the transmission link
ensures that power imported into California from out-of-state EGUs must account for the cost of CO2
emissions represented by its generation, such that the model may clear the California market in a
manner consistent with AB32 policy treatment of CO2 emissions.
Federal CO2 standards for existing sources are not modeled, given ongoing litigation and regulatory
review.35 For new fossil fuel-fired sources, EPA Platform v6 continues to include the Standards of
Performance for Greenhouse Gas Emissions from New, Modified, and Reconstructed Stationary Sources:
Electric Generating Units (New Source Rule).36 Although this rule is also being reviewed,37 the standards
of performance are legally in effect until such review is completed and/or revised. In addition, state level
CO2 standards were implemented in Colorado (HB21-1266), Massachusetts (Massachusetts Senate Bill
9), North Carolina (North Carolina House Bill 951), Oregon (Oregon House Bill 2021), and Washington
(Washington state SB5126).
35 EPA Memorandum: "Status of Affordable Clean Energy Rule and Clean Power Plan," February 12, 2021. Available
at https://www.epa.gov/sites/default/files/2021-02/documents/ace letter 021121.doc siqned.pdf.
36 80 FR 64510
37 82 FR 16330
3-29
-------
3.10.6 Non-Air Regulations Impacting EGUs
Cooling Water Intakes (316(b)) Rule
Section 316(b) of the Clean Water Act requires that National Pollutant Discharge Elimination System
(NPDES) permits for facilities with cooling water intake structures ensure that the location, design,
construction, and capacity of the structures reflect the best technology available to minimize harmful
impacts on the environment. Under a 1995 consent decree with environmental organizations, EPA
divided the section 316(b) rulemaking into three phases. All new facilities except offshore oil and gas
exploration facilities were addressed in Phase I in December 2001; all new offshore oil and gas
exploration facilities were later addressed in June 2006 as part of Phase III. This final rule also removes
a portion of the Phase I rule to comply with court rulings. Existing large electric-generating facilities were
addressed in Phase II in February 2004. Existing small electric-generating and all manufacturing facilities
were addressed in Phase III (June 2006). However, Phase II and the existing facility portion of Phase III
were remanded to EPA for reconsideration because of legal proceedings. This final rule combines these
remands into one rule and provides a holistic approach to protecting aquatic life impacted by cooling
water intakes. The rule covers roughly 1,065 existing facilities that are designed to withdraw at least 2
million gallons per day of cooling water. EPA estimates that 544 power plants are affected by this rule.
The final regulation has three components for affected facilities: 1) reduce fish impingement through a
technology option that meets best technology available requirements, 2) conduct site-specific studies to
help determine whether additional controls are necessary to reduce entrainment, and 3) meet
entrainment standards for new units at existing facilities when additional capacity is added. EPA Platform
v6 includes the cost of complying with this rule. The cost assumptions and analysis for 316(b) can be
found in Chapter 8.7 of the Rule's Technical Development Document for the Final Section 316(b) Existing
Facilities Rule at https://www.epa.qov/sites/production/files/2015~04/documents/coolinq-water phase-
4 tdd 2014.pdf.
For more information on 316(b), go to https://www.epa.qov/coolinq-water-intakes.
Combustion Residuals from Electric Utilities (CCR)
In December of 2014, EPA finalized national regulations to provide a comprehensive set of requirements
for the safe disposal of coal combustion residuals (CCRs), commonly known as coal ash, from coal-fired
power plants. The final rule is the culmination of extensive study on the effects of coal ash on the
environment and public health. The rule establishes technical requirements for CCR landfills and surface
impoundments under Subtitle D of the Resource Conservation and Recovery Act.
EPA Platform v6 includes cost of complying with this rule's requirements by taking the estimated plant-
level compliance cost identified for the CCR final rule and apportioning them into unit-level cost38. Three
categories of unit-level cost were quantified: capital cost, fixed operating and maintenance cost (FOM),
and variable operating and maintenance (VOM) cost. The method for apportioning these costs to the
unit-level for inclusion in EPA Platform is discussed in the Addendum to the RIA for EPA's 2015 Coal
Combustion Residuals (CCR) Final Rule. The initial plant-level cost estimates are discussed in the Rule's
Regulatory Impact Analysis.
In September of 2017, EPA granted petitions to reconsider some provisions of the rule. In granting the
petitions, EPA determined that it was appropriate, and in the public's interest to reconsider specific
provisions of the final CCR rule based in part on the authority provided through the Water Infrastructure
38 CCR related cost adders were not applied to units with CCR-based retirement dates no later than
12/31/2028.
3-30
-------
for Improvements to the Nation (WIIN) Act. At time of this modeling update, EPA had not committed to
changing any part of the rule or agreeing with the merits of the petition - the Agency is simply granting
petitions to reconsider specific provisions. Should EPA decide to revise specific provisions of the final
CCR rule, it will go through notice and comment period, and the rules corresponding model specification
would be subsequently changed in future base case platforms.
On July 29, 2020, the U.S. Environmental Protection Agency (EPA) finalized several changes to the
regulations for this rule to implement the court's vacatur of certain closure requirements. In response to
court rulings, this final rule specified that all unlined surface impoundments are required to retrofit or
close, not just those that have detected groundwater contamination above regulatory levels. The rule also
changed the classification of compacted-soil lined or "clay-lined" surface impoundments from "lined" to
"unlined," which means that formerly defined clay-lined surface impoundments are no longer considered
lined surface impoundments and need to be retrofitted or closed. These changes, and corresponding
requirements and cost, are reflected in this version of IPM using the same methodology described in the
Addendum for the RIA for EPA's 2015 CCR Rule mentioned above.
For more information on CCR, go to http://www.epa.gov/coalash/coal-ash-rule.
Effluent Limitation and Guidelines (ELG)
In September of 2015, EPA finalized a rule revising the regulations for Steam Electric Power Generating
category (40 CFR Part 423).39 The rule established federal limits on the levels of toxic metals in
wastewater that can be discharged from power plants. The rule established or updated standards for
wastewater streams from flue gas desulfurization, fly ash, bottom ash, flue gas mercury control, and
gasification of fuels.
On October 13, 2020 - EPA published a reconsideration rule that revised the requirements for flue gas
desulfurization (FGD) wastewater and bottom ash (BA) transport water; revised the voluntary incentives
program for FGD wastewater; added subcategories; and established new compliance dates. These
changes, and corresponding requirements and cost, are reflected in EPA Platform v6. EPA reflects this
rule in this base case by apportioning the estimated total capital and FOM costs to likely affected units
based on controls and capacity. The cost adders are reflected in the model inputs and were applied
starting in 2025, by which point the requirements were expected to be fully implemented.
On July 26, 2021, EPA announced it was initiating a supplemental rulemaking to strengthen certain
discharge limits in the Steam Electric Power Generating category. EPA undertook a science-based
review of the 2020 Steam Electric Reconsideration Rule under Executive Order 13990, finding that
opportunities for improvement exist. EPA intends to issue a proposed rule for public comment in the fall of
2022. The current rule will continue to be implemented (and reflected in IPM) and any additional or
updated requirements from this supplemental rulemaking will be incorporated when final.
For more information on ELG, go to https://www.epa.gov/eq/effluent-quidelines-plan.
3.10.7 State-Specific Environmental Regulations
EPA Platform v6 represents enacted laws and regulations in states affecting emissions from the electricity
sector. Table 3-31 summarizes the provisions of state laws and regulations that are represented in EPA
Platform v6.
39 https://vww.eDa.aov/eq/steam-electric-power-qeneratina-effluent-guidelines-2015-final-rule
3-31
-------
3.10.8 New Source Review (NSR) Settlements
New Source Review (NSR) settlements refer to legal agreements with companies resulting from the
permitting process under the CAAA which requires industry to undergo an EPA pre-construction review of
proposed environmental controls either on new facilities or as modifications to existing facilities where
there would result a "significant increase" in a regulated pollutant. A summary of the units affected and
how the settlements were modeled can be found in Table 3-32.
State settlements and citizen settlements are also represented in EPA Platform v6. These are
summarized in Table 3-33 and Table 3-34 respectively.
3.10.9 Emission Assumptions for Potential (New) Units
There are no location-specific variations in the emission and removal rate capabilities of potential new
units. In IPM, potential new units are modeled as additional capacity and generation that may come
online in each model region. Across all model regions, the emission and removal rate capabilities of
potential new units are the same, and they reflect applicable federal emission limitations on new sources.
The specific assumptions regarding the emission and removal rates of potential new units in EPA
Platform v6 are presented in Table 3-26. (Note: Nuclear, wind, solar, and fuel cell technologies are not
included in Table 3-26 because they do not emit any of the listed pollutants.) For additional details on the
modeling of potential new units, see Chapter 4.
3.10.10 Renewable Portfolio Standards and Clean Energy Standards
Renewable Portfolio Standards (RPS) generally refer to various state-level policies that require renewable
generation to meet a specified share of generation or sales. In EPA Platform v6, the state RPS
requirements are represented at a state level based on existing requirements. Table 3-18 and Table 3-19
show the state-level RPS and solar carve-out requirements.
Table 3-18 Renewable Portfolio Standards in v6
State
2028 2030 2035 2040 2045 2050 2055
Arizona
California
Colorado
Connecticut
District of Columbia
Delaware
Iowa
Illinois
Massachusetts
Maryland
Maine
Michigan
Minnesota
Missouri
Montana
North Carolina
New Hampshire
New Jersey
New Mexico
Nevada
New York
Ohio
Oregon
Pennsylvania
Rhode Island
Texas
8.6% 8.6% 8.6% 8.6% 8.6% 8.6% 8.6%
52.0% 57.3% 70.7% 84.0% 97.3% 100.0% 100.0%
21.2% 21.2% 21.2% 21.2% 21.2% 21.2% 21.2%
40.0% 44.0% 44.0% 44.0% 44.0% 44.0% 44.0%
73.0% 87.0% 100.0% 100.0% 100.0% 100.0% 100.0%
18.9% 20.0% 28.5% 28.5% 28.5% 28.5% 28.5%
0.6% 0.6% 0.6% 0.6% 0.5% 0.5% 0.5%
32.5% 40.0% 45.0% 50.0% 50.0% 50.0% 50.0%
28.5% 30.5% 35.5% 40.5% 45.5% 50.5% 50.5%
47.5% 50.0% 50.0% 50.0% 50.0% 50.0% 50.0%
71.0% 80.0% 85.0% 90.0% 95.0% 100.0% 100.0%
35.0% 35.0% 35.0% 35.0% 35.0% 35.0% 35.0%
28.5% 28.5% 28.5% 28.5% 28.5% 28.5% 28.5%
10.5% 10.5% 10.5% 10.5% 10.5% 10.5% 10.5%
10.4% 10.4% 10.4% 10.4% 10.4% 10.4% 10.4%
6.9% 6.9% 6.9% 6.9% 6.9% 6.9% 6.9%
23.0% 23.0% 23.0% 23.0% 23.0% 23.0% 23.0%
46.5% 52.5% 52.5% 52.5% 52.5% 52.5% 52.5%
41.6% 45.2% 57.2% 69.2% 70.7% 72.3% 72.3%
34.8% 41.4% 41.4% 41.4% 41.4% 41.4% 41.4%
61.2% 70.0% 70.0% 70.0% 70.0% 70.0% 70.0%
7.6% 7.6% 7.6% 7.6% 7.6% 7.6% 7.6%
21.6% 27.6% 36.1% 41.1% 42.6% 42.6% 42.6%
8.0% 8.0% 8.0% 8.0% 8.0% 8.0% 8.0%
57.0% 73.5% 100.0% 100.0% 100.0% 100.0% 100.0%
3.9% 3.8% 3.6% 3.4% 3.2% 3.0% 3.0%
3-32
-------
Virginia 27.1% 32.0% 46.2% 62.6% 78.9% 81.6% 81.6%
Vermont 74.6% 79.8% 85.0% 85.0% 85.0% 85.0% 85.0%
Washington 12.2% 12.2% 12.2% 12.2% 12.2% 12.2% 12.2%
Wisconsin 9.6% 9.6% 9.6% 9.6% 9.6% 9.6% 9.6%
Notes:
The Renewable Portfolio Standard percentages are applied to modeled electricity sale projections.
North Carolina standards are adjusted to account for swine waste and poultry waste set-asides.
Table 3-19 State RPS Solar Carve-outs in v6
State
2028
2030
2035
2040
2045
2050
2055
District of Columbia
4.5%
5.0%
7.0%
9.0%
10.0%
10.0%
10.0%
Delaware
3.0%
3.6%
7.1%
7.1%
7.1%
7.1%
7.1%
Illinois
1.5%
1.5%
1.5%
1.5%
1.5%
1.5%
1.5%
Massachusetts
0.2%
0.2%
0.3%
0.3%
0.4%
0.4%
0.4%
Maryland
14.5%
14.5%
14.5%
14.5%
14.5%
14.5%
14.5%
Minnesota
1.2%
1.2%
1.2%
1.2%
1.2%
1.2%
1.2%
Missouri
0.2%
0.2%
0.2%
0.2%
0.2%
0.2%
0.2%
North Carolina
0.1%
0.1%
0.1%
0.1%
0.1%
0.1%
0.1%
New Hampshire
0.7%
0.7%
0.7%
0.7%
0.7%
0.7%
0.7%
New Jersey
3.7%
2.2%
1.1%
1.1%
1.1%
1.1%
1.1%
Pennsylvania
0.5%
0.5%
0.5%
0.5%
0.5%
0.5%
0.5%
Clean Energy Standards require a certain percentage of electricity sales be met through zero carbon
resources, such as renewables, nuclear, and hydropower. Several states, including California, New
Mexico, Nevada, New York, and Washington, have recently implemented clean energy standards. These
requirements are summarized in Table 3-20. In addition, multiple U.S. states have recently adopted
offshore wind energy policies, which are summarized in Table 3-21. Thermal generation limits are
imposed in states where RPS or CES standards exceed 50% of sales to ensure that the states do not
generate excess thermal power to satisfy exports.
Table 3-22 summarizes the limits imposed in EPA Platform v6. These limits are not provided in affected
PJM and New England states as these states can meet their RPS requirements within PJM or ISONE.
Table 3-20 Clean Energy Standards in v6
State
2028
2030
2035
2040
2045
2050
2055
Colorado
-
-
-
-
-
53%
53%
Oregon
-
-
-
100%
100%
100%
100%
Illinois
-
-
-
-
-
100%
100%
Massachusetts
36%
40%
50%
60%
70%
80%
80%
Connecticut
0%
40%
70%
100%
100%
100%
100%
California
-
-
-
-
-
100%
100%
New Mexico
-
-
-
-
70%
90%
90%
Nevada
-
-
-
-
-
100%
100%
New York
-
-
-
100%
100%
100%
100%
Washington
-
100%
100%
100%
100%
100%
100%
Table 3-21 Offshore Wind Mandates in v6
State
Bill/Act
Mandate Specifications
Implementation Year
Maryland
Senate Bill 516
400 MW, 800 MW, and 1,200
MW of offshore wind capacity by
2026, 2028 and 2030
respectively
2030
Maryland Offshore Wind
Energy Act of 2013
368 MW of offshore wind
capacity (248 MW of US Wind,
2023
3-33
-------
State
Bill/Act
Mandate Specifications
Implementation Year
Inc. and 120 MW of Skipjack
Offshore Energy, LLC projects)
New Jersey
Executive Order No. 92
7,500 MW of offshore wind
capacity by 2035
2035
Connecticut
House Bill 7156
2,000 MW of offshore wind
capacity by 2030
2030
Massachusetts
Massachusetts Energy
Diversity Act
4,000 MW of offshore wind
capacity by 2027
2028
New York
Climate Leadership and
Community Protection Act
9,000 MW of offshore wind
capacity by 2035
2035
Virginia
Virginia Clean Economy
Act
development by Dominion
Energy Virginia of qualified
offshore wind projects having an
aggregate rated capacity of not
less than 5,200 megawatts by
January 1, 2034
2035
Maine
Final Report of the Ocean
Energy Task Force, 2009
Goal of 5,000 MW of offshore
wind capacity by 2030
Not implemented
California
3,500 MW by 2030 and 25,000
MW by 2045
2030
Table 3-22 Fossil Generation Limits (GWh) in v6
State
2028
2030
2035
2040
2045
2050
2055
California
145,220
134,260
105,555
73,248
36,634
29,508
31,920
Colorado
-
-
-
-
-
48,590
51,956
Illinois
-
-
-
92,227
96,868
11,378
11,942
Nevada
-
-
-
-
-
5,501
5,942
New Mexico
-
-
13,578
11,228
11,603
6,044
6,470
New York
67,773
55,820
58,212
12,484
13,303
14,227
15,129
Oregon
-
-
-
6,546
7,054
7,676
8,283
Virginia
-
-
-
64,131
42,621
40,915
43,366
Washington
-
10,693
11,409
12,289
13,242
14,409
15,549
3.10.11 Canada CO2 and Renewable Regulations
Several CO2 regulations in Canada are represented in EPA Platform v6. Under the Reduction of Carbon
Dioxide Emissions from Coal-fired Generation of Electricity Regulations, the CO2 standard of 420 tonne
/GWh of electricity produced applies to both coal-fired electricity generating units commissioned after July
1, 2015, and existing coal units that have reached their end-of-life date as defined by the regulation. EPA
Platform v6 also models British Columbia's carbon tax, Manitoba's Emissions Tax on Coal and Petroleum
Coke Act, and the Ontario and Quebec's participation in Western Climate Initiative (WCI) cap-and-trade
program. Coming into force on January 1, 2012, Manitoba's Emissions Tax on Coal and Petroleum Coke
Act requires a tax rate of $10 per tonne of CO2 equivalent emissions on coal-fired and petroleum coke-
fired units. Ontario and Quebec's participation in WCI is modeled through the application of the CO2
allowance price from CA AB32. EPA Platform v6 also models the province level renewable electricity
programs in Canada. Table 3-23 shows the province level renewable electricity requirements as a
percentage of electricity sales.
3-34
-------
Table 3-23 Canada Renewable Electricity Requirements (%) in v6
Province
2028
2030
2035
2040
2045
2050
2055
British Columbia
93.0%
93.0%
93.0%
93.0%
93.0%
93.0%
93.0%
Alberta
30.0%
30.0%
30.0%
30.0%
30.0%
30.0%
Saskatchewan
40.0%
50.0%
50.0%
50.0%
50.0%
50.0%
50.0%
New Brunswick
40.0%
40.0%
40.0%
40.0%
40.0%
40.0%
40.0%
Nova Scotia
40.0%
40.0%
40.0%
40.0%
40.0%
40.0%
40.0%
Prince Edward Island
30.0%
30.0%
30.0%
30.0%
30.0%
30.00%
30.00%
3.11 Emissions Trading and Banking
Several environmental air regulations included in EPA Platform v6 involve regional trading and banking of
emission allowances. This includes the five programs of the Cross-State Air Pollution Rule (CSAPR) -
SO2 Group 1, SO2 Group 2, NOx Annual, NOx Ozone Season Group 1, NOx Ozone Season Group 2, and
NOx Ozone Season Group 3; the Regional Greenhouse Gas Initiative (RGGI) for CO2; the SIP Call Ozone
Season NOx; and the West Region Air Partnership's (WRAP) program regulating SO2 (adopted in
response to the federal Regional Haze Rule).
Table 3-24 and Table 3-25 summarize the key parameters of these trading and banking programs as
incorporated in EPA Platform v6. EPA Platform v6 does not include any explicit assumptions on the
allocation of emission allowances among model plants under any of the programs.
3.11.1 Intertemporal Allowance Price Calculation
Under a perfectly competitive cap-and-trade program that allows banking (with a single, fixed future cap,
and full banking allowed), the allowance price always increases by the discount rate between periods if
affected sources have allowances banked between those two periods. This is a standard economic result
for cap-and-trade programs and is consistent with producing a least-cost solution.
EPA Platform v6 uses the same discount rate assumption that governs all intertemporal economic
decision-making in the model. The approach assumes that allowance trading is a standard activity
engaged in by generation asset owners and that their intertemporal investment decisions as related to
allowance trading will not fundamentally differ from other investment decisions. For more information on
how this discount rate was calculated, see Section 10.4.
Table 3-24 Trading and Banking Rules in v6 - Part 1
SIP Call - Ozone Season NOx
WRAP- SO2
RGGI - CO2
Coverage
All fossil units > 25 MW1
All fossil units > 25 MW2
All fossil units > 25 MW3
Timing
Ozone Season (May - September)
Annual
Annual
Size of Initial Bank
(MTons)
The bank starting in 2016 is assumed
to be zero
The bank starting in 2018
is assumed to be zero
2023: 113,656
Total Allowances
(MTons)
2016-2059: 72.845
2018-2059: 89.6
2023
2024
2025
2026
2027
2028
2029
2030
112,458
108,803
105,148
101,493
97,838
94,183
90,528
-2059: 86,873
Notes:
1 Rhode Island, Connecticut, Delaware, District of Columbia, Massachusetts, North Carolina, and South Carolina are the NOx SIP
Call states not covered by the CSAPR Ozone Season program.
2 New Mexico, Utah, and Wyoming.
3 Connecticut, Delaware, Maine, New Hampshire, New York, Vermont, Rhode Island, Massachusetts, Maryland, Virginia, and New
Jersey.
3-35
-------
Table 3-25 CASPR Trading and Banking Rules in v6 - Part 2
CSAPR-
so2 -
CSAPR-
so2 -
CSAPR-
Annual
CSAPR
Update Rule
- Ozone
CSAPR Update
Rule - Ozone
Season NOx -
Group 2
Revised CSPR
Update Rule -
Ozone Season -
Region 1
Region 2
NOx
Season NOx
- Group 1
Group 3
Coverage
All fossil
units > 25
MW1
All fossil
units > 25
MW2
All fossil
units > 25
MW3
All fossil
units > 25
MW5
All fossil units >
25 MW4
All fossil units > 25
MW6
Timing
Annual
Annual
Annual
Ozone
Season
(May -
September)
Ozone Season
(May -
September)
Ozone Season (May
- September)
The bank
The bank
The bank
The bank
The bank starting in
2021 is 21% of the
starting aggregate
state budgets
Size of
Initial Bank
starting in
2023 is
starting in
2023 is
starting in
2021 is
starting in
2021 is
The cap in 2021
includes 21% of
(MTons)
assumed
to be zero
assumed to
be zero
assumed to
be zero
assumed to
be zero
banking
Total
2023 -
2023 -
2023-
2023 -2059:
24.041
2023 -2059:
143.408
2023-100,526
Allowances
2059:
2059:
2059:
2024 through 2059 -
(MTons)
1372.631
597.579
1069.256
96,975
Notes:
1 Illinois, Indiana, Iowa, Kentucky, Maryland, Michigan, Missouri, New Jersey, New York, North Carolina,
Ohio, Pennsylvania, Tennessee, Virginia, West Virginia, and Wisconsin.
2 Alabama, Georgia, Kansas, Minnesota, Nebraska, and South Carolina.
3 Alabama, Georgia, Illinois, Indiana, Iowa, Kansas, Kentucky, Maryland, Michigan, Minnesota, Missouri,
Nebraska, New Jersey, New York, North Carolina, Ohio, Pennsylvania, South Carolina, Tennessee,
Virginia, West Virginia, and Wisconsin.
4 Alabama, Arkansas, Iowa, Illinois, Indiana, Kansas, Kentucky, Louisiana, Maryland, Michigan, Missouri,
Mississippi, New Jersey, New York, Ohio, Oklahoma, Pennsylvania, Tennessee, Texas, Virginia,
Wisconsin, and West Virginia.
5 Georgia.
6 Illinois, Indiana, Kentucky, Louisiana, Maryland, Michigan, New Jersey, New York, Ohio, Pennsylvania, Virginia, and West
Virginia.
3-36
-------
Table 3-26 Emission and Removal Rate Assumptions for Potential (New) Units in v6
Controls,
Removal,
and
Emissions
Rates
Ultra-
Supercritical
Pulverized
Coal
Ultra-
Supercritical
Pulverized
Coal with
30% CCS
Ultra-
Supercritical
Pulverized
Coal with
90% CCS
Advanced
Combined Cycle
Advanced
Combined Cycle
with CCS
Advanced
Combustion
Turbine
Biomass
Geothermal
Landfill
Gas
CM
o
CO
Removal /
Emissions
Rate
98% with a
floor of 0.06
Ibs/MMBtu
98% with a
floor of 0.06
Ibs/MMBtu
98% with a
floor of 0.06
Ibs/MMBtu
None
None
None
0.08
Ibs/MMBtu
None
None
NO„
Emission
Rate
0.07
Ibs/MMBtu
0.07
Ibs/MMBtu
0.07
Ibs/MMBtu
0.011 Ibs/MMBtu
0.011 Ibs/MMBtu
0.011 Ibs/MMBtu
0.02
Ibs/MMBtu
None
0.09
Ibs/MMBtu
Hg
Removal /
Emissions
Rate
90%
90%
90%
Natural Gas:
0.000138
Ibs/MMBtu
Oil:
0.483 Ibs/MMBtu
Natural Gas:
0.000138
Ibs/MMBtu
Oil:
0.483 Ibs/MMBtu
Natural Gas:
0.000138
Ibs/MMBtu
Oil:
0.483 Ibs/MMBtu
0.57
Ibs/MMBtu
3.70
None
co2
Removal /
Emissions
Rate
202.8-215.8
Ibs/MMBtu
30%
90%
Natural Gas:
117.08 Ibs/MMBtu
Oil:
161.39 Ibs/MMBtu
90%
Natural Gas:
117.08 Ibs/MMBtu
Oil:
161.39 Ibs/MMBtu
None
None
None
HCL
Removal /
Emissions
Rate
99% with a
floor of 0.001
Ibs/MMBtu
99% with a
floor of 0.001
Ibs/MMBtu
99% with a
floor of 0.001
Ibs/MMBtu
3-37
-------
Table 3-27 Recalculated NOx Emission Rates for SCR Equipped Units Sharing Common Stacks
with Non-SCR Units in v6
NOx Post-
SCR
Mode 1 NOx
Mode 2 NOx
Mode 3 NOx
Mode 4 NOx
UniquelD_
Capacity
Comb
Online
Rate
Rate
Rate
Rate
Plant Name
Final
(MW)
Control
Year
(Ibs/MMBtu)
(Ibs/MMBtu)
(Ibs/MMBtu)
(Ibs/MMBtu)
Ghent
1356 B 2
495
0.305
0.305
0.305
0.305
Ghent
1356 B 3
485
SCR
2004
0.075
0.075
0.075
0.075
Cooper
1384 B 1
116
0.273
0.273
0.199
0.199
Cooper
1384 B 2
225
SCR
2012
0.075
0.075
0.075
0.075
J H Campbell
1710 B 1
260
0.179
0.179
0.179
0.179
J H Campbell
1710 B 2
348
SCR
2013
0.047
0.047
0.047
0.047
W H Sammis
2866 B 5
290
SNCR
0.245
0.245
0.199
0.199
W H Sammis
2866 B 6
600
SCR
2010
0.075
0.075
0.075
0.075
W H Sammis
2866 B 7
600
SCR
2010
0.075
0.075
0.075
0.075
Crist
641 B 4
75
SNCR
0.406
0.119
0.147
0.1
Crist
641 B 5
75
SNCR
0.376
0.116
0.147
0.1
Crist
641 B 6
299
SCR
2012
0.248
0.068
0.248
0.068
Crist
641 B 7
475
SCR
2005
0.062
0.062
0.062
0.062
Clifty Creek
983 B 4
196
SCR
2003
0.075
0.075
0.075
0.075
Clifty Creek
983 B 5
196
SCR
2002
0.075
0.075
0.075
0.075
Clifty Creek
983_B_6
196
0.667
0.3
0.667
0.3
3.12 45Q - Credit for Carbon Dioxide Sequestration
Inflation Reduction Act of 2022, Section 45Q - which amended a Credit for Carbon Dioxide Sequestration
originally passed in 2008 (hereafter referred to as the 45Q tax credit) is implemented in EPA Platform v6.
The updated 45Q tax credit offers increased monetary incentives through a tax credit for the capture and
geologic storage of CO2 that electric power plants and other industrial sources in the United States would
otherwise emit. The essential features of the tax credit are as follows:
• $60 per metric ton in 2022 for CO2 captured and injected into existing oil wells for enhanced oil
recovery (EOR). The credit is adjusted for inflation post-2026.
• $85 per metric ton in 2022 for CO2 captured and sequestrated in geologic formation (non-EOR).
The credit is adjusted for inflation post-2026.
• The difference in the amounts of credit between EOR and Non-EOR is designed to recognize that
the EOR captured CO2 can be used to produce oil that may not otherwise be recovered while the
non-EOR stored CO2 does not bring additional revenue.
• Credits are available to plants that start construction or begin a retrofit before January 1, 2033,
and are assumed to be applied for the first 12 years of operation. Due to an assumed
construction lead time of 5 plus years for CCS retrofits, CCS retrofits in 2030 and 2035 run years
are assumed to qualify for the tax credit.
The 45Q tax credit is implemented by applying the value of the credit through an adjustment to the step
prices in the CO2 storage cost curves.40 The process involves converting the dollar amounts of credit into
40 For more information on the CO2 storage cost curves, see Chapter 6- CO2 Capture, Storage, and Transport in the
Documentation for EPA's Power Sector Modeling Platform v6 Using Integrated Planning Model. The documentation
is available online at https://vwwy.epa.gov/airmarkets/documentation-ipm-platform-v8-all-chapters.
3-38
-------
2019 real dollars, calculating weighted average tax credits by run year, and applying the weighted
average tax credits to the individual step prices in the CO2 storage cost curves.
List of tables and attachments that are uploaded directly to the web:
Table 3-28 Regional Net Internal Demand in EPA Platform v6 Post-IRA 2022 Reference Case
Table 3-29 Annual Transmission Capabilities of U.S. Model Regions in EPA Platform v6 Post-IRA 2022
Reference Case
Table 3-30 Turndown Assumptions for Coal Steam Units in EPA Platform v6 Post-IRA 2022 Reference
Case
Table 3-31 State Power Sector Regulations included in EPA Platform v6 Post-IRA 2022 Reference Case
Table 3-32 New Source Review (NSR) Settlements in EPA Platform v6 Post-IRA 2022 Reference Case
Table 3-33 State Settlements in EPA Platform v6 Post-IRA 2022 Reference Case
Table 3-34 Citizen Settlements in EPA Platform v6 Post-IRA 2022 Reference Case
Table 3-35 Availability Assumptions in EPA Platform v6 Post-IRA 2022 Reference Case
Table 3-36 BART Regulations included in EPA Platform v6 Post-IRA 2022 Reference Case
Attachment 3-1 Incremental Demand Accounting for on-the-books EPA OTAQ GHG Final Rule (not
reflected in AEO2021) in EPA Platform v6 Post-IRA 2022 Reference Case
Attachment 3-2 NOx Rate Development in EPA Platform v6 Post-IRA 2022 Reference Case
3-39
-------
4. Generating Resources
Existing, planned-committed, and potential are the three types of generating units modeled in EPA
Platform v6 Post-IRA 2022 Reference Case (EPA Platform v6). Electric generating units currently in
operation are termed as existing units. Units that are anticipated to be in operation in the near future, for
having broken ground or secured financing, are planned-committed units. Potential units refer to new
generating options that IPM builds to meet industry capacity expansion projections. Existing and
planned-committed units enter IPM as exogenous inputs, whereas potential units are endogenous to IPM
in that the model determines the location and size of the potential units to build.
This chapter is organized as follows.
i) Section 4.1 provides background information on the National Electric Energy Data System
(NEEDS), the database that serves as the repository for information on existing and planned-
committed electric generating units modeled,
ii) Section 4.2 provides detailed information on existing non-nuclear generating units,
iii) Section 4.3 provides detailed information on planned-committed units,
iv) Section 4.4 provides detailed information on potential units, and
v) Section 4.6 describes assumptions pertaining to existing and potential nuclear units.
4.1 National Electric Energy Data System (NEEDS)
EPA Platform v6 uses the NEEDS v6 database as its source for data on all existing and planned-
committed units. Section 4.2 discusses the sources used in developing data on existing units. The
population of existing units in the NEEDS v6 represents electric generating units that were in operation
through the end of 2021. Section 4.3 discusses the sources used in developing data on planned-
committed units. The population of planned-committed includes units online or scheduled to come online
from 2022 through June 30, 2028.
4.2 Existing Units
The sections below describe the procedures for determining the population of existing units in NEEDS v6,
as well as the capacity, location, and configuration information of each unit in the population.
4.2.1 Population of Existing Units
The capacity data for existing units in NEEDS v6 was obtained from the sources reported in Table 4-1.
The September 2019 EIA Form 860M is the primary data source on existing units. Table 4-2 specifies
the screening rules applied to the data source to ensure data consistency and adaptability for use in EPA
Platform v6.
4-1
-------
Table 4-1 Data Sources for NEEDS v6
Data Source1
Data Source Documentation
ElA's Form EIA-860 is both a monthly and annual survey of utility and non-utility power
plants at the generator level. It contains data such as summer, winter and nameplate
capacity, location (state and county), operating status, prime mover, energy sources
and in-service date of existing and proposed generators. NEEDS v6 uses EIA Form 860
(September 2019 monthly version and 2018 annual release) data as primary generator
data inputs.
ElA's Form EIA-860
ElA's Form EIA-860 also collects data of steam boilers such as energy sources, boiler
identification, location, operating status, and design information; and associated
environmental equipment such as NOx combustion and post-combustion control, FGD
scrubber, mercury control and particulate collector device information. Note that boilers
in plants with less than 10 MW do not report all data elements. The association
between boilers and generators is also provided. Note that boilers and generators are
not necessarily in a one-to-one correspondence. NEEDS v6 uses EIA Form 860 (2018
annual release) data as one of the primary boiler data inputs.
ElA's Annual Energy
Outlook (AEO)
The Energy Information Administration (EIA) Annual Energy Outlook presents annually
updated projections of energy supply, demand and prices covering a 20-25 year time
horizon. The projections are based on results from ElA's National Energy Modeling
System (NEMS). Information from AEO 2020 Reference Case such as heat rates and
capacity for nuclear units was used in NEEDS v6.
EPA's Emission
Tracking System
The Emission Tracking System (ETS) database is updated quarterly. It contains boiler-
level information such as primary fuel, heat input, SO2, NOx, Mercury, and HCI controls,
and SO2 and NOx emissions. NEEDS v6 uses annual and seasonal ETS (2019) data as
one of the primary data inputs for NOx rate development and environmental equipment
assignment.
Utility and Regional
EPA Office
Comments
Comments from utilities and regional EPA offices, and EPA research regarding the
population in NEEDS as of Spring 2022 (e.g., retirements and new units) as well as unit
characteristics were incorporated in NEEDS v6.
Note:
1 Shown in Table 4-1 are the primary issue dates of the indicated data sources used. Other vintages of these data
sources were also used in instances where data were not available for the indicated issued date, or where there were
methodological reasons for using other vintages of the data.
Table 4-2 Rules Used in Populating NEEDS v6
Scope
Rule
Capacity
Excluded units that had reported summer capacity, winter capacity, and nameplate capacity of
zero or blank.
Status
Excluded units that were out of service for three consecutive years (i.e., generators or boilers
with status codes "OS41" or "OA42" in the latest three reporting years) and units that were no
longer in service and not expected to be returned to service (i.e., generators or boilers with
status codes of "RE43"). Status of boiler(s) and associated generator(s) were considered for
determining operation status.
Planned or
Committed
Units
For plant types other than wind, solar and energy storage, included planned units that had
broken ground and were expected to be online by June 30, 2028.
For wind, solar and energy storage units, included planned units that had broken ground, had
received, had pending regulatory approvals or had planned for installation and were expected to
41 OS - Out of service and was not used for some or all of the reporting period and is NOT expected to be returned to
service in the next calendar year.
42 OA - Out of service and was not used for some or all of the reporting period but is expected to be returned to
service in the next calendar year.
43 RE - Retired and no longer in service and not expected to be returned to service.
4-2
-------
Scope
Rule
be online by June 30, 2028. Also included one solar PV unit at Alira plant with a capacity of
222.8 megawatt that has pending regulatory approval and is scheduled to come online in 2030.
Firm/Non-firm
Electric Sales
Excluded non-utility onsite generators that did not produce electricity for sale to the grid on a net
basis.
The NEEDS v6 includes steam units at the boiler level and non-steam units at the generator level
(nuclear units are also at the generator level). A unit in NEEDS v6, therefore, refers to a boiler in the
case of a steam unit and a generator in the case of a non-steam unit.
Table 4-3 provides a summary of the population and capacity of the existing units included in NEEDS v6
through 2021. The final population of existing units is supplemented based on information from other
sources. These include comments from utilities, submissions to EPA's Emission Tracking System, Annual
Energy Outlook, and other research.
EPA Platform v6 removes units from the NEEDS inventory based on public announcements of future
closures. The removal of such units pre-empts IPM from making any further decisions regarding the
operational status or configuration of the units. These units are removed from the NEEDS inventory only if
a high degree of certainty could be assigned to future implementation of the announced action and are
identified from reviewing several data sources, including:
i) Reviewing unit retirement list from EIA Electric Generator Capacity data (EIA Form 860M),
December 2021
ii) PJM Future Deactivation Requests and PJM Generator Deactivations, March 2022 (updated
frequently)
iii) ERCOT Generator Interconnection Status Report, March 2022 (updated frequently)
iv) MISO Generation Interconnection Queue, March 2022 (updated frequently). Units that have been
cleared by a regional transmission operator (RTO) or independent system operator (ISO) to retire
before June 30, 2028, or whose RTO/ISO clearance to retire is contingent on actions that can be
completed before June 30, 2028
v) Units that have committed specifically to retire before June 30, 2028, under federal or state
enforcement actions or regulatory requirements
vi) Research by EPA and ICF staff as of Spring 2022
Research includes:
• Reviewing utility company Integrated Resource Plan (IRP), Sustainably, Climate and ESG
Reports, along with company news releases, to capture retirement or repowering data on
owned fleet.
• Reviewing investor news released by company that outlines closure or repowering of owned
fleet
• Referencing EIA Electric Power Monthly Report Table 6.6 Planned U.S. Electric Generation
Unit Retirements.
• Reviewing outside news articles that capture closure or repowering of individual Electricity
Generating Units (EGU), or reports released from utility companies.
Units required to retire pursuant to enforcement actions or state rules on July 1, 2028, or later are
retained in NEEDS v6. Such July 1, 2028-or-later retirements are captured as constraints on those units
in IPM modeling, and the units are retired in future year projections per the terms of the related
requirements.
4-3
-------
The "Capacity Dropped" and the "Retired Through 2028" worksheets in NEEDS list all units that are
removed from the NEEDS v6 inventory.
Table 4-3 Summary Population (through 2021) of Existing Units in NEEDS v6
Plant Type
Number of Units
Capacity (MW)
Biomass
167
3,436
Coal Steam
366
144,889
Combined Cycle
1,880
274,569
Combustion Turbine
5,783
145,443
Energy Storage
376
6,148
Fossil Waste
62
1,382
Fuel Cell
162
268
Geothermal
158
2,472
Hydro
3,817
79,307
IGCC
5
815
Landfill Gas
1,484
1,754
Municipal Solid Waste
150
1,935
Non-Fossil Waste
223
2,299
Nuclear
91
93,485
O/G Steam
387
55,799
Offshore Wind
2
41
Onshore Wind
1,503
137,129
Pumped Storage
152
22,820
Solar PV
5,124
62,459
Solar Thermal
11
1,486
Tires
2
52
US Total
21,905
1,037,987
4.2.2 Capacity
The unit capacity data implemented in NEEDS v6 reflects net summer dependable capacity.44 Table 4-4
summarizes the hierarchy of data sources used in compiling capacity data. In other words, capacity
values are taken from a particular source only if the sources listed above it do not provide adequate data
for the unit in question.
Table 4-4 Hierarchy of Data Sources for Capacity in NEEDS v6
Sources Presented in Hierarchy
Net Summer Capacity from Comments / ICF Research
AEO 2020 Nuclear Capacity in 2023
September 2019 EIA Form 860 monthly Net Summer Capacity
2018 EIA Form 860 Net Summer Capacity
Notes:
Presented in hierarchical order that applies.
If the capacity of a unit is zero MW, the unit is excluded from NEEDS population.
As noted earlier, NEEDS v6 includes boiler-level data for steam units and generator-level data for non-
steam units. Capacity data in EIA Form 860 are generator-specific, not boiler-specific. Therefore, it was
necessary to develop an algorithm for parsing generator-level capacity to the boiler level for steam
producing units.
44 As used here, net summer dependable capacity is the net capability of a generating unit in megawatts (MW) for
daily planning and operation purposes during the summer peak season, after accounting for station or auxiliary
services.
4-4
-------
The capacity-parsing algorithm used for steam units in NEEDS v6 considered boiler-generator mapping.
Fossil steam electric units have boilers attached to generators that produce electricity. There are
generally four types of links between boilers and generators: one boiler to one generator, one boiler to
many generators, many boilers to one generator, and many boilers to many generators.
The capacity-parsing algorithm used for steam units in NEEDS v6 utilizes steam flow data with the boiler-
generator mapping. Under EIA Form 860, steam units report the maximum steam flow from the boiler to
the generator. There is, however, no further data on the steam flow of each boiler-generator link. Instead,
EIA Form 860 contains only the maximum steam flow for each boiler. Table 4-5 summarizes the algorithm
used for parsing capacity with data on maximum steam flow and boiler-generator mapping. In Table 4-5,
MFb, refers to the maximum steam flow of boiler /' and MWg; refers to the capacity of generator j. The
algorithm uses the available data to derive the capacity of a boiler, referred to as MWb; in Table 4-5.
Table 4-5 Capacity-Parsing Algorithm for Steam Units in NEEDS v6
Type of Boiler-Generator Links
For Boiler B1 to BN linked
to Generators G1 to GN
One-to-One
One-to-Many
Many-to-One
Many-to-Many
MWBi =
MWgj
MWBi =
ZjMWcj
MWBi =
(MFei / ZiMFei) * MWgj
MWBi =
(MFBi / ZiMFei) * ZjMWcj
Notes:
MFb; = maximum steam flow of boiler i
MWgj = electric generation capacity of generator j
Since EPA Platform v6 uses net energy for load as demand, NEEDS includes only generators that sell
most of their power to the electric grid. The approach is intended to be broadly consistent with the
generating capacity used in the AEO projections where demand is net energy for load. The generators
that should be in NEEDS v6 by this qualification are determined from the 2018 EIA Form 923 non-utility
source and disposition data set.
4.2.3 Plant Location
The physical location of each unit in NEEDS is represented by the unit's model region, state, and county
data.
State and County
NEEDS v6 uses the state and county data from the September 2019 EIA Form 860M.
Model Region
For each unit, the associated model region was derived based on NERC assessment regions reported in
EIA Form 860 and ISO/RTO reports. For units with no NERC assessment region data, state and county
data were used to derive associated model regions. Table 3-1 in Chapter 3 provides a summary of the
mapping between NERC assessment regions and EPA Platform v6 model regions.
4.2.4 Online Year
EPA Platform v6 uses online year to capture when a unit entered service. NEEDS includes online years
for all units in the population. Online years for boilers were from the 2018 EIA Form 860, and online
years for generators were derived primarily from reported in-service dates in the September 2019 EIA
Form 860M.
EPA Platform v6 includes constraints to set the retirement year for generating units that are firmly
committed to retiring after June 30, 2028, based on state or federal regulations, enforcement actions, and
announcements.
4-5
-------
Economic retirement options are also provided to coal, oil and gas steam, combined cycle, combustion
turbines, and biomass units to allow the model the option to retire a unit if it finds economical to do so. In
IPM, a retired unit ceases to incur fixed O&M and variable O&M costs. The unit, however, continues to
make annualized capital cost payment on any previously incurred capital cost for model-installed retrofits
projected prior to retirement.
4.2.5 Unit Configuration
Unit configuration refers to the physical specification of a unit's design. Unit configuration in EPA
Platform v6 drives model plant aggregation and modeling of pollution control options and mercury
emission modification factors. NEEDS v6 contains for each unit, data on the firing and bottom type, as
well as existing and committed emission controls the unit has. Table 4-6 shows the hierarchy of data
sources used in determining a unit configuration. The sources listed below are also supplemented by
recent ICF and EPA research to ensure the unit configuration data in NEEDS is the most comprehensive
and up-to-date possible.
Table 4-6 Data Sources for Unit Configuration in NEEDS v6
Unit
Component
Primary Data
Source
Secondary Data Source
Tertiary Data
Source
Other
Sources
Default
Firing Type
2018 EIA 860
EPA's Emission Tracking
System (ETS)-2019
-
-
-
Bottom Type
2018 EIA 860
EPA's Emission Tracking
System (ETS)-2019
-
-
Dry
SO2 Pollution
Control
2018 EIA 860
EPA's Emission Tracking
System (ETS)-2019
NSR Settlement
or Comments
-
No
Control
NOx Pollution
Control
2018 EIA 860
EPA's Emission Tracking
System (ETS)-2019
NSR Settlement
or Comments
-
No
Control
Particulate
Matter Control
2018 EIA 860
EPA's Emission Tracking
System (ETS)-2019
NSR Settlement
or Comments
-
-
Mercury Control
2018 EIA 860
EPA's Emission Tracking
System (ETS)-2019
NSR Settlement
or Comments
-
-
HCL Control
2018 EIA 860
EPA's Emission Tracking
System (ETS)-2019
NSR Settlement
or Comments
-
-
4.2.6 Model Plant Aggregation
While EPA Platform v6 using IPM is comprehensive in representing all the units contained in NEEDS v6,
an aggregation scheme is used to combine existing units with similar characteristics into model plants.
The aggregation scheme serves to reduce the size of the model, making the model manageable while
capturing the essential characteristics of the generating units. The aggregation scheme is designed so
that each model plant represents only generating units from a single model region and state. The design
makes it possible to obtain state-level results directly from IPM outputs. In addition, the aggregation
scheme supports the modeling of plant-level emission limits on fossil generation.
The aggregation scheme encompasses different categories including location, size, technology, heat rate,
fuel choices, unit configuration, SO2 emission rates, and environmental regulations among others. Units
are aggregated together only if they match on all the different categories specified for the aggregation.
The 11 major categories used for the aggregation scheme in EPA Platform v6 are the following.
i) Facility (ORIS) for all fossil units except combustion turbine units smaller than or equal to 25 MW
ii) Model Region
iii) State
iv) Unit Technology Type
v) Unit Configuration
4-6
-------
vi) Cogen
vii) Fuel Category
viii) Fuel Demand Region
ix) Applicable Environmental Regulations
x) Heat Rates
xi) Size
Table 4-7 shows the number of actual units by generation technology type and the related number of
aggregated model plants in the EPA Platform v6. For each plant type, the table shows the number of
generating units and the number of model plants representing the generating units.45
Table 4-7 Aggregation Profile of Model Plants as Provided at Set up of v6
Existing and Planned/Committed Units
Plant Type
Number of Units
Number of IPM Model Plants
Biomass
314
116
Coal Steam
387
293
Combined Cycle
1,991
720
Combustion Turbine
5,816
1,237
Distributed Solar PV
130
130
Energy Storage
162
66
Fossil Waste
68
31
Fuel Cell
99
17
Geothermal
153
10
Hydro
5,502
200
IGCC
5
2
IMPORT
1
1
Landfill Gas
1,482
94
Municipal Solid Waste
151
53
Non-Fossil Waste
250
89
Nuclear
106
106
O/G Steam
469
291
Offshore Wind
1
1
Onshore Wind
1,731
89
Pumped Storage
159
27
Solar PV
4,290
97
Solar Thermal
12
5
Tires
2
1
Total
23,281
3,676
New Units
Plant Type
Number of Units
Number of IPM Model Plants
New Battery Storage
-
1008
New Biomass
—
134
New Combined Cycle
—
86
New Combined Cycle with CCS
-
192
45 (1) The "Number of IPM Model Plants" shown for many of the "Plant Types" in the "Retrofits" block in Table 4-7
exceeds the "Number of IPM Model Plants" shown for "Plant Type" "Coal Steam" in the block labeled "Existing and
Planned - Committed Units", because a particular retrofit "Plant Type" can include multiple technology options and
multiple timing options (e.g., Technology A in Stage 1 + Technology B in Stage 2 + Technology C in Stage 3, the
reverse timing, or multiple technologies simultaneously in Stage 1).
(2) Since only a subset of coal plants is eligible for certain retrofits, many of the "Plant Types" in the "Retrofits" block
that represent only a single retrofit technology (e.g., "Retrofit Coal with SNCR") have a "Number of IPM Model Plants"
that is a smaller than the "Number of IPM Model Plants" shown for "Plant Type" "Coal Steam".
(3) The total number of model plants representing different types of new units often exceeds the 67 U.S. model
regions and varies from technology to technology for several reasons. First, some technologies have multiple
vintages (i.e., different cost and/or performance parameters depending on which run year in which the unit is
created), which must be represented by separate model plants in each IPM region. Second, some technologies are
not available in particular regions (e.g., geothermal is geographically restricted to certain regions).
4-7
-------
New Combustion Turbine
—
115
New Fuel Cell
—
75
New Geothermal
—
61
New Hydro
—
153
New Landfill Gas
—
379
New Nuclear
—
68
New Offshore Wind
—
582
New Onshore Wind
—
3,087
New Small Modular Reactor
—
64
New Solar PV
—
3,165
New Solar Thermal
—
248
New Ultrasupercritical Coal with 30% CCS
—
192
New Ultrasupercritical Coal with 90% CCS
-
192
New Ultrasupercritical Coal without CCS
—
5
Total
—
9,806
Retrofits
Plant Type
Number of Units
Number of IPM Model Plants
Retro
t Coal w
th ACI
—
2
Retro
t Coal w
th ACI + DSI
—
4
Retro
t Coal w
th ACI + DSI + HRI
—
4
Retro
t Coal w
th ACI + DSI + HRI + SCR
—
3
Retro
t Coal w
th ACI + DSI + HRI + SCR + Scrubber
—
2
Retro
t Coal w
th ACI + DSI + HRI + Scrubber
—
4
Retro
t Coal w
th ACI + DSI + HRI + Scrubber + SNCR
—
2
Retro
t Coal w
th ACI + DSI + HRI + SNCR
—
3
Retro
t Coal w
th ACI + DSI + SCR
—
3
Retro
t Coal w
th ACI + DSI + SCR + Scrubber
—
2
Retro
t Coal w
th ACI + DSI + Scrubber
—
4
Retro
t Coal w
th ACI + DSI + Scrubber + SNCR
—
2
Retro
t Coal w
th ACI + DSI + SNCR
—
3
Retro
t Coal w
th ACI + HRI
—
2
Retro
t Coal w
th ACI + HRI + SCR
—
2
Retro
t Coal w
th ACI + HRI + SCR + Scrubber
—
2
Retro
t Coal w
th ACI + HRI + Scrubber
—
2
Retro
t Coal w
th ACI + HRI + Scrubber + SNCR
—
2
Retro
t Coal w
th ACI + HRI + SNCR
—
2
Retro
t Coal w
th ACI + SCR
—
2
Retro
t Coal w
th ACI + SCR + Scrubber
—
2
Retro
t Coal w
th ACI + Scrubber
—
2
Retro
t Coal w
th ACI + Scrubber + SNCR
—
2
Retro
t Coal w
th ACI + SNCR
—
2
Retro
t Coal w
th C2G
—
233
Retro
t Coal w
th C2G + SCR
—
233
Retro
t Coal w
th CCS
—
784
Retro
t Coal w
th CCS + HRI
—
624
Retro
t Coal w
th CCS + HRI + SCR
—
184
Retro
t Coal w
th CCS + HRI + SCR + Scrubber
—
128
Retro
t Coal w
th CCS + HRI + Scrubber
—
176
Retro
t Coal w
th CCS + HRI + Scrubber + SNCR
—
120
Retro
t Coal w
th CCS + HRI + SNCR
—
116
Retro
t Coal w
th CCS + SCR
—
196
Retro
t Coal w
th CCS + SCR + Scrubber
—
128
Retro
t Coal w
th CCS + Scrubber
—
176
Retro
t Coal w
th CCS + Scrubber + SNCR
—
120
Retro
t Coal w
th CCS + SNCR
—
128
Retro
t Coal w
th DSI
—
6
Retro
t Coal w
th DSI + HRI
—
31
Retro
t Coal w
th DSI + HRI + SCR
—
25
Retro
t Coal w
th DSI + HRI + SCR + Scrubber
—
3
Retro
t Coal w
th DSI + HRI + Scrubber
—
3
Retro
t Coal w
th DSI + HRI + SNCR
—
25
Retro
t Coal w
th DSI + SCR
-
40
4-8
-------
Retrofit Coal
w
th DSI + SCR + Scrubber
—
9
Retrofit Coal
w
th DSI + Scrubber
—
6
Retrofit Coal
w
th DSI + SNCR
—
40
Retrofit Coal
w
th HRI
—
366
Retrofit Coal
w
th HRI + SCR
—
210
Retrofit Coal
w
th HRI + SCR + Scrubber
—
204
Retrofit Coal
w
th HRI + Scrubber
—
233
Retrofit Coal
w
th HRI + Scrubber + SNCR
—
188
Retrofit Coal
w
th HRI + SNCR
—
168
Retrofit Coal
w
th SCR
—
133
Retrofit Coal
w
th SCR + Scrubber
—
258
Retrofit Coal
w
th Scrubber
—
114
Retrofit Coal
w
th Scrubber + SNCR
—
245
Retrofit Coal
w
th SNCR
—
102
Retrofit Combined Cycle with CCS
—
2388
Retrofit Oil/Gas steam with SCR
—
147
Total
—
8,350
Retirements
Plant Type
Number of Units
Number of IPM Model Plants
Biomass Retirement
—
116
CC Retirement
—
3,108
Coal Retirement
—
4,549
CT Retirement
—
1,237
Geothermal Retirement
—
10
Hydro Retirement
—
107
IGCC Retirement
—
2
Landfill Gas Retirement
—
94
Nuke Retirement
—
106
Oil/Gas steam Retirement
—
671
Total
—
10,000
Grand Total (Existing and Planned/Committed + New + Retrofits + Early Retirements): 31,832
4.2.7 Cost and Performance Characteristics of Existing Units46
In EPA Platform v6, the cost and performance characteristics of an existing unit are determined by the
unit's heat rates, emission rates, variable operation, and maintenance cost (VOM), and fixed operation
and maintenance costs (FOM). For existing units, only the cost of maintaining (FOM) and running (VOM)
the unit are modeled because capital costs and all related carrying capital charges are sunk, and hence,
economically irrelevant for projecting least-cost investment and operational decisions going forward. The
section below discusses the cost and performance assumptions for existing units used in the EPA
Platform v6.
Variable Operating and Maintenance Cost (VOM)
VOM represents the non-fuel variable cost associated with producing electricity. If the generating unit
contains pollution control equipment, VOM includes the cost of operating the control equipment. Table
4-8 below summarizes VOM assumptions used in EPA Platform v6. The following further discusses the
components of VOM costs and the VOM modeling methodology.
Variable O&M Approach: EPA Platform v6 uses a modeling construct termed as Segmental VOM for
combined cycle units to capture the variability in operation and maintenance costs that are treated as a
function of the unit's dispatch pattern. All other technologies are assigned static VOM assumptions.
The VOM for combustion turbines are differentiated by the turbine technology. The VOM for combined
cycles and combustion turbine units includes the costs of both major maintenance and consumables
46 All units excluding nuclear units.
4-9
-------
while for coal steam and oil/gas steam units includes only the cost of consumables. The VOM cost of
various emission control technologies is also incorporated.
Major maintenance: Major maintenance costs are those required to maintain a unit at its delivered
performance specifications and whose terms are usually dictated through its long-term service agreement
(LTSA). The three main areas of maintenance for gas turbines include combustion inspection, hot gas
path inspection, and major inspections. All these costs are driven by the hours of operation and the
number of starts that are incurred within that time period of operation. In a cycling or mid-merit type mode
of operation, there are many starts, accelerating the approach of an inspection. As more starts are
incurred compared to the generation produced, cost per generation increases. For base load operation
there are fewer starts spread over more generation, lowering the cost per generation. While this
nomenclature is for gas-turbine based systems, steam turbine-based systems have a parallel construct.
Consumables: The model captures consumable costs, as purely a function of output and does not vary
across the segmented time-period. In other words, the consumables cost component is held constant
over both peak and off-peak segments. Consumables include chemicals, lube oils, make-up water,
wastewater disposal, reagents, and purchased electricity.
Data Sources for Gas-Turbine Based Prime Movers:
ICF has engaged its deep expertise in operation & maintenance costs for these types of prime movers to
develop generic variable O&M costs as a function of technology. As mentioned above the variable O&M
for gas-turbine based systems tracks LTSA costs, start-up, and consumables.
Data Sources for Stand-Alone Steam Turbine Based Prime Movers:
The value levels of non-fuel variable O&M data for stand-alone steam turbine plants are based on ICF
expertise. The VOM cost adders of various emission control technologies are based on cost functions
described in Chapter 5.
Table 4-8 VOM Assumptions in v6
Capacity Type
SO2 Control
NOx Control
Hg Control
Variable O&M
(2019$/mills/kWh)
Biomass
-
-
-
7.56
No NOx Control
No Hg Control
1.52
AC I
3.08
No SO2 Control
SCR
No Hg Control
2.4
AC I
3.96
SNCR
No Hg Control
2.3
AC I
3.86
No NOx Control
No Hg Control
3.55
AC I
5.11
Coal Steam
Dry FGD
SCR
No Hg Control
4.43
AC I
5.99
SNCR
No Hg Control
4.33
AC I
5.89
No NOx Control
No Hg Control
4.18
AC I
5.73
Wet FGD
SCR
No Hg Control
5.06
AC I
6.62
SNCR
No Hg Control
4.96
AC I
6.52
4-10
-------
Capacity Type
SO2 Control
NOx Control
Hg Control
Variable O&M
(2019$/mills/kWh)
DSI
No NOx Control
No Hg Control
7.75
AC I
9.31
SCR
No Hg Control
8.63
AC I
10.19
SNCR
No Hg Control
8.53
AC I
10.09
Combined Cycle
No SO2 Control
No NOx Control
No Hg Control
2.14-4.02
SCR
2.28-4.16
SNCR
2.81 -4.69
Combustion Turbine
No SO2 Control
No NOx Control
No Hg Control
4.61 -6.52
SCR
4.72-6.63
SNCR
4.72-6.63
Fuel Cell
-
-
-
45.07
Geothermal
-
-
-
1.16
Hydro
-
-
-
1.39
IGCC
-
-
-
2.42-4.29
Landfill Gas / Municipal
Solid Waste
-
-
-
6.94
Oil/gas Steam
No SO2 Control
No NOx Control
No Hg Control
0.88
SCR
1.03
SNCR
1.55
Pumped Storage
-
-
-
0.02
Solar
-
-
-
0
Wind
-
-
-
0
Fixed Operation and Maintenance Cost (FOM)
FOM represents the annual fixed cost of maintaining a unit. FOM costs are incurred independent of
generation levels and signify the fixed cost of operating and maintaining the unit's availability to provide
generation. Table 4-9 summarizes the FOM assumptions.47 Note that FOM varies by the age of the unit,
and the total FOM cost incurred by a unit depends on its capacity size. The values appearing in the table
include the cost of maintaining any associated pollution control equipment. The values in Table 4-9 are
based on FERC (Federal Energy Regulatory Commission) Form 1 data maintained by SNL and ICF
research. The following further discusses the procedure for developing the FOM costs.
Stand Alone - Steam Turbines Based Prime Movers
O&M cost data for existing coal and oil/gas steam units were developed starting with FERC Form 1 data
sets from the years 2011 to 2016. The FERC Form-1 database does not explicitly report separate fixed
and variable O&M expenses. In deriving Fixed O&M costs, generic variable O&M costs are assigned to
each individual power plant. Next, the assumed variable O&M cost is subtracted from the total O&M
reported by FERC Form-1 to calculate a starting point for fixed O&M. Thereafter, other cost items which
are not reported by FERC Form-1 are added to the raw FOM starting point. These unreported cost items
are selling, general, and administrative expenses (SG&A), property taxes, insurance, and routine capital
expenditures. A detailed description of the fixed O&M derivation methodology is provided below.
47 Cogen units whose primary purpose is to provide process heat are called as bottoming cycle units and are
identified based on Form EIA 860. Such units are provided a FOM of zero in EPA Platform v6. This is to
acknowledge the fact that the economics of such a unit cannot be comprehensively modeled in a power sector
focused model.
4-11
-------
Figure 4-1 Derivation of Plant Fixed O&M Data
i) Assign generic VOM cost to each unit in FERC Form 1 based on the control configuration.
Subtract this VOM from the total O&M cost from FERC Form 1 to calculate raw FOM cost.
The FOM cost of operating the existing controls is estimated based on cost functions in
Chapter 5. and deducted from the raw FOM cost. Aggregate this unit level raw FOM cost
data into age-based categories. The weighted average raw FOM costs for uncontrolled units
by age group is the output of this step and is used as the starting point for subsequent steps.
ii) An owner/operator fee for SG&A services in the range of 20-30% is added to raw fixed O&M
figures in step 1.
iii) Property tax and insurance cost estimates in $/kW-year are also added. These figures vary
by plant type.
iv) A generic percentage value to cover routine capex is added to raw fixed O&M figures in step
1. The percentage varies by prime mover and is based on a review of FERC Form 1 data
v) Finally, generic FOM cost adders for various emission control technologies are estimated
using cost functions described in Chapter 5. Based on the emission control configuration of
each unit in NEEDS, the appropriate emission control cost adder is added to the FOM cost of
an uncontrolled unit from step iv.
The fixed O&M derivation approach relies on top-down calculation of fixed costs based on FERC Form-1
data and ICF's own non-fuel variable O&M, SG&A, routine capital expenditures, property tax, and
insurance.
Gas-Turbine Based Prime Movers
Similar to the stand-alone steam turbine based prime movers, the fixed O&M for gas-turbine based
systems tracks: labor, routine maintenance, property taxes, insurance, owner/operator SG&A, and routine
capital expenditures. These generic fixed O&M costs as a function of technology are based on ICF's
expertise in fixed O&M costs for these types of prime movers.
Table 4-9 FOM Assumptions in v6
Plant Type
SO2 Control
NOx Control
Hg Control
Age of Unit
FOM (2019$ /kW-Yr)
Biomass
-
-
-
All Years
149.3
0 to 40 Years
30.1
No Hg Control
40 to 50 Years
34.42
No NOx Control
Greater than 50 Years
44.22
0 to 40 Years
30.19
Coal Steam
No SO2 Control
AC I
40 to 50 Years
34.51
Greater than 50 Years
44.31
0 to 40 Years
30.93
SCR
No Hg Control
40 to 50 Years
35.25
Greater than 50 Years
45.05
AC I
0 to 40 Years
31.01
4-12
-------
Plant Type
SO2 Control
NOx Control
Hg Control
Age of Unit
FOM (2019$ /kW-Yr)
40 to 50 Years
35.33
Greater than 50 Years
45.14
0 to 40 Years
30.39
No Hg Control
40 to 50 Years
34.71
SNCR
Greater than 50 Years
44.52
0 to 40 Years
30.48
AC I
40 to 50 Years
34.8
Greater than 50 Years
44.6
0 to 40 Years
39.18
No Hg Control
40 to 50 Years
43.5
No NOx Control
Greater than 50 Years
53.3
0 to 40 Years
39.26
AC I
40 to 50 Years
43.58
Greater than 50 Years
53.39
0 to 40 Years
40
No Hg Control
40 to 50 Years
44.32
Dry FGD
SCR
Greater than 50 Years
54.13
0 to 40 Years
40.09
AC I
40 to 50 Years
44.41
Greater than 50 Years
54.21
0 to 40 Years
39.47
No Hg Control
40 to 50 Years
43.79
SNCR
Greater than 50 Years
53.59
0 to 40 Years
39.55
AC I
40 to 50 Years
43.87
Greater than 50 Years
53.68
0 to 40 Years
40.95
No Hg Control
40 to 50 Years
45.28
No NOx Control
Greater than 50 Years
55.08
0 to 40 Years
41.04
AC I
40 to 50 Years
45.36
Greater than 50 Years
55.16
0 to 40 Years
41.78
No Hg Control
40 to 50 Years
46.1
Wet FGD
SCR
Greater than 50 Years
55.9
0 to 40 Years
41.87
AC I
40 to 50 Years
46.19
Greater than 50 Years
55.99
0 to 40 Years
41.25
No Hg Control
40 to 50 Years
45.57
SNCR
Greater than 50 Years
55.37
0 to 40 Years
41.33
AC I
40 to 50 Years
45.65
Greater than 50 Years
55.46
0 to 40 Years
31.44
No Hg Control
40 to 50 Years
35.76
No NOx Control
Greater than 50 Years
45.57
DSI
0 to 40 Years
31.53
AC I
40 to 50 Years
35.85
Greater than 50 Years
45.65
SCR
No Hg Control
0 to 40 Years
32.27
40 to 50 Years
36.59
4-13
-------
Plant Type
SO2 Control
NOx Control
Hg Control
Age of Unit
FOM (2019$ /kW-Yr)
Greater than 50 Years
46.39
AC I
0 to 40 Years
32.36
40 to 50 Years
36.68
Greater than 50 Years
46.48
SNCR
No Hg Control
0 to 40 Years
31.73
40 to 50 Years
36.05
Greater than 50 Years
45.86
AC I
0 to 40 Years
31.82
40 to 50 Years
36.14
Greater than 50 Years
45.95
Combined Cycle
No SO2 Control
No NOx Control
No Hg Control
-
30.18
SCR
No Hg Control
-
31.59
SNCR
No Hg Control
-
30.92
Combustion Turbine
No SO2 Control
No NOx Control
No Hg Control
-
19.73
SCR
No Hg Control
-
21.84
SNCR
No Hg Control
-
20.15
Fuel Cell
-
-
-
All Years
0
Geothermal
-
-
-
All Years
100.74
Hydro
-
-
-
All Years
15.81
Integrated
Gasification
Combined Cycle
No SO2 Control
No NOx Control
-
All Years
108.71
Landfill Gas /
Municipal Solid
Waste
-
-
-
All Years
259.23
Oil/gas Steam
No SO2 Control
No NOx Control
No Hg Control
0 to 40 Years
17.99
40 to 50 Years
27.32
Greater than 50 Years
35.6
SCR
No Hg Control
0 to 40 Years
19.34
40 to 50 Years
28.67
Greater than 50 Years
36.94
SNCR
No Hg Control
0 to 40 Years
18.22
40 to 50 Years
27.55
Greater than 50 Years
35.83
Pumped Storage
-
-
-
All Years
18.29
Solar Photovoltaics
-
-
-
All Years
31.6
Solar Thermal
-
-
-
All Years
82.65
Wind
-
-
-
All Years
35.26
Heat Rates
Heat Rates describe the efficiency of the unit expressed as BTUs per kWh. The treatment of heat rates is
discussed in Section 3.9.
Lifetimes
Unit lifetime assumptions are detailed in Sections 3.8 and 4.2.8.
SO? Rates
Section 3.10.1 contains a detailed discussion of SO2 rates for existing units.
4-14
-------
NOy Rates
Section 3.10.3 contains a detailed discussion of NOx rates for existing units.
Mercury Emission Modification Factors (EMF)
Mercury EMF refers to the ratio of mercury emissions (mercury outlet) to the mercury content of the fuel
(mercury inlet). Section 5.7.2 contains a detailed discussion of the EMF assumptions in EPA Platform v6.
Coqeneration Units
For cogeneration units, the dispatch decisions in IPM are only based on the benefits obtained from the
electric portion of a cogeneration unit. In IPM, a cogeneration unit uses a net heat rate, which is
calculated by dividing heat content of fuel consumed for power generation by electricity generated from
this fuel. To capture the total emissions from the cogeneration unit, a multiplier is applied to the power
only emissions. The multiplier is calculated as a ratio between the total heat rate and the net heat rate,
where the total heat rate is calculated by dividing the heat content of fuel consumed for power and steam
generation by electricity generated from this fuel.
Coal Switching
Recognizing that boiler modifications and fuel handling enhancements may be required for unrestricted
switching from bituminous to subbituminous coal, and vice versa, the following procedure applies in EPA
Platform v6 to coal units that have the option to burn both bituminous and subbituminous coals.
(i) An examination of the EIA Form 923 coal delivery data for the period 2010-2019 is conducted for each
unit to determine the unit's historical maximum share of bituminous coal and that of subbituminous coal.
For example, if in at least one year during the period 2010-2019 a unit burned 90% or less subbituminous
coal, its historical maximum share of subbituminous coal is set at 90%.
(ii) The following rules then apply.
Blending Subbituminous Coal:
If a unit's historical maximum share of subbituminous coal is greater than 90%, the unit incurs no fuel
switching cost adder to increase its subbituminous coal burn. The unit is assumed to have already made
the fuel handling and boiler investments needed to burn up to 100% subbituminous coal. It would
therefore face no additional cost. In addition, the unit's heat rate is assumed to reflect the impact of
burning the corresponding proportion of subbituminous coal.
If a unit's historical maximum share of subbituminous coal is less than 90%, the unit incurs a heat rate
penalty of 5% and a fuel switching cost adder. The heat rate penalty reflects the impact of the higher
moisture content subbituminous coal on the unit's heat rate. And the cost adder is designed to cover
boiler modifications, or alternative power purchases in lieu of capacity deratings that would otherwise be
associated with burning subbituminous coal with its lower heating value relative to bituminous coal. The
cost adder is determined as follows:
• If the unit's historical maximum share of subbituminous coal is less than 20%, the unit can burn
up to 20% subbituminous coal at no cost adder. Burning beyond 20% subbituminous coal, the
unit incurs a cost adder of 286 (2019$ per kW).
• If the unit's historical maximum share of subbituminous coal is greater than 20% but less than
90%, the unit can burn up to its historical maximum share of subbituminous coal at no cost adder.
4-15
-------
Burning beyond its historical maximum share of subbituminous coal, the unit incurs a cost adder
calculated by the following equation:
Fuel Switching Cost Adder (2019$ per kW) =
f (100 — Historical Maximum Share of Subbituminous))
286 x\- ; -\
[ (100 -20) J
Blending Bituminous Coal:
If a unit's historical maximum share of bituminous coal is greater than 90%, the unit incurs no fuel
switching cost adder.
If a unit's historical maximum share of bituminous coal is less than 90%, the unit incurs a fuel switching
cost adder determined as follows:
• If the unit's historical maximum share of bituminous coal is less than 20%, the unit can burn up to
20% bituminous coal at no cost adder. Burning beyond 20% bituminous coal, the unit incurs a
cost adder of 57 (2019$ per kW).
• If the unit's historical maximum share of bituminous coal is greater than 20% but less than 90%,
the unit can burn up to its historical maximum share of bituminous coal at no cost adder. Burning
beyond its historical maximum share of bituminous coal, the unit incurs a cost adder calculated by
the following equation:
Fuel Switching Cost Adder (2019$ per kW) =
((100 — Historical Maximum Share of Bituminous))
57 X( (100 -20) J
4.2.8 Life Extension Costs for Existing Units
The modeling time horizon in EPA Platform v6 extends to 2059 and covers a period of almost 30 years.
This time horizon requires consideration of investments, beyond routine maintenance, necessary to
extend the life of existing units. The life extension costs for different unit types are summarized in Table
4-10 below. Each unit has the option to retire or incorporate the life extension costs. These costs were
based on a review of 2007-2016 FERC Form 1 data maintained by SNL regarding reported annual capital
expenditures made by older units. The life extension costs were added once the unit reaches its
assumed lifespan. Life extension costs for nuclear units are discussed in Section 4.6.1.
Table 4-10 Life Extension Cost Assumptions Used in v6
Lifespan without Life
Life Extension
Capital Cost of
Life Extension Cost
Plant Type
Extension
Cost
New Unit
as Proportion of New
Expenditures
(2019$/kW)
(2019$/kW)
Unit Capital Cost (%)
Biomass
40
253
3,853
6.6
Coal Steam
40
203
3,481
5.84
Combined Cycle
30
82
901
9.06
Combustion Turbine
30
242
667
36.3
IC Engine
30
226
1,713
13.2
Oil/Gas Steam
40
174
3,169
5.5
IGCC
40
258
3,481
7.4
Landfill Gas
20
135
1,480
9.1
Notes:
Life extension expenditures double the lifespan of the unit.
4-16
-------
4.3 Planned-Committed Units
EPA Platform v6 includes all planned-committed units that are likely to come online because ground has
been broken, financing obtained, or other demonstrable factors indicate a high probability that the unit will
be built before June 30, 2028.
In addition, wind, solar, and energy storage units that had received, had pending regulatory approvals, or
were flagged as planned for installation per the December 2021 version of EIA Form 860 monthly and
were expected to be online by June 30, 2028, were also included.
4.3.1 Population and Model Plant Aggregation
Table 4-11 summarizes the extent of the inventory of planned-committed units represented by unit types
and generating capacity. Table 4-34 gives a breakdown of planned-committed units by IPM region, plant
type, and capacity.
Table 4-11 Summary of Planned-Committed Units in NEEDS v6
Type
Capacity (MW)
Year Range Described
Renewables/Non-conventional
Energy Storage
11,339
2022 - 2025
Fuel Cell
16
2022 - 2022
Geothermal
17
2022 - 2022
Hydro
4
2022 - 2022
Landfill Gas
3
2022 - 2022
Offshore Wind
3,285
2024 - 2027
Onshore Wind
16,604
2022 - 2026
Solar PV
47,265
2022 - 2030
Subtotal
78,533
Fossil/Conventional
Combined Cycle
12,312
2022 - 2024
Combustion Turbine
1,126
2022 - 2024
Nuclear
2,200
2023 - 2023
Subtotal
15,638
Grand Total
94,171
Note:
Any unit in NEEDS v6 that has an online year of 2022 or laterwas considered a Planned/Committed Unit.
4.3.2 Capacity
The capacity data of planned-committed units in NEEDS v6 was obtained from the December 2021
version of EIA Form 860 monthly.
4.3.3 State and Model Region
State location data for the planned-committed units in NEEDS v6 came from the December 2021 version
of EIA Form 860 monthly. The state-county information was then used to assign planned-committed units
to their respective model regions.
4.3.4 Online and Retirement Year
As noted above, planned-committed units included in NEEDS v6 are only those likely to come on-line
before June 30, 2028, as 2028 is the first analysis year in the EPA Platform v6. All planned-committed
units were assigned an online year and given a default retirement year of 9999.
4-17
-------
4.4 Potential Units
The EPA Platform v6 includes options for developing a variety of potential units that may be built at a
future date in response to electricity demand and the constraints represented in the model. Defined by
region, technology, and the year available, potential units with an initial capacity of zero MW are inputs
into IPM. When the model is run, the capacity of certain potential units is raised from zero to meet
demand and other system and operating constraints. This results in the model's projection of new
capacity.
In Table 4-7, the block labeled "New Units" provides the type and number of potential units available in
EPA Platform v6. The following sections describe the cost and performance assumptions for the potential
units represented in the EPA Platform v6.
4.4.1 Methodology for Deriving the Cost and Performance Characteristics of Conventional
Potential Units
The cost and performance characteristics of conventional potential units in EPA Platform v6 are derived
primarily from assumptions used in the Annual Energy Outlook (AEO) 2021 published by the U.S.
Department of Energy's Energy Information Administration.
4.4.2 Cost and Performance for Potential Conventional Units
Table 4-12 shows the cost and performance assumptions for potential conventional units. The cost and
performance assumptions are based on the size (i.e., net electrical generating capacity in MW) indicated
in the table. However, the total new capacity that is added in each model run for these technologies is
not restricted to these capacity levels.
The table includes several components of cost. The total installed cost of developing and building a new
unit is captured through capital cost. It includes expenditures on pollution control equipment that new
units are assumed to install to satisfy air regulatory requirements. The capital costs shown are typically
referred to as overnight capital costs. They include engineering, procurement, construction, startup, and
owner's costs (for such items as land, cooling infrastructure, administration and associated buildings, site
works, switchyards, project management, and licenses). The capital costs of new units are increased to
account for the cost of maintaining and expanding the transmission network. This cost based on AEO
2021 is equal to 103 2019$/kW outside of WECC and NY regions and 154 2019$/kW within these
regions. The capital costs do not include interest during construction (IDC). IDC is added to the capital
costs during the set-up of an IPM run. Calculation of IDC is based on the construction profile of the build
option and the discount rate. Details on the discount rate used in the EPA Platform v6 are provided in
Chapter 10 of this documentation.
Table 4-12 also shows fixed operating and maintenance (FOM) and variable operating and maintenance
(VOM) components of cost. FOM is the annual cost of maintaining a generating unit. It represents
expenses incurred regardless of the extent that the unit is run. It is expressed in units of $ per kW per
year. VOM represents the non-fuel variable costs incurred in running an electric generating unit. It is
proportional to the electrical energy produced and is expressed in units of $ per MWh.
In addition to the three components of cost, Table 4-12 indicates the first run year available, lead time,
vintage periods, heat rate, and availability for each type of unit. Lead time represents the construction
time needed for a unit to come online. Vintage periods are used to capture the cost and performance
improvements resulting from technological advancement and learning-by-doing. Mature technologies and
technologies whose first year available are not at the start of the modeling time horizon may have only
one vintage period, whereas newer technologies may have several vintage periods. Heat rate indicates
the efficiency of the unit and is expressed in units of energy consumed (Btus) per unit of electricity
generated (kWh). Availability indicates the percentage of time that a generating unit is available to
provide electricity to the grid once it is online. Availability considers estimates of the time consumed by
4-18
-------
planned maintenance and forced outages. The emission characteristics of the potential units can be
found in Table 3-26.
4.4.3 Short-Term Capital Cost Adder
In addition to the capital costs shown in Table 4-12 and Table 4-15, EPA Platform v6 includes a short-
term capital cost adder that takes effect if the new capacity deployed in a specific model run year exceeds
certain upper bounds. This adder reflects the added cost incurred due to short-term competition for
scarce labor and materials. Table 4-13 shows the cost adders for each type of potential unit for model run
years through 2035. The adder is not imposed after 2035, assuming markets for labor and materials
have sufficient time to respond to changes in demand.
The column labeled "Step 1" in Table 4-13 indicates the total capacity of a particular plant type that can
be built in a given model run year without incurring a cost adder. However, if the Step 1 upper bound is
exceeded, then either the Step 2 or Step 3 cost adder is incurred by the entire capacity deployed, where
the level of the cost adder depends upon the total new capacity added in that run year. For example, the
Step 1 upper bound in 2030 for landfill gas potential units is 375 MW. If no more than this total new
landfill gas capacity is built in 2030, only the capital cost shown in Table 4-15 is incurred. If the model
builds between 375 and 652 MW, the Step 2 cost adder of $652/kW applies to the entire capacity
deployed. If the total new landfill gas capacity exceeds the Step 2 upper bound of 652 MW, the Step 3
capacity adder of $2,095/kW is incurred by the entire capacity deployed in that run year. The short-term
capital cost adders shown in Table 4-13 were based on AEO assumptions. The short-term capital cost
adder step widths for renewable technologies are increased by 21%, 29%, and 50% in 2028, 2030, and
2035 run years respectively to reflect the impact of IRA's Advanced Manufacturing Production Tax Credit
(45X). The scalars are linearly interpolated in between 2023 (no increase) and 2035 (50% increase).
4.4.4 Regional Cost Adjustment
The capital costs reported in Table 4-12 are generic. Before implemented, the capital cost values are
converted to region-specific costs by applying regional cost adjustment factors that capture regional
differences in labor, material, and construction costs and ambient conditions. These factors are
calculated by multiplying the regional cost and ambient condition multipliers. The regional cost multipliers
are based on county level estimates developed by the Energy Institute at the University of Texas at
Austin.48 The ambient condition multipliers are from AEO 2017. Table 4-14 summarizes the regional cost
adjustment factors at the IPM region and technology level. The factors are applied to both conventional
technologies shown in Table 4-12 and renewable and nonconventional technologies shown in Table 4-15.
However, they are not applied to hydro and geothermal technologies as site-specific costs are used for
these two technologies.
48 New U.S. Power Costs: by County, with Environmental Externalities, University of Texas at Austin, Energy Institute. July 2016
4-19
-------
Table 4-12 Performance and Unit Cost Assumptions for Potential (New) Capacity from Conventional Technologies in v6
Combined Cycle
Combined Cycle
Combustion Turbine
Combustion Turbine
Advanced
Small Modular
Ultra-supercritical
- Single Shaft
- Multi Shaft
- Industrial Frame
- Aeroderivative
Nuclear
Reactor
Coal without CCS
Size (MW)
418
1083
237
105
2156
600
650
First Year Available
2028
2028
2028
2028
2028
2028
2028
Lead Time (Years)
3
3
2
2
6
6
4
Availability
87%
87%
93%
93%
90%
90%
85%
Vintage #1 (2028
Heat Rate (Btu/kWh)
6,431
6,370
9,905
9,124
10,455
10,455
8,638
Capital (2019$/kW)
1,007
891
638
1,051
5,823
6,399
3,454
Fixed O&M (2019$/kW/yr)
13.99
12.10
6.95
16.17
120.69
94.25
40.27
Variable O&M (2019$/MWh)
2.53
1.86
4.46
4.66
2.35
2.98
4.46
Vintage #2 (2030
Heat Rate (Btu/kWh)
6,431
6,370
9,905
9,124
10,455
10,455
8,638
Capital (2019$/kW)
977
864
616
1,016
5,620
6,176
3,334
Fixed O&M (2019$/kW/yr)
13.99
12.10
6.95
16.17
120.69
94.25
40.27
Variable O&M (2019$/MWh)
2.53
1.86
4.46
4.66
2.35
2.98
4.46
Vintage #3 (2035
Heat Rate (Btu/kWh)
6,431
6,370
9,905
9,124
10,455
10,455
8,638
Capital (2019$/kW)
905
800
568
936
5,140
5,650
3,050
Fixed O&M (2019$/kW/yr)
13.99
12.10
6.95
16.17
120.69
94.25
40.27
Variable O&M (2019$/MWh)
2.53
1.86
4.46
4.66
2.35
2.98
4.46
Vintage #4 (2040
Heat Rate (Btu/kWh)
6,431
6,370
9,905
9,124
10,455
10,455
8,638
Capital (2019$/kW)
845
747
527
869
4,733
5,205
2,810
Fixed O&M (2019$/kW/yr)
13.99
12.10
6.95
16.17
120.69
94.25
40.27
Variable O&M (2019$/MWh)
2.53
1.86
4.46
4.66
2.35
2.98
4.46
Vintage #5 (2045
Heat Rate (Btu/kWh)
6,431
6,370
9,905
9,124
10,455
10,455
8,638
Capital (2019$/kW)
789
698
490
807
4,355
4,792
2,587
Fixed O&M (2019$/kW/yr)
13.99
12.10
6.95
16.17
120.69
94.25
40.27
Variable O&M (2019$/MWh)
2.53
1.86
4.46
4.66
2.35
2.98
4.46
Vintage #6 (2050-2055)
Heat Rate (Btu/kWh)
6,431
6,370
9,905
9,124
10,455
10,455
8,638
Capital (2019$/kW)
732
648
452
746
3,973
4,374
2,361
Fixed O&M (2019$/kW/yr)
13.99
12.10
6.95
16.17
120.69
94.25
40.27
Variable O&M (2019$/MWh)
2.53
1.86
4.46
4.66
2.35
2.98
4.46
Notes:
a Capital cost represents overnight capital cost.
b IPM regions in urban areas (NENGREST, NY_Z_J, NY_Z_K, PJM_SMAC, PJM_COMD, WEC_LADW, WEC_SDGE, and WEC_BANC) are assigned "Combined Cycle - Single
Shaft" and "Combustion Turbine - Aeroderivative" technologies. All other regions are assigned "Combined Cycle - Multi Shaft" and "Combustion Turbine - Industrial Frame"
technologies.
0 The ultra-supercritical coal plant without CCS is not compliant with 80 FR 64510.
4-20
-------
Table 4-13 Short-Term Capital Cost Adders for New Power Plants in v6 (2019$)
Plant Type
2028
2030
2035
Step 1
Step 2
Step 3
Step 1
Step 2
Step 3
Step 1
Step 2
Step 3
Biomass
Upper Bound (MW)
4,471
7,738
No limit
1,263
2,196
No limit
3,157
5,490
No limit
Adder ($/kW)
-
1,229
3,905
-
1,695
5,384
-
1,551
4,925
Coal Steam - UPC
Upper Bound (MW)
Adder ($/kW)
38,189
66,416
1,579
No limit
5,014
10,911
18,976
1,524
No limit
4,840
27,278
47,440
1,394
No limit
4,428
Coal Steam - UPC30
Upper Bound (MW)
Adder ($/kW)
38,189
66,416
1,966
No limit
6,246
10,911
18,976
1,895
No limit
6,020
27,278
47,440
1,727
No limit
5,487
Coal Steam - UPC90
Upper Bound (MW)
Adder ($/kW)
38,189
66,416
2,546
No limit
8,088
10,911
18,976
2,451
No limit
7,785
27,278
47,440
2,225
No limit
7,069
Combined Cycle
Upper Bound (MW)
317,116
552,171
No limit
90,862
158,020
No limit
227,154
395,050
No limit
Adder ($/kW)
-
402
1,276
-
389
1,235
-
360
1,142
Combustion Turbine
Upper Bound (MW)
Adder ($/kW)
146,018
252,463
283
No limit
900
41,147
71,560
272
No limit
864
102,868
178,900
249
No limit
791
Fuel Cell
Upper Bound (MW)
4,056
7,031
No limit
1,150
2,000
No limit
2,875
5,000
No limit
Adder ($/kW)
-
2,504
7,954
-
2,371
7,530
-
2,057
6,534
Geothermal
Upper Bound (MW)
1,559
2,660
No limit
454
789
No limit
1,320
2,295
No limit
Adder ($/kW)
-
2,772
8,806
-
2,757
8,757
-
2,741
8,706
Landfill Gas
Upper Bound (MW)
1,317
2,287
No limit
375
652
No limit
937
1,630
No limit
Adder ($/kW)
-
389
1,234
-
661
2,098
-
613
1,947
Nuclear
Upper Bound (MW)
10,929
19,007
No limit
3,329
5,790
No limit
9,677
16,830
No limit
Adder ($/kW)
-
1,937
6,153
-
1,870
5,939
-
1,710
5,432
Solar Thermal
Upper Bound (MW)
7,919
13,772
No limit
2,412
4,195
No limit
7,012
12,195
No limit
Adder ($/kW)
-
1,339
4,252
-
1,257
3,992
-
1,172
3,724
Solar PV
Upper Bound (MW)
148,217
239,176
No limit
37,485
65,191
No limit
108,968
189,510
No limit
Adder ($/kW)
-
217
688
-
154
489
-
134
426
Onshore Wind
Upper Bound (MW)
192,613
314,378
No limit
50,181
87,271
No limit
145,875
253,695
No limit
Adder ($/kW)
-
220
700
-
176
559
-
142
451
Offshore Wind
Upper Bound (MW)
5,702
9,301
No limit
9,675
10,772
No limit
21,300
24,488
No limit
Adder ($/kW)
-
531
1,686
-
552
1,754
-
475
1,508
Hydro
Upper Bound (MW)
Adder ($/kW)
4,870
8,470
582
No limit
1,848
1,484
2,580
582
No limit
1,848
4,313
7,500
582
No limit
1,848
4-21
-------
Table 4-14 Regional Cost Adjustment Factors for Conventional and Renewable Generating Technologies in v6
Regional Multiplier
Model
Region
Combined
Cycle
Combined
Cycle with
Carbon
Capture
Combustion
Turbine
Hydro
Nuclear
Biomass
Geothermal
Landfill
Gas
Offshore
Wind
Onshore
Wind
Solar
PV
Solar
Thermal
Fuel
Cell
Ultra-
supercritical
Coal without
CCS
Ultra-
supercriti
cal Coal
with 30%
CCS
Ultra-
supercriti
cal Coal
with 90%
CCS
ERC PHDL
1.006
1.006
1.042
1.000
0.979
0.922
1.000
0.920
1.002
1.002
0.961
0.916
0.937
1.005
1.005
0.992
ERC REST
0.977
0.977
1.027
1.000
0.969
0.922
1.000
0.920
0.968
0.968
0.935
0.889
0.937
0.981
0.981
0.969
ERC WEST
0.999
0.999
1.038
1.000
0.976
0.922
1.000
0.920
0.989
0.989
0.952
0.909
0.937
0.997
0.997
0.985
FRCC
0.983
0.983
1.033
1.000
0.976
0.948
1.000
0.949
0.961
0.961
0.936
0.899
0.960
1.001
1.001
0.991
MIS AMSO
0.955
0.955
1.015
1.000
0.963
0.930
1.000
0.933
0.949
0.949
0.917
0.865
0.946
0.958
0.958
0.947
MIS AR
0.977
0.977
1.022
1.000
0.977
0.930
1.000
0.933
0.977
0.977
0.950
0.914
0.946
0.995
0.995
0.987
MIS MS
0.958
0.958
1.013
1.000
0.968
0.930
1.000
0.933
0.958
0.958
0.929
0.884
0.946
0.972
0.972
0.962
MIS IA
1.001
1.001
1.017
1.000
0.999
0.968
1.000
0.968
1.041
1.041
1.011
0.993
0.975
1.013
1.013
1.008
MIS IL
1.000
1.000
1.016
1.000
0.999
1.017
1.000
1.019
1.014
1.014
0.999
0.990
1.017
1.021
1.021
1.020
MIS INKY
0.987
0.987
1.007
1.000
0.998
1.010
1.000
0.994
1.003
1.003
0.987
0.972
0.997
1.009
1.009
1.008
MIS LA
0.958
0.958
1.013
1.000
0.967
0.930
1.000
0.933
0.957
0.957
0.926
0.879
0.946
0.968
0.968
0.956
MIS LMI
1.009
1.009
1.015
1.000
1.016
0.995
1.000
0.997
1.024
1.024
1.007
1.002
0.999
1.025
1.025
1.022
MIS MAPP
0.970
0.970
1.003
1.000
0.986
0.968
1.000
0.968
1.035
1.035
0.985
0.945
0.975
0.976
0.976
0.967
MIS MIDA
0.996
0.996
1.015
1.000
0.997
0.968
1.000
0.968
1.040
1.040
1.007
0.984
0.975
1.007
1.007
1.000
MIS MNWI
1.006
1.006
1.020
1.000
1.000
0.968
1.000
0.968
1.050
1.050
1.021
1.008
0.975
1.015
1.015
1.010
MIS MO
0.995
0.995
1.015
1.000
0.995
1.017
1.000
1.019
1.016
1.016
0.996
0.981
1.017
1.013
1.013
1.009
MIS WOTA
0.956
0.956
1.010
1.000
0.966
0.930
1.000
0.933
0.956
0.956
0.923
0.875
0.946
0.964
0.964
0.952
MIS WUMS
1.028
1.028
1.032
1.000
1.013
1.010
1.000
0.994
1.045
1.045
1.029
1.029
0.997
1.046
1.046
1.044
NENG CT
1.181
1.181
1.146
1.000
1.068
1.030
1.000
1.009
1.081
1.081
1.076
1.103
1.009
1.112
1.112
1.116
NENG ME
1.064
1.064
1.074
1.000
1.042
1.030
1.000
1.009
1.065
1.065
1.017
0.993
1.009
1.048
1.048
1.047
NENG REST
1.115
1.115
1.105
1.000
1.053
1.030
1.000
1.009
1.068
1.068
1.038
1.034
1.009
1.075
1.075
1.075
NY Z A
1.061
1.061
1.072
1.000
1.039
1.034
1.000
0.999
1.021
1.021
1.000
0.988
0.995
1.050
1.050
1.046
NY Z B
1.076
1.076
1.081
1.000
1.043
1.034
1.000
0.999
1.027
1.027
1.004
0.992
0.995
1.058
1.058
1.054
NY Z C&E
1.110
1.110
1.111
1.000
1.056
1.034
1.000
0.999
1.038
1.038
1.015
1.005
0.995
1.080
1.080
1.078
NY Z D
1.076
1.076
1.092
1.000
1.045
1.034
1.000
0.999
1.043
1.043
1.008
0.986
0.995
1.056
1.056
1.053
NY Z F
1.129
1.129
1.122
1.000
1.055
1.034
1.000
0.999
1.060
1.060
1.039
1.040
0.995
1.085
1.085
1.085
NY Z G-l
1.195
1.195
1.161
1.000
1.068
1.034
1.000
0.999
1.079
1.079
1.085
1.130
0.995
1.119
1.119
1.122
NY Z J
1.257
1.257
1.205
1.000
1.074
1.227
1.000
1.260
1.093
1.093
1.123
1.216
1.212
1.157
1.157
1.162
NY Z K
1.241
1.241
1.196
1.000
1.073
1.227
1.000
1.260
1.092
1.092
1.104
1.163
1.212
1.153
1.153
1.158
PJM AP
1.073
1.073
1.088
1.000
1.034
1.010
1.000
0.994
1.008
1.008
0.982
0.961
0.997
1.072
1.072
1.069
PJM ATSI
1.031
1.031
1.046
1.000
1.018
1.010
1.000
0.994
1.007
1.007
0.988
0.974
0.997
1.043
1.043
1.039
PJM_COMD
1.022
1.022
1.026
1.000
1.009
1.010
1.000
0.994
1.040
1.040
1.033
1.042
0.997
1.039
1.039
1.039
PJM Dom
1.144
1.144
1.153
1.000
1.046
0.913
1.000
0.911
1.018
1.018
0.988
0.964
0.932
1.130
1.130
1.127
PJM EMAC
1.209
1.209
1.179
1.000
1.073
1.065
1.000
1.033
1.066
1.066
1.063
1.090
1.027
1.144
1.144
1.148
PJM PENE
1.097
1.097
1.105
1.000
1.047
1.065
1.000
1.033
1.024
1.024
1.002
0.988
1.027
1.083
1.083
1.081
PJM SMAC
1.155
1.155
1.144
1.000
1.063
1.065
1.000
1.033
1.036
1.036
1.008
0.990
1.027
1.118
1.118
1.118
PJM West
0.991
0.991
1.019
1.000
1.004
1.010
1.000
0.994
0.989
0.989
0.965
0.939
0.997
1.012
1.012
1.008
PJM WMAC
1.151
1.151
1.144
1.000
1.060
1.065
1.000
1.033
1.043
1.043
1.024
1.018
1.027
1.113
1.113
1.113
S C KY
0.981
0.981
1.015
1.000
0.990
0.934
1.000
0.933
0.979
0.979
0.953
0.919
0.948
1.006
1.006
1.004
S C TVA
0.957
0.957
1.003
1.000
0.979
0.934
1.000
0.933
0.968
0.968
0.939
0.899
0.948
0.981
0.981
0.975
S D AECI
0.989
0.989
1.014
1.000
0.992
1.017
1.000
1.019
1.013
1.013
0.990
0.971
1.017
1.005
1.005
0.999
S SOU
0.963
0.963
1.020
1.000
0.969
0.925
1.000
0.925
0.953
0.953
0.922
0.873
0.942
0.982
0.982
0.972
S VACA
1.015
1.015
1.059
1.000
1.003
0.913
1.000
0.911
0.975
0.975
0.940
0.896
0.932
1.033
1.033
1.025
4-22
-------
Regional Multiplier
Ultra-
Ultra-
Combined
Ultra-
supercriti
supercriti
Model
Cycle with
supercritical
cal Coal
cal Coal
Combined
Carbon
Combustion
Landfill
Offshore
Onshore
Solar
Solar
Fuel
Coal without
with 30%
with 90%
Region
Cycle
Capture
Turbine
Hydro
Nuclear
Biomass
Geothermal
Gas
Wind
Wind
PV
Thermal
Cell
CCS
CCS
CCS
SPP N
1.000
1.000
1.032
1.000
0.986
0.973
1.000
0.975
1.016
1.016
0.980
0.948
0.979
1.009
1.009
0.998
SPP NEBR
0.976
0.976
1.009
1.000
0.988
0.968
1.000
0.968
1.029
1.029
0.984
0.945
0.975
0.982
0.982
0.971
SPP SPS
0.992
0.992
1.028
1.000
0.980
0.956
1.000
0.952
1.005
1.005
0.963
0.920
0.962
0.991
0.991
0.979
SPP WAUE
0.974
0.974
1.006
1.000
0.987
0.968
1.000
0.968
1.034
1.034
0.986
0.947
0.975
0.979
0.979
0.970
SPP WEST
0.978
0.978
1.020
1.000
0.978
0.956
1.000
0.952
0.991
0.991
0.957
0.918
0.962
0.989
0.989
0.978
WEC BANC
1.232
1.232
1.173
1.000
1.072
1.076
1.000
1.055
1.124
1.124
1.098
1.112
1.045
1.208
1.208
1.203
WEC CALN
1.230
1.230
1.172
1.000
1.071
1.076
1.000
1.055
1.123
1.123
1.096
1.109
1.045
1.207
1.207
1.201
WEC LADW
1.183
1.183
1.141
1.000
1.055
1.076
1.000
1.055
1.104
1.104
1.074
1.076
1.045
1.167
1.167
1.151
WEC SDGE
1.154
1.154
1.120
1.000
1.046
1.076
1.000
1.055
1.084
1.084
1.054
1.049
1.045
1.141
1.141
1.123
WECC AZ
1.187
1.187
1.190
1.000
1.011
1.000
1.000
0.982
1.035
1.035
0.998
0.970
0.986
1.181
1.181
1.166
WECC CO
1.157
1.157
1.194
1.000
0.988
0.936
1.000
0.947
1.027
1.027
0.976
0.932
0.958
1.156
1.156
1.142
WECC ID
1.045
1.045
1.070
1.000
1.004
1.002
1.000
0.982
1.048
1.048
1.000
0.965
0.989
1.066
1.066
1.058
WECC IID
1.262
1.262
1.236
1.000
1.036
1.000
1.000
0.982
1.069
1.069
1.038
1.028
0.986
1.252
1.252
1.233
WECC MT
1.021
1.021
1.054
1.000
0.992
1.002
1.000
0.982
1.039
1.039
0.990
0.953
0.989
1.037
1.037
1.030
WECC NM
1.131
1.131
1.161
1.000
0.990
1.000
1.000
0.982
1.018
1.018
0.977
0.938
0.986
1.129
1.129
1.115
WECC NNV
1.157
1.157
1.137
1.000
1.040
1.002
1.000
0.982
1.087
1.087
1.053
1.045
0.989
1.157
1.157
1.147
WECC PNW
1.123
1.123
1.109
1.000
1.035
1.002
1.000
0.982
1.074
1.074
1.042
1.032
0.989
1.145
1.145
1.144
WECC SCE
1.180
1.180
1.139
1.000
1.054
1.076
1.000
1.055
1.100
1.100
1.070
1.071
1.045
1.163
1.163
1.144
WECC SNV
1.230
1.230
1.220
1.000
1.030
1.000
1.000
0.982
1.071
1.071
1.044
1.042
0.986
1.237
1.237
1.219
WECC UT
1.050
1.050
1.075
1.000
1.002
1.002
1.000
0.982
1.043
1.043
0.997
0.962
0.989
1.063
1.063
1.051
WECC WY
1.016
1.016
1.055
1.000
0.987
1.002
1.000
0.982
1.031
1.031
0.976
0.927
0.989
1.024
1.024
1.012
4-23
-------
Table 4-15 Performance and Unit Cost Assumptions for Potential (New) Renewable and Non-Conventional Technologies in v6
Geothermal
Biomass
Landfill Gas
LGHI
Fuel Cells
Solar
Photovoltaic
Solar
Thermal
Onshore
Wind
Offshore
Wind
Battery
Storage
(4 Hours)
Battery
Storage
(8 Hours)
Size (MW)
50
50
36
10
100
104
200
1,000
60
60
First Year Available
2028
2028
2028
2028
2028
2028
2028
2028
2028
2028
Lead Time (Years)
4
4
3
3
1
3
3
3
1
1
Availability
80% - 90%
83%
90%
87%
90%
90%
95%
95%
96.4%
96.4%
Generation Capability
Economic
Dispatch
Economic
Dispatch
Economic
Dispatch
Economic
Dispatch
Generation
Profile
Economic
Dispatch
Generation
Profile
Generation
Profile
Economic
Dispatch
Economic
Dispatch
Vintage #1
(2028-2054) Vintage #1 (2028)
Heat Rate (Btu/kWh)
Capital (2019$/kW)
Fixed O&M (2019$/kW/yr)
Variable O&M (2019$/MWh)
30,000
3,233 - 43,097
101 - 1,067
0.00
13,500
3,835
124.74
4.79
8,513
1,507
19.94
6.15
6,469
5,573
30.54
0.58
0
877
17.98
0.00
0
4,628
53.82
2.89
0
995
39.69
0.00
0
1,666
85.77
0.00
0
853
21.32
0.00
0
1,508
37.71
0.00
Vintage #2 (2030)
Heat Rate (Btu/kWh)
Capital (2019$/kW)
Fixed O&M (2019$/kW/yr)
Variable O&M (2019$/MWh)
13,500
3,701
124.74
4.79
8,513
1,465
19.94
6.15
6,469
5,275
30.54
0.58
0
759
16.64
0.00
0
4,409
50.45
2.89
0
910
38.95
0.00
0
1,559
83.01
0.00
0
784
19.60
0.00
0
1,371
34.28
0.00
Vintage #3 (2035)
Heat Rate (Btu/kWh)
Capital (2019$/kW)
Fixed O&M (2019$/kW/yr)
Variable O&M (2019$/MWh)
13,500
3,386
124.74
4.79
8,513
1,364
19.94
6.15
6,469
4,578
30.54
0.58
0
726
16.22
0.00
0
4,119
50.45
2.89
0
865
37.49
0.00
0
1,439
77.63
0.00
0
735
18.38
0.00
0
1,277
31.93
0.00
Vintage #4 (2040)
Heat Rate (Btu/kWh)
Capital (2019$/kW)
Fixed O&M (2019$/kW/yr)
Variable O&M (2019$/MWh)
13,500
3,119
124.74
4.79
8,513
1,280
19.94
6.15
6,469
3,971
30.54
0.58
0
692
15.80
0.00
0
4,067
50.45
2.89
0
819
36.03
0.00
0
1,348
73.58
0.00
0
686
17.15
0.00
0
1,183
29.57
0.00
Vintage #5 (2045)
Heat Rate (Btu/kWh)
Capital (2019$/kW)
Fixed O&M (2019$/kW/yr)
Variable O&M (2019$/MWh)
13,500
2,871
124.74
4.79
8,513
1,202
19.94
6.15
6,469
3,414
30.54
0.58
0
658
15.39
0.00
0
4,055
50.45
2.89
0
774
34.57
0.00
0
1,275
70.33
0.00
0
637
15.93
0.00
0
1,089
27.22
0.00
Vintac
e #6 (2050-2055)
Heat Rate (Btu/kWh)
Capital (2019$/kW)
Fixed O&M (2019$/kW/yr)
Variable O&M (2019$/MWh)
13,500
2,620
124.74
4.79
8,513
1,120
19.94
6.15
6,469
2,878
30.54
0.58
0
624
14.99
0.00
0
4,043
50.45
2.89
0
728
33.11
0.00
0
1,214
67.62
0.00
0
588
14.70
0.00
0
995
24.87
0.00
4-24
-------
4.4.5 Cost and Performance for Potential Renewable Generating and Non-Conventional
Technologies
Table 4-15 summarizes the cost and performance assumptions in EPA Platform v6 for potential
renewable and non-conventional technology generating units. The parameters shown in the table are
based on AEO 2021 for biomass, landfill gas, and fuel cell. For battery storage, onshore wind, offshore
wind, solar PV, and solar thermal technologies, the parameters shown are based on the National
Renewable Energy Laboratory's (NREL's) 2021 Annual Technology Baseline (ATB) moderate case. The
geothermal assumptions are based on ATB 2019. The size (MW) shown in Table 4-15 represents the
capacity on which unit cost estimates were developed and does not indicate the total potential capacity
that the model can build of a given technology. Due to the distinctive nature of generation from
renewable resources, some of the values shown are averages or ranges that are discussed in further
detail in the following subsections. The short-term capital cost adder in Table 4-13 and the regional cost
adjustment factors in Table 4-14 apply equally to the renewable and non-conventional generation
technologies as to the conventional generation technologies.
Wind Generation
EPA Platform v6 includes onshore wind, offshore-fixed, and offshore-floating wind generation
technologies. The following sections describe key aspects of the representation of wind generation: wind
quality and resource potential, distance to transmission, generation profiles, reserve margin contribution,
and capital cost calculation.
Wind Quality and Resource Potential: The NREL resource base for onshore wind is represented by ten
wind speed class categories (Class 1 - Class 10). EPA Platform v6 only models the categories Class 1 -
Class 9. The NREL resource base for offshore wind is represented by fixed (Class 1 - Class 7), and
floating (Class 8 - Class 14) categories. EPA Platform v6 models the categories Class 1 - Class 12.
Table 4-36, Table 4-16, and Table 4-17 present the onshore, offshore fixed, and offshore floating wind
resource assumptions. The resource class field in the tables further subdivides the wind speed class
categories based on wind speed.
Table 4-16 Offshore Fixed Regional Potential Wind Capacity (MW) by Wind Class, Resource Class, and
Cost Class in v6
IPM Region
State
Wind
Resource
Cost Class
Class
Class
1
2
3
4
5
6
ERC_REST
TX
Class 5
Class 6
6
5
2,976
2,622
693
3,245
3,035
3,052
3,004
4,243
FRCC
FL
Class 6
5
2,900
3,091
2,636
3,362
2,810
9,172
MIS AMSO
LA
Class 6
5
885
909
858
900
920
12,957
MIS LA
LA
Class 6
5
31
MIS LMI
Ml
Class 2
7
154
MIS_WOTA
LA
Class 6
5
871
922
903
903
875
36,861
TX
Class 6
5
519
1,038
1,038
781
1,049
15,042
MIS_WUMS
Ml
Class 3
7
237
Wl
Class 4
6
0
NENG ME
ME
Class 1
8
12
NENGREST
MA
Class 1
8
1,418
2,118
4,236
2,118
2,118
8,708
Rl
Class 1
8
14
NY_Z_K
NY
Class 1
Class 2
8
7
165
685
212
PJM ATSI
OH
Class 3
7
1,560
1,606
1,491
NC
Class 2
7
2,597
2,545
841
PJM_Dom
VA
Class 2
Class 4
7
6
2,390
2
1,022
PJM_EMAC
DE
Class 1
Class 2
8
7
2,894
2,987
274
4-25
-------
IPM Region
State
Wind
Resource
Cost Class
Class
Class
1
2
3
4
5 6
MD
Class 2
7
2,423
NJ
Class 1
8
2,945
3,010
3,004
2,922
Class 2
7
2,968
2,475
VA
Class 2
7
2,983
3,014
14
AL
Class 6
5
2,950
3,040
983
s_sou
FL
Class 6
5
29
GA
Class 6
5
2,980
3,020
357
MS
Class 6
5
2,435
NC
Class 3
7
2,971
2,393
S_VACA
Class 5
6
2,767
2,645
3,586
2,307
SC
Class 5
6
2,647
2,885
3,299
2,978
3,162 20,234
Class 6
5
2,957
2,996
Table 4-17 Offshore Floating Regional Potential Wind Capacity (MW) by Wind Class, Resource
Class, and Cost Class in v6
IPM Region
State
Wind
Resource
Cost Class
Class
Class
1
2
3
4
5
6
MIS LMI
Ml
Class 12
7
2,154
MIS WUMS
Ml
Class 12
7
113
NENG_ME
ME
Class 8
8
330
330
330
330
85,755
Class 11
7
397
397
397
6,940
MA
Class 8
8
2,176
2,888
1,444
3,882
2,528
370,283
NENGREST
Class 11
7
1,450
Rl
Class 8
8
376
NY Z J
NY
Class 11
7
8,509
NY_Z_K
NY
Class 9
8
608
696
796
694
663
74,310
Class 11
7
397
794
794
789
588
PJM_Dom
NC
Class 12
7
2,509
2,681
2,595
1,782
2,515
4,918
VA
Class 12
7
1,986
DE
Class 10
8
2,978
992
Class 11
7
496
MD
Class 10
8
397
PJM EMAC
Class 11
7
2,846
2,846
2,846
2,846
2,846
27,846
NJ
Class 10
8
2,717
3,194
2,577
3,376
3,022
33,803
Class 11
7
2,942
3,031
1,539
3,839
1,919
34,612
VA
Class 12
7
2,978
2,796
3,200
2,600
S VACA
NC
Class 12
7
397
3,176
3,176
3,176
3,176
321,572
WEC_CALN
CA
Class 12
7
2,984
2,800
3,210
2,762
3,177
513,613
Class 8
8
2,222
3,640
3,640
3,640
360,347
CA
Class 8
8
2,780
3,197
2,774
1,646
WECC_PNW
OR
Class 8
Class 12
8
7
2,754
3,175
3,064
2,908
2,383
43,714
345,408
WA
Class 12
7
2,646
2,646
2,646
2,646
2,646
74,215
WECC SCE
CA
Class 12
7
1,312
3,772
3,772
3,772
72,915
Generation Profiles: Unlike other generation technologies, which dispatch on an economic basis subject
to their availability constraint, wind, and solar technologies dispatch only when the wind blows and the
sun shines. To represent intermittent renewable generating sources such as wind and solar, EPA
Platform v6 uses hourly generation profiles. All wind and solar photovoltaic units are provided with hourly
generation profiles. The profiles are customized for each resource class within an IPM region and state
combination.
4-26
-------
The generation profile indicates the amount of generation (kWh) per MW of available capacity. The wind
generation profiles were prepared with data from NREL. Table 4-37 shows the generation profiles for
onshore and offshore wind units in all model region, state, and class combinations for vintage 2028.
Improvements in onshore wind and offshore wind capacity factors overtime are modeled through three
vintages (2028, 2030, and 2040) of potential wind units.
To obtain the seasonal generation for the units in a particular resource class in a specific region, the
installed capacity is multiplied by the number of hours in the season and the seasonal capacity factor.
Capacity factor is the average "kWh of generation per MW" from the applicable generation profile. The
annual capacity factors for wind generation that are used in EPA Platform v6 were obtained from NREL
and are shown in Table 4-35, Table 4-18, and Table 4-19.
Table 4-18 Offshore Fixed Average Capacity Factor by Wind Class and Resource Class in v6
Wind
Class
Resource
Class
Capacity Factor (%)
IPM Region
State
Vintage #1 (2028-
2059)
Vintage #2 (2030-
2059)
Vintage #3 (2040-
2059)
ERC_REST
TX
Class 5
6
47%
48%
48%
Class 6
5
42%
43%
43%
FRCC
FL
Class 6
5
37%
38%
38%
MIS AMSO
LA
Class 6
5
36%
37%
37%
MIS LA
LA
Class 6
5
38%
39%
39%
MIS LMI
Ml
Class 2
7
47%
48%
49%
MIS_WOTA
LA
Class 6
5
39%
40%
40%
TX
Class 6
5
41%
42%
42%
MIS_WUMS
Ml
Class 3
7
48%
49%
50%
Wl
Class 4
6
48%
49%
50%
NENG ME
ME
Class 1
8
49%
50%
51%
NENGREST
MA
Class 1
8
49%
50%
50%
Rl
Class 1
8
46%
47%
48%
NY_Z_K
NY
Class 1
8
46%
47%
48%
Class 2
7
48%
49%
50%
PJM ATSI
OH
Class 3
7
47%
48%
48%
NC
Class 2
7
45%
46%
47%
PJM Dom
VA
Class 2
7
45%
46%
46%
Class 4
6
46%
47%
48%
DE
Class 1
8
45%
46%
47%
Class 2
7
48%
49%
49%
PJM_EMAC
MD
Class 2
7
47%
48%
48%
NJ
Class 1
8
46%
47%
47%
Class 2
7
47%
49%
49%
VA
Class 2
7
45%
46%
47%
AL
Class 6
5
36%
37%
37%
S_SOU
FL
Class 6
5
36%
36%
37%
GA
Class 6
5
42%
43%
43%
MS
Class 6
5
36%
37%
37%
NC
Class 3
7
46%
47%
48%
S_VACA
Class 5
6
47%
48%
48%
SC
Class 5
6
45%
46%
46%
Class 6
5
42%
43%
43%
4-27
-------
Table 4-19 Offshore Floating Average Capacity Factor by Wind Class and Resource Class in v6
Wind
Class
Resource
Class
Capacity Factor (%)
IPM Region
State
Vintage #1
(2028-2059)
Vintage #2 (2030-
2059)
Vintage #3 (2040-
2059)
MIS LMI
Ml
Class 12
7
47%
48%
48%
MIS WUMS
Ml
Class 12
7
46%
46%
47%
NENG_ME
ME
Class 8
8
52%
53%
53%
Class 11
7
49%
49%
49%
MA
Class 8
8
51%
51%
52%
NENGREST
Class 11
7
51%
51%
51%
Rl
Class 8
8
52%
52%
52%
NY Z J
NY
Class 11
7
50%
51%
51%
NY_Z_K
NY
Class 9
8
51%
52%
52%
Class 11
7
50%
51%
51%
PJM_Dom
NC
Class 12
7
45%
46%
46%
VA
Class 12
7
45%
46%
46%
DE
Class 10
8
50%
50%
51%
Class 11
7
50%
51%
51%
MD
Class 10
8
50%
50%
50%
PJM EMAC
Class 11
7
49%
50%
50%
NJ
Class 10
8
51%
51%
51%
Class 11
7
50%
50%
51%
VA
Class 12
7
45%
46%
46%
S VACA
NC
Class 12
7
46%
46%
46%
WEC_CALN
CA
Class 8
Class 12
8
7
55%
49%
56%
49%
56%
50%
CA
Class 8
8
47%
47%
47%
WECC_PNW
OR
Class 8
Class 12
8
7
51%
46%
51%
46%
51%
46%
WA
Class 12
7
44%
44%
45%
WECC SCE
CA
Class 12
7
48%
49%
49%
Reserve Margin Contribution (also referred to as capacity credit): EPA Platform v6 uses reserve margins,
discussed in detail in Section 3.6, to model reliability. Each region has a reserve margin requirement
which is used to determine the total capacity needed to reliably meet peak demand. The ability of a unit
to assist a region in meeting its reliability requirements is modeled through the unit's contribution to
reserve margin. If the unit has 100 percent contribution towards reserve margin, then the entire capacity
of the unit is counted towards meeting the region's reserve margin requirement. However, if any unit has
less than a 100 percent contribution towards reserve margin, then only the designated share of the unit's
capacity counts towards the reserve margin requirement.
All units except those that depend on intermittent resources have 100% contributions toward reserve
margin. Intermittent resources such as wind and solar have limited (less than 100 percent) contributions
toward reserve margins requirements.
Capacity credit assumptions for onshore wind, offshore wind, and solar PV units are estimated as the
function of penetration of solar and wind. A two-step approach is developed to estimate the capacity
credit at a unit level. In the first step, the method estimates the sequence of solar and wind units to build
in each ISO/NERC assessment region. Table 3-11 provides the mapping between the ISO/NERC
assessment region and the IPM region. To do so, each solar and wind unit in an ISO/NERC assessment
region is sorted from cheapest to most expensive in terms of cost and potential revenue generation. Unit
level capital costs, FOM costs, capital charge rate, and average energy price in each IPM region are
used. In the second step, capacity credit is estimated for each unit in the sequence as the ratio between
the MW of peak reduced and the capacity of the unit. Unit level hourly generation profiles and ISO/NERC
assessment region level hourly load curves are used. The approach allows the EPA Platform v6 to
4-28
-------
endogenously account for the decline of capacity credit for intermittent resources with their rising
penetration.
Table 4-20, Table 4-21, and Table 4-22 present the reserve margin contributions apportioned to new wind
units in the EPA Platform v6.
Table 4-20 Onshore Reserve Margin Contribution by Wind Class in v6
Wind Class
Vintage #1 (2028-2059)
Vintage #2 (2030-2059)
Vintage #3 (2040-2059)
Class 1
0% - 77%
0% - 79%
0% - 79%
Class 2
16%
16%
16%
Class 3
0% - 84%
0% - 87%
0% - 88%
Class 4
0% - 82%
0% - 86%
0% - 87%
Class 5
0% - 81%
0% - 84%
0% - 86%
Class 6
0% - 37%
0% - 39%
0% - 40%
Class 7
0% - 83%
0% - 87%
0% - 89%
Class 8
0% - 51%
0% - 53%
0% - 54%
Class 9
0% - 86%
0% - 91%
0% - 93%
Table 4-21 Offshore Fixed Reserve Margin Contribution by Wind Class in v6
Wind Class
Vintage #1 (2028-2059)
Vintage #2 (2030-2059)
Vintage #3 (2040-2059)
Class 1
0.3% - 80%
0.3% - 82%
0.3% - 83%
Class 2
0.1%-85%
0.1%-87%
0.1%-88%
Class 3
0% - 30%
0% - 30%
0% - 31%
Class 4
6.6% - 7.6%
6.8%-7.7%
6.9%-7.9%
Class 5
1.4% - 36%
1.4% - 37%
1.4% - 37%
Class 6
0% - 63%
0% - 64%
0% - 65%
Table 4-22 Offshore Floating Reserve Margin Contribution by Wind Class in v6
Wind Class
Vintage #1 (2028-2059)
Vintage #2 (2030-2059)
Vintage #3 (2040-2059)
Class 8
0% - 86.4%
0% - 87.2%
0% - 87.8%
Class 9
1.9% - 89%
1.9% - 90%
1.9%-91%
Class 10
1.4%-2.9%
1.5%-2.9%
1.5%-2.9%
Class 11
0% - 32%
0% - 32%
0% - 32%
Class 12
0% - 33%
0% - 33%
0% - 33%
Capital cost calculation: Capital costs for wind units include spur-line transmission costs. The resources
for wind and solar are highly sensitive to location. These spur-line costs represent the cost of needed
spur lines and are based on an estimated distance to transmission infrastructure. NREL develops these
supply curves based on a geographic-information-system analysis, which estimates the resource
accessibility costs in terms of supply curves based on the expected cost of linking renewable resource
sites to the high-voltage, long-distance transmission network. For IPM modeling purposes, the NREL
spur line cost curves are aggregated into a piecewise step curve for each resource class within each
model region and state combination. The sizes of the initial steps are based on the model region load,
while the last step holds the residual resource. The wind class and resource class level spur line cost
curves for each model region and state combination are aggregated into a six-step cost curve for onshore
wind and offshore wind units. To obtain the capital cost for a particular new wind model plant, the capital
cost adder applicable to the new plant by resource and cost class shown in Table 4-23, Table 4-24, and
Table 4-38, is added to the base capital cost shown in Table 4-15.
4-29
-------
Table 4-23 Capital Cost Adder (2019$/kW) for New Offshore Fixed Wind Plants in v6
IPM Region
State
Wind
Resource
Cost Class
Class
Class
1
2
3
4
5
6
ERC_REST
TX
Class 5
Class 6
6
5
124
27
918
28
31
41
47
97
FRCC
FL
Class 6
5
19
20
26
31
47
132
MIS AMSO
LA
Class 6
5
41
50
118
176
183
358
MIS LA
LA
Class 6
5
4,541
MIS LMI
Ml
Class 2
7
4,795
MIS_WOTA
LA
Class 6
5
61
84
101
106
112
312
TX
Class 6
5
25
25
25
26
27
95
MIS_WUMS
Ml
Class 3
7
9,713
Wl
Class 4
6
117,699
NENG ME
ME
Class 1
8
5,420
NENGREST
MA
Class 1
8
13
157
157
157
157
421
Rl
Class 1
8
12,392
NY_Z_K
NY
Class 1
Class 2
8
7
245
3
183
PJM ATSI
OH
Class 3
7
262
404
1,486
NC
Class 2
7
39
130
371
PJM_Dom
VA
Class 2
Class 4
7
6
59
15,579
353
DE
Class 1
8
63
Class 2
7
44
387
PJM_EMAC
MD
Class 2
7
180
NJ
Class 1
8
31
79
109
186
Class 2
7
3
198
VA
Class 2
7
287
216,051
3,560,858
AL
Class 6
5
103
217
636
S_SOU
FL
Class 6
5
1,096
GA
Class 6
5
51
119
610
MS
Class 6
5
208
NC
Class 3
7
67
466
S_VACA
Class 5
6
8
59
65
205
SC
Class 5
6
5
11
15
18
20
91
Class 6
5
19
130
Table 4-24 Capital Cost Adder (2019$/kW) for New Offshore Floating Wind Plants in v6
IPM Region
State
Wind
Resource
Cost Class
Class
Class
1
2
3
4
5
6
MIS LMI
Ml
Class 12
7
771
MIS WUMS
Ml
Class 12
7
4,453
NENG_ME
ME
Class 11
Class 8
00 -vl
59
59
59
59
59
59
59
222
590
MA
Class 11
7
118
NENGREST
Class 8
8
8
10
10
11
59
338
Rl
Class 8
8
1,116
NY Z J
NY
Class 11
7
118
NY_Z_K
NY
Class 11
7
92
92
92
92
93
Class 9
8
2
3
6
12
43
222
PJM_Dom
NC
Class 12
7
45
65
98
206
235
282
VA
Class 12
7
89
DE
Class 10
8
49
92
PJM_EMAC
Class 11
7
167
MD
Class 10
8
51
Class 11
7
69
69
69
69
69
174
4-30
-------
IPM Region
State
Wind
Resource
Cost Class
Class
Class
1
2
3
4
5
6
NJ
Class 10
8
18
39
68
71
75
125
Class 11
7
50
54
64
69
69
108
VA
Class 12
7
69
224
465
154,476
S VACA
NC
Class 12
7
59
62
62
62
62
216
WEC_CALN
CA
Class 12
Class 8
00 -vl
00
26
70
37
52
70
67
70
318
379
CA
Class 8
8
254
274
639
1,226
WECC_PNW
OR
Class 12
Class 8
00 -vl
33
36
41
63
65
60
153
WA
Class 12
7
45
45
45
45
45
239
WECC SCE
CA
Class 12
7
56
81
81
81
526
As an illustrative example, Table 4-25 shows the calculations that would be performed to derive the
potential electric generation, reserve margin contribution, and cost of potential (new) onshore capacity in
wind class 1, resource class 7, and cost class 1 in the WECC_CO model region in run year 2023.
Table 4-25 Example Calculations of Wind Generation, Reserve Margin Contribution, and Capital Cost
for Onshore Wind in WECC_CO for Wind Class 7, Resource Class 5, and Cost Class 1.
Reauired Data
Table 4-35
Table 4-36
Table 4-36
Table 4-36
Potential wind capacity (C) =
Winter average generation (Gw) per available MW =
Winter Shoulder average generation (Gws) per available MW =
Summer average generation (Gs) per available MW =
Hours in Winter (Hw) season (December - February) =
Hours in Winter Shoulder (Hws) season (Mar, Apr, Oct. Nov.) =
Hours in Summer (Hs) season (May - September) =
1,876 MW
277 kWh/MW
330 kWh/MW
363 kWh/MW
2,160 hours
2,928 hours
3,672 hours
Table 4-20
Reserve Margin Contribution (RM) =
3.64 percent
Table 4-15
Table 4-37
Table 4-14
Capital Cost (Cap2o2s) in vintage range for year 2028 =
Capital Cost Adder (CCA0n,ci) for onshore cost class 1 =
Regional Factor (RF)
$995/kW
$33/kW
1.027
Calculations
Generation Potential = C x Gw x Hw + C x Gws x Hws + C x Gs x Hs
= 1,876 MW x 277kWh/MW x 2,160 hours +
1,876 MW x 330kWh/MW x 2,928 hours +
1,876 MW x 363kWh/MW x 3,672 hours
= 5,437 GWh
Reserve Margin Contribution = RM x C
= 3.64% x 1,876 MW
= 68 MW
Capital Cost =
(Cap2o28 x RF + CCA0n,ci) x C
= ($995/kW x 1.027 + $33/kW) x 1,876MW
= $1,978,741
4-31
-------
Solar Generation
EPA Platform v6 includes solar photovoltaics and solar thermal generation technologies. The following
sections describe four key aspects of the representation of solar generation: solar resource potential,
generation profiles, reserve margin contribution, and capital cost calculation.
Solar Resource Potential: The resource potential estimates for solar photovoltaics and solar thermal
technologies were developed by NREL by model region, state, and resource class. The NREL resource
base for solar photovoltaics is represented by ten resource classes. In EPA Platform v6, the top eight
resource classes are primarily modeled for solar photovoltaics. The NREL resource base for solar
thermal is represented by twelve resource classes. In EPA Platform v6, the top eight resource classes
are modeled for solar thermal. The solar thermal technology has a ten-hour thermal energy storage (TES)
and is considered a dispatchable resource for modeling purposes. These are summarized in Table 4-39
and Table 4-40.
Generation Profiles: Table 4-41 shows the generation profiles for solar photovoltaics units in all model
region, state, and resource combinations. The capacity factors for solar generation that are used in EPA
Platform v6 were obtained from NREL and are shown in Table 4-44 and Table 4-45.
Reserve margin contribution (also referred to as capacity credit): The reserve margin contribution section
for wind units summarizes the approach followed for calculating the reserve margin contribution for solar
photovoltaics units. Table 4-26 presents the reserve margin contributions apportioned to new solar
photovoltaics units in the EPA Platform v6. The solar thermal units are assumed to have 10-hour TES
and are assigned 100% reserve margin contribution.
Table 4-26 Solar Photovoltaic Reserve Margin Contribution by Resource Class in v6
Resource
Class
Vintage #1 (2028-2059)
Vintage #2 (2030-2059)
Vintage #3 (2040-2059)
Class 1
0% -19%
0% -19%
0% - 20%
Class 2
0% - 94%
0% - 98%
0%-100%
Class 3
0% - 93%
0% - 98%
0%-100%
Class 4
0% - 94%
0% - 98%
0%-100%
Class 5
0% - 49%
0% - 52%
0% - 53%
Class 6
0% - 67%
0% - 70%
0% - 72%
Class 7
0% - 71%
0% - 75%
0% - 77%
Class 8
0% - 91%
0% - 95%
0% - 98%
Class 9
0% - 3%
0% - 3%
0% - 3%
Class 10
0% - 56%
0% - 59%
0% - 61%
Capital Costs: Similar to wind units, capital costs for solar units include transmission spur line cost
adders. The resource class level spur line cost curves for each model region and state combination are
aggregated into a seven-step cost curve. Table 4-42 and Table 4-43 illustrate the capital cost adder by
resource and cost class for new solar units.
Geothermal Generation
Geothermal Resource Potential: Twelve model regions in EPA Platform v6 have geothermal potential.
The potential resource in each of these regions is shown in Table 4-27 and is based on NREL ATB 2019.
GEO-Hydro Flash49, GEO-Hydro Binary, GEO-NF EGS Flash, and GEO-NF EGS Binary are the included
technologies.
49 In dual flash systems, high temperature water (above 400DF) is sprayed into a tank held at a much lower pressure
than the fluid. This causes some of the fluid to "flash," i.e., rapidly vaporize to steam. The steam is used to drive a
turbine, which, in turn, drives a generator. In the binary cycle technology, moderate temperature water (less than
4-32
-------
Table 4-27 Regional Assumptions on Potential Geothermal Electric Capacity in v6
IPM Model Region
Capacity (MW)
WEC CALN
498
WECC AZ
26
WECC CO
21
WECC ID
237
WECC 11D
2,832
WECC MT
29
WECC NM
22
WECC NNV
1,421
WECC PNW
633
WECC SCE
496
WECC UT
208
WECC WY
39
Grand Total
6,461
Cost Calculation: EPA Platform v6 does not contain a single capital cost, but multiple geographically
dependent capital costs for geothermal generation. The assumptions for geothermal were developed
using NREL 2019 ATB cost and performance estimates for 152 sites. Both dual flash and binary cycle
technologies were represented. The 152 sites were aggregated into 61 different options based on
geographic location and cost and performance characteristics of geothermal sites in each of the 12
eligible IPM regions where geothermal generation opportunities exist. Table 4-28 shows the potential
geothermal capacity and cost characteristics for applicable model regions.
Table 4-28 Potential Geothermal Capacity and Cost Characteristics by Model Region in v6
Region
Net Capacity (MW)
Capital Cost (2019$/kW)
FOM (2019$/kW-yr)
6
8
15,793
21,606
491
595
11
13,488
385
WEC CALN
29
4,259
123
29
6,161
199
82
25,178
614
333
11,235
214
WECC AZ
26
20,826
577
WECC_CO
8
12
21,628
15,192
596
429
10
17,924
501
14
22,689
612
WECCJD
CO CO
CM CM
19,847
43,097
555
1,067
44
12,753
360
112
9,567
266
74
3,325
114
85
27,086
657
WECC_IID
91
137
5,803
4,600
189
147
257
11,351
208
2,188
4,207
101
WECC_MT
7
22
21,996
17,782
603
497
400DF) vaporizes a secondary, working fluid, which drives a turbine and generator. Due to its use of more plentiful,
lower temperature geothermal fluids, these systems tend to be most cost effective and are expected to be the most
prevalent future geothermal technology.
4-33
-------
Region
Net Capacity (MW)
Capital Cost (2019$/kW)
FOM (2019$/kW-yr)
WECC_NM
9
13
21,542
14,961
594
386
45
15,833
434
50
6,275
190
66
7,541
219
67
19,429
536
77
13,502
392
WECC_NNV
92
93
27,121
3,833
679
128
103
3,233
102
138
9,360
281
148
4,088
137
264
23,460
589
279
4,627
152
6
20,197
581
12
7,984
252
15
16,701
490
15
21,804
599
17
18,588
535
19
16,096
446
WECC_PNW
23
23
13,123
16,899
370
474
41
5,379
176
48
9,807
292
57
12,345
344
101
6,679
205
124
3,270
109
132
7,602
230
25
24,214
628
WECC_SCE
27
155
16,230
11,009
457
200
289
3,233
101
1
31,401
520
WECC_UT
2
86
22,476
3,233
535
111
120
19,296
470
WECC WY
39
14,104
398
Landfill Gas Electricity Generation
Landfill Gas Resource Potential: Estimates of potential electric capacity from landfill gas are based on
the AEO 2019 inventory. EPA Platform v6 represents the "high", "low", and "very low" categories of
potential landfill gas units. The categories refer to the amount and rate of methane production from the
existing landfill site. Table 4-46 summarizes potential electric capacity from landfill gas.
There are several things to note about Table 4-46. The AEO 2019 NEMS region level estimates of the
potential electric capacity from new landfill gas units are disaggregated to IPM regions based on
electricity demand. The limits listed in Table 4-46 apply to the IPM regions indicated in column 1. In EPA
Platform v6, the new landfill gas electric capacity in the corresponding IPM regions shown in column 1
cannot exceed the limits shown in columns 3-5. As noted, the capacity limits for three categories of
potential landfill gas units are distinguished in the table based on the rate of methane production at three
categories of landfill sites: LGHI = high rate of landfill gas production, LGLo = low rate of landfill gas
production, and LGLVo = very low rate of landfill gas production. The values shown in Table 4-46
represent an upper bound on the amount of new landfill capacity that can be added in each of the
4-34
-------
indicated model regions and states for each of the three landfill categories. The cost and performance
assumptions for adding new capacity in each of the three landfill categories are presented in Table 4-15.
Small Hydro
EPA Platform v6 models resource potential from non-powered dams (NPD) and new stream development
(NSD) categories of new small hydro. While NPD are existing dams that do not currently have
hydropower, NSD are greenfield hydropower developments along previously undeveloped waterways.
Table 4-29 and Table 4-30 summarize the assumptions for NPD and NSD.
Table 4-29 Potential Non-Powered Dam in v6
IPM Region
State
Capacity
(MW)
Capacity
Factor (%) -
Winter
Capacity Factor (%)
-Winter Shoulder
Capacity
Factor (%) -
Summer
Capital
Cost (2019
$/kW)
FOM
(2019
$/kW)
ERC REST
TX
338
55.1%
57.5%
48.7%
2,195
16.51
ERC WEST
TX
27
45.0%
53.0%
49.4%
2,191
51.88
FRCC
FL
126
56.6%
60.4%
66.6%
2,336
25.88
MIS AMSO
LA
158
66.8%
61.1%
43.5%
1,646
23.34
MIS AR
AR
786
61.3%
63.7%
53.9%
1,630
11.27
MIS IA
IA
383
49.4%
71.4%
75.5%
1,756
15.61
MIS IL
IL
630
55.1%
71.9%
72.7%
1,548
12.46
MISJNKY
IN
65
68.4%
65.5%
52.2%
2,804
34.89
KY
536
75.2%
68.6%
46.1%
1,308
13.41
MIS LA
LA
643
66.7%
61.0%
43.3%
1,610
12.35
MIS LMI
Ml
24
75.4%
76.5%
60.8%
3,889
54.60
MIS_MAPP
MT
17
42.5%
61.6%
80.2%
2,222
55.55
ND
15
32.2%
59.8%
67.1%
2,622
65.55
MIS MIDA
IA
150
49.4%
71.3%
75.5%
1,761
23.84
Ml
0.02
68.6%
77.9%
72.0%
5,143
128.58
MIS MNWI
MN
123
54.0%
71.8%
74.8%
2,292
26.13
Wl
94
52.1%
74.5%
76.7%
1,921
29.45
MIS_MO
IA
4
49.1%
70.9%
75.3%
1,860
46.50
MO
159
52.7%
71.4%
74.8%
1,456
23.29
MIS MS
MS
102
73.4%
63.1%
45.1%
2,006
28.42
MIS_WOTA
LA
23
66.8%
61.1%
43.5%
1,777
44.42
TX
123
60.4%
59.2%
46.1%
1,501
26.10
MIS_WUMS
Ml
4
71.1%
77.3%
67.8%
4,415
110.38
Wl
111
53.7%
75.4%
77.2%
1,857
27.32
NENG CT
CT
59
74.3%
75.0%
54.7%
3,019
36.55
NENG ME
ME
15
66.7%
73.8%
61.6%
5,040
67.42
MA
53
74.2%
73.5%
51.1%
4,663
38.19
NENGREST
NH
56
70.2%
75.5%
58.3%
3,134
37.45
Rl
11
76.3%
72.3%
48.7%
4,552
77.86
VT
13
69.5%
74.7%
56.3%
3,228
72.42
NY Z A
NY
12
74.2%
72.7%
50.6%
2,371
59.28
NY Z B
NY
8
74.2%
72.7%
50.6%
2,437
60.92
NY Z C&E
NY
66
74.2%
72.7%
50.6%
2,532
34.61
NY Z D
NY
49
74.2%
72.7%
50.6%
2,508
39.65
NY Z F
NY
78
74.2%
72.7%
50.6%
2,550
32.04
NY Z G-l
NY
28
74.2%
72.7%
50.6%
2,341
50.93
MD
13
70.2%
68.5%
49.5%
2,767
69.17
PJM_AP
PA
236
78.3%
71.4%
47.7%
2,042
19.44
VA
3
68.9%
68.9%
50.1%
3,576
89.40
WV
138
73.7%
68.1%
48.1%
1,982
24.78
PJM_ATSI
OH
64
70.2%
67.3%
52.0%
2,793
35.08
PA
43
77.9%
71.4%
48.2%
1,896
42.12
PJM COMD
IL
198
57.5%
72.6%
71.9%
1,868
21.07
PJM Dom
NC
2
68.6%
65.7%
49.4%
2,134
53.36
4-35
-------
Capacity
Capacity
Capital
FOM
Capacity
Factor (%) -
Capacity Factor (%)
Factor (%) -
Cost (2019
(2019
IPM Region
State
(MW)
Winter
-Winter Shoulder
Summer
$/kW)
$/kW)
VA
13
68.9%
68.8%
50.1%
3,025
71.99
DE
1
71.3%
71.7%
56.7%
4,790
119.74
PJM_EMAC
MD
13
72.8%
72.9%
58.5%
2,456
61.41
NJ
17
75.7%
73.6%
56.3%
4,415
63.49
PA
9
74.9%
71.3%
50.7%
2,548
63.69
PJM PENE
PA
316
77.7%
71.4%
48.2%
2,084
17.05
PJM_SMAC
DC
1
72.8%
72.9%
58.5%
3,055
76.37
MD
15
72.5%
72.6%
57.9%
3,182
68.01
IN
8
69.6%
65.8%
53.4%
2,615
65.37
KY
375
74.8%
68.3%
46.5%
1,493
15.77
PJM West
OH
170
70.2%
67.1%
51.1%
2,614
22.55
VA
8
69.2%
68.2%
49.4%
2,544
63.61
WV
37
70.5%
67.0%
46.1%
2,229
45.18
PJM WMAC
PA
49
74.9%
71.2%
50.1%
2,725
39.81
S C KY
KY
134
70.4%
63.5%
40.0%
2,252
25.11
AL
118
74.5%
62.7%
41.3%
1,675
26.59
GA
30
75.8%
71.3%
61.9%
1,815
45.39
KY
1,022
76.6%
69.8%
48.3%
1,194
10.01
S C TVA
MS
94
75.3%
64.0%
43.4%
2,008
29.56
NC
2
72.7%
70.0%
57.4%
3,752
93.79
TN
12
75.4%
66.1%
48.4%
2,390
59.74
VA
1
69.2%
68.2%
49.3%
2,540
63.50
S D AECI
MO
92
53.5%
71.8%
73.1%
1,637
29.84
AL
723
74.5%
63.7%
43.8%
1,362
11.71
S_SOU
FL
11
72.5%
70.7%
64.4%
2,374
59.35
GA
51
75.8%
71.3%
61.9%
1,966
38.93
MS
12
74.1%
63.4%
44.5%
2,030
50.75
GA
0.09
75.8%
71.3%
61.9%
2,241
56.03
S VACA
NC
91
68.9%
66.0%
50.0%
2,416
29.95
SC
43
75.5%
71.9%
62.4%
3,059
41.93
SPP_N
KS
36
40.3%
52.9%
58.5%
2,299
45.64
MO
10
63.9%
63.9%
50.5%
2,551
63.78
SPP NEBR
KS
3
40.3%
52.9%
58.5%
2,476
61.91
SPP SPS
NM
26
40.6%
62.0%
75.7%
2,444
52.62
AR
343
61.3%
63.6%
53.8%
1,567
16.41
LA
24
66.8%
61.1%
43.5%
1,661
41.53
SPP WEST
MO
0.40
53.5%
57.3%
48.4%
2,890
72.25
OK
312
48.5%
57.8%
54.6%
1,869
17.13
TX
20
59.7%
51.5%
35.0%
2,237
55.94
WEC BANC
CA
0.09
62.6%
69.0%
61.6%
3,551
88.78
WEC CALN
CA
111
62.7%
69.0%
61.6%
2,637
27.38
WEC LADW
CA
27
55.6%
72.2%
77.5%
2,051
51.27
WECC AZ
AZ
58
67.3%
73.7%
72.8%
2,234
36.72
WECC CO
CO
146
47.5%
65.5%
80.4%
1,914
24.15
WECC ID
ID
6
65.8%
74.0%
72.1%
3,644
91.11
WECC IID
CA
0.38
55.6%
72.2%
77.5%
1,758
43.94
WECC MT
MT
54
52.8%
66.4%
79.5%
2,914
37.90
WECC_NM
NM
63
37.8%
67.3%
82.1%
2,416
35.49
TX
15
36.6%
67.1%
83.0%
2,514
62.86
WECC NNV
NV
12
50.0%
65.6%
69.2%
4,128
75.57
CA
4
74.8%
76.9%
68.5%
3,338
83.45
WECC_PNW
ID
1
47.5%
64.3%
74.2%
3,071
76.79
OR
87
79.1%
72.2%
56.1%
2,631
30.60
WA
70
83.9%
72.6%
61.4%
2,536
33.69
WECC SCE
CA
34
55.6%
72.2%
77.4%
1,966
46.99
WECC SNV
NV
2
88.1%
84.7%
81.7%
3,609
90.24
WECC UT
UT
29
55.5%
69.2%
78.4%
2,382
50.58
4-36
-------
IPM Region
State
Capacity
(MW)
Capacity
Factor (%) -
Winter
Capacity Factor (%)
-Winter Shoulder
Capacity
Factor (%) -
Summer
Capital
Cost (2019
$/kW)
FOM
(2019
$/kW)
WECC WY
WY
36
43.8%
64.8%
76.2%
2,162
45.59
Table 4-30 Potential New Stream Development in v6
IPM Region
State
Capacity
(MW)
Capacity
Factor (%) -
Winter
Capacity Factor (%) -
Winter Shoulder
Capacity Factor
(%) - Summer
Capital Cost
(2019 $/kW)
FOM
(2019
$/kW)
MIS MO
MO
639
51.7%
69.0%
75.2%
3,567
12.39
NENG ME
ME
406
65.4%
73.2%
62.7%
5,917
15.20
MA
13
75.3%
74.7%
53.6%
5,603
72.74
NENGREST
NH
117
71.1%
76.2%
59.9%
4,979
26.69
VT
58
69.9%
74.9%
57.4%
5,837
36.73
PJM AP
PA
7
74.6%
71.1%
48.3%
4,614
93.17
PJM_EMAC
NJ
27
75.7%
74.2%
56.6%
4,974
51.62
PA
30
74.8%
71.2%
48.3%
4,614
49.68
PJM PENE
PA
239
74.8%
71.2%
48.3%
4,179
19.34
PJM SMAC
MD
79
69.8%
69.7%
50.6%
5,003
31.94
PJM WMAC
PA
622
74.8%
71.2%
48.2%
4,062
12.53
S VACA
SC
51
76.0%
72.3%
61.5%
5,629
38.88
SPP N
MO
350
49.7%
70.0%
79.6%
3,527
16.27
WECC NNV
NV
13
47.5%
65.8%
71.7%
6,731
71.25
WECC_PNW
OR
48
51.3%
72.3%
86.5%
4,585
40.14
WA
394
64.8%
71.0%
72.3%
3,986
15.42
Energy Storage
Energy storage is the capture of energy produced at one time for use at a later time. Presently, the most
common energy storage technologies are pumped storage and lithium-ion battery storage. EPA Platform
v6 includes both existing and new battery storage by IPM region and state. While EPA Platform v6
models existing pumped storage, it does not model new pumped storage options.
The cost and performance assumptions for new 4-hour and 8-hour battery storage units in EPA platform
v6 are based on NREL ATB 2021 and are summarized in Table 4-15. Energy storage options in EPA
Platform v6 are assigned capacity credits that are a function of penetration. Using a heuristic approach, a
capacity credit curve is independently calculated for both 4-hour and 8-hour battery storage options at an
IPM model region level. It estimates how much storage is needed to reduce net peak demand at different
levels of storage penetration. For each model region, 300 storage power capacities (sized from 0 to 30%
of the annual peak in 0.1% increments) are simulated. The amount of stored energy required to reduce
the episodic peak demand by the storage power capacity is determined for each storage power capacity.
The capacity credit is calculated as the ratio between the storage duration (4/8 hours) and the episode
length with the most storage requirement. Hourly load curves adjusted for hourly generation from existing
solar and wind units are used for the analysis. Four steps of storage options are provided in each IPM
region. The first step is assigned 100% capacity credit for 4-hour storage options, and the second step
100% capacity credit for 8-hour storage options. The sum of step widths for the first and second steps
equals the step width of the 100% capacity credit step of 8-hour energy storage options. The other two
steps are assigned lower than 100% capacity credits based on the capacity credit curve for 8-hour
storage options. Table 4-31 summarizes these assumptions.
4-37
-------
Table 4-31 Bounds and Reserve Margin Contribution for Potential (New) Battery Storage in v6
IPM Region
Bound (MW)
Reserve Margin Contribution (%)
Step 1
Step 2
Step 3
Step 4
Step 1
Step 2
Step 3
Step 4
ERC_REST
3,500
11,585
3,500
6,417
100%
100%
0%
0%
ERC_WEST
982
641
196
144
100%
100%
0%
0%
FRCC
5,088
12,556
264
1,916
100%
100%
0%
0%
MIS_AMSO
358
451
684
840
100%
100%
0%
0%
MIS_AR
576
668
1,100
740
100%
100%
0%
0%
MIS_IA
313
828
143
58
100%
100%
0%
0%
MIS IL
665
1,307
745
722
100%
100%
0%
0%
MIS_INKY
995
1,631
1,472
1,870
100%
100%
0%
0%
MIS_LA
727
453
1,108
1,286
100%
100%
0%
0%
MIS_LMI
965
2,773
2,822
675
100%
100%
1%
0%
MIS_MAPP
182
229
39
86
100%
100%
0%
0%
MIS_MIDA
346
1,278
642
1,141
100%
100%
0%
0%
MIS_MNWI
1,727
2,247
836
761
100%
100%
0%
0%
MIS_MO
456
1,302
865
228
100%
100%
0%
0%
MIS_MS
463
450
688
384
100%
100%
0%
0%
MIS_WOTA
470
287
765
869
100%
100%
0%
0%
MIS_WUMS
805
1,293
833
1,379
100%
100%
0%
0%
NENG_CT
1,632
1,026
1,427
1,266
100%
100%
0%
0%
NENG_ME
180
308
113
150
100%
100%
0%
0%
NENGREST
4,558
2,601
2,148
239
100%
100%
0%
0%
NY_Z_A
451
329
18
116
100%
100%
0%
0%
NY_Z_B
433
299
98
62
100%
100%
0%
0%
NY_Z_C&E
725
376
55
220
100%
100%
0%
0%
NY_Z_D
47
107
9
55
100%
100%
0%
0%
NY_Z_F
407
125
141
158
100%
100%
0%
0%
NY_Z_G-I
408
532
398
101
100%
100%
0%
0%
NY_Z_J
824
1,181
824
1,290
100%
100%
0%
0%
NY_Z_K
755
529
215
154
100%
100%
0%
0%
PJM_AP
778
525
364
1,363
100%
100%
0%
0%
PJM_ATSI
766
1,454
1,532
938
100%
100%
0%
0%
PJM_COMD
1,354
3,675
1,185
1,040
100%
100%
0%
0%
PJM_Dom
1,301
663
3,189
2,501
100%
100%
8%
0%
PJM_EMAC
3,132
4,753
2,137
1,032
100%
100%
0%
0%
PJM_PENE
291
233
308
195
100%
100%
0%
0%
PJM_SMAC
1,395
1,170
870
1,065
100%
100%
0%
0%
PJM_West
1,504
3,181
1,633
6,576
100%
100%
0%
0%
PJM_WMAC
950
654
605
1,493
100%
100%
48%
0%
S_C_KY
382
835
148
976
100%
100%
0%
0%
S_C_TVA
3,396
156
3,122
5,035
100%
100%
0%
0%
S_D_AECI
571
346
105
185
100%
100%
0%
0%
S_SOU
5,151
2,546
6,250
3,414
100%
100%
0%
0%
S_VACA
5,828
4,898
4,154
3,720
100%
100%
0%
0%
SPP_N
739
2,770
1,902
129
100%
100%
0%
0%
SPP_NEBR
430
874
68
676
100%
100%
0%
0%
SPP_SPS
723
1,184
77
323
100%
100%
0%
0%
SPP_WAUE
377
191
156
576
100%
100%
0%
0%
SPP_WEST
850
3,530
2,422
928
100%
100%
8%
0%
WEC_BANC
823
266
157
87
100%
100%
0%
0%
WEC_CALN
9,589
2,484
283
220
100%
100%
0%
0%
WEC_LADW
2,499
2,335
500
785
100%
100%
0%
0%
WEC_SDGE
1,249
766
187
211
100%
100%
0%
0%
WECC_AZ
3,827
3,856
2,216
1,611
100%
100%
0%
0%
WECC_CO
1,394
4,359
1,076
229
100%
100%
0%
0%
4-38
-------
IPM Region
Bound (MW)
Reserve Margin Contribution (%)
Step 1
Step 2
Step 3
Step 4
Step 1
Step 2
Step 3
Step 4
WECCJD
1,089
716
324
361
100%
100%
0%
0%
WECC_IID
442
277
1
1
100%
100%
0%
0%
WECC_MT
585
350
100
841
100%
100%
0%
0%
WECC_NM
1,244
1,229
136
250
100%
100%
0%
0%
WECC_NNV
521
528
274
66
100%
100%
0%
0%
WECC_PNW
6,672
3,296
2,331
3,778
100%
100%
0%
0%
WECC_SCE
9,416
3,546
3,516
1,865
100%
100%
0%
0%
WECC_SNV
772
1,131
496
358
100%
100%
0%
0%
WECC_UT
1,276
1,593
661
196
100%
100%
0%
0%
WECC_WY
1,031
426
715
133
100%
100%
0%
0%
CN_AB
663
398
809
2,108
100%
100%
0%
0%
CN_BC
886
450
1,998
886
100%
100%
13%
0%
CN_MB
294
162
441
574
100%
100%
0%
0%
CN_NB
173
31
165
257
100%
100%
0%
0%
CN_NF
70
40
116
25
100%
100%
0%
0%
CN_NL
159
51
214
52
100%
100%
0%
0%
CN_NS
183
181
147
177
100%
100%
0%
0%
CN_ON
947
2,307
3,158
874
100%
100%
7%
0%
CN_PE
54
50
35
5
100%
100%
0%
0%
CN_PQ
5,167
699
932
4,856
100%
100%
0%
0%
CN_SK
212
87
295
586
100%
100%
0%
0%
Multiple U.S. states have instituted standalone targets and mandates for energy storage procurement.
Table 4-33 summarizes the state-specific energy storage mandates in EPA platform v6. Under Assembly
Bill No. 2514 and Assembly Bill No. 2868, the California Public Utilities Commission (CPUC) established
energy storage targets for the state's three investor-owned utilities (lOUs), namely, Pacific Gas and
Electric Company, Southern California Edison, and San Diego Gas & Electric. The California state
mandates are therefore modeled at the utility level.
4.5 Inflation Reduction Act Impacts on New Units
The tax credits for new renewable technology investments provided under the Inflation Reduction Act of
2022 are implemented in EPA Platform v6 as a reduction to capital costs. A production tax credit (PTC) of
1.5 cents/kWh in 1992 dollars or an investment tax credit (ITC) of 30 percent are applied to renewable
technologies, The 1.5 cents PTC and 30 percent ITC is the rate for units that meet the wage and
apprenticeship requirements. While a 10% energy community tax credit is provided to all new energy
storage technologies, the 10% energy community tax credit is prorated based on the share of the total
IPM regional land area that qualifies as an energy community for solar and wind units. Table 4-32
summarizes the PTC/ITC Energy Community Tax Credit increment allocated to each IPM region.
The tax credits are applied to investments made in all run years during the 2028-2055 period, as the
power sector CO2 emissions do not reduce by 75% below the 2021 level of 1,551 million metric tonnes.
4-39
-------
Table 4-32 Energy Community Tax Credit Increment for Solar and Wind Units
IPM Region
PTC/ITC increment (%)
IPM Region
PTC/ITC increment (%)
ERC_PHDL
10
PJM_Dom
2.5
ERC REST
5
PJM EMAC
2.5
ERC_WEST
10
PJM_PENE
10
FRCC
2.5
PJM SMAC
2.5
MIS_AMSO
7.5
PJM_West
5
MIS AR
0
PJM WMAC
7.5
MIS_D_MS
0
S_C_KY
5
MIS IA
2.5
S C TVA
2.5
MIS IL
7.5
S_D_AECI
2.5
MIS INKY
5
S SOU
2.5
MIS_LA
5
S_VACA
2.5
MIS LMI
2.5
SPP N
2.5
MIS_MAPP
2.5
SPP_NEBR
0
MIS_MIDA
2.5
SPP_SPS
10
MIS MNWI
2.5
SPP WAUE
2.5
MIS_MO
2.5
SPP_WEST
2.5
MIS WOTA
7.5
WEC BANC
0
MIS_WUMS
2.5
WEC_CALN
0
NENG CT
0
WEC LADW
0
NENG_ME
0
WEC_SDGE
0
NENGREST
0
WECC AZ
5
NY_Z_A
2.5
WECC_CO
7.5
NY Z B
2.5
WECC ID
0
NY_Z_C&E
2.5
WECC_IID
0
NY Z D
0
WECC MT
2.5
NY_Z_F
0
WECC_NM
7.5
NY Z G-l
0
WECC NNV
2.5
NY_Z_J
0
WECC_PNW
2.5
NY_Z_K
2.5
WECC_SCE
5
PJM AP
7.5
WECC SNV
7.5
PJM_ATSI
5
WECC_UT
5
PJM COMD
2.5
WECC WY
5
4-40
-------
Table 4-33 Energy Storage Mandates in v6
State/Region
Bill
Mandate Type
Mandate Specifications
Implementation
Status
California
Assembly Bill No. 2514
Target in MW
Energy storage target of 1,325 megawatts for Pacific Gas
and Electric Company, Southern California Edison, and San
Diego Gas & Electric by 2020, with installations required no
later than the end of 2024.
2025
LADWP adopted a resolution setting its 2021 energy storage
target at 178 MW.
New York
New York State
Energy Storage Target
Target in MW
1,500 Megawatts by 2025 and up to 3,000 megawatts by
2030.
2025
New Jersey
Assembly Bill No. 3723
Target in MW
600 megawatts of energy storage by 2021 and 2,000
megawatts of energy storage by 2030.
2021
Oregon
House Bill 2193
Target in MWh
per electric
company
An electric company shall procure one or more qualifying
energy storage systems that have the capacity to store at
least five megawatt hours of energy on or before January 1,
2020.
2020
Massachusetts
Chapter 188
Target in MWh
200 Megawatt hour (MWh) energy storage target for electric
distribution companies to procure viable and cost-effective
energy storage systems to be achieved by January 1, 2020.
2020
House Bill 4857
Target in MWh
Goal of 1,000 MWh of energy storage by the end of 2025.
2025
Virginia
Virginia Clean
Economy Act
Target in MW
Requires, by 2035, American Electric Power and Dominion
Energy Virginia to construct or acquire 400 and 2,700
megawatts of energy storage capacity, respectively
2035
Connecticut
Target in MW
300 MW by 2025, 650 MW by 2028, and 1,000 MW by 2031
2025
Minnesota
Target in MW
400 MW by 2030
2030
Nevada
Order No. 44671
Target in MW
1,000 MW by 2030
2030
4-41
-------
4.6 Nuclear Units
4.6.1 Existing Nuclear Units
Population. Plant Location, and Unit Configuration: To provide maximum granularity in forecasting the
behavior of existing nuclear units, all 91 nuclear units in EPA Platform v6 are represented by separate
model plants. As noted in Table 4-7, the 93 nuclear units include 91 currently operating units plus Vogtle
Units 3 and 4, which are scheduled to come online post 2022. All units are listed in Table 4-47. The
population characteristics, plant location, and unit configuration data in the NEEDS v6 were obtained
primarily from EIA Form 860 and AEO 2020.
Capacity: Nuclear units are baseload power plants with high fixed (capital and fixed O&M) costs and
relatively low variable (fuel and variable O&M) costs. Due to their low variable costs, nuclear units are
typically projected to dispatch up to their assumed availability (the maximum extent possible).
Consequently, a nuclear unit's capacity factor is equivalent to its availability. Thus, EPA Platform v6 uses
capacity factor assumptions to define the upper bound on generation from nuclear units. Nuclear
capacity factor assumptions in EPA Platform v6 are based on an Annual Energy Outlook projection
algorithm. The nuclear capacity factor projection algorithm is described below:
• For each reactor, the capacity factor over time depends on the reactor's age.
• Capacity factors increase initially due to learning and decrease in the later years due to aging.
• For individual reactors, vintage classifications (older and newer) are used.
• For the older vintage (starting before 1982) nuclear power plants, the performance peaks at 25 years:
o Before 25 years: Performance increases by 0.5 percentage point per year;
o 25- years: Performance remains flat; and
• For the newer vintage (starting in or after 1982) nuclear power plants, the performance peaks at 30
years:
o Before 30 years: Performance increases by 0.7 percentage points per year;
o 30- years: Performance remains flat; and
• A maximum capacity factor of 90 percent is assumed unless a capacity factor above 90 percent was
observed for the unit. Given that historical capacity factors are above 90 percent, the assumed
annual capacity factors range from 60 percent to 96 percent.
Cost and Performance: Unlike non-nuclear existing conventional units discussed in Section 4.2.7,
emission rates are not needed for nuclear units since there are no SO2, NOx, CO2, or mercury emissions
from nuclear units.
As with other generating resources, EPA Platform v6 uses heat rate, variable O&M costs, and fixed O&M
costs from AEO 2020 to characterize the cost of operating existing nuclear units. In addition, the fixed
O&M costs from the AEO are increased by 20% to reflect general and administrative (G&A) costs. The
data are shown in Table 4-47.
EPA Platform v6 also imposes lifetime extension costs for nuclear units (see Section 4.2.8). Nuclear units
are not assumed to have a maximum lifetime (see Section 3.8).
As nuclear units have aged, some units have been retired from service or are planning to retire over the
modeled time horizon. For a list of operational nuclear units, see the NEEDS v6 database In EPA
Platform v6, the retirement dates of Diablo Canyon units 1 and 2 were extended by 5 years.
4-42
-------
Zero Emission Credit (ZEQ Programs: New York and Illinois passed legislation in 2017 to provide
support to selected existing nuclear units that could be at risk of early closure due to declining profitability.
The New York Clean Energy Standard for a 12-year period creates ZECs that are currently applicable for
Fitzpatrick, Ginna, and Nine Mile Point nuclear power plants. The New York load-serving entities (LSEs)
are responsible for purchasing ZECs equal to their share of the statewide load, providing an additional
revenue stream to the nuclear power plants holding the ZECs. Similar to the New York program, the
Illinois Future Energy Jobs Bill creates a ZEC program covering a 10-year term for Clinton and Quad
Cities nuclear power plants.
EPA Platform v6 implicitly models the effect of ZECs by disabling the retirement options for Fitzpatrick,
Ginna, Nine Mile Point nuclear power plants in the 2028 run years.
New Jersey has established a ZEC program. As a result, Salem Harbor 1 & 2 and Hope Creek nuclear
units are eligible to receive payments during the year of implementation plus the three following years and
may be considered for additional three-year renewal periods thereafter.
Ohio passed House Bill 6 which includes a provision to collect $150 million per year
through 2027 into a Nuclear Generation Fund to be distributed to qualifying nuclear generating units
located in Ohio at a rate of $9 per MWh credit. Due to the ongoing uncertainty of this provision, EPA
Platform v6 does not model the impact of this provision on the Perry and Davis Besse nuclear plants.
Nuclear Retirement Limits: In EPA Platform v6, endogenous retirements of nuclear units are not allowed.
Nuclear units are retired per a predetermined retirement schedule. Single-unit plants owned by regulated
and nonregulated entities and multiple-unit plants owned by nonregulated entities are assumed to have a
lifetime of 60 years. In addition, multiple-unit plants owned by regulated entities are assumed to have a
lifetime of 80 years.
Life-Extension Costs: Attachment 4-1 summarizes the approach to estimating unit-level life extension
costs for existing nuclear units. Unlike other plant types, life-extension costs for nuclear units are
calculated as a function of age and are applied starting in the 2028 run year. The life-extension costs are
calculated as 17 + 1.25 multiplied by the age of the unit before 50 years of age. After the age of 50 years,
the life-extension costs are assumed to be 70 $/kW-yr.
To reflect the improvements made through the life extension investments, the FOM costs are reduced by
25 $/kW-yr starting age of 51 years.
4.6.2 Potential Nuclear Units
The cost and performance assumptions for nuclear potential units that the model has the option to build
are shown in Table 4-12. The cost assumptions are from AEO 2021.
List of tables that are uploaded directly to the web:
Table 4-34 Planned-Committed Units by Model Region in NEEDS for EPA Platform v6 Post-IRA 2022
Reference Case
Table 4-35 Onshore Average Capacity Factor by Wind Class, Resource Class, and Vintage in EPA
Platform v6 Post-IRA 2022 Reference Case
Table 4-36 Onshore Regional Potential Wind Capacity (MW) by Wind Class, Resource Class, and Cost
Class in EPA Platform v6 Post-IRA 2022 Reference Case
Table 4-37 Wind Generation Profiles in EPA Platform v6 Post-IRA 2022 Reference Case (kWh of
Generation per MW of Capacity)
4-43
-------
Table 4-38 Capital Cost Adder (2019$/kW) for New Onshore Wind Plants by Resource and Cost Class in
EPA Platform v6 Post-IRA 2022 Reference Case
Table 4-39 Solar Photovoltaic Regional Potential Capacity (MW) by Resource and Cost Class in EPA
Platform v6 Post-IRA 2022 Reference Case
Table 4-40 Solar Thermal Regional Potential Capacity (MW) by Resource and Cost Class in EPA
Platform v6 Post-IRA 2022 Reference Case
Table 4-41 Solar Photovoltaic Generation Profiles in EPA Platform v6 Post-IRA 2022 Reference Case
(kWh of Generation per MW of Capacity)
Table 4-42 Solar Photovoltaic Regional Capital Cost Adder (2019$/kW) for Potential Units by Resource
and Cost Class in EPA Platform v6 Post-IRA 2022 Reference Case
Table 4-43 Solar Thermal Regional Capital Cost Adder (2019$/kW) for Potential Units by Resource and
Cost Class in EPA Platform v6 Post-IRA 2022 Reference Case
Table 4-44 Solar Photovoltaic Average Capacity Factor by Resource Class and Vintage in EPA Platform
v6 Post-IRA 2022 Reference Case
Table 4-45 Solar Thermal Capacity Factor by Resource Class and Season in EPA Platform v6 Post-IRA
2022 Reference Case
Table 4-46 Potential Electric Capacity from New Landfill Gas Units in EPA Platform v6 Post-IRA 2022
Reference Case (MW)
Table 4-47 Characteristics of Existing Nuclear Units in EPA Platform v6 Post-IRA 2022 Reference Case
Attachment 4-1 Nuclear Power Plant Life Extension Cost Development Methodology in EPA Platform v6
Post-IRA 2022 Reference Case
4-44
-------
5. Emission Control Technologies
This chapter describes the emission control technology assumptions implemented in the EPA Platform v6
Post-IRA 2022 Reference Case (EPA Platform v6). EPA uses retrofit emission control cost models
developed for EPA by the engineering firm Sargent & Lundy. EPA Platform v6 includes assumptions
regarding control options for sulfur dioxide (SO2), nitrogen oxides (NOx), mercury (Hg), carbon dioxide
(CO2), and hydrogen chloride (HCI). The options are listed in Table 5-1. They are available in EPA
Platform v6 for meeting existing and potential federal, regional, and state emission limits. Besides the
options shown in Table 5-1 and described in this chapter, EPA Platform v6 offers other compliance
options for meeting emission limits. These include switching fuel, adjusting the level of dispatch, and
retiring.
Table 5-1 Retrofit Emission Control Options in v6
SO2 Control
Technology
Options
NOx Control
Technology
Options
Mercury Control
Technology Options
CO2 Control
Technology
Options
HCI Control
Technology
Options
Limestone Forced
Oxidation (LSFO)
Scrubber
Selective Catalytic
Reduction (SCR)
System
Activated Carbon
Injection (ACI)
System
CO2 Capture and
Sequestration
Limestone Forced
Oxidation (LSFO)
Scrubber
Lime Spray Dryer
(LSD) Scrubber
Selective Non-
Catalytic Reduction
(SNCR) System
SO2 and NOx Control
Technology Removal
Co-benefits
Coal-to-Gas
Lime Spray Dryer
(LSD) Scrubber
Dry Sorbent
Injection (DSI)
Heat Rate
Improvement
Dry Sorbent
Injection (DSI)
Detailed reports and example calculation worksheets for Sargent & Lundy retrofit emission control cost
models used by EPA are available in Attachment 5-1 through Attachment 5-9.
5.1 Sulfur Dioxide Control Technologies - Scrubbers
Two commercially available Flue Gas Desulfurization (FGD) scrubber technology options for removing the
SO2 produced by coal-fired power plants are offered in EPA Platform v6: Limestone Forced Oxidation
(LSFO) — a wet FGD technology and Lime Spray Dryer (LSD) — a semi-dry FGD technology which
employs a spray dryer absorber (SDA). In wet FGD systems the polluted gas stream is brought into
contact with a liquid alkaline sorbent (typically limestone) by forcing it through a pool of the liquid slurry or
by spraying it with the liquid. In dry FGD systems the polluted gas stream is brought into contact with the
alkaline sorbent in a semi-dry state through use of a spray dryer. The removal efficiency for SDA drops
steadily for coals whose SO2 content exceeds 3 lbs S02/MMBtu, the technology is therefore provided to
only plants which have the option to burn coals with sulfur content no greater than 3 lbs S02/MMBtu. In
EPA Platform v6 when a unit retrofits with an LSD SO2 scrubber, it loses the option of burning certain high
sulfur content coals (see Table 5-2).
The LSFO and LSD SO2 emission control technologies are available to existing unscrubbed units. They
are also available to existing scrubbed units with reported removal efficiencies of less than 50%. Such
units are considered to have an injection technology and are classified as unscrubbed for modeling
purposes in the NEEDS v6 database. The scrubber retrofit costs for these units are the same as those
for regular unscrubbed units retrofitting with a scrubber.
Default SO2 removal rates for wet and dry FGD were based on data reported in EIA 860 (2018). These
default removal rates were the average of all SO2 removal rates for a dry or wet FGD as reported in EIA
860 (2018) for the FGD installation year.
To reduce the incidence of implausibly high outlier removal rates, the following adjustment is made. Units
for which reported EIA Form 860 (2018) SO2 removal rates are higher than the average of the upper
5-1
-------
quartile of SO2 removal rates across all scrubbed units are assigned the upper quartile average. The
adjustment is not made, however, if a unit's reported removal rate was recently confirmed by utility
comments. Furthermore, one upper quartile removal rate is calculated across all installation years and
replaces any reported removal rate that exceeds it no matter the installation year.
Existing units not reporting FGD removal rates in EIA Form 860 (2018) are assigned the default SO2
removal rate for a dry or wet FGD for that installation year.
The FGD removal efficiencies in South Carolina are based on efficiencies realized during the 2015-2018
period. In addition, the S02 rate floor values for existing coal units with FGD's are calculated as follows.
• Dry FGD - minimum (0.08, minimum reported ETS SO2 rate for the 2014-2018 period)
• Wet FGD - minimum (0.06, minimum reported ETS SO2 rate for the 2014-2018 period)
As shown in Table 5-2, for FGD retrofits installed by the model, the assumed SO2 removal rates will be
98% for wet FGD and 95% for dry FGD.
The procedures used to derive the cost of each scrubber type are discussed in detail in the following
sections.
Table 5-2 Retrofit SO2 Emission Control Performance Assumptions in v6
Performance Assumptions
Limestone Forced Oxidation (LSFO)
Lime Spray Dryer (LSD)
Percent Removal*
98%
with a floor of 0.06 Ibs/MMBtu
95%
with a floor of 0.08 Ibs/MMBtu
Capacity Penalty
Calculated based on characteristics of
the unit:
See Table 5-3
Calculated based on characteristics of
the unit:
See Table 5-3
Heat Rate Penalty
Cost (2019$)
Applicability
Units >25 MW
Units > 25 MW
Sulfur Content Applicability
Coals < 3 lbs SCh/MMBtu
Applicable Coal Types
BA, BB, BD, BE, BG, BH, SA, SB, SD,
SE, LD, LE, LG, LH, PK, and WC
BA, BB, BD, BE, SA, SB, SD, SE, LD,
and LE
* If the SO2 permit rate of the unit is lower than the floor rate, the SO2 permit rate is used as the floor rate.
Potential (new) coal-fired units built by IPM are also assumed to be constructed with a wet scrubber
achieving a removal efficiency of 98%. Further, the costs of potential new coal units include the cost of
scrubbers.
5.1.1 Methodology for Obtaining SO2 Controls Costs
Sargent & Lundy's updated performance/cost models for wet and dry SO2 scrubbers are implemented in
EPA Platform v6 to develop the capital, fixed O&M (FOM), and variable O&M (VOM) components of cost.
For details of Sargent & Lundy Wet FGD and SDA FGD cost models, see Attachment 5-1 and Attachment
5-2.
Capacity and Heat Rate Penalties: In IPM the amount of electrical power required to operate a retrofit
emission control device is represented through a reduction in the amount of electricity available for sale to
the grid. For example, if 1.6% of a unit's electrical generation is needed to operate a scrubber, the unit's
capacity is reduced by 1.6%. The reduction in the unit's capacity is called the capacity penalty. At the
same time, to capture the total fuel used in generation both for sale to the grid and for internal load (i.e.,
for operating the control device), the unit's heat rate is scaled up such that a comparable reduction (1.6%
5-2
-------
in the example) in the new higher heat rate yields the original heat rate.50 The factor used to scale up the
original heat rate is called the heat rate penalty. It is a modeling procedure only and does not represent
an increase in the unit's actual heat rate (i.e., a decrease in the unit's generation efficiency).51 In EPA
Platform v6, specific LSFO and LSD heat rate and capacity penalties are calculated for each installation
based on equations from the Sargent & Lundy models that consider the rank of coal burned, its
uncontrolled SO2 rate, and the heat rate of the model plant.
Table 5-3 presents the LSFO and LSD capital, fixed O&M, and variable O&M costs as well as capacity
and heat rate penalties for representative capacities and heat rates.
5.1.2 SO2 Controls for Units with Capacities from 25 MW to 100 MW (25 MW < capacity <100
MW)
In EPA Platform v6, coal units with capacities between 25 MW and 100 MW are offered the same SO2
control options as larger units. However, for modeling purposes, the costs of controls for these units are
assumed to be equivalent to that of a 50 MW for Dry FGD and 100 MW for Wet FGD. These
assumptions are based on several considerations. First, to achieve economies of scale, several units
within this size range are likely to be ducted to share a single common control, so the minimum capacity
cost equivalency assumption, though generic, would be technically plausible. Second, single units within
this size range that are not grouped to achieve economies of scale are likely to switch to a lower sulfur
coal, repower or convert to natural gas firing, use dry sorbent injection, and/or reduce operating hours.
Illustrative scrubber costs for 25-100 MW coal units with a range of heat rates can be found by referring to
the LSFO 100 MW and LSD 100MW "Capital Costs ($/kW)" and "Fixed O&M" columns in Table 5-3. The
Variable O&M cost component, which applies to units regardless of size, can be found in the fifth column
in this table.
50 Mathematically, the relationship ofthe heat rate and capacity penalties (both expressed as positive percentage
values) can be represented as follows:
Heat Rate Penalty =
1
Capacity Penalty
100
51 The NEEDS heat rate is an unmodified, original heat rate to which this retrofit-based heat rate penalty procedure is
applied. The procedure is limited to units at which IPM adds a retrofit in the model.
5-3
-------
Table 5-3 Illustrative Scrubber Costs (2019$) for Representative Capacities and Heat Rates in v6
Scrubber Type
Heat Rate
Capacity
Heat
Variable
Capacity (MW)
(Btu/kWh)
Penalty
(%)
Rate
Penalty
(%)
O&M
(mills/kWh)
100
300
500
700
1000
Capital
Cost
Fixed
O&M
Capital
Cost
Fixed
O&M
Capital
Cost
Fixed
O&M
Capital
Cost
Fixed
O&M
Capital
Cost
Fixed
O&M
($/kW)
($/kW-yr)
($/kW)
($/kW-yr)
($/kW)
($/kW-yr)
($/kW)
($/kW-yr)
($/kW)
($/kW-yr)
LSFO
Minimum Cutoff: > 25 MW
9,000
-1.60
1.63
2.42
949
26.3
689
12.5
594
9.3
539
8.6
486
7.1
Maximum Cutoff: None
10,000
-1.78
1.82
2.67
994
26.7
722
12.9
622
9.6
564
8.9
509
7.4
11,000
-1.96
2.00
2.92
1,036
27.2
752
13.2
649
9.9
588
9.1
531
7.6
LSD
Minimum Cutoff: > 25 MW
9,000
-1.18
1.20
2.79
801
19.2
587
9.6
507
7.3
455
6.2
455
5.7
Maximum Cutoff: None
10,000
-1.32
1.33
3.11
839
19.6
614
9.9
531
7.6
477
6.4
477
5.9
11,000
-1.45
1.47
3.42
875
19.9
640
10.2
554
7.8
497
6.6
497
6.1
Note 1: The above cost estimates assume a boiler burning 3 Ib/MMBtu S02 Content Bituminous Coal for LSFO and 2 Ib/MMBtu S02 Content Bituminous Coal for LSD.
Note 2: The Variable O&M costs in this table do not include the cost of additional auxiliary power (VOMP) component in the Sargent & Lundy cost models. For modeling purposes, IPM reflects the
auxiliary power consumption through capacity penalty.
5-4
-------
5.2 Nitrogen Oxides Control Technology
There are two main categories of NOx reduction technologies: combustion and post-combustion controls.
Combustion controls reduce NOx emissions during the combustion process by regulating flame
characteristics such as temperature and fuel-air mixing. Post-combustion controls operate downstream of
the combustion process and remove NOx emissions from the flue gas. All the technologies included in
EPA Platform v6 are commercially available and currently in use in numerous power plants.
5.2.1 Combustion Controls
EPA Platform v6 does not model combustion control upgrades as a retrofit option. The decision was
based on two considerations, the relatively low cost of combustion controls compared with that of post
combustion NOx controls and the possible impact on model size. EPA identified units in NEEDS that
have not employed state-of-the-art combustion controls. EPA then estimated the NOx rates for such units
based on an analysis of historical rates of units with state-of-the-art NOx combustion controls. Emission
rates provided by state-of-the-art combustion controls are presented in Attachment 3-2.
5.2.2 Post-combustion NOx Controls
EPA Platform v6 provides two post-combustion retrofit NOx control technologies for existing coal units:
Selective Catalytic Reduction (SCR) and Selective Non-Catalytic Reduction (SNCR). Oil/gas steam units,
on the other hand, are provided with only SCR retrofits. NOx reduction in a SCR system takes place by
injecting ammonia (Nhb) vapor into the flue gas stream where the NOx is reduced to nitrogen (N2) and
water (H2O) abetted by passing over a catalyst bed typically containing titanium, vanadium oxides,
molybdenum, and/or tungsten. As its name implies, SNCR operates without a catalyst. In a SNCR
system, a nitrogenous reducing agent (reagent), typically urea or ammonia, is injected into, and mixed
with, hot flue gas where it reacts with the NOx in the gas stream reducing it to nitrogen gas and water
vapor. Due to the presence of a catalyst, SCR can achieve greater NOx reductions than SNCR.
However, SCR costs are higher than SNCR costs.
Table 5-4 summarizes the performance and applicability assumptions for each post-combustion NOx
control technology and provides a cross-reference to information on cost assumptions.
Table 5-4 Retrofit NOx Emission Control Performance Assumptions in v6
Control Performance
Assumptions
Selective Catalytic Reduction
(SCR)
Selective Non-Catalytic Reduction
(SNCR)
Unit Type
Coal
Oil/Gas
Coal
Output Rate
0.05 Ib/MMBtu
0.03 Ib/MMBtu
-
Percent Removal
-
-
Pulverized Coal: 25% (25-200 MW), 20%
(200-400 MW), 15% (>400 MW)
Fluidized Bed: 50%
Rate Floor
-
-
Pulverized Coal: 0.1 Ib/MMBtu
Fluidized Bed: 0.08 Ib/MMBtu
Size Applicability
Units >25 MW
Units > 25 MW
Units >25 MW
Costs (2019$)
See Table 5-5
See Table 5-6
See Table 5-5
5-5
-------
5.2.3 Methodology for Obtaining SCR and SNCR Costs for Coal Steam Units
Sargent & Lundy SCR and SNCR cost models are implemented to develop the capital, fixed O&M, and
variable O&M costs. For details of Sargent & Lundy SCR and SNCR cost models, see Attachment 5-3,
Attachment 5-4, Attachment 5-5, and Attachment 5-6.
In the Sargent & Lundy's cost models for SNCR, the NOx removal efficiency varies by unit size and burner
type as summarized in Table 5-4. Additionally, the capital, fixed, and variable operating and maintenance
costs of SNCR on circulating fluidized bed (CFB) units are distinguished from the corresponding costs for
other boiler types (e.g., cyclone and wall fired). -An air heater modification cost applies for plants that
burn bituminous coal whose SO2 content is 3 Ibs/MMBtu or greater.
Table 5-5 presents the SCR and SNCR capital, fixed O&M, and variable O&M costs as well as capacity
and heat rate penalties for coal steam units of representative capacities and heat rates.
5-6
-------
Table 5-5 Illustrative Post Combustion NOx Control Costs (2019$) for Coal Plants for Representative Sizes and Heat Rates under the
Assumptions in v6
Capacit
Heat
Rate
Penalty
(%)
Variable
O&M
(mills/kWh
\
Capacity (MW)
Heat Rate
(Btu/kWh)
100
300
500
700
1000
Control Type
y
Penalty
(%)
Capita
I Cost
Fixed
O&M
Capita
I Cost
Fixed
O&M
Capita
I Cost
Fixed
O&M
Capita
I Cost
Fixed
O&M
Capita
I Cost
Fixed
O&M
i
($/kW)
($/kW-yr)
($/kW)
($/kW-yr)
($/kW)
($/kW-yr)
($/kW)
($/kW-yr)
($/kW)
($/kW-yr)
SCR
9,000
-0.54
0.54
1.34
430
2.12
352
0.94
326
0.81
311
0.74
298
0.69
Minimum Cutoff: > 25
MW
10,000
-0.56
0.56
1.45
468
2.25
384
1.01
357
0.87
341
0.80
327
0.75
Maximum Cutoff: None
11,000
-0.58
0.59
1.56
505
2.38
417
1.08
388
0.94
371
0.87
355
0.81
SNCR - Tangential,
25% Removal
9,000
1.12
59
0.52
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
Efficiency
Minimum Cutoff: > 25
MW
10,000
-0.05
0.78
1.25
60
0.54
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
Maximum Cutoff: 200
MW
11,000
1.37
62
0.55
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
SNCR - Tangential,
20% Removal
9,000
0.90
N/A
N/A
31
0.28
N/A
N/A
N/A
N/A
N/A
N/A
Efficiency
Minimum Cutoff: > 200
MW
10,000
-0.05
0.63
1.00
N/A
N/A
32
0.28
N/A
N/A
N/A
N/A
N/A
N/A
Maximum Cutoff: 400
MW
11,000
1.10
N/A
N/A
33
0.29
N/A
N/A
N/A
N/A
N/A
N/A
SNCR - Tangential,
15% Removal
9,000
0.67
N/A
N/A
N/A
N/A
23
0.21
19
0.17
16
0.14
Efficiency
Minimum Cutoff: > 400
MW
10,000
-0.05
0.49
0.75
N/A
N/A
N/A
N/A
23
0.21
19
0.17
16
0.14
Maximum Cutoff: None
11,000
0.82
N/A
N/A
N/A
N/A
24
0.22
20
0.17
16
0.14
SNCR - Fluidized Bed
9,000
2.26
47
0.41
25
0.23
19
0.17
16
0.14
13
0.12
Minimum Cutoff: > 25
MW
10,000
-0.05
1.51
2.52
48
0.43
26
0.23
20
0.17
16
0.14
13
0.12
Maximum Cutoff: None
11,000
2.77
49
0.44
27
0.24
20
0.17
17
0.14
14
0.12
Note 1: Assumes Bituminous Coal, NOx rate: 0.5 Ib/MMBtu, and SO2 rate: 2.0 Ib/MMBtu.
Note 2: The Variable O&M costs in this table do not include the cost of additional auxiliary power (VOMP) component in the Sargent & Lundy cost models. For modeling purposes, IPM reflects the auxiliary
power consumption through capacity penalty.
Note 3: Heat rate penalty includes the effect of capacity penalty.
5-7
-------
5.2.4 Methodology for Obtaining SCR Costs for Oil/Gas Steam Units
The cost calculations for SCR described in section 5.2.3 apply to coal units. Table 5-6 presents the SCR and SNCR capital, fixed O&M, and
variable O&M costs as well as capacity and heat rate penalties for oil/gas steam units of representative capacities and heat rates.
Table 5-6 Post-Combustion NOx Controls Costs (2019$) for Oil/Gas Steam for Representative Sizes and Heat Rates under the
Assumptions in v6
Capacity (MW)
Capacity
Penalty
(%)
Heat
Variable
O&M
(mills/kWh)
100
300
500
700
1000
Control Type
Heat Rate
(Btu/kWh)
Rate
Penalty
(%)
Capital
Cost
($/kW)
Fixed
O&M
($/kW-
yr)
Capital
Cost
($/kW)
Fixed
O&M
($/kW-
yr)
Capital
Cost
($/kW)
Fixed
O&M
($/kW-
yr)
Capital
Cost
($/kW)
Fixed
O&M
($/kW-
yr)
Capital
Cost
($/kW)
Fixed
O&M
($/kW-yr)
SCR
Minimum Cutoff: >
25 MW
Maximum Cutoff:
None
Assuming Natural
Gas
9,000
-0.27
0.27
0.64
181
1.25
137
0.49
124
0.38
117
0.33
110
0.29
NOx rate: 0.3
Ib/MMBtu
10,000
-0.28
0.28
0.71
195
1.30
149
0.52
135
0.41
127
0.35
120
0.31
11,000
-0.29
0.29
0.78
209
1.35
161
0.54
146
0.43
138
0.38
131
0.34
SCR
Minimum Cutoff: >
25 MW
Maximum Cutoff:
None
Assuming Oil
9,000
-0.27
0.28
0.65
188
1.27
143
0.51
130
0.40
122
0.34
116
0.30
NOx rate: 0.3
Ib/MMBtu
10,000
-0.29
0.29
0.72
203
1.32
156
0.53
141
0.42
134
0.37
126
0.33
11,000
-0.30
0.30
0.79
217
1.38
168
0.56
153
0.44
145
0.39
137
0.35
5-8
-------
5.3 Biomass Co-firing
Biomass co-firing is provided as an option for those coal-fired units in EPA Platform v6 that per EIA Form
923 had co-fired biomass during the 2015-2019 period. Table 5-7 lists the units provided with the co-
firing option and the limit on share of the biomass co-firing. The remaining coal power plants are not
provided with this choice as logistics and boiler engineering considerations place limits on the extent of
biomass that can be fired. The logistical considerations arise primarily because biomass is only economic
to transport a limited distance from where it is grown due to its relatively low energy density. In addition,
the extent of storage that can be devoted at a power plant to such a fuel is another limiting factor. Boiler
efficiency and other engineering considerations, largely driven by the relatively higher moisture content
and lower heat content of biomass compared to fossil fuel, also plays a role in limiting the potential
adoption of co-firing.
Table 5-7 Coal Units with Biomass Co-firing Option in v6
Plant Name
Unit ID
Biomass Co-Firing Share Limit (%)52
Virginia City Hybrid Energy Center
1
16.3
University of Iowa Main Power Plant
BLR11
45.3
University of Iowa Main Power Plant
BLR10
97.6
Northampton Generating Company LP
BLR1
0.7
TES Filer City Station
2
4.4
TES Filer City Station
1
4.4
Pixelle Specialty Solutions LLC - Spring Grove Facility
5PB036
32.9
Manitowoc
9
18.2
Schiller
6
2.0
Schiller
4
1.9
Hibbing
4
99.7
5.4 Mercury Control Technologies
For any power plant, mercury emissions depend on the mercury content of the fuel used, the combustion
and physical characteristics of the unit, and the emission control technologies deployed. In the absence
of activated carbon injection (ACI), mercury emission reductions below the mercury content of the fuel are
strictly due to characteristics of the combustion process and incidental removal resulting from other
pollution control technologies, e.g., the SO2, NOx, and particulate matter controls. The following
discussion is divided into three parts. Sections 5.4.1 and 5.4.2 explain the two factors that determine
mercury emissions that result from unit configurations lacking ACI. Section 5.4.1 discusses how mercury
content of fuel is modeled. Section 5.4.2 looks at the procedure to capture the mercury reductions
resulting from different unit and (non-mercury) control configurations. Section 5.4.3 explains the mercury
emission control options that are available. Each section indicates the data sources and methodology
used.
5.4.1 Mercury Content of Fuels
Coal
Assumptions pertaining to the mercury content of coal (and the majority of emission modification factors
discussed below in Section 5.4.2) are derived from EPA's "Information Collection Request for Electric
Utility Steam Generating Unit Mercury Emissions Information Collection Effort" (ICR).53 A two-year effort
52 In EPA Platform v6, the limit on biomass co-firing is expressed as the percentage of the facility (ORIS code) level fuel input
that is produced from biomass.
53 Data from the ICR can be found at http://www.epa.qov/ttn/atw/combust/utiItox/mercury.html. In 2009, EPA
collected some additional information regarding mercury through the Collection Effort for New and Existing Coal- and
5-9
-------
initiated in 1998 and completed in 2000, the ICR had three main components: (1) identifying all coal-fired
units owned and operated by publicly-owned utility companies, federal power agencies, rural electric
cooperatives, and investor-owned utility generating companies, (2) obtaining "accurate information on the
amount of mercury contained in the as-fired coal used by each electric utility steam generating unit with a
capacity greater than 25 megawatts electric [MWe]), as well as accurate information on the total amount
of coal burned by each such unit," and (3) obtaining data by coal sampling and stack testing at selected
units to characterize mercury reductions from representative unit configurations.
The ICR resulted in more than 40,000 data points indicating the coal type, sulfur content, mercury content
and other characteristics of coal burned at coal-fired utility units greater than 25 MW. To make this data
usable, these data points were first grouped by IPM coal types and IPM coal supply regions. IPM coal
types divide bituminous, subbituminous, and lignite coal into different grades based on sulfur content.
Oil, natural gas, and waste fuels
Assumptions pertaining to the mercury content for oil, gas, and waste fuels are based on data derived
from previous EPA analysis of mercury emissions from power plants.54 Table 5-8 provides a summary of
the assumptions on the mercury content for oil, gas, and waste fuels.
Table 5-8 Mercury Concentration Assumptions for Non-Coal Fuels in v6
Fuel Type
Mercury Concentration (Ibs/TBtu)
Oil
0.48
Natural Gas
0.00 a
Petroleum Coke
2.66
Biomass
0.57
Municipal Solid Waste
71.85
Geothermal Resource
2.97-3.7
Note:
a The values appearing in this table are rounded to two decimal places. The zero-value shown for natural gas is
based on an EPA study that found a mercury content of 0.000138 Ibs/TBtu. Values for geothermal resources
represent a range.
5.4.2 Mercury Emission Modification Factors
Emission Modification Factors (EMFs) represent the mercury reductions attributable to the specific burner
type and configuration of SO2, NOx, and particulate matter control devices at an electric generating unit.
An EMF is the ratio of outlet mercury concentration to inlet mercury concentration, and depends on the
unit's burner type, particulate control device, post-combustion NOx control and SO2 scrubber control. In
other words, the mercury reduction achieved (relative to the inlet) during combustion and flue-gas
treatment process is (1-EMF), such that the lower the EMF, the greater the mercury reduction. If the EMF
is 0.25, then 25% of the inlet mercury concentration is emitted as outlet mercury concentration, and
therefore the unit has achieved a 75% reduction in mercury that would otherwise be emitted without the
properties influencing the EMF. The EMF varies by the type of coal (i.e., bituminous, subbituminous, and
lignite) used during the combustion process.
Deriving EMFs involves obtaining mercury inlet data by coal sampling and mercury emission data by
stack testing at a representative set of coal units. As noted, EPA's EMFs were initially based on 1999
mercury ICR emission test data. More recent testing conducted by the EPA, DOE, and industry
Oil-Fired Electricity Utility Steam Generating Units (EPA ICR No.2362.01 (OMB Control Number 2060-0631),
however the information collected was not similarly comprehensive and was thus not used to update mercury
assumptions in EPA Platform v6.
54 Analysis of Emission Reduction Options for the Electric Power Industry, Office of Air and Radiation, U.S. EPA,
March 1999.
5-10
-------
participants55 has provided a better understanding of mercury emissions from electric generating units
and mercury capture in pollution control devices. Overall, the 1999 ICR data revealed higher levels of
mercury capture for bituminous coal-fired plants than for subbituminous and lignite coal-fired plants, and
significant capture of ionic Hg in wet-FGD scrubbers. Additional mercury testing indicates that for
bituminous coals, SCR systems can convert elemental Hg into ionic Hg and thus allow easier capture in a
downstream wet-FGD scrubber. This understanding of mercury capture with SCRs is incorporated in
EPA Platform v6 mercury EMFs for unit configurations with SCR and wet scrubbers.
Table 5-9 provides a summary of EMFs used in EPA Platform v6. Table 5-10 provides definitions of
acronyms for existing controls that appear in Table 5-9. Table 5-11 provides a key to the burner type
designations appearing in Table 5-9.
Table 5-9 Mercury Emission Modification Factors Used in v6
Burner
Particulate Control
Post-
Post-
Bituminous
Subbituminous
Lignite
Type
combustion
Control -
NO„
combustion
Control -
so2
EMF
EMF*
EMF
FBC
Cold Side ESP
No SCR
None
0.65
0.1
0.62
FBC
Cold Side ESP
No SCR
Dry FGD
0.64
0.1
1
FBC
Cold Side ESP + FF
No SCR
None
0.05
0.1
0.43
FBC
Cold Side ESP + FF
No SCR
Dry FGD
0.05
0.1
1
FBC
Fabric Filter
No SCR
None
0.05
0.1
0.43
FBC
Fabric Filter
No SCR
Dry FGD
0.05
0.1
0.43
FBC
Hot Side ESP + FGC
No SCR
None
1
0.1
1
FBC
Hot Side ESP + FGC
No SCR
Dry FGD
0.6
0.1
1
FBC
No Control
No SCR
None
1
0.1
1
Non FBC
Cold Side ESP
SCR
None
0.64
0.1
1
Non FBC
Cold Side ESP
SCR
Wet FGD
0.1
0.1
0.56
Non FBC
Cold Side ESP
SCR
Dry FGD
0.64
0.1
1
Non FBC
Cold Side ESP
No SCR
None
0.64
0.1
1
Non FBC
Cold Side ESP
No SCR
Wet FGD
0.05
0.1
0.56
Non FBC
Cold Side ESP
No SCR
Dry FGD
0.64
0.1
1
Non FBC
Cold Side ESP + FF
SCR
None
0.2
0.1
1
Non FBC
Cold Side ESP + FF
SCR
Wet FGD
0.1
0.1
0.56
Non FBC
Cold Side ESP + FF
SCR
Dry FGD
0.05
0.1
1
Non FBC
Cold Side ESP + FF
No SCR
None
0.2
0.1
1
Non FBC
Cold Side ESP + FF
No SCR
Wet FGD
0.05
0.1
0.56
Non FBC
Cold Side ESP + FF
No SCR
Dry FGD
0.05
0.1
1
Non FBC
Cold Side ESP + FGC
SCR
None
0.64
0.1
1
Non FBC
Cold Side ESP + FGC
SCR
Wet FGD
0.1
0.1
0.56
Non FBC
Cold Side ESP + FGC
SCR
Dry FGD
0.64
0.1
1
Non FBC
Cold Side ESP + FGC
No SCR
None
0.64
0.1
1
Non FBC
Cold Side ESP + FGC
No SCR
Wet FGD
0.05
0.1
0.56
Non FBC
Cold Side ESP + FGC
No SCR
Dry FGD
0.64
0.1
1
Non FBC
Cold Side ESP + FGC
+ FF
SCR
None
0.2
0.1
1
Non FBC
Cold Side ESP + FGC
+ FF
SCR
Wet FGD
0.1
0.1
0.56
Non FBC
Cold Side ESP + FGC
+ FF
SCR
Dry FGD
0.05
0.1
1
Non FBC
Cold Side ESP + FGC
+ FF
No SCR
None
0.2
0.1
1
Non FBC
Cold Side ESP + FGC
+ FF
No SCR
Wet FGD
0.05
0.1
0.56
55 For a detailed summary of emissions test data see Control of Emissions from Coal-Fired Electric Utility Boilers: An
Update, EPA/Office of Research and Development, February 2005. The report can be found at
https://cfpub.epa.gov/si/si_public_record_report.cfm?Lab=NRMRL&dirEntryld=219113.
5-11
-------
Burner
Particulate Control
Post-
Post-
Bituminous
Subbituminous
Lignite
Type
combustion
combustion
EMF
EMF*
EMF
Control -
Control -
NO„
so2
Non FBC
Cold Side ESP + FGC
No SCR
Dry FGD
0.05
0.1
1
+ FF
Non FBC
Fabric F
Iter
SCR
None
0.11
0.1
1
Non FBC
Fabric F
Iter
SCR
Wet FGD
0.1
0.1
0.56
Non FBC
Fabric F
Iter
SCR
Dry FGD
0.05
0.1
1
Non FBC
Fabric F
Iter
No SCR
None
0.11
0.1
1
Non FBC
Fabric F
Iter
No SCR
Wet FGD
0.1
0.1
0.56
Non FBC
Fabric F
Iter
No SCR
Dry FGD
0.05
0.1
1
Non FBC
Hot Side ESP
SCR
None
0.9
0.1
1
Non FBC
Hot Side ESP
SCR
Wet FGD
0.1
0.1
1
Non FBC
Hot Side ESP
SCR
Dry FGD
0.6
0.1
1
Non FBC
Hot Side ESP
No SCR
None
0.9
0.1
1
Non FBC
Hot Side ESP
No SCR
Wet FGD
0.05
0.1
1
Non FBC
Hot Side ESP
No SCR
Dry FGD
0.6
0.1
1
Non FBC
Hot Side ESP + FF
SCR
None
0.11
0.1
1
Non FBC
Hot Side ESP + FF
SCR
Wet FGD
0.1
0.1
0.56
Non FBC
Hot Side ESP + FF
SCR
Dry FGD
0.05
0.1
1
Non FBC
Hot Side ESP + FF
No SCR
None
0.11
0.1
1
Non FBC
Hot Side ESP + FF
No SCR
Wet FGD
0.03
0.1
0.56
Non FBC
Hot Side ESP + FF
No SCR
Dry FGD
0.05
0.1
1
Non FBC
Hot Side ESP + FGC
SCR
None
0.9
0.1
1
Non FBC
Hot Side ESP + FGC
SCR
Wet FGD
0.1
0.1
1
Non FBC
Hot Side ESP + FGC
SCR
Dry FGD
0.6
0.1
1
Non FBC
Hot Side ESP + FGC
No SCR
None
0.9
0.1
1
Non FBC
Hot Side ESP + FGC
No SCR
Wet FGD
0.05
0.1
1
Non FBC
Hot Side ESP + FGC
No SCR
Dry FGD
0.6
0.1
1
Non FBC
Hot Side ESP + FGC +
PP
SCR
Dry FGD
0.05
0.1
1
Non FBC
r r
Hot Side ESP + FGC +
pp
No SCR
None
0.11
0.1
1
Non FBC
r r
Hot Side ESP + FGC +
pp
No SCR
Dry FGD
0.05
0.1
1
Non FBC
r r
No Control
SCR
None
1
0.1
1
Non FBC
No Control
SCR
Wet FGD
0.1
0.1
1
Non FBC
No Control
SCR
Dry FGD
0.6
0.1
1
Non FBC
No Control
No SCR
None
1
0.1
1
Non FBC
No Control
No SCR
Wet FGD
0.58
0.1
1
Non FBC
No Control
No SCR
Dry FGD
0.6
0.1
1
Non FBC
PM Scrubber
SCR
None
0.9
0.1
1
Non FBC
PM Scrubber
SCR
Wet FGD
0.1
0.1
1
Note: 2017 annual emissions data suggests that, with subbituminous coal, many configurations are now achieving at least 90%
removal of mercury. This table was updated from previous versions to reflect this recent observation. For 2017 emissions data,
see: https://ampd.epa.gov.
Table 5-10 Definition of Acronyms for Existing Controls
Acronym
Description
ESP
Electrostatic Precipitator - Cold Side
HESP
Electrostatic Precipitator - Hot Side
ESP/O
Electrostatic Precipitator - Other
FF
Fabric Filter
FGD
Flue Gas Desulfurization - Wet
DS
Flue Gas Desulfurization - Dry
SCR
Selective Catalytic Reduction
PMSCRUB
Particulate Matter Scrubber
5-12
-------
Table 5-11 Key to Burner Type Designations in Table 5-9
"PC" refers to conventional pulverized coal boilers. Typical configurations include wall-fired and tangentially
fired boilers (also called T-fired boilers). In wall-fired boilers the burner's coal and air nozzles are mounted on a
single wall or opposing walls. In tangentially fired boilers the burner's coal and air nozzles are mounted in each
corner of the boiler.
"Cyclone" refers to cyclone boilers where air and crushed coal are injected tangentially into the boiler through a
"cyclone burner" and "cyclone barrel" which create a swirling motion allowing smaller coal particles to be burned
in suspension and larger coal particles to be captured on the cyclone barrel wall where they are burned in molten
slag.
"Stoker" refers to stoker boilers where lump coal is fed continuously onto a moving grate or chain, which moves
the coal into the combustion zone in which air is drawn through the grate and ignition takes place. The carbon
gradually burns off, leaving ash which drops off at the end into a receptacle, from which it is removed for
disposal.
"FBC" refers to "fluidized bed combustion" where solid fuels are suspended on upward-blowing jets of air,
resulting in a turbulent mixing of gas and solids and a tumbling action which provides especially effective
chemical reactions and heat transfer during the combustion process.
"Other" refers to miscellaneous burner types including cell burners and arch, roof, and vertically-fired burner
configurations.
5.4.3 Mercury Control Capabilities
EPA Platform v6 offers two options for mercury pollution control: (1) combinations of SO2, NOx, and
particulate controls which deliver mercury reductions as a co-benefit; and (2) Activated Carbon Injection
(ACI), a retrofit option specifically designed for mercury control. The options are discussed below.
Mercury Control through SO2 and NOx Retrofits
Units that install SO2, NOx, and particulate controls reduce mercury emissions as a byproduct of these
retrofits. Section 5.4.2 described how EMFs are used to capture mercury emissions depending on the rank
of coal burned, the generating unit's combustion characteristics, and the specific configuration of SO2,
NOx, and particulate controls (i.e., hot and cold-side electrostatic precipitators (ESPs), fabric filters (also
called "baghouses"), and particulate matter (PM) scrubbers).
Activated Carbon Injection (ACQ
The technology used for mercury control in EPA Platform v6 is Activated Carbon Injection (ACI) downstream
of the combustion process in coal fired units. Sargent & Lundy's updated cost and performance
assumptions for ACI are used (and are described further below).
Three alternative ACI options are represented as capable of providing 90% mercury removal for all possible
configurations of boiler, emission controls, and coal types used in the U.S. electric power sector. The
three ACI options differ, based on whether they are used in conjunction with an electrostatic precipitator
(ESP) or a fabric filter (also called a baghouse). The three ACI options are:
• ACI with Existing ESP
• ACI with Existing Baghouse
• ACI with an Additional Baghouse (also referred to as Toxecon)
In the third option listed above the additional baghouse is installed downstream of the pre-existing
particulate matter device and the activated carbon is injected after the existing controls. This
configuration allows the fly ash to be removed before it is contaminated by the mercury.
5-13
-------
For modeling purposes, EPA assumes that all three configurations use brominated ACI, where a small
amount of bromine is chemically bonded to the powdered carbon, which is injected into the flue gas
stream. EPA recognizes that amended silicates and possibly other non-carbon, non-brominated
substances are in development and may become available as alternatives to brominated carbon as a
mercury sorbent.
The applicable ACI option depends on the coal type burned, its SO2 content, the boiler and particulate
control type, and in some instances consideration of whether an SO2 scrubber (FGD) system and SCR
NOx post-combustion control are present. Table 5-12 shows the ACI assignment scheme used to achieve
90% mercury removal. EPA Platform v6 does not explicitly model ACI retrofit options.
5-14
-------
Table 5-12 Assignment Scheme for Mercury Emissions Control Using Activated Carbon Injection in v6
Air pollution controls
Bituminous Coal
Subbituminous Coal
Lignite Coal
Burner Type
Particulate Control Type
SCR
FGD
ACI
Toxecon
Sorbent Inj
ACI
Toxecon
Sorbent Inj
ACI
Toxecon
Sorbent Inj
System
System
Required?
Required?
Rate
(lb/million
acfm)
Required?
Required?
Rate
(lb/million
acfm)
Required?
Required?
Rate
(lb/million
acfm)
FBC
Cold Side ESP + Fabric Filter without FGC
-
-
Yes
No
2
Yes
No
2
Yes
No
2
FBC
Cold Side ESP without FGC
-
-
Yes
No
5
Yes
No
5
Yes
No
5
FBC
Fabric Filter
-
Dry FGD
No
No
0
Yes
No
2
Yes
No
2
FBC
Fabric Filter
-
-
Yes
No
2
Yes
No
2
Yes
No
2
FBC
Hot Side ESP with FGC
-
-
Yes
Yes
2
Yes
Yes
2
Yes
Yes
2
Non-FBC
Cold Side ESP + Fabric Filter with FGC
-
Dry FGD
Yes
No
2
Yes
No
2
Yes
No
2
Non-FBC
Cold Side ESP + Fabric Filter with FGC
-
-
Yes
No
2
Yes
No
2
Yes
No
2
Non-FBC
Cold Side ESP + Fabric Filter with FGC
-
Wet FGD
Yes
No
2
Yes
No
2
Yes
No
2
Non-FBC
Cold Side ESP + Fabric Filter with FGC
SCR
-
Yes
No
2
Yes
No
2
Yes
No
2
Non-FBC
Cold Side ESP + Fabric Filter with FGC
SCR
Dry FGD
Yes
No
2
Yes
No
2
Yes
No
2
Non-FBC
Cold Side ESP + Fabric Filter with FGC
SCR
Wet FGD
Yes
No
2
Yes
No
2
Yes
No
2
Non-FBC
Cold Side ESP + Fabric Filter without FGC
-
Dry FGD
Yes
No
2
Yes
No
2
Yes
No
2
Non-FBC
Cold Side ESP + Fabric Filter without FGC
-
-
Yes
No
2
Yes
No
2
Yes
No
2
Non-FBC
Cold Side ESP + Fabric Filter without FGC
-
Wet FGD
Yes
No
2
Yes
No
2
Yes
No
2
Non-FBC
Cold Side ESP + Fabric Filter without FGC
SCR
-
Yes
No
2
Yes
No
2
Yes
No
2
Non-FBC
Cold Side ESP + Fabric Filter without FGC
SCR
Dry FGD
Yes
No
2
Yes
No
2
Yes
No
2
Non-FBC
Cold Side ESP + Fabric Filter without FGC
SCR
Wet FGD
Yes
No
2
Yes
No
2
Yes
No
2
Non-FBC
Cold Side ESP with FGC
-
Dry FGD
Yes
Yes
2
Yes
Yes
2
Yes
Yes
2
Non-FBC
Cold Side ESP with FGC
-
-
Yes
Yes
2
Yes
Yes
2
Yes
Yes
2
Non-FBC
Cold Side ESP with FGC
-
Wet FGD
Yes
Yes
2
Yes
Yes
2
Yes
Yes
2
Non-FBC
Cold Side ESP with FGC
SCR
-
Yes
Yes
2
Yes
Yes
2
Yes
Yes
2
Non-FBC
Cold Side ESP with FGC
SCR
Dry FGD
Yes
Yes
2
Yes
Yes
2
Yes
Yes
2
Non-FBC
Cold Side ESP with FGC
SCR
Wet FGD
Yes
Yes
2
Yes
Yes
2
Yes
Yes
2
Non-FBC
Cold Side ESP without FGC
-
-
Yes
No
5
Yes
No
5
Yes
No
5
Non-FBC
Cold Side ESP without FGC
-
Wet FGD
Yes
No
5
Yes
No
5
Yes
No
5
Non-FBC
Cold Side ESP without FGC
SCR
-
Yes
No
5
Yes
No
5
Yes
No
5
Non-FBC
Cold Side ESP without FGC
SCR
Dry FGD
Yes
No
5
Yes
No
5
Yes
No
5
Non-FBC
Cold Side ESP without FGC
SCR
Wet FGD
Yes
No
5
Yes
No
5
Yes
No
5
Non-FBC
Fabric Filter
-
Dry FGD
Yes
No
2
Yes
No
2
Yes
No
2
Non-FBC
Fabric Filter
-
-
Yes
No
2
Yes
No
2
Yes
No
2
Non-FBC
Fabric Filter
-
Wet FGD
Yes
No
2
Yes
No
2
Yes
No
2
Non-FBC
Fabric Filter
SCR
Dry FGD
Yes
No
2
Yes
No
2
Yes
No
2
Non-FBC
Fabric Filter
SCR
-
Yes
No
2
Yes
No
2
Yes
No
2
Non-FBC
Fabric Filter
SCR
Wet FGD
Yes
No
2
Yes
No
2
Yes
No
2
Non-FBC
Hot Side ESP + Fabric Filter with FGC
-
-
Yes
No
2
Yes
No
2
Yes
No
2
Non-FBC
Hot Side ESP + Fabric Filter with FGC
-
Wet FGD
Yes
No
2
Yes
No
2
Yes
No
2
Non-FBC
Hot Side ESP + Fabric Filter with FGC
-
Dry FGD
No
No
0
Yes
No
2
Yes
No
2
Non-FBC
Hot Side ESP + Fabric Filter with FGC
SCR
Wet FGD
Yes
No
2
Yes
No
2
Yes
No
2
Non-FBC
Hot Side ESP + Fabric Filter with FGC
SCR
Dry FGD
No
No
0
Yes
No
2
Yes
No
2
Non-FBC
Hot Side ESP + Fabric Filter with FGC
SCR
-
Yes
No
2
Yes
No
2
Yes
No
2
Non-FBC
Hot Side ESP + Fabric Filter without FGC
-
Dry FGD
No
No
0
Yes
No
2
Yes
No
2
Non-FBC
Hot Side ESP + Fabric Filter without FGC
-
-
Yes
No
2
Yes
No
2
Yes
No
2
Non-FBC
Hot Side ESP + Fabric Filter without FGC
-
Wet FGD
Yes
No
2
Yes
No
2
Yes
No
2
Non-FBC
Hot Side ESP + Fabric Filter without FGC
SCR
Dry FGD
No
No
0
Yes
No
2
Yes
No
2
Non-FBC
Hot Side ESP + Fabric Filter without FGC
SCR
-
Yes
No
2
Yes
No
2
Yes
No
2
5-15
-------
Air pollution controls
Bituminous Coal
Subbituminous Coal
Lignite Coal
Burner Type
Particulate Control Type
SCR
FGD
ACI
Toxecon
Sorbent Inj
ACI
Toxecon
Sorbent Inj
ACI
Toxecon
Sorbent Inj
System
System
Required?
Required?
Rate
(lb/million
acfm)
Required?
Required?
Rate
(lb/million
acfm)
Required?
Required?
Rate
(lb/million
acfm)
Non-FBC
Hot Side ESP + Fabric Filter without FGC
SCR
Wet FGD
Yes
No
2
Yes
No
2
Yes
No
2
Non-FBC
Hot Side ESP with FGC
-
Dry FGD
Yes
Yes
2
Yes
Yes
2
Yes
Yes
2
Non-FBC
Hot Side ESP with FGC
-
-
Yes
Yes
2
Yes
Yes
2
Yes
Yes
2
Non-FBC
Hot Side ESP with FGC
-
Wet FGD
Yes
Yes
2
Yes
Yes
2
Yes
Yes
2
Non-FBC
Hot Side ESP with FGC
SCR
Dry FGD
Yes
Yes
2
Yes
Yes
2
Yes
Yes
2
Non-FBC
Hot Side ESP with FGC
SCR
-
Yes
Yes
2
Yes
Yes
2
Yes
Yes
2
Non-FBC
Hot Side ESP with FGC
SCR
Wet FGD
Yes
Yes
2
Yes
Yes
2
Yes
Yes
2
Non-FBC
Hot Side ESP without FGC
-
Dry FGD
Yes
Yes
2
Yes
Yes
2
Yes
Yes
2
Non-FBC
Hot Side ESP without FGC
-
-
Yes
Yes
2
Yes
Yes
2
Yes
Yes
2
Non-FBC
Hot Side ESP without FGC
-
Wet FGD
Yes
Yes
2
Yes
Yes
2
Yes
Yes
2
Non-FBC
Hot Side ESP without FGC
SCR
Dry FGD
Yes
Yes
2
Yes
Yes
2
Yes
Yes
2
Non-FBC
Hot Side ESP without FGC
SCR
-
Yes
Yes
2
Yes
Yes
2
Yes
Yes
2
Non-FBC
Hot Side ESP without FGC
SCR
Wet FGD
Yes
Yes
2
Yes
Yes
2
Yes
Yes
2
Non-FBC
No Control
-
Dry FGD
Yes
Yes
2
Yes
Yes
2
Yes
Yes
2
Non-FBC
No Control
»
--
Yes
Yes
2
Yes
Yes
2
Yes
Yes
2
Non-FBC
No Control
-
Wet FGD
Yes
Yes
2
Yes
Yes
2
Yes
Yes
2
Non-FBC
No Control
SCR
Dry FGD
Yes
Yes
2
Yes
Yes
2
Yes
Yes
2
Non-FBC
No Control
SCR
-
Yes
Yes
2
Yes
Yes
2
Yes
Yes
2
Non-FBC
No Control
SCR
Wet FGD
Yes
Yes
2
Yes
Yes
2
Yes
Yes
2
Non-FBC
PM Scrubber
-
Dry FGD
Yes
Yes
2
Yes
Yes
2
Yes
Yes
2
Non-FBC
PM Scrubber
-
-
Yes
Yes
2
Yes
Yes
2
Yes
Yes
2
Non-FBC
PM Scrubber
-
Wet FGD
Yes
Yes
2
Yes
Yes
2
Yes
Yes
2
Non-FBC
PM Scrubber
SCR
Dry FGD
Yes
Yes
2
Yes
Yes
2
Yes
Yes
2
Non-FBC
PM Scrubber
SCR
-
Yes
Yes
2
Yes
Yes
2
Yes
Yes
2
Non-FBC
PM Scrubber
SCR
Wet FGD
Yes
Yes
2
Yes
Yes
2
Yes
Yes
2
Note: In the table above "Toxecon" refers to the option described as "ACI System with an Additional Baghouse" and "ACI + Full Baghouse with a Sorbent Injection (Inj) Rate of 2 lbs/million acfm" elsewhere
in this chapter.
5-16
-------
5.4.4 Methodology for Obtaining ACI Control Costs
The ACI model developed by Sargent & Lundy in 2017 assumes that the carbon feed rate dictates the
size of the equipment and resulting costs. The feed rate in turn is a function of the required removal (in
this case 90%) and the type of particulate control device. The model assumes a carbon feed rate of 5
pounds of carbon injected for every 1,000,000 actual cubic feet per minute (acfm) of flue gas would
provide the stipulated 90% mercury removal rate for units shown in Table 5-13 as qualifying for ACI
systems with existing ESP. For generating units with fabric filters a lower injection rate of 2 pound per
million acfm is required. Alternative sets of costs were developed for each of the three ACI options: ACI
systems for units with existing ESPs, ACI for units with existing fabric filters (baghouses), and the
combined cost of ACI plus an additional baghouse for units that either have no existing particulate control
or that require ACI plus a baghouse in addition to their existing particulate control. There are various
reasons that a combined ACI plus additional baghouse would be required. These include situations
where the existing ESP cannot handle the additional particulate load associated with the ACI or where S03
injection is currently in use to condition the flue gas for the ESP. Another cause for combined ACI and
baghouse is use of PRB coal whose combustion produces mostly elemental mercury, not ionic mercury,
due to this coal's low chlorine content.
For the combined ACI and fabric filter option a full-size baghouse with an air-to-cloth (A/C) ratio of 4.0 is
assumed, as opposed to a polishing baghouse with a 6.0 A/C ratio.56
Table 5-13 presents the capital, fixed O&M, and variable O&M costs as well as the capacity and heat rate
penalties for the three ACI options represented in EPA Platform v6. For each ACI option, values are
shown for an illustrative set of generating units with a representative range of capacities and heat rates.
For details of Sargent & Lundy ACI cost model, see Attachment 5-8.
5.5 Hydrogen Chloride (HCI) Control Technologies
The following sub-sections describe how HCI emissions from coal are represented, the emission control
technologies available for HCI removal, and the cost and performance characteristics of these
technologies in EPA Platform v6.
5.5.1 Chlorine Content of Fuels
HCI emissions from the power sector result from the chlorine content of the coal that is combusted by
electric generating units. Data on chlorine content of coals had been collected as part of EPA's 1999
"Information Collection Request for Electric Utility Steam Generating Unit Mercury Emissions Information
Collection Effort" (ICR 1999) described above in section 5.4.1. This data is incorporated into the model to
provide the capability for EPA Platform v6 to project HCI emissions. The procedures used for this are
presented below.
Western subbituminous coal (such as that mined in the Powder River Basin), and lignite coal contain
natural alkalinity in the form of non-glassy calcium oxide (CaO) and other alkaline and alkaline earth
oxides. This fly ash (classified as 'Class C' fly ash) has a natural pH of 9 and higher and the natural
alkalinity can effectively neutralize much of the HCI in the flue gas stream prior to the primary control
device.
56 The air-to-cloth (A/C) ratio is the volumetric flow, (typically expressed in Actual Cubic Feet per Minute, ACFM) of
flue gas entering the baghouse divided by the areas (typically in square feet) of fabric filter cloth in the baghouse.
The lower the A/C ratio, e.g., A/C = 4.0 compared to A/C = 6.0, the greater area of the cloth required and the higher
the cost for a given volumetric flow.
5-17
-------
Table 5-13 Illustrative Activated Carbon Injection (ACI) Costs (2019$) for Representative Sizes and Heat Rates under the Assumptions in v6
Control Type
Heat Rate
(Btu/kWh)
Capacity
Penalty
(%)
Heat
Rate
Penalty
(%)
Variable
O&M cost
(mllls/kWh)
Capacity (MW)
100
300
500
700
1000
Capital
Cost
($/kW)
Fixed
O&M
($/kW-yr)
Capital
Cost
($/kW)
Fixed
O&M
($/kW-yr)
Capital
Cost
($/kW)
Fixed
O&M
($/kW-yr)
Capital
Cost
($/kW)
Fixed
O&M
($/kW-yr)
Capital
Cost
($/kW)
Fixed
O&M
($/kW-yr)
ACI System with an
Existing ESP ACI with a
Sorbent Injection Rate of
5 lbs/million acfm
assuming Bituminous
Coal
9,000
-0.02
0.02
2.36
42.58
0.34
16.74
0.13
10.85
0.09
8.15
0.07
6.02
0.05
10,000
-0.02
0.02
2.62
43.28
0.35
17.01
0.14
11.02
0.09
8.28
0.07
6.11
0.05
11,000
-0.02
0.02
2.88
43.90
0.35
17.25
0.14
11.18
0.09
8.40
0.07
6.20
0.05
ACI System with an
Existing Baghouse ACI
with a Sorbent Injection
Rate of 2 lbs/million
acfm Assuming
Bituminous Coal
9,000
-0.02
0.02
1.69
37.12
0.30
14.60
0.12
9.45
0.08
7.10
0.06
5.24
0.04
10,000
-0.02
0.02
1.88
37.72
0.30
14.82
0.12
9.60
0.08
7.21
0.06
5.33
0.04
11,000
-0.02
0.02
2.07
38.27
0.31
15.04
0.12
9.74
0.08
7.32
0.06
5.40
0.04
ACI System with an
Additional Baghouse
ACI + Full Baghouse
with a Sorbent Injection
Rate of 2 lbs/million
acfm Assuming
Bituminous Coal
9,000
-0.62
0.62
0.50
313.92
1.10
236.83
0.83
210.55
0.74
195.47
0.68
181.05
0.63
10,000
-0.62
0.62
0.56
338.75
1.18
256.69
0.90
228.50
0.80
212.28
0.74
196.74
0.69
11,000
-0.62
0.62
0.62
363.09
1.27
276.17
0.97
246.12
0.86
228.77
0.80
212.12
0.74
Note 1: The above cost estimates assume bituminous coal consumption.
Note 2: The Variable O&M costs in this table do not include the cost of additional auxiliary power (VOMP) component in the Sargent & Lundy cost models. For modeling purposes,
IPM reflects the auxiliary power consumption through capacity penalty.
5-18
-------
Eastern bituminous coals, by contrast, tend to produce fly ash with lower natural alkalinity. Though
bituminous fly ash (classified as 'Class F' fly ash) may contain calcium, it tends to be present in a glassy
matrix and unavailable for acid-base neutralization reactions.
To assess the extent of expected natural neutralization, resulting in large part from the alkalinity of the fly
ash, the 2010 ICR57 data was examined. According to that data, units burning some of the subbituminous
coals without operating acid gas control technology emitted substantially lower HCI than would otherwise
be expected if the emissions were based solely on the chlorine content of those coals. Comparing the
assumed chlorine content of the subbituminous coals modeled in EPA Platform v6 with the estimated
values based on responses to the 2010 ICR supports the EPA Platform v6 assumption that combustion of
subbituminous and lignite coals results in a 95% reduction in HCI emissions relative to the assumed
chlorine content of the coal.
5.5.2 HCI Removal Rate Assumptions for Existing and Potential Units
SO2 emission controls on existing and new (potential) units provide the HCI reductions indicated in Table
5-14. New supercritical pulverized coal units (column 3) that the model builds include FGD (wet or dry)
which is assumed to provide a 99% removal rate for HCI. For existing conventional pulverized coal units
with pre-existing FGD (column 5), the HCI removal rate is assumed to be 5% higher than the reported SO2
removal rate up to a maximum of 99% removal. In addition, for fluidized bed combustion units (column 4)
with no FGD and no fabric filter, the HCI removal rate is assumed to be the same as the SO2 removal rate
up to a maximum of 95%. FBCs with fabric filters are assumed to have an HCI removal rate of 95%.
Table 5-14 HCI Removal Rate Assumptions for Potential (New) and Existing Units in v6
Potential (New)
Existing Units with FGD
Gas
Controls ==>
Ultra-Supercritical
Pulverized Coal with
30%/90% CCS
Fluidized Bed Combustion (FBC)
Conventional Pulverized Coal
(CPC) with Wet or Dry FGD
HCI
Removal
Rate
99%
Without fabric filter:
Same as reported SO2 removal
rate up to a maximum of 95%
With fabric filter: 95%
Reported SO2 removal rate +
5% up to a maximum of 99%
5.5.3 HCI Retrofit Emission Control Options
The retrofit options for HCI emission control are discussed in detail in the following sub-sections and
summarized in Table 5-15.
Wet and Dry FGD
In addition to providing SO2 reductions, wet scrubbers (Limestone Forced Oxidation, LSFO) and dry
scrubbers (Lime Spray Dryer, LSD) reduce HCI as well. For both LSFO and LSD the HCI removal rate is
assumed to be 99% with a floor of 0.001 Ibs/MMBtu. This is summarized in columns 2-5 of Table 5-15.
57 Collection Effort for New and Existing Coal- and Oil-Fired Electricity Utility Steam Generating Units (EPA ICR
No.2362.01 (OMB Control Number 2060-0631)
5-19
-------
Table 5-15 Retrofit HCI and SO2 Emission Control Performance Assumptions in v6
Performance
Assumptions
Limestone Forced Oxidation
(LSFO)
Lime Spray Dryer (LSD)
Dry Sorbent Injection (DSI)
SO2
HCI
SO2
HCI
SO2
HCI
Percent
Removal
98%
with a floor of
0.06
Ibs/MMBtu
99%
with a floor of
0.001
Ibs/MMBtu
95%
with a floor
of
0.08
Ibs/MMBtu
99%
with a floor
of
0.001
Ibs/MMBtu
50%
98%
with a floor of
0.002 Ibs/MMBtu
Capacity
Penalty
Calculated based on
characteristics of the unit:
See Table 5-3
Calculated based on
characteristics of the unit:
See Table 5-3
Calculated based on
characteristics of the unit:
See Excerpt from Table 5-17
Heat Rate
Penalty
Applicability
Units > 25 MW
Units > 25 MW
Units >25 MW
Sulfur Content
Applicability
Coals < 3.0 lbs of
S02/MMBtu
Coals < 2.0 lbs of
S02/MMBtu
Applicable
Coal
Types
BA, BB, BD, BE, BG, BH, SA,
SB, SD, SE, LD, LE, LG, LH, PK,
and WC
BA, BB, BD, BE, SA, SB,
SD, SE, LD, and LE
BA, BB, BD, SA, SB, SD,
and LD
Dry Sorbent Injection
EPA Platform v6 includes dry sorbent injection (DSI) as a retrofit option for achieving (in combination with a
particulate control device) both SO2 and HCI removal. In DSI for HCI reduction, a dry sorbent is injected
into the flue gas duct where it reacts with the HCI and SO2 in the flue gas to form compounds that are then
captured in a downstream fabric filter or electrostatic precipitator (ESP) and disposed of as waste. A
sorbent is a material that takes up another substance by either adsorption on its surface or absorption
internally or in solution. A sorbent may also chemically react with another substance. The sorbent
assumed in the cost and performance characterization discussed in this section is Trona (sodium
sesquicarbonate), a sodium-rich material with major underground deposits found in Sweetwater County,
Wyoming. Trona is typically delivered with an average particle size of 30 jjm diameter but can be reduced
to about 15 jjm through onsite in-line milling to increase its surface area and capture capability. While the
Sargent & Lundy description of the DSI technology includes references to the hydrated lime option, only
the Trona option is implemented in EPA Platform v6.
Removal rate assumptions: The removal rate assumptions for DSI are summarized in Table 5-15. The
assumptions shown in the last two columns of Table 5-15 were derived from assessments by EPA
engineering staff in consultation with Sargent & Lundy. As indicated in this table, the assumed SO2
removal rate for DSI + fabric filter is 50%. The retrofit DSI option on an existing unit with existing ESP is
always provided in combination with a fabric filter (Toxecon configuration).
Methodology for Obtaining DSI Control Costs: The cost and performance model for DSI was updated by
Sargent & Lundy. The model is used to derive the cost of DSI retrofits with two alternatives, associated
particulate control devices, i.e., ESP and fabric filter. The cost model notes that the cost drivers of DSI are
quite different from those of wet or dry FGD. Whereas plant size and coal sulfur rates are key underlying
determinants of FGD cost, sorbent feed rate and fly ash waste handling are the main drivers of the capital
cost of DSI, with plant size and coal sulfur rates playing a secondary role.
Furthermore, the DSI sorbent feed rate and variable O&M costs are based on assumptions that a fabric
filter and in-line Trona milling are used, and that the SO2 removal rate is 50%. The corresponding HCI
removal effect is estimated to be 98% for units with fabric filter.
The cost of fly ash waste handling, which is the other key contributorto DSI cost, is a function of the type of
particulate capture device and the flue gas SO2.
5-20
-------
Total waste production involves the production of both reacted and unreacted sorbent and fly ash.
Sorbent waste is a function of the sorbent feed rate with an adjustment for excess sorbent feed. Use of
sodium-based DSI may make the fly ash unsalable, which would mean that any fly ash produced must be
landfilled along with the reacted and unreacted sorbent waste. Typical ash contents for each fuel are
used to calculate a total fly ash production rate. The fly ash production is added to the sorbent waste to
account for the total waste stream for the variable O&M analysis.
For purposes of modeling, the total variable O&M includes the first two component costs noted in the
previous paragraph, i.e., the costs for sorbent usage and the costs associated with waste production and
disposal.
Table 5-16 presents the capital, fixed O&M, variable O&M costs as well as the capacity and heat rate
penalties of a DSI retrofit for an illustrative and representative set of generating units with the capacities
and heat rates indicated. For details of Sargent & Lundy DSI cost model, see Attachment 5-7.
5.6 Fabric Filter (Baghouse) Cost Development
Fabric filters are not endogenously modeled as a separate retrofit option. In EPA Platform v6, an existing
or new fabric filter particulate control device is a pre-condition for installing a DSI retrofit, and the cost of
these retrofits at plants without an existing fabric filter include the cost of installing a new fabric filter. This
cost was added to the DSI costs discussed in section 5.5. The costs associated with a new fabric filter
retrofit are derived from the cost and performance updated by Sargent & Lundy. Similarly, dry scrubber
retrofit costs also include the cost of a fabric filter.
The engineering cost analysis is based on a pulse-jet fabric filter which collects particulate matter on a
fabric bag and uses air pulses to dislodge the particulate from the bag surface and collect it in hoppers for
removal via an ash handling system to a silo. This is a mature technology that has been operating
commercially for more than 25 years. "Baghouse" and "fabric filters" are used interchangeably to refer to
such installations.
Capital Cost: The major driver of fabric filter capital cost is the air-to-cloth (A/C) ratio. The A/C ratio is
defined as the volumetric flow, (typically expressed in Actual Cubic Feet per Minute, ACFM) of flue gas
entering the baghouse divided by the areas (typically in square feet) of fabric filter cloth in the baghouse.
The lower the A/C ratio, e.g., A/C = 4.0 compared to A/C = 6.0, the greater the area of the cloth required
and the higher the cost for a given volumetric flow. An A/C ratio of 4.0 is used in EPA Platform v6, and it
is assumed that the existing ESP remains in place and active.
Table 5-17 presents the capital, fixed O&M, variable O&M costs for fabric filters as represented in EPA
Platform v6 for an illustrative set of generating units with a representative range of capacities and heat
rates. See Attachment 5-9 for details of the Sargent & Lundy fabric filter PM control cost model.
5-21
-------
Table 5-16 Illustrative Dry Sorbent Injection (DSI) Costs (2019$) for Representative Sizes and Heat Rates in v6
Control
Type
Heat Rate
(Btu/kWh)
so2
Rate
(lb/
MMBtu)
Capacity
Penalty
(%)
Heat
Rate
Penalty
(%)
Variable
O&M
(mills/kWh)
Capacity (MW)
100
300
500
700
1000
Capital
Cost
($/kW)
Fixed
O&M
($/kW-
yr)
Capital
Cost
($/kW)
Fixed
O&M
($/kW-
yr)
Capital
Cost
($/kW)
Fixed
O&M
($/kW-
yr)
Capital
Cost
($/kW)
Fixed
O&M
($/kW-
yr)
Capital
Cost
($/kW)
Fixed
O&M
($/kW-
yr)
DSI
9,000
2.0
-0.37
0.37
6.24
132.4
3.80
60.3
1.40
41.8
0.88
32.9
0.66
25.5
0.48
Assuming
10,000
2.0
-0.41
0.41
6.94
136.5
3.84
62.2
1.41
43.1
0.89
33.9
0.66
26.3
0.49
Bituminous
Coal
11,000
2.0
-0.45
0.45
7.64
140.3
3.87
63.9
1.43
44.3
0.90
34.8
0.67
27.0
0.49
Note 1: A S02 removal efficiency of 50% is assumed in the above calculations.
Note: The Variable O&M costs in this table do not include the cost of additional auxiliary power (VOMP) component in the Sargent & Lundy cost models. For modeling purposes, IPM
reflects the auxiliary power consumption through capacity penalty.
Table 5-17 Illustrative Particulate Controls Costs (2019$) for Representative Sizes and Heat Rates in v6
Coal Type
Heat Rate
(Btu/kWh)
Capacity
Penalty
(%)
Heat
Rate
Penalty
(%)
Variable
O&M
(mills/kWh)
Capacity (MW)
100
300
500
700
1000
Capital
Cost
($/kW)
Fixed
O&M
($/kW-
yr)
Capital
Cost
($/kW)
Fixed
O&M
($/kW-
yr)
Capital
Cost
($/kW)
Fixed
O&M
($/kW-
yr)
Capital
Cost
($/kW)
Fixed
O&M
($/kW-
yr)
Capital
Cost
($/kW)
Fixed
O&M
($/kW-
yr)
9,000
0.06
271
0.9
220
0.8
200
0.7
187
0.7
175
0.6
Bituminous
10,000
-0.60
0.60
0.07
295
1.0
240
0.8
217
0.8
204
0.7
191
0.7
11,000
0.07
319
1.1
259
0.9
235
0.8
220
0.8
206
0.7
Note: The Variable O&M costs in this table do not include the cost of additional auxiliary power (VOMP) component in the Sargent & Lundy cost models. For modeling purposes, IPM
reflects the auxiliary power consumption through capacity penalty.
5-22
-------
5.7 Coal-to-Gas Conversions58
In EPA Platform v6, existing coal plants are given the option to burn natural gas by investing in a coal-to-
gas retrofit. There are two components of cost in this option: boiler modification costs and the cost of
extending natural gas lateral pipeline spurs from the boiler to a natural gas main pipeline. These two
components of cost and their associated performance implications are discussed in the following sections.
5.7.1 Boiler Modifications for Coal-To-Gas Conversions
Enabling natural gas firing in a coal boiler typically involves installation of new gas burners and
modifications to the ducting, windbox (i.e., the chamber surrounding a burner through which pressurized
air is supplied for fuel combustion), and possibly to the heating surfaces used to transfer energy from the
exiting hot flue gas to steam (referred to as the convection pass). It may also involve modification of
environmental equipment. Engineering studies are performed to assess operating characteristics like
furnace heat absorption and exit gas temperature; material changes affecting piping and components like
superheaters, reheaters, economizers, and recirculating fans; and operational changes to soot blowers,
spray flows, air heaters, and emission controls.
The following table summarizes the cost and performance assumptions for coal-to-gas boiler modifications
as incorporated in EPA Platform v6. The values in the table were developed by EPA's engineering staff
based on technical papers59 and discussions with industry engineers familiar with such projects. They
were designed to be broadly applicable across the existing coal fleet (with the exceptions noted in the
table). Coal-to-gas retrofit options in EPA Platform v6 force a permanent change in fuel type from coal to
natural gas. Coal therefore can no longer be fired.
Table 5-18 Cost and Performance Assumptions for Coal-to-Gas Retrofits in v6
Factor
Description
Notes
Applicability:
Existing pulverized coal (PC)
fired and cyclone boiler units of
a size greater than 25 MW:
Not applicable forfluidized bed combustion
(FBC) and stoker boilers.
Capacity Penalty:
None
The furnace of a boiler designed to burn coal is
oversized for natural gas, and coal boilers include
equipment, such as coal mills, that are not needed for
gas. As a result, burning gas should have no impact
on net power output.
Heat Rate
Penalty:
+ 5%
When gas is combusted instead of coal, the stack
temperature is lower and the moisture loss to stack is
higher. This reduces efficiency, which is reflected in
an increase in the heat rate.
Incremental
Capital Cost:
PC units: (2019$)/kW =
305.71*(75/MW)a0.35
Cyclone units: (2019$)/kW =
427.99*(75/MW)a0.35
The cost function covers new gas burners and piping,
windbox modifications, air heater upgrades, gas
recirculating fans, and control system modifications.
Example for 50 MW PC unit:
$/kW = 305.71 *(75/50)A0.35 = 352.32
Incremental
Fixed O&M:
-33% FOM cost of the existing
coal unit
Due to reduced needs for operators, maintenance
materials, and maintenance staff when natural gas
combusted, FOM costs decrease by 33%.
58 As discussed here coal-to-gas conversion refers to the modification of an existing boiler to allow it to fire natural gas.
It does not refer to the addition of a gas turbine to an existing boiler cycle, the replacement of a coal boiler with a new
natural gas combined cycle plant, or to the gasification of coal for use in a natural gas combustion turbine.
59 For an example see Babcock and Wilcox's White Paper MS-14 "Natural Gas Conversions of Exiting Coal-Fired
Boilers" 2010 (https://slidex.tips/download/natural-qas-conversions-of-existinq-coal-fired-boilers).
5-23
-------
Factor
Description
Notes
Incremental
Variable O&M:
-25% VOM cost of the existing
coal unit
Due to reduced waste disposal and miscellaneous
other costs, VOM costs decrease by 25%.
Fuel Cost:
Natural Gas
To obtain natural gas the unit incurs the cost of
extending lateral pipeline spurs from the boiler to the
local transmission mainline. See Section 5.7.2.
NOx emission rate:
50% of existing coal unit NOx
emission rate, with a floor of
0.05 Ibs/MMBtu
The 0.05 Ibs/MMBtu floor is the same as the NOx rate
floor for new retrofit SCR on units burning
subbituminous coal.
SO2 emissions:
Zero
5.7.2 Natural Gas Pipeline Requirements for Coal-To-Gas Conversions
For every individual coal facility in the U.S., the distance and associated cost of constructing pipeline
laterals from each facility to the interstate natural gas pipeline system was determined. Table 5-21 shows
the pipeline costing results for each qualifying existing coal-fired unit in EPA Platform v6.
The lateral costs represent the minimum cost to connect coal power plants to the closest pipelines so that
the plants can use natural gas. The estimated costs include both the cost for the lateral based on its
mileage and size and the compression needed to support the movement of incremental gas needed for
cofiring. They do not, however, include costs for mainline transport beyond those represented by the gas
basis in EPA Platform v6. Thus, it is implicit that all gas needed to fire the plants would be purchased on
a spot basis, and mainline expansion will not be needed to support the transport of incremental gas
associated with cofiring beyond the amounts included in the EPA Base Case. This assumption will hold
so long as the gas needed to support coal-to-gas conversion is not overly concentrated at specific
locations during specific times of the year on gas pipeline systems in those areas are being highly
utilized.
The process for estimating the lateral costs is shown in Figure 1 below. A general description of the
process follows.
Figure 5-1 Process for Lateral Cost Estimation
First, the raw data for pipelines is extracted from National Pipeline Mapping System (NPMS) shapefiles
that contain maps of pipelines throughout the United States, published by the U.S. Department of
Transportation, Pipeline and Hazardous Materials Safety Administration (PHMSA). The NPMS shapefiles
contain thousands of data points that are used to digitally map over 300 pipelines across the U.S. The
5-24
-------
NPMS shapefiles are preprocessed along with ABB Velocity Suite data provided by Hitachi Energy and
with data from the PHMSA Natural and Other Gas Transmission and Gathering Pipeline Systems Annual
Report Part H database to provide pipeline distance and diameter information. This initial step of the
process extracts the necessary raw data from the three different data sets, including the pipeline ID,
location, and diameter of the closest 50 points to each power plant.
The second step defines the necessary data for each power plant considered in the analysis. The most
relevant data includes the location, size, and heat rate of the power plant, the amount of gas needed by
the plant for converting to natural gas, and the lateral cost factor on a dollar-per-inch-mile basis.
The third step defines lateral assumptions for each plant. Broad assumptions have been made as well as
direct assumptions for the configuration of laterals for each power plant included in the analysis. They
include assumptions for the maximum distance that can be considered for each lateral connection and
the potential offtake from each pipeline point.
Using the raw data, power plant information, and the general assumptions from steps 1, 2, and 3, the
fourth step in the process finalizes the distances to and pipeline capacity of the pipelines closest to each
power plant. This step of the process defines values for the matrix of mileage and lateral capacity for up
to 20 laterals for each plant that are subsequently applied in the optimization analysis.
Step five sets up the matrix for all lateral options. The analysis assumes that up to two laterals may be
applied for each power plant, and the capacity and costs for the lateral combinations for each power plant
are defined. The matrix of lateral combinations considers both the distance and size of each lateral.
Diameters from 4" to 32" in 2" increments are considered in the size matrix, yielding a total of up to
43,050 combinations of laterals that could serve each plant. This matrix includes 300 single lateral
options, i.e., 15 different lateral diameters for each of the 20 potential pipeline connections, plus 42,750 2-
lateral options that work through all combinations of lateral diameter and pipeline connections applying
two different laterals to serve each plant.
After the lateral option matrix has been fully populated, the option from the 43,050 combinations that
satisfies a power plant's natural gas need at the lowest cost is selected.
5.8 Retrofit Assignments
In IPM, model plants that represent existing generating units have the option of maintaining their current
system configuration, retrofitting with pollution controls, or retiring. The decision to retrofit or retire is
endogenous to IPM and based on the least cost approach to meeting demand subject to modeled system
and operational constraints. IPM is capable of modeling retrofits and retirements at each applicable
model unit at three different points in time, referred to as three stages. At each stage a retrofit set may
consist of a single retrofit (e.g., LSFO Scrubber) or pre-specified combinations of retrofits (e.g., ACI +
LSFO Scrubber + SCR). In EPA Platform v6, first stage retrofit options are provided to existing coal-
steam and oil/gas steam plants. These plants, along with others such as combined cycle, combustion
turbines, biomass, and nuclear plants, are also given retirement as an option in stage one. Third stage
retrofit options are offered to coal-steam plants only.
5-25
-------
Table 5-19 presents the first stage retrofit options available by plant type. Table 5-20 presents the
second and third stage retrofit options available to coal-steam plants. The cost of multiple retrofits on the
same model plant, whether installed in one or multiple stages, is additive. In EPA Platform v6, projections
of pollution control equipment capacity and retirements are limited to the pre-specified combinations listed
in Table 5-19 and Table 5-20.
Table 5-19 First Stage Retrofit Assignment Scheme in v6
Plant Type
Retrofit Option 1st Stage
Criteria
Coal Steam
Coal Retirement
All coal steam boilers
Coal Steam SCR
All coal steam boilers that are 25 MW or larger and do not possess an
existing SCR control option
Coal Steam SNCR - Non FBC
Boilers
All non FBC coal steam boilers that are 25 MW or larger and do not possess
an existing post combustion NOx control option
Coal Steam SNCR - FBC
Boilers
All coal FBC units that are 25 MW or larger and do not possess an existing
post combustion NOx control option
LSD Scrubber
All unscrubbed coal steam boilers 25 MW or larger and burning less than 3
Ibs/MMBtu SO2 coal
LSFO Scrubber
All unscrubbed and non FBC coal steam boilers 25 MW or larger
CO2 Capture and Storage
All scrubbed coal steam boilers 400 MW or larger
ACI - Hg Control Option
(with and without Toxecon)
All coal steam boilers larger than 25 MW that do not have an ACI and have
an Hg EMF greater than 0.1. Actual ACI technology type will be based on
the boilers fuel and technology configuration. See discussion in Chapter 5.
LSD Scrubber + SCR
Combination options - Individual technology level restrictions apply
LSD Scrubber + SNCR
LSFO Scrubber + SCR
LSFO Scrubber + SNCR
ACI + SCR
ACI + SNCR
ACI + LSD Scrubber
ACI + LSFO Scrubber
ACI + LSD Scrubber + SCR
ACI + LSFO Scrubber + SCR
ACI + LSD Scrubber + SNCR
ACI + LSFO Scrubber + SNCR
DSI
All unscrubbed and non FBC coal steam boilers 25 MW or larger with Fabric
Filter and burning less than 2 Ibs/MMBtu SO2 coal.
DSI + Fabric Filter
All unscrubbed and non FBC-coal steam boilers 25 MW or larger without
Fabric Filter, with CESP or HESP, and burning less than 2 Ibs/MMBtu SO2
coal.
DSI + SCR
Combination options - Individual technology level restrictions apply
DSI + SNCR
ACI + DSI
ACI + DSI + SCR
ACI + DSI + SNCR
Heat Rate Improvement
All coal steam boilers with a heat rate larger than 9,500 Btu/kWh
Coal-to-Gas
All coal steam boilers that are 25 MW or larger
Integrated Gasification Combined Cycle
IGCC Retirement
All integrated gasification combined cycle units
Combined Cycle
CC Retirement
All combined cycle units
CO2 Capture and Storage
All combined cycle sets 400 MW or larger
Combustion Turbine
CT Retirement
All combustion turbine units
Nuclear
5-26
-------
Plant Type
Retrofit Option 1st Stage
Criteria
Nuclear Retirement
All nuclear power units
Oil and Gas Steam
Oil/Gas Retirement
All oil/gas steam boilers
Oil/Gas Steam SCR
All oil/gas steam boilers 25 MW or larger that do not possess an existing post
combustion NOx control option
Table 5-20 Second and Third Stage Retrofit Assignment Schemes in v6
Plant Type
Retrofit Option 1st Stage
Retrofit Option 2nd Stage
Retrofit Option 3rd Stage
Coal Steam
SO2 Control Option
Heat Rate Improvement
HCI Control Option
Heat Rate Improvement
NOx Control Option1
CO2 Control Option
None
Heat Rate Improvement
CO2 Control Option
Coal Retirement
None
NOx Control Option
Heat Rate Improvement
SO2 Control Option2
CO2 Control Option
None
Heat Rate Improvement
CO2 Control Option
Coal Retirement
None
NOx Control Option
Heat Rate Improvement
SO2 Control Option
Heat Rate Improvement
Hg Control Option3
HCI Control Option
Heat Rate Improvement
CO2 Control Option
None
Heat Rate Improvement
CO2 Control Option
Coal Retirement
None
CO2 Control Option4
None
None
NOx Control Option1 + SO2
Control Option2
CO2 Control Option
None
Heat Rate Improvement
CO2 Control Option
Coal Retirement
None
SO2 Control Option
Heat Rate Improvement
NOx Control Option1 + Hg
Control Option3
HCI Control Option
Heat Rate Improvement
CO2 Control Option
None
Heat Rate Improvement
CO2 Control Option
Coal Retirement
None
NOx Control Option
Heat Rate Improvement
SO2 Control Option2 + Hg
CO2 Control Option
None
Control Option3
Heat Rate Improvement
CO2 Control Option
Coal Retirement
None
NOx Control Option1 + SO2
CO2 Control Option
None
Control Option2 + Hg Control
Heat Rate Improvement
CO2 Control Option
Option3
Coal Retirement
None
NOx Control Option
Heat Rate Improvement
HCI Control Option5
SO2 Control Option
Heat Rate Improvement
Heat Rate Improvement
None
Coal Retirement
None
NOx Control Option1 + HCI
Control Option5
SO2 Control Option
Heat Rate Improvement
Heat Rate Improvement
None
Coal Retirement
None
NOx Control Option
Heat Rate Improvement
Hg Control Option3 + HCI
SO2 Control Option
Heat Rate Improvement
Control Option5
Heat Rate Improvement
None
Coal Retirement
None
SO2 Control Option
Heat Rate Improvement
Heat Rate Improvement
None
5-27
-------
Plant Type
Retrofit Option 1st Stage
Retrofit Option 2nd Stage
Retrofit Option 3rd Stage
NOx Control Option1 + HCI
Control Option5 + Hg Control
Option3
Coal Retirement
None
Heat Rate Improvement
NOx Control Option
None
SO2 Control Option
None
HCI Control Option
None
CO2 Control Option
None
Coal Retirement
None
Coal-to-Gas
NOx Control Option
None
Oil/Gas Retirement
None
Coal Retirement
None
None
Combined Cycle
CO2 Capture and Storage
CC Retirement
None
Oil and Gas Steam
NOx Control Option1
Oil/Gas Retirement
None
Oil/Gas Retirement
None
None
Notes:
1"NOx Control Option" implies that a model plant may be retrofitted with one of the following NOx control technologies: SCR,
SNCR - non-FBC, or SNCR - FBC
2"S02 Control Option" implies that a model plant may be retrofitted with one of the following SO2 control technologies: LSFO
scrubber or LSD scrubber
3"Hg Control Option" implies that a model plant may be retrofitted with one of the following activated carbon injection
technology options for reduction of mercury emissions: ACI or ACI + Toxecon
4"C02 Control Option" implies that a model plant may be retrofitted with carbon capture
and storage technology
5"HCI Control Option" implies that a model plant may be retrofitted with a DSI (with milled
Trona)
List of tables and attachments that are directly uploaded to the web:
Table 5-21 Cost of Building Pipelines to Coal Plants in EPA Platform v6 Post-IRA 2022 Reference Case
Attachment 5-1 Wet FGD Cost Methodology
Attachment 5-2 SDA FGD Cost Methodology
Attachment 5-3 SCR Cost Methodology for Coal-Fired Boilers
Attachment 5-4 SCR Cost Methodology for Oil-Gas-Fired Boilers
Attachment 5-5 SNCR Cost Methodology for Coal-Fired Boilers
Attachment 5-6 SNCR Cost Methodology for Oil-Gas-Fired Boilers
Attachment 5-7 DSI Cost Methodology
Attachment 5-8 Hg Cost Methodology
Attachment 5-9 PM Cost Methodology
Attachment 5-10 Combustion Turbine NOx Control Technology Memo
5-28
-------
6. C02 Capture, Storage, and Transport
6.1 CO2 Capture
The EPA Platform v6 Post-IRA 2022 Reference Case (EPA Platform v6) allows for the building of
potential (new) Ultra-Supercritical Coal (USC) and Natural Gas Combined Cycle (NGCC) Electric
Generating Units (EGUs) with Carbon Capture and Storage (CCS) technology.60 CCS is also available as
a retrofit option to existing coal-fired and NGCC EGUs.
6.1.1 CO2 Capture for Potential EGUs
Potential USC EGUs are provided with two CCS options, namely, a 30-percent carbon dioxide (CO2)
capture efficiency option and a 90-percent CO2 capture efficiency option. Potential NGCC EGUs, on
the other hand, are provided with only the 90-percent CO2 capture efficiency option. The CCS cost and
performance assumptions provided in Table 6-1 are based on the Annual Energy Outlook 2021 (AEO
2021). The assumptions represent an amine-based, post-combustion CO2 capture system.
Table 6-1 Cost and Performance Assumptions for Potential USC and NGCC with and without
Carbon Capture in v6
Combined
Cycle -
Single
Shaft
Combined
Cycle -
Multi Shaft
Combined
Cycle with
CCS
Ultra-
supercritical
Coal without
CCS61
Ultra-
supercritical
Coal with
30% CCS
Ultra-
supercritical
Coal with
90% CCS
Size (MW)
418
1083
377
650
650
650
First Year Available
2028
2028
2030
2028
2030
2030
Lead Time (Years)
3
3
3
4
4
4
Availability
87%
87%
87%
85%
85%
85%
Vintag
e #1 (2028)
Heat Rate (Btu/kWh)
Capital (2019$/kW)
Fixed O&M (2019$/kW/yr)
Variable O&M (2019$/MWh)
6,431
1,007
13.99
2.53
6,370
891
12.10
1.86
7,124
2,358
27.38
5.79
8,638
3,454
40.27
4.46
9,751
4,303
53.87
7.02
12,507
5,572
59.08
10.89
Vintag
e #2 (2030)
Heat Rate (Btu/kWh)
Capital (2019$/kW)
Fixed O&M (2019$/kW/yr)
Variable O&M (2019$/MWh)
6,431
977
13.99
2.53
6,370
864
12.10
1.86
7,124
2,268
27.38
5.79
8,638
3,334
40.27
4.46
9,751
4,147
53.87
7.02
12,507
5,363
59.08
10.89
Vintag
e #3 (2035)
Heat Rate (Btu/kWh)
Capital (2019$/kW)
Fixed O&M (2019$/kW/yr)
Variable O&M (2019$/MWh)
6,431
905
13.99
2.53
6,370
800
12.10
1.86
7,124
2,060
27.38
5.79
8,638
3,050
40.27
4.46
9,751
3,780
53.87
7.02
12,507
4,870
59.08
10.89
Vintag
e #4 (2040)
Heat Rate (Btu/kWh)
Capital (2019$/kW)
Fixed O&M (2019$/kW/yr)
Variable O&M (2019$/MWh)
6,431
845
13.99
2.53
6,370
747
12.10
1.86
7,124
1,883
27.38
5.79
8,638
2,810
40.27
4.46
9,751
3,469
53.87
7.02
12,507
4,450
59.08
10.89
Vintag
e #5 (2045)
Heat Rate (Btu/kWh)
Capital (2019$/kW)
Fixed O&M (2019$/kW/yr)
Variable O&M (2019$/MWh)
6,431
789
13.99
2.53
6,370
698
12.10
1.86
7,124
1,719
27.38
5.79
8,638
2,587
40.27
4.46
9,751
3,180
53.87
7.02
12,507
4,061
59.08
10.89
Vintag
e #6 (2050)
Heat Rate (Btu/kWh)
Capital (2019$/kW)
Fixed O&M (2019$/kW/yr)
Variable O&M (2019$/MWh)
6,431
732
13.99
2.53
6,370
648
12.10
1.86
7,124
1,556
27.38
5.79
8,638
2,361
40.27
4.46
9,751
2,889
53.87
7.02
12,507
3,672
59.08
10.89
Vintage #7 (2055)
60 The term carbon capture refers to removing CO2 from the flue gases emitted by fossil fuel-fired EGUs.
61 The ultra-supercritical coal plant without CCS is not compliant with 80 FR 64510.
6-1
-------
Combined
Cycle -
Single
Shaft
Combined
Cycle -
Multi Shaft
Combined
Cycle with
CCS
Ultra-
supercritical
Coal without
CCS61
Ultra-
supercritical
Coal with
30% CCS
Ultra-
supercritical
Coal with
90% CCS
Heat Rate (Btu/kWh)
6,431
6,370
7,124
8,638
9,751
12,507
Capital (2019$/kW)
732
648
1,556
2,361
2,889
3,672
Fixed O&M (2019$/kW/yr)
13.99
12.10
27.38
40.27
53.87
59.08
Variable O&M (2019$/MWh)
2.53
1.86
5.79
4.46
7.02
10.89
6.1.2 C02 Capture for Existing EGUs with CCS retrofit
As noted, EPA Platform v6 offers the option of adding CCS to existing coal-fired and NGCC EGUs as a
retrofit option starting in 2030. The option comes with a CO2 capture efficiency of 90 percent. As in the
case of potential EGUs with CCS, the CO2 capture assumptions for CCS retrofit represent an amine-
based, post-combustion CO2 capture system.
The cost and performance assumptions provided in Table 6-2 are based on the Sargent & Lundy62 cost
algorithm (Attachment 6-1 summarizes the study)63. One issue that must be addressed when installing
an amine-based, post-combustion CO2 capture system is that sulfur oxides (e.g., sulfur dioxide (SO2)
and sulfur trioxide (SO3)) in the EGU flue gas can degrade the amine-based solvent that absorbs the
CO2. Since the amine will preferentially absorb SO2 before CO2, it will be necessary to treat the EGU
flue gas to lower the sulfur oxide concentration to 10 parts per million by volume or less. Meeting this
constraint will require installing a supplemental Wet Flue Gas Desulfurization (FGD) technology or
retrofitting an existing FGD. However, existing FGDs are not retrofitted in v6. The supplemental FGDs
are also not implemented in the v6.
Table 6-2 Performance and Unit Cost (2019 $) Assumptions for Carbon Capture in v6
Technology
Capacity
(MW)
Heat Rate
(Btu/kWh)
Capital
Cost
$/kW)
Fixed
O&M
($/kW-yr)
Variable
O&M
(mills/kWh)2
Capacity
Penalty
(%)
Heat Rate
Penalty
(%)
9,000
1,915
27.9
4.3
27.6
38.1
400
10,000
2,222
31.3
5.0
30.7
44.3
11,000
2,557
35.0
5.8
33.7
50.9
9,000
1,915
23.9
4.3
27.6
38.2
Coal Steam
700
10,000
2,222
27.2
5.0
30.7
44.3
11,000
2,557
30.7
5.8
33.8
51.0
9,000
1,915
22.3
4.3
27.6
38.2
1,000
10,000
2,222
25.5
5.0
30.7
44.3
11,000
2,557
28.9
5.8
33.8
50.9
7,000
1,043
18.1
1.7
15.2
18.0
400
8,000
1,223
20.0
2.0
17.4
21.1
9,000
1,414
22.1
2.3
19.6
24.4
Combined
7,000
1,043
14.7
1.7
15.2
18.0
Cycle
700
8,000
1,224
16.6
2.0
17.4
21.1
9,000
1,414
18.5
2.3
19.6
24.4
1,000
7,000
1,043
13.3
1.7
15.2
18.0
8,000
1,224
15.2
2.0
17.4
21.1
62 Sargent & Lundy. "IPM Model - Updates to Cost and Performance forAPC Technologies - CO2 Reduction Retrofit
Cost Development Methodology." Project 13527-002; January 2023.
63 The capital cost of the CCS retrofit options on coal steam units is assumed to reduce by 5% starting in 2030 and by
10% starting in 2040. Similarly, the capital cost of the CCS retrofit options on combined cycle units is assumed to
reduce by 5%, 7%, 10%, and 15% starting in 2028, 2030, 2035, and 2040 respectively. These reductions are
expected due to lessons learned and experience gained from demonstrations based on 45Q incentivized projects and
movement toward competitive bidding projects with multiple executed projects for each supplier.
6-2
-------
| | 9,000 | 1,414 17.1 23 19_6 24.4 |
Note:
incremental costs are applied to the derated (i.e., after retrofit) capacity.
2The C02 Transportation, Storage, and Monitoring portion of the variable O&M has been removed from
Sargent & Lundy cost method and modeled separately.
The capacity-derating penalty and associated heat rate penalty are an output of the Sargent & Lundy model.
(See Section 5.1.1 for further details.)
6.2 CO2 Storage
The capacity and cost assumptions for CO2 storage in EPA Platform v6 Post-IRA 2022 Reference Case
are the same as in the EPA Platform v6 Summer 2021 Reference Case. The assumptions are based on the
Geosequestration Cost Analysis Tool (GeoCAT) - a spreadsheet model developed for the U.S. EPA by ICF
in support of the U.S. EPA's Underground Injection Control (UIC) Program forCC>2 Geologic Storage
Wells.64 In an earlier version of the EPA Platform v6, the EPA Platform v6 November 2018 Reference
Case, ICF updated the major cost components in the GeoCAT model, including revising onshore and
offshore injection and monitoring costs to reflect 2016 industry drilling, equipment, and service costs.65 In
addition to updating costs, ICF updated storage capacity, well injectivity, and other assumptions by state
and offshore area using data from the research program conducted at DOE/NETL. Assumptions for the
amount of carbon dioxide injected for enhanced oil recovery (EOR) was updated using 1972 to 2016
performance data for U.S. carbon dioxide miscible flood projects.
The GeoCAT model combines detailed characteristics of sequestration capacity by state and geologic
setting for the U.S. with costing algorithms for individual components of CO2 geologic sequestration.
The model outputs are regional sequestration cost curves that indicate how much potential storage
capacity is available at different lifecycle CO2 storage cost points in units of dollars per metric ton
stored.
The GeoCAT model includes three modules:
i) A unit cost specification module
ii) A project scenario costing module
iii) A geologic and regional cost curve module
The unit cost specification module includes data and assumptions for 120 cost elements falling within the
following categories:
i) Geologic site characterization
ii) Area of review and corrective action (including fluid flow and reservoir modeling during and
after injection and identification, evaluation, and remediation of existing wells within the area of
review)
iii) Injection well and other facilities construction
iv) Well operation
v) Monitoring the movement of CO2 in the subsurface
vi) Mechanical integrity testing
64 Federal Requirements Under the UIC Program for CO2 Geologic Sequestration Wells, Federal Register, December
10, 2010 (Volume 75, Number 237), pages 77229-77303.
65 The major data sources for updating costs was the Bureau of Labor Statistics (BLS) Producers Price Index (PPI)
for various products and services related to oil and gas well drilling (https://vwwy.bls.gov/ppi/), the "Joint Association
Survey of Drilling Costs" published by the American Petroleum Institute (http://vwwy.api.org/products-and-
services/statistics#tab overview). and the "Well Cost Study" published by the Petroleum Services Association of
Canada (https://wvw.psac.ca/resources/well-cost-studv-overview/).
6-3
-------
vii) Financial responsibility (to maintain sufficient resources for activities related to closing and
remediation of the site)
viii) Post injection site care
ix) Site closure
x) General and administrative
Of the ten cost categories for geologic CO2 sequestration listed above, the largest cost drivers (in
roughly descending order of magnitude) are well operation, injection well and other facilities construction,
and monitoring the movement of CO2 in the subsurface. The cost estimates are consistent with the
requirements for geologic storage facilities under the UIC Class VI rule66 and Greenhouse Gas (GhG)
Reporting Program Subpart RR67. The price of oil assumed for the calculation of EOR economics is
$75/barrel.
The costs derived in the unit cost specification module are used in the GeoCAT project scenario
costing module to develop commercial scale costs for eight sequestration scenarios compliant with UIC
Class VI standards and GhG Reporting Program Subpart RR:
i) Deep saline formations
ii) Depleted gas fields
iii) Depleted oil fields
iv) Enhanced oil recovery
v) Enhanced coal bed methane recovery
vi) Enhanced shale gas
vii) Basalt storage
viii) Unmineable coal seams
EPA's GeoCAT application for CO2 sequestration includes only storage capacity for the first four
sequestration scenarios. The last four reservoir types are not included because they are not considered
technically mature enough to allow CO2 storage in the foreseeable future.
The current GeoCAT model includes the DOE analysis of the lower-48 states CO2 sequestration
capacities from the "Carbon Sequestration Atlas of the United States and Canada Version 5."68 ICF
enhanced these assessments to include additional details needed for economic modeling such as the
distribution of capacity by state, drilling depth, injectivity, etc. The geologic and regional cost curve
module applies regionalized unit cost factors to these geologic characterizations to develop regional
geologic storage cost curves.69 The analysis of storage volumes is carried out by regional carbon
sequestration partnerships as overseen by NETL in Morgantown, West Virginia. State-level onshore
and offshore capacity volumes are reported for storage in oil and gas reservoirs and deep saline
formations. The great majority of storage volume is in deep saline formations, which are present in
many states and in most states with oil and gas production. In the version of the Atlas used here,
66 Supra Note 59.
67 Title 40 of the Code of Federal Regulations (CFR), Part 98 (Mandatory GhG Reporting), Subpart RR (Geologic
Sequestration of CO2). See https://ecfr.io/Title-40/sp40.23.98.rr.
68 Carbon Sequestration Atlas of the United States and Canada - Version 5 (2015), U.S. Department of Energy,
National Energy Technology Laboratory, Morgantown, WV https://www.netl.dpe.gpv/researoh/coal/oarbon-
storage/atlasv. Accessed mid-October 2016 with data updates through 2015.
69 Detailed discussions of the GeoCAT model and its application for EPA can be found in U.S. Environmental
Protection Agency, Office of Water, "Geologic CO2 Sequestration Technology and Cost Analysis, Technical Support
Document" (EPA 816-B-08-009) June 2008, https://www.epa.gov/sites/production/files/201S-
07/documents/support uic co2 technologyandcostanalvsis.pdf and Harry Vidas, Robert Hugman and Christa Clapp,
"Analysis of Geologic Sequestration Costs for the United States and Implications for Climate Change Mitigation,"
Science Digest, Energy Procedia, Volume 1, Issue 1, February 2009, Pages 4281-4288. Available online at
https://vww.sciencedirect.com/science/article/pii/S18766102090Q8832.
6-4
-------
offshore storage volumes have also been broken out by DOE into the Gulf of Mexico, Atlantic, and
Pacific Outer Continental Shelf (OCS) regions. ICF carried out a separate analysis to break out CO2
EOR storage potential from the total potential in oil and gas reservoirs reported in NATCARB.
Efficiency Assumptions for EOR Uses of CO2
Relying on performance data from 1972 to 2016, the geologic storage cost curve for EOR is based on an
average EOR efficiency of 10 thousand cubic feet of CO2 per incremental barrel of crude oil (Mcf/bbl).
The NETL CO2 EOR Primer70 shows that from the start of CO2 floods in 1972 to 2008 the average
efficiency was 7.66 Mcf/bbl. Data for the most recent seven year has shown a lower average efficiency of
over 10.32 Mcf/bbl. Taken together, the data implies an average of 8.62 Mcf/bbl for all years from 1972
to 2016.
Historical CO2 EOR: 1972-2008
Billion cubic feet of CO2 11,000
Million barrels of crude oil 1,437
Mcf/barrel 7.66
Source: NETL, "Carbon Dioxide Enhanced Oil
Recovery", 2010
Historical CO2 EOR: 2009-2016
Billion cubic feet of CO2 8,339
Million barrels of crude oil 808
Mcf/barrel 10.32
Source: ICF estimates based on EPA GHG Inventory
and Oil & Gas Journal Annual EOR Survey
Historical CO2 EOR: 1972-2016
Billion cubic feet of CO2 19,339
Million barrels of crude oil 2,244
Mcf/barrel 8.62
Source: Sum of prior two tables
The average of all historical and ongoing EOR projects through the end of their lifetimes is likely to
exceed 9.0 Mcf/bbl as they continue to operate at ratios above 10 Mcf/bbl.71 ICF has chosen a calibration
point of 10 Mcf/bbl for the average of potential future CO2 EOR under the belief that the quality of future
projects would likely be worse (i.e., require more CO2 per unit of incremental oil production) than historical
projects. The revised average efficiency value of 10 Mcf/bbl is approximately 15 percent higher than the
original version of GeoCAT, which was calibrated to the older historical data.
The results of the project scenario costing module are taken as inputs into the geologic and regional
cost curve module of GeoCAT, which generates national and regional cost curves indicating the volume
of sequestration capacity in each region and state in the U.S. as a function of total cost per ton of CO2
including all capital and operating costs. The result is a database of sequestration capacity by state,
geologic reservoir type, and cost step.
Table 6-3 shows the NATCARB V storage volumes for the U.S. Lower-48 as allocated to GeoCAT
categories. Total Lower-48 capacity is assessed at 8,216 gigatonnes. There are no volumes in the
current model for potential storage in depleted gas field reservoirs because these are not reported in
NATCARB.
70 National Energy Technology Laboratory, "Carbon Dioxide Enhanced Oil Recovery", 2010,
https://vww.netl.doe.qov/file%20library/research/oil-qas/C02 EOR Primer.pdf
71 For example, assuming an average of 10 years of future operation at the 2016 ratios leads to a lifetime average for
all historical and ongoing CO2 EOR project of 9.09 Mcf/bbl.
6-5
-------
For EPA Platform v6, GeoCAT represents storage opportunities in 37 of the lower 48 continental
states.72 Louisiana and Texas have both onshore and offshore state-level storage cost curves. In
addition, because NATCARB does not provide state-level data, there are multi-state Atlantic offshore and
Pacific offshore storage cost curves. The result is 41 storage cost curves shown in Table 6-4.
Table 6-3 Lower-48 C02 Sequestration Capacity by Region (Gigatonnes) in v6
Offshore Allocation in GeoCAT
Onshore
Offshore
Total
Louisiana
Texas
GOM Total
Pacific
Atlantic
Total
C02 Enhanced Oil Recovery
Low
11.2
1.1
12.3
Mid
15.0
1.5
16.4
1.5
0.0
1.5
0.0
0.0
1.5
High
22.5
2.2
24.7
Depleted Oil
Low
128.0
11.8
139.8
Mid
170.7
15.7
186.4
12.7
3.0
15.7
0.1
0.0
15.7
High
256.0
23.6
279.6
Unmineable Coal
Low
47.8
2.0
49.8
Mid
63.7
2.6
66.4
0.0
0.0
0.0
2.6
0.0
2.6
High
95.6
4.0
99.5
Saline
Low
4,252
1,708
5,960
Mid
5,669
2,277
7,947
1,240
798
2,038
37
202
2,277
High
12,477
3,416
15,893
Totals
2,297
Low
4,439
1,723
6,162
Mid
5,919
2,297
8,216
1,254
801
2,055
40
202
High
12,851
3,446
16,297
Low
139.2
12.9
152.1
Mid
185.6
17.2
202.8
14.16
2.97
17.13
0.05
0.00
High
278.5
25.8
304.2
Oil Subtotal
17.18
Note: Individual values may not sum to reported totals due to rounding.
The cost curves in Table 6-4 are in the form of step functions. In any given year within the IPM model, a
specified amount of storage is available at a particular step price until either the annual storage limit or
the total storage capacity is reached. In determining whether the total storage capacity has been
reached, the model tracks the cumulative storage used up through the current year. Once the
cumulative storage used equals the total storage capacity at that price step, no more storage is available
going forward at that particular step price and, so, higher priced steps must be used.
C02 storage opportunities are relevant not just to power sector sources, but also to sources in other
industrial sectors. Therefore, before being incorporated as a supply representation into EPA Platform
v6, the original C02 storage capacity in each storage region was reduced by an estimate of the
storage that would be occupied by C02 generated by other industrial sector sources at the relevant
level of cost effectiveness (represented by $/ton C02 storage cost).
To do this, ICF first estimated the level of industrial demand for C02 storage in each C02 storage
region in a scenario where the value of abating C02 emissions is assumed to be $50 per ton (this
abatement value is relevant not only to willingness to pay for storage but also for the cost of capture
72 The states without identified storage opportunities in EPA Platform v6 are Connecticut, Iowa, Maine,
Massachusetts, Minnesota, Nevada, New Hampshire, New Jersey, Rhode Island, Vermont, and Wisconsin. These
states were either not assessed or were found to not have storage opportunities in NATCARB for the four
sequestration scenarios included in EPA's inventory, (i.e., deep saline formations, depleted gas fields, depleted oil
fields, and enhanced oil recovery).
6-6
-------
and transportation of the abated CO2).73 The quantity of industrial sequestration economic at $50/ton
represent the "high quality" industrial sources that have high CO2 purity and would be easiest to
capture, rehydrate, and compress. They are made up of ethanol plants, hydrogen production at
refineries and merchant plants and gas processing plants where CO2 is removed from the natural gas.
This amount was calculated as 128 million tons per year.
Then, for each region, ICF calculated the ratio of the industrial demand to total storage capacity
available for a storage price of less than zero dollars per ton (that is, the parts of storage cost curves
made up of EOR opportunities where the benefit of incremental oil production exceeded the storage
costs). An upper limit of $0.00 per ton was chosen under the belief that the earliest uses of CO2 from
industrial sources most likely would continue the current practice of targeting EOR opportunities. Converting
this quantity of capacity reserved for industrial CCS to a percent value and subtracting from 100 percent,
ICF obtained the percent of storage capacity available to the electricity sector at less than zero dollars
per ton. Finally, the Annual Step Bound (MMTons) and Total Storage Capacity (MMTons) was
multiplied by this percentage value for each step below zero dollars74 in the cost curves for the region to
obtain the reduced storage capacity that went into the storage cost curves for the electric sector in EPA
Platform v6. Thus, the values shown in Table 6-4 represent the storage available specifically to the
electric sector after subtracting an amount that might be used by the industrial sector.
The price steps in the Table 6-4 are the same from region to region. (That is, STEP9 [column 2] has a
step cost value of $9.64/Ton [column 3] across all storage regions [column 1], This across-region
price equivalency holds for every step.) However, the amount of storage available in any given year
(labeled Annual Step Bound (MMTons) in column 4) and the total storage available over all years
(labeled Total Storage Capacity (MMTons) in column 5) vary from region to region. In any given
region, the cost curves are the same for every run year, indicating that over the modeling time horizon
no new storage is being identified to augment the current storage capacity estimates. This
assumption is not meant to imply that no additional potential storage capacity could be identified by
NATCARB or another organization. Such future capacity discoveries could be represented in the
model if the model runs exhaust key components of the currently estimated storage capacity.
6.3 CO2 Transport
Each of the 64 IPM model regions can send C02 to the 41 regions represented by the storage cost curves
in Table 6-4. The associated transport costs (in 2019$/Ton) are shown in Table 6-5. For the model, ICF
has also updated assumptions about the costs of CO2 pipelines. These costs were derived by first
calculating the pipeline distance from each of the C02 Production Regions to each of the C02 Storage
Regions listed in Table 6-4. CO2 transportation costs are based on a pipeline cost of $228,000 per inch-
mile which is consistent with the EPA Platform v6 natural gas supply curve and basis differential
assumptions from GMM. The costs also assume a 12-inch pipeline with a minimum distance of 100
miles.
73 The approach that ICF employed to estimate industrial demand forC02 storage is described in ICF International,
"Methodology and Results for Initial Forecast of Industrial CCS Volumes," January 2009.
74 Zero and negative cost steps represent storage available from enhanced oil recovery (EOR) where oil
producers either pay or offer free storage for CO2 that is injected into mature oil wells to enhance the amount of oil
recovered. The value of the CO2 for EOR is calculated using the average price of crude oil of $75/bbl. There is
also a small market for CO2 injection in enhanced coal bed methane (ECBM) production. ECBM is excluded from
EPA's inventory as discussed earlier.
6-7
-------
List of tables that are uploaded directly to the web:
Table 6-4 CO2 Storage Cost Curves in EPA Platform v6 Post-IRA 2022 Reference Case
Table 6-5 C02 Transportation Matrix in EPA Platform v6 Post-IRA 2022 Reference Case
Attachment 6-1 CO2 Reduction Retrofit Cost Development Methodology
6-8
-------
7. Coal
The next three chapters cover the representation and underlying assumptions for fuels in EPA Platform
v6 Post-IRA 2022 Reference Case (EPA Platform v6). Chapter 7 focuses on coal, Chapter 8 on natural
gas, and Chapter 9 on other fuels (fuel oil, biomass, nuclear fuel, and waste fuels).
This chapter presents four main topics. The first topic discusses how the coal market is represented.
Included are discussions of coal supply and demand regions, coal quality characteristics, and the
assignment of coal to power plants.
The second topic concerns coal supply curves which were developed using a bottom-up, mine-based
approach. The approach depicts the coal choices and associated prices that power plants face over the
modeling time horizon. Included are discussions of the methods and data used to quantify the
economically recoverable coal reserves, characterize their cost, and build the 71 coal supply curves
implemented in EPA Platform v6. Also, step-by-step illustrative examples of the approach are provided.
The third topic covers coal transportation. Included are discussions of the transport network, the
methodology used to assign costs to the links in the network, and the geographic, infrastructure, and
regulatory considerations that come into play in developing specific rail, barge, and truck transport rates.
Finally, issues concerning competition among sources of coal supply and demand are addressed.
Competition on the supply side includes imported coal that arrives from non-U.S. or non-Canadian
basins. Competition on the demand side includes demand for international thermal exports, as well as
domestic industrial, residential, and commercial demand for thermal coal. These assumptions are
discussed in Section 7.4.
The assumptions for the coal supply curves and coal transportation were finalized in January 2021, and
were developed through a collaborative process with EPA supported by the following independent team
of coal experts (with key areas of responsibility noted in parenthesis): ICF (IPM model integration and
team coordination), Wood Mackenzie (coal supply curve development), and Hellerworx (coal
transportation cost development).
7.1 Coal Market Representation
Coal supply, coal demand, coal quality, and the assignment of specific types of coals to individual coal-
fired generating units are the four key components of the endogenous coal market modeling framework in
EPA Platform v6. The modeling representation attempts to reflect the actual options available to each
existing coal-fired power plant while aggregating data sufficiently to keep the model size and solution time
within acceptable limits.
Each coal-fired power plant modeled is reflected as its own coal demand region. The demand regions
are defined to reflect the coal transportation options, including rail, barge, truck, and conveyer belt, that
are available to the plant. These demand regions are interconnected by a transportation network to at
least one of the 34 geographically dispersed coal supply regions. The model's supply-demand region
links reflect actual on-the-ground transportation pathways. Each coal supply region can supply and each
coal demand region can demand at least one grade of coal. Based on historical and engineering data (as
described in Section 7.1.5), each coal-fired power plant is also assigned several coal grades which the
plant may use if available within its demand region.
The endogenous demand for coal is generated by coal-fired power plants interacting with a set of
exogenous supply curves (see Table 7-26 for coal supply curve data) for each coal grade in each supply
region. The curves show the supply of coal (by coal supply region and coal grade) that is available to
meet the demand at a given price. The supply and demand for each coal grade is linked to and affected
by the supply and demand for every other coal grade across supply and demand regions. The
transportation network, which is also called as coal transportation matrix, in Table 7-25 provides delivery
7-1
-------
cost to move coal from a free-on-board point of sale in the coal basin to the end-use power plant. The
transportation cost combined with the free-on-board supply cost reflects the delivered cost a plant
considers when making its coal selection. IPM derives the equilibrium coal consumption and prices that
result when the entire electric system is operating at least cost while meeting emission constraints and
other operating requirements over the modeling time horizon.
7.1.1 Coal Supply Regions
There are 34 coal supply regions, each representing geographic aggregations of coal-mining areas that
supply one or more coal grades. Coal supply regions may differ from one another in the types and quality
of coal they can supply. Table 7-1 lists the coal supply regions included in EPA Platform v6. Figure 7-1
provides a map showing the location of both the coal supply regions listed in Table 7-1 and the broader
supply basins commonly used when referring to U.S. coal reserves.
Table 7-1 Coal Supply Regions in EPA Platform v6
Region
State
Supply Region
Central Appalachia
Kentucky, East
KE
Central Appalachia
Tennessee
TN
Central Appalachia
Virginia
VA
Central Appalachia
West Virginia, South
WS
Dakota Lignite
Montana, East
ME
Dakota Lignite
North Dakota
ND
East Interior
Indiana
IN
East Interior
Kentucky, West
KW
East Interior
Illinois
IL
Gulf Lignite
Texas
TX
Gulf Lignite
Louisiana
LA
Gulf Lignite
Mississippi
MS
Northern Appalachia
Maryland
MD
Northern Appalachia
Ohio
OH
Northern Appalachia
Pennsylvania, Central
PC
Northern Appalachia
Pennsylvania, West
PW
Northern Appalachia
West Virginia, North
WN
Rocky Mountains
Utah
UT
Rocky Mountains
Colorado, Green River
CG
Rocky Mountains
Colorado, Raton
CR
Rocky Mountains
Colorado, Uinta
CU
Southern Appalachia
Alabama
AL
Southwest
Arizona
AZ
Southwest
New Mexico, San Juan
NS
West Interior
Oklahoma
OK
Western Montana
Montana, Bull Mountains
MT
Western Montana
Montana, Powder River
MP
Western Wyoming
Wyoming, Green River
WG
Wyoming Northern PRB
Wyoming, Powder River Basin (8800)
WH
Wyoming Southern PRB
Wyoming, Powder River Basin (8400)
WL
Alaska
Alaska
AK
Alberta
Alberta
AB
British Columbia
British Columbia
BC
Saskatchewan
Saskatchewan
SK
7-2
-------
Figure 7-1 Map of the Coal Supply Regions in v6
Coal Supply Regions
7.1.2 Coal Demand Regions
Coal demand regions are designed to reflect coal transportation options available to power plants. Each
existing coal-fired power plant is reflected as its own individual demand region. The transportation
infrastructure (i.e., rail, barge, truck, or conveyor belt), proximity to mine (i.e., mine mouth or not mine
mouth), and transportation competitiveness levels (i.e., non-competitive, low-cost competitive, or high-
cost competitive) are developed specific to each plant (demand region).
IPM determines the amount and type of new generation capacity to add within each of the 67 U.S. IPM
model regions. The model regions reflect the administrative, operational, and transmission geographic
structure of the U.S. electricity grid. Since new plants could be located at various locations within a
region, a generic transportation cost for different coal types is developed for these new plants. The
methodology for deriving that cost is described in Section 7.3.
7.1.3 Coal Quality Characteristics
Coal varies by heat content, SO2 content, HCI content, and mercury content among other characteristics.
To capture differences in the sulfur and heat content of coal, a two-letter coal grade nomenclature is
used. The first letter indicates the coal rank (i.e., bituminous, subbituminous, or lignite) with their
associated heat content ranges (as shown in Table 7-2). The second letter indicates their sulfur grade,
(i.e., the SO2 ranges associated with a given type of coal). The sulfur grades and associated SO2 ranges
are shown in Table 7-3.
7-3
-------
Table 7-2 Coal Rank Heat Content Ranges
Coal Type
Heat Content (Btu/lb)
Classification
Bituminous
>10,260- 13,000
B
Subbituminous
> 7,500- 10,260
S
Lignite
less than 7,500
L
Table 7-3 Coal Grade SO2 Content Ranges
SO2 Grade
SO2 Content Range (Ibs/MMBtu)
A
0.00-0.80
B
0.81 - 1.20
D
1.21 - 1.66
E
1.67-3.34
G
3.35-5.00
H
> 5.00
The EPA Platform v6 assumptions on the heat, HCI, mercury, SO2, and ash contents of coal are derived
from EPA's Information Collection Request for Electric Utility Steam Generating Unit Mercury Emissions
Information Collection Effort (ICR).75
A two-year effort initiated in 1998 and completed in 2000, the ICR had three main components: (1)
identifying all coal-fired units owned and operated by publicly-owned utility companies, federal power
agencies, rural electric cooperatives, and investor-owned utility generating companies, (2) obtaining
"accurate information on the amount of mercury contained in the as-fired coal used by each electric utility
steam generating unit with a capacity greater than 25 megawatts electric, as well as accurate information
on the total amount of coal burned by each such unit,", and (3) obtaining data by coal sampling and stack
testing at selected units to characterize mercury reductions from representative unit configurations. Data
regarding the SO2, chlorine, and ash contents of the coal used was obtained along with mercury content.
The ICR captured the origin of the coal burned, and thus provided a pathway for linking emission
properties to coal basins.
The 1998-2000 ICR resulted in more than 40,000 data points indicating the coal type, sulfur content,
mercury content, ash content, chlorine content, and other characteristics of coal burned at coal-fired utility
boilers greater than 25 MW.
Annual fuel characteristic and delivery data reported on EIA Form 923 also provide continual data points
on coal heat content, sulfur content, and geographic origin, which are used as a check against
characteristics initially identified through the ICR.
7.1.4 Coal Emission Factors
To make the data usable in EPA Platform v6, the ICR data points were first grouped by IPM coal grades
and IPM coal supply regions. Using the grouped ICR data, the average heat, SO2, mercury, HCI, and ash
contents were calculated for each combination of coal grade and supply region. In instances where no
data was available for a particular coal grade in a specific supply region, the national average SO2 and
mercury values for the coal grade were used. The coal characteristics of Canadian coal supply regions
are based on the coal characteristics of the adjacent U.S. coal supply regions. The resulting values are
shown in Table 7-4. The C02 values were derived from data in the Energy Information Administration's
Annual Energy Outlook 2016.
75 Data from the ICR can be found at http://www.epa.qov/ttn/atw/combust/utiItox/mercury.html
7-4
-------
Table 7-4 Coal Quality Characteristics by Supply Region and Coal Grade in v6
Coal Supply Region
Coal
Grade
SO2
Content
Mercury
Content
Ash
Content
HCI
Content
CO2
Content
Cluster
Number
(Ibs/MMBtu)
(Ibs/TBtu)
(Ibs/MMBtu)
(Ibs/MMBtu)
(Ibs/MMBtu)
SA
0.59
5.29
5.47
0.009
215.5
1
AB
SB
0.94
6.06
6.94
0.013
215.5
4
SD
1.43
5.35
11.60
0.008
215.5
1
AK
SA
0.59
5.29
5.47
0.009
216.1
1
BB
1.09
4.18
9.76
0.012
204.7
4
AL
BD
1.35
7.28
10.83
0.029
204.7
1
BE
2.68
12.58
10.70
0.028
204.7
1
AZ
BB
1.05
5.27
7.86
0.067
207.1
2
BC
BD
1.40
6.98
8.34
0.096
216.1
3
CG
BB
0.90
4.09
8.42
0.021
209.6
4
SB
0.93
2.03
7.06
0.007
212.8
1
CR
BB
1.05
5.27
7.86
0.067
209.6
2
CU
BB
0.86
4.01
7.83
0.009
209.6
4
BE
2.25
6.52
6.61
0.214
203.1
2
IL
BG
4.56
6.53
8.09
0.113
203.1
3
BH
5.58
5.43
9.06
0.103
203.1
1
BE
2.31
5.21
7.97
0.036
203.1
3
IN
BG
4.27
7.20
8.22
0.028
203.1
3
BH
6.15
7.11
8.63
0.019
203.1
3
BB
1.04
4.79
6.41
0.112
206.4
5
KE
BD
1.44
5.97
7.45
0.087
206.4
2
BE
2.12
7.93
7.71
0.076
206.4
4
KW
BG
4.46
6.90
8.01
0.097
203.1
3
BH
5.73
8.16
10.21
0.053
203.1
3
LA
LE
2.49
7.32
17.15
0.014
212.6
1
MD
BE
2.78
15.62
11.70
0.072
204.7
5
BG
3.58
16.64
16.60
0.018
204.7
5
ME
LE
1.83
11.33
11.69
0.019
219.3
2
MP
SA
0.62
4.24
3.98
0.007
215.5
1
SD
1.49
4.53
10.13
0.006
215.5
1
MS
LE
2.76
12.44
21.51
0.018
216.5
3
MT
BB
1.05
5.27
7.86
0.067
215.5
2
ND
LE
2.27
8.30
12.85
0.014
219.3
1
SB
0.89
4.60
14.51
0.014
209.2
2
NS
SD
1.55
7.54
23.09
0.007
209.2
2
SE
1.90
8.65
23.97
0.008
209.2
1
BE
3.08
18.70
7.08
0.075
204.7
6
OH
BG
3.99
18.54
8.00
0.071
204.7
5
BH
6.43
13.93
9.13
0.058
204.7
4
OK
BG
4.65
26.07
13.54
0.051
202.8
4
BE
2.57
17.95
9.23
0.096
204.7
6
PC
BG
3.79
21.54
9.59
0.092
204.7
2
BH
6.29
34.71
13.89
0.148
204.7
5
BE
2.51
8.35
5.37
0.090
204.7
4
PW
BG
3.69
8.56
6.48
0.059
204.7
1
BH
7.78
16.46
11.56
0.046
204.7
2
SK
LD
1.51
7.53
11.57
0.014
219.3
1
LE
2.76
12.44
21.51
0.018
219.3
3
7-5
-------
Coal Supply Region
Coal
Grade
so2
Content
Mercury
Content
Ash
Content
HCI
Content
co2
Content
Cluster
Number
(Ibs/MMBtu)
(Ibs/TBtu)
(Ibs/MMBtu)
(Ibs/MMBtu)
(Ibs/MMBtu)
TN
BB
1.14
3.78
10.35
0.083
206.4
3
BE
2.13
8.43
6.47
0.043
206.4
4
LE
3.00
14.65
25.65
0.020
212.6
4
TX
LG
3.91
14.88
25.51
0.036
212.6
1
LH
5.67
30.23
23.95
0.011
212.6
1
BA
0.67
4.37
7.39
0.015
209.6
1
UT
BB
0.94
3.93
8.58
0.016
209.6
4
BD
1.37
4.38
10.50
0.026
209.6
3
BE
2.34
9.22
7.41
0.095
209.6
4
BB
1.05
4.61
6.97
0.054
206.4
5
VA
BD
1.44
5.67
7.97
0.028
206.4
2
BE
2.09
8.40
8.05
0.028
206.4
4
BB
1.13
1.82
5.58
0.005
214.3
3
WG
SB
1.06
4.22
8.72
0.009
214.3
3
SD
1.33
4.33
10.02
0.008
214.3
1
WH
SA
0.52
5.61
5.51
0.010
214.3
2
WL
SA
0.71
5.61
7.09
0.010
214.3
3
SB
0.93
6.44
7.92
0.012
214.3
4
WN
BE
2.55
10.28
7.89
0.092
204.7
7
BH
6.09
8.82
9.62
0.045
204.7
3
BB
1.09
5.75
9.15
0.091
206.4
1
WS
BD
1.32
8.09
9.25
0.098
206.4
4
BE
1.94
8.83
9.89
0.102
206.4
4
Next, a clustering algorithm was used to further aggregate the data for model size management
purposes. The clustering analysis was performed on the S02, mercury, and HCI content data shown in
Table 7-4 using the SAS statistical software package. Clustering analysis places objects into groups or
clusters, such that data in a given cluster tend to be similar to each other and dissimilar to data in other
clusters. The clustering analysis involved two steps. First, the number of clusters of S02, mercury, and
HCI contents for each coal grade was determined based on the range in S02, mercury, and HCI contents
across all coal supply regions. Each coal grade used one to seven clusters. The number of clusters for
each coal grade was limited to keep the model size and run time within acceptable limits. Second, for
each coal grade, the clustering procedure was applied to all the regional S02, mercury, and HCI contents
shown in Table 7-4. Using the SAS cluster procedure, each of the constituent regional contents was
assigned to a cluster and the cluster average S02, mercury, and HCI contents were estimated. The
resulting contents are shown in Table 7-5 through Table 7-9.
7-6
-------
Table 7-5 Coal Clustering by Coal Grade - S02 Emission Factors (Ibs/MMBtu)
Coal Type by Sulfur Grade
S02 Emission Factors (Ibs/MMBtu)
Cluster #1
Cluster #2
Cluster #3
Cluster #4
Cluster# 5
Cluster #6
Cluster# 7
Cluster Value
Data Range
Cluster
Value
Data Range
Cluster
Value
Data Range
Cluster
Value
Data Range
Cluster
Value
Data Range
Cluster
Value
Data Range
Cluster
Value
Data Range
Low
High
Low
High
Low
High
Low
High
Low
High
Low
High
Low
High
Low Sulfur Bituminous (BA)
0.67
0.67
0.67
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
Low Sulfur Bituminous (BB)
1.10
1.09
1.10
1.05
1.05
1.05
1.14
1.13
1.14
0.95
0.86
1.09
1.04
1.04
1.05
-
-
-
-
-
-
Low Medium Sulfur Bituminous (BD)
1.35
1.35
1.35
1.44
1.44
1.44
-
-
-
1.39
1.37
1.40
1.32
1.32
1.32
-
-
-
-
-
-
Medium Sulfur Bituminous (BE)
2.68
2.68
2.68
2.25
2.25
2.25
2.31
2.31
2.31
2.19
1.94
2.51
2.78
2.78
2.78
2.82
2.57
3.08
2.55
2.55
2.55
High Sulfur Bituminous (BG)
3.69
3.69
3.69
3.79
3.79
3.79
4.43
4.27
4.56
-
-
-
-
-
-
4.65
4.65
4.65
3.78
3.58
3.99
High Sulfur Bituminous (BH)
5.58
5.58
5.58
7.78
7.78
7.78
5.99
5.73
6.15
6.43
6.43
6.43
6.29
6.29
6.29
-
-
-
-
-
-
Low Sulfur Subbituminous (SA)
0.60
0.59
0.62
0.52
0.52
0.52
0.71
0.71
0.71
-
-
-
-
-
-
-
-
-
-
-
-
Low Sulfur Subbituminous (SB)
0.93
0.93
0.93
-
-
-
0.89
0.89
0.89
1.06
1.06
1.06
0.94
0.93
0.94
-
-
-
-
-
-
Low Medium Sulfur Subbituminous (SD)
1.42
1.33
1.49
1.55
1.55
1.55
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
Medium Sulfur Subbituminous (SE)
1.90
1.90
1.90
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
Low Medium Sulfur Lignite (LD)
1.51
1.51
1.51
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
Medium Sulfur Lignite (LE)
2.38
2.27
2.49
1.83
1.83
1.83
2.76
2.76
2.76
3.00
3.00
3.00
-
-
-
-
-
-
-
-
-
High Sulfur Lignite (LG)
3.91
3.91
3.91
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
High Sulfur Lignite (LH)
5.67
5.67
5.67
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
Table 7-6 Coal Clustering by Coal Grade - Mercury Emission Factors (Ibs/TBtu)
Mercury Emission Factors (Ibs/TBtu)
Coal Type by Sulfur Grade
Cluster #1
Cluster #2
Cluster #3
Cluster #4
Cluster# 5
Cluster #6
Cluster #7
Cluster
Data Range
Cluster
Data Range
Cluster
Data Range
Cluster
Data Range
Cluster
Data Range
Cluster
Data Range
Cluster
Data Range
Value
Low
High
Value
Low
High
Value
Low
High
Value
Low
High
Value
Low
High
Value
Low
High
Value
Low
High
Low Sulfur Bituminous (BA)
4.37
4.37
4.37
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
Low Sulfur Bituminous (BB)
6.74
5.75
7.74
5.27
5.27
5.27
2.80
1.82
3.78
4.05
3.93
4.18
4.70
4.61
4.79
-
-
-
-
-
-
Low Medium Sulfur Bituminous (BD)
7.28
7.28
7.28
5.82
5.67
5.97
-
-
-
5.68
4.38
6.98
8.09
8.09
8.09
-
-
-
-
-
-
Medium Sulfur Bituminous (BE)
12.58
12.58
12.58
6.52
6.52
6.52
5.21
5.21
5.21
8.53
7.93
9.22
15.62
15.62
15.62
18.33
17.95
18.70
10.28
10.28
10.28
High Sulfur Bituminous (BG)
8.56
8.56
8.56
21.54
21.54
21.54
6.88
6.53
7.20
-
-
-
-
-
-
26.07
26.07
26.07
17.59
16.64
18.54
High Sulfur Bituminous (BH)
5.43
5.43
5.43
16.46
16.46
16.46
8.03
7.11
8.82
13.93
13.93
13.93
34.71
34.71
34.71
-
-
-
-
-
-
Low Sulfur Subbituminous (SA)
4.94
4.24
5.29
5.61
5.61
5.61
5.61
5.61
5.61
-
-
-
-
-
-
-
-
-
-
-
-
Low Sulfur Subbituminous (SB)
2.03
2.03
2.03
-
-
-
4.60
4.60
4.60
4.22
4.22
4.22
6.25
6.06
6.44
-
-
-
-
-
-
Low Medium Sulfur Subbituminous (SD)
4.74
4.33
5.35
7.54
7.54
7.54
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
Medium Sulfur Subbituminous (SE)
8.65
8.65
8.65
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
Low Medium Sulfur Lignite (LD)
7.53
7.53
7.53
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
Medium Sulfur Lignite (LE)
7.81
7.32
8.30
11.33
11.33
11.33
12.44
12.44
12.44
14.65
14.65
14.65
-
-
-
-
-
-
-
-
-
High Sulfur Lignite (LG)
High Sulfur Lignite (LH)
14.88
30.23
14.88
30.23
14.88
30.23
--
—
--
--
—
--
--
—
--
—
—
—
—
—
—
—
—
—
7-7
-------
Table 7-7 Coal Clustering by Coal Grade - Ash Emission Factors (Ibs/MMBtu)
Ash Emission Factors (Ibs/MMBtu)
Coal Type by Sulfur Grade
Cluster #1
Cluster #2
Cluster #3
Cluster #4
Cluster #5
Cluster #6
Cluster# 7
Cluster
Data Range
Cluster
Data Range
Cluster
Data Range
Cluster
Data Range
Cluster
Data Range
Cluster
Data Range
Cluster
Data Range
Value
Low
High
Value
Low
High
Value
Low
High
Value
Low
High
Value
Low
High
Value
Low
High
Value
Low
High
Low Sulfur Bituminous (BA)
7.39
7.39
7.39
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
Low Sulfur Bituminous (BB)
6.98
4.81
9.15
7.86
7.86
7.86
7.97
5.58
10.35
8.65
7.83
9.76
6.69
6.41
6.97
-
-
-
-
-
-
Low Medium Sulfur Bituminous (BD)
10.83
10.83
10.83
7.71
7.45
7.97
-
-
-
9.42
8.34
10.50
9.25
9.25
9.25
-
-
-
-
-
-
Medium Sulfur Bituminous (BE)
10.70
10.70
10.70
6.61
6.61
6.61
7.97
7.97
7.97
7.48
5.37
9.89
11.70
11.70
11.70
8.16
7.08
9.23
7.89
7.89
7.89
High Sulfur Bituminous (BG)
6.48
6.48
6.48
9.59
9.59
9.59
8.10
8.01
8.22
-
-
-
-
-
-
13.54
13.54
13.54
12.30
8.00
16.60
High Sulfur Bituminous (BH)
9.06
9.06
9.06
11.56
11.56
11.56
9.49
8.63
10.21
9.13
9.13
9.13
13.89
13.89
13.89
-
-
-
-
-
-
Low Sulfur Subbituminous (SA)
4.97
3.98
5.47
5.51
5.51
5.51
7.09
7.09
7.09
-
-
-
-
-
-
-
-
-
-
-
-
Low Sulfur Subbituminous (SB)
7.06
7.06
7.06
-
-
-
14.51
14.51
14.51
8.72
8.72
8.72
7.43
6.94
7.92
-
-
-
-
-
-
Low Medium Sulfur Subbituminous (SD)
10.58
10.02
11.60
23.09
23.09
23.09
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
Medium Sulfur Subbituminous (SE)
23.97
23.97
23.97
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
Low Medium Sulfur Lignite (LD)
11.57
11.57
11.57
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
Medium Sulfur Lignite (LE)
15.00
12.85
17.15
11.69
11.69
11.69
21.51
21.51
21.51
25.65
25.65
25.65
-
-
-
-
-
-
-
-
-
High Sulfur Lignite (LG)
High Sulfur Lignite (LH)
25.51
23.95
25.51
23.95
25.51
23.95
--
—
--
--
—
--
--
—
--
—
—
—
—
—
—
—
—
--
Table 7-8 Coal Clustering by Coal Grade - HCI Emission Factors (Ibs/MMBtu)
HCI Emission Factors (Ibs/MMBtu)
Cluster #1
Cluster #2
Cluster #3
Cluster #4
Cluster# 5
Cluster #6
Cluster #7
Coal Type by Sulfur Grade
Cluster
Data Range
Cluster
Data Range
Cluster
Data Range
Cluster
Data Range
Cluster
Data Range
Cluster
Data Range
Cluster
Data Range
\/_i...
\/_i...
\ #„i,,.
\/_i...
\ #„i,,.
\ #„i,,.
Low
High
Low
High
Low
High
Low
High
Low
High
Low
High
Low
High
Low Sulfur Bituminous (BA)
0.015
0.015
0.015
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
Low Sulfur Bituminous (BB)
0.054
0.018
0.091
0.067
0.067
0.067
0.044
0.005
0.083
0.015
0.009
0.021
0.083
0.054
0.112
-
-
-
-
-
-
Low Medium Sulfur Bituminous (BD)
0.029
0.029
0.029
0.057
0.028
0.087
-
-
-
0.061
0.026
0.096
0.098
0.098
0.098
-
-
-
-
-
-
Medium Sulfur Bituminous (BE)
0.028
0.028
0.028
0.214
0.214
0.214
0.036
0.036
0.036
0.072
0.028
0.102
0.072
0.072
0.072
0.085
0.075
0.096
0.092
0.092
0.092
High Sulfur Bituminous (BG)
0.059
0.059
0.059
0.092
0.092
0.092
0.079
0.028
0.113
-
-
-
-
-
-
0.051
0.051
0.051
0.045
0.018
0.071
High Sulfur Bituminous (BH)
0.103
0.103
0.103
0.046
0.046
0.046
0.039
0.019
0.053
0.058
0.058
0.058
0.148
0.148
0.148
-
-
-
-
-
-
Low Sulfur Subbituminous (SA)
0.008
0.007
0.009
0.010
0.010
0.010
0.010
0.010
0.010
-
-
-
-
-
-
-
-
-
-
-
-
Low Sulfur Subbituminous (SB)
0.007
0.007
0.007
-
-
-
0.014
0.014
0.014
0.009
0.009
0.009
0.013
0.012
0.013
-
-
-
-
-
-
Low Medium Sulfur Subbituminous (SD)
0.007
0.006
0.008
0.007
0.007
0.007
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
Medium Sulfur Subbituminous (SE)
0.008
0.008
0.008
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
Low Medium Sulfur Lignite (LD)
0.014
0.014
0.014
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
Medium Sulfur Lignite (LE)
0.014
0.014
0.014
0.019
0.019
0.019
0.018
0.018
0.018
0.020
0.020
0.020
-
-
-
-
-
-
-
-
-
High Sulfur Lignite (LG)
0.036
0.036
0.036
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
High Sulfur Lignite (LH)
0.011
0.011
0.011
-
-
-
-
-
-
-
-
-
-
-
-
--
--
--
--
--
--
7-8
-------
Table 7-9 Coal Clustering by Coal Grade - C02 Emission Factors (Ibs/MMBtu)
Coal Type by Sulfur Grade
C02 Emission Factors (Ibs/MMBtu)
Cluster #1
Cluster #2
Cluster #3
Cluster #4
Cluster# 5
Cluster #6
Cluster# 7
Cluster
Data Range
Cluster
Data Range
Cluster
Data Range
Cluster
Data Range
Cluster
Data Range
Cluster
Data Range
Cluster
Data Range
Value
Low
High
Value
Low
High
Value
Low
High
Value
Low
High
Value
Low
High
Value
Low
High
Value
Low
High
Low Sulfur Bituminous (BA)
209.6
209.6
209.6
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
Low Sulfur Bituminous (BB)
206.8
206.4
207.1
210.7
207.1
215.5
210.4
206.4
214.3
208.4
204.7
209.6
206.4
206.4
206.4
-
-
-
-
-
-
Low Medium Sulfur Bituminous (BD)
204.7
204.7
204.7
206.4
206.4
206.4
-
-
-
212.9
209.6
216.1
206.4
206.4
206.4
-
-
-
-
-
-
Medium Sulfur Bituminous (BE)
204.7
204.7
204.7
203.1
203.1
203.1
203.1
203.1
203.1
206.7
204.7
209.6
204.7
204.7
204.7
204.7
204.7
204.7
204.7
204.7
204.7
High Sulfur Bituminous (BG)
204.7
204.7
204.7
204.7
204.7
204.7
203.1
203.1
203.1
-
-
-
-
-
-
202.8
202.8
202.8
204.7
204.7
204.7
High Sulfur Bituminous (BH)
203.1
203.1
203.1
204.7
204.7
204.7
203.6
203.1
204.7
204.7
204.7
204.7
204.7
204.7
204.7
-
-
-
-
-
-
Low Sulfur Subbituminous (SA)
215.7
215.5
216.1
214.3
214.3
214.3
214.3
214.3
214.3
-
-
-
-
-
-
-
-
-
-
-
-
Low Sulfur Subbituminous (SB)
212.8
212.8
212.8
-
-
-
209.2
209.2
209.2
214.3
214.3
214.3
214.9
214.3
215.5
-
-
-
-
-
-
Low Medium Sulfur Subbituminous (SD)
215.1
214.3
215.5
209.2
209.2
209.2
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
Medium Sulfur Subbituminous (SE)
209.2
209.2
209.2
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
Low Medium Sulfur Lignite (LD)
219.3
219.3
219.3
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
Medium Sulfur Lignite (LE)
216.0
212.6
219.3
219.3
219.3
219.3
217.9
216.5
219.3
212.6
212.6
212.6
-
-
-
-
-
-
-
-
-
High Sulfur Lignite (LG)
212.6
212.6
212.6
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
High Sulfur Lignite (LH)
212.6
212.6
212.6
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
7-9
-------
7.1.5 Coal Grade Assignments
The grades of coal that may be used by specific generating units were determined by an expert
assessment of the ranks of coal that a unit had used in the past, the removal efficiency of the installed
FGD, and the SO2 permit rate of the unit. Examples of the coal grade assignments made for individual
plants in EPA Platform v6 are shown in Table 7-10. Not all the coal grades allowed to a plant by the coal
grade assignment are necessarily available in the plant's assigned coal demand region (due to
transportation limitations). IPM endogenously selects the coal consumed by a plant by considering both
the constraint of the plant's coal grade assignment and the constraint of the coals available within a
plant's coal demand region.
Table 7-10 Example of Coal Assignments Made in v6
Plant Name
Unit
Permit Rate
(Ibs/MMBtu)
Scrubber?
Fuels Allowed
Mt Storm
3
0.15
Yes
BA, BB, BD
Mitchell
1
1.2
Yes
BA, BB, BD, BE, BG, BH
Scherer
1
1.2
Yes
SA, SB, SD, SE
Newton
1
0.5
No
SA, SB, SD, SE
Limestone
LIM1
0.6
Yes
LD, LE, LG, LH, SA, SB, SD, SE
San Miguel
SM-1
1.2
Yes
LD, LE, LG, LH
7.2 Coal Supply Curves
7.2.1 Nature of Supply Curves Developed for EPA Platform v6
In keeping with IPM's data-driven bottom-up modeling framework, a bottom-up approach (relying heavily
on detailed economic and resource geology data and assessments) was used to prepare the thermal coal
supply curves for EPA Platform v6. EPA utilized Wood Mackenzie to develop the curves based on their
extensive experience in preparing mine-by-mine estimates of cash operating costs for operating mines in
the U.S., their access to both public and proprietary data sources, and their active updating of the data
through both research and interviews.
In order to establish consistent nomenclature, Wood Mackenzie first mapped its internal list of coal
regions and qualities to EPA's 34 coal supply regions (described above in sections 7.1.1) and the 14 coal
rank/grades combinations (described above in section 7.1.3). The combined code list is shown in Table
7-11 below with the IPM coal supply regions appearing in the rows and the coal grades in the columns.
Wood Mackenzie then created supply curves for each region and coal-grade combination (indicated by
the "x" in Table 7-11) for forecast years 2023, 2025, 2028, 2030, 2035, 2040, 2045, and 2050.
Table 7-11 Basin-Level Groupings Used in Preparing v6 Coal Supply Curves
Bituminous
Lignite
Subbituminous
Coal
Supply
Region
Geo
Region
Geo.
Sub-
Region
BA
BB
BD
BE
BG
BH
LD
L
E
LG
LH
SA
SB
S
D
SE
AB
Canada
Alberta, Canada
X
X
X
AK
Alaska
Alaska |
X
AL
Appalachi
a
Southern
Appalachia
X
X
X
AZ
West
Southwest
X
BC
Canada
British Columbia
X
CG
West
Rocky Mountain
X
X
CR
West
Rocky Mountain
X
CU
West
Rocky Mountain
X
IL
Interior
East Interior (Illinois
Basin)
X
X
X
7-10
-------
Bituminous
Lignite
Subbituminous
Coal
Supply
Region
Geo
Region
Geo.
Sub-
Region
BA
BB
BD
BE
BG
BH
LD
L
E
LG
LH
SA
SB
S
D
SE
IN
Interior
East Interior (Illinois
Basin)
X
X
X
KE
Appalachi
a
Central
Appalachia
X
X
X
KW
Interior
East Interior (Illinois
Basin)
X
X
LA
Interior
Gulf Lignite
X
MD
Appalachi
a
Northern
Appalachia
X
X
ME
West
Dakota Lignite
X
MP
West
Powder River
Basin
X
X
MS
Gulf
Gulf Lignite
Coast
X
MT
West
Western
Montana
X
ND
West
Dakota Lignite
X
NS
West
Southwest
X
X
X
OH
Appalachi
a
Northern
Appalachia
X
X
X
OK
West
West Interior
X
PC
Appalachi
a
Northern
Appalachia
X
X
X
PW
Appalachi
a
Northern
Appalachia
X
X
X
SK
Canada
Saskatchewan
X
X
TN
Appalachi
a
Central
Appalachia
X
X
TX
Interior
Gulf Lignite
X
X
X
UT
West
Rocky
Mountai
n
X
X
X
X
VA
Appalachi
a
Central
Appalachia
X
X
X
WG
West
Western
Wyoming
X
X
X
WH
West
Powder River
Basin
X
WL
West
Powder River
Basin
X
X
WN
Appalachi
a
Northern
Appalachia
X
X
WS
Appalachi
a
Central
Appalachia
X
X
X
7.2.2 Cost Components in the Supply Curves
Costs are represented as total cash costs, which is a combination of a mine's operating cash costs plus
royalty & levies. These costs are estimated on a Free on Board (FOB) basis at the point of sale. Capital
costs (either expansionary or sustaining) are not included in the cash cost estimate for existing mines.
For projects, the expansionary capital is spread across the mine life and included in the costs. The total
cash cost is the best metric for the supply curves as coal prices tend to be ultimately determined by the
incremental cost of production (i.e., total cash cost).
Operating cash cost
These are the direct operating cash costs and include, where appropriate, mining, coal preparation,
product transport, and overheads. No capital cost component or depreciation & amortization charge is
included for operating mines. Expansionary capital is included for new greenfield projects. Operating
cash costs consist of the following elements:
7-11
-------
Mining costs - Mining costs are the direct cost of mining coal and associated waste material for surface
and underground operations. It includes any other mine site costs, such as ongoing rehabilitation /
reclamation, security, community development costs. It also includes the cost of transporting raw coal
from the mining location to the raw coal stockpile at the coal preparation plant.
Coal preparation - The cost of coal preparation includes raw coal stockpile reclaim, crushing and
screening, washing and marketable coal product stockpiling (if applicable).
Transport - This covers all transport costs of product coal to point of sale. Transport routes with multiple
modes (e.g., truck and rail) are shown as total cost per marketable ton for all stages of the transport
route. Loading charges are included in this cost if relevant.
Overheads - This is any non-production related general and administration overheads that are essential
to the production and sale of a mine's coal product. Examples would be mine site staff not related to
mining, essential corporate management or a sales and marketing charge.
It is important to note that although the formula for calculating mine costs is consistent across regions,
some tax rates and fees vary by state and mine type. In general, there are two mine types: underground
(deep) or surface mines. Underground mining is categorized as being either a longwall (LW) or a
continuous room-and-pillar mine (CM). Geologic conditions and characteristics of the coal seams
determine which method will be used. Surface mines are typically categorized by the type of mining
equipment used in their operation such as draglines (DL), or truck & shovels (TS). These distinctions are
important because the equipment used by the mine affects productivity measures and ultimately mine
costs. Further information on operating cost methodology and assumptions can be found in Attachment
7-1.
Royalties and Levies
These include, where appropriate, coal royalties, mine safety levies, health levies, industry research
levies and other production taxes. These taxes, fees and levies vary on a regional basis.
7.2.3 Procedures Employed in Determining Mining Costs
The total cash costs of mines have been estimated in current year terms using public domain information
including; geological reports, reported statistics on production, labor and input costs, and company
reports. The estimates have been validated by reference to information gained by visits to operations,
and discussions with industry participants.
Because the estimates are based only on public information and analysis, and do not represent private
knowledge of an operation's actual costs, there may be deviations from actual costs. In instances where
confidential information is held by Wood Mackenzie, it has not been used to produce the published
estimates. Several methods are employed for cost estimation depending on the availability of information
and the diversity of mining operations. When possible, Wood Mackenzie analysts developed detailed
lists of mine-related costs. Costs such as employee wages & benefits, diesel fuel, spare parts, roof bolts,
and explosives among a host of others are summed to form a mine's operating cash costs.
Where information is incomplete, cost items are grouped into categories that can be compared with
industry averages by mine type and location. These averages can be adjusted up or down based on new
information or added assumptions. The adjustments take the form of cost multipliers or parameter
values. Specific cost multipliers are developed with the aid of industry experts and proprietary formulas.
This method is at times used to convert materials and supplies, on-site trucking costs and mine and
division overhead categories into unit removal costs by equipment type. To check the accuracy of these
7-12
-------
cost estimates, cash flow analysis of publicly traded companies is used. Mine cash-costs are extracted
from corporate cash flows and compared with the initial estimates. Adjustments for discrepancies are
made on a case-by-case basis.
Many of the cost assumptions associated with labor and productivity were taken from the Mine Safety
Health Administration (MSHA) database. All active mines report information specific to production levels,
number of employees and employee hours worked. Wood Mackenzie supplements the basic MSHA data
with information obtained from mine personnel interviews and industry contacts. Phone conversations
and conferences with industry professionals provide additional non-reported information such as work
schedules, equipment types, percentages of washed coal, and trucking distances from the mine to wash-
plants and load-out terminals.
For each active or proposed mine, Wood Mackenzie reports the estimated cost to take coal from the mine
to a logical point-of-sale. The logical point-of-sale may be a truck or railcar load-out or even a barge
facility. This is done to produce a consistent cost comparison between mines. Any transport costs
beyond the point-of-sale terminal are not part of this analysis and are not reflected in the supply curves
themselves.
7.2.4 Procedure Used in Determining Mine Productivity
Projected production and stripping ratios are the key determinants of surface mine productivity. Wood
Mackenzie assumes mining costs increase as stripping ratios increase. The stripping ratio is the quantity
of overburden removed relative to the quantity of coal recovered. Assuming that reserves are developed
where they are easiest to mine and deliver to market, general theory suggests that as the easy reserves
are depleted, greater amounts of overburden must be handled for the same amount of coal production;
thus causing a decrease in mining productivity. However, some productivity loss may be offset by
technological improvements in labor saving equipment.
In order to calculate the amount of employee hours, and therefore the labor cost, of future production
Wood Mackenzie uses a multi-step process. First, employee hours associated with coal production for
each mine are obtained from MSHA. Total production is then divided by these hours to calculate
productivity, measured in short tons per employee hour. Future production levels are divided by this
productivity measurement to obtain future employee hours needed to produce that volume of coal. From
there, the total staffing level can be determined, and the associated cost calculated.
A similar approach is used for underground mines. First, as background, the specific factors affecting
productivity at such mines are identified. For example, underground mines do not have stripping ratios.
Productivity estimates for these mines largely depend on the type of mining technique used (which is a
function of the region's geology). For instance, longwall-mines can produce a high volume of low-cost
coal but geologic constraints like small reserve blocks and the occurrence of faulting tends to limit this
technique to certain regions. In addition to geologic constraints, there are variables that can impact
underground-mine productivity that are often difficult to quantify and forecast.
7.2.5 Procedure to Determine Total Recoverable Reserves by Region and Type
Before mine operators are allowed to mine coal, they must request various permits, conduct
environmental impact studies (EIS) and, in many cases, notify corporate shareholders. In each of these
instances, mine operators are asked to estimate annual production and total recoverable reserves. Wood
Mackenzie uses the mine operators' statements as the starting point for production and reserves
forecasts. If no other material is available, interviews with company personnel will provide an estimate.
7-13
-------
Region and coal type determinations for unlisted reserves are based on public information reported for
similarly located mines. Classifying reserves this way means considering not only a mine's geographic
location but also its geologic conditions such as depth and type of overburden and the specific identity of
the coal seam(s) being mined. For areas where public information is not available or is incomplete, Wood
Mackenzie engineers and geologists estimate reserve amounts based on land surveys and reports of
coal depth and seam thickness provided by the U.S. Geologic Service (USGS). This information is then
used to extrapolate reserve estimates from known coal sources to unknown sources. Coal quality
determinations for unknown reserves are assigned in much the same way.
Once a mine becomes active, actual production numbers reported in corporate SEC filings and MSHA
reports are subtracted from the total reserve number to arrive at current reserve amounts. Wood
Mackenzie consistently updates the reserves database when announcements of new or amended
reserves are made public. As a final check, the Wood Mackenzie supply estimates are balanced against
the Demonstrated Reserve Base (DRB)76 estimates to ensure that they do not exceed the DRB
estimates.
7.2.6 New Mine Assumptions
New mines have been included based on information that Wood Mackenzie maintains on each supply
region. They include announced projects, coal lease applications and unassigned reserves reported by
mining companies. Where additional reserves are known to exist, additional incremental steps have been
added and designated with the letter "N" in the "Step Name" field of the supply curves. These
incremental steps were added based on characteristics of the specific region, typical mine size, and cost
trends. They do not necessarily imply a specific mine or mine type.
Wood Mackenzie has also identified technical coal reserves that may be commercial in the longer-term,
but would most likely not be developed until after the completion of mine development already underway
or announced. These reserves are often the "last step" in a coal supply curve due to the more difficult
geologic conditions and have been designated using the above methodology.
In addition to new mines, Wood Mackenzie also identifies extension mines. These are denoted with the
letter "A", "B", "C" or "D" at the end of an existing mine step name (e.g., E2A). These mine steps reflect
the extension of a particular mine operating through a new lease covering tracts not previously
recoverable under the existing mine operation. These mine expansions, like new mines, include the
capital expansionary component in their cost of production.
7.2.7 Other Notable Procedures
Currency Assumptions
For consistency with the cost basis used in EPA Platform v6 Post-IRA 2022 Reference Case, costs are
converted to real 2019$.
Future Cost Adjustments
Changes in mine productivity are a key factor impacting the evolution of costs overtime. In general, mine
productivity is expected to continue to decline - in large part due to worsening geology and more difficult
to mine reserves. Productivity has declined at a -0.94% compound annual growth rate (CAGR) from
2000-2019 as shown in Figure 7-2.
76 Posted by the Energy Information Administration (EIA) in its Coal Production Report.
7-14
-------
Figure 7-2 Coal Mine Productivity (2000-2019)
-0.94% CAGR
Source: U.S. Department of Labor, Mine Safety and Health Administration
Source: U.S. Department of Labor, Mine Safety and Health Administration
Figure 7-3 Average Annual Cost Growth Assumptions by Region
(2021-2050)
2.0%
Figure 7-3 shows the compounded average annual growth rate (CAGR) of mining costs by basin over the
forecast period. It should be noted that cost increases would ultimately be linked to market demand (as
demand grows, the faster the rate of depletion of lower cost reserves). Costs in some supply basins are
7-15
-------
expected to increase more quickly than others due to issues such as mining conditions, productivity,
infrastructure limitations, etc. Region-specific information can be found in section 7.2.9.
Supply Growth Limitations
To the maximum extent possible, the IPM model is set up to determine the optimal volume of coal supply,
which can be profitably supplied. For two of the lower cost basins (Powder River and Illinois basins),
maximum production capacities are included as constraints (production ceilings) to reflect more
accurately the upper bound of what could be produced in a given year. Those limits, represented in
millions of tons per year, are shown in Figure 7-4. These ceilings, while not binding in EPA's reference
case, are necessary to guard against modeling excess annual production capacity in certain basins under
sensitivity scenarios. For instance, in the PRB, several of the "new" mines reflect expansion mines that
would not be developed until the initial mine is further depleted. In this case, the production ceiling helps
safeguard against a modeling scenario that would simultaneously produce from both of these mines.
Figure 7-4 Maximum Annual Coal Production Capacity per Year (Million Short Tons)
2023
2025
2028
2030
2035
2040
2045
2050
ILB
200
220
240
240
240
240
240
240
PRB
500
520
560
560
600
600
600
600
7.2.8 Cumulative Supply Curve Development
The description below describes the depicts the cumulative supply curve. Table 7-26 shows the actual
coal supply curves.
Once costs are estimated for all new or existing mines, they are sorted by cash cost, lowest to highest,
and plotted cumulatively by production to form a supply curve. The supply curve then represents all
mines - new or existing as well as both underground and surface mines- irrespective of market demand
Mines located toward the bottom of the curve have the lowest cost and are most likely to be developed
while the mines at the top of the curve are higher cost and will likely wait to be developed. The process
for developing a cumulative supply curve is illustrated in Figure 7-5 and Figure 7-6.
7-16
-------
Figure 7-5 Illustration of Preliminary Step in Developing a Cumulative Coal Supply Curve
Key
E = EXISTING MINE
N = NEW MINE
U = UNDERGROUND MINE
S = SURFACE MINE
New oi
Existing?
Mine
Type
Cost
P induction
N
A
S
f 30
2
E
8
U
$ 20
4
N
C
s
1 32
1
N
D
s
$ 36
0.5
E
E
s
1 29
2
N
F
s
$ 26
25
E
G
u
t 25
5
E
H
u
% 23
4
E
1
u
S 27
3
N
J
s
1 35
0.25
In the table and graph above, mine costs and production are sorted alphabetically by mine name. To
develop a supply curve from the above table the values must be sorted by mine costs from lowest to
highest. A new column for cumulative production is added, and then a supply curve graph is created
which shows the costs on the '¥' axis and the cumulative production on the 'X' axis. Notice below that the
curve contains all mines - new or existing as well as both underground and surface mines. The resulting
curve is a continuous supply curve but can be modified to show costs as a stepped supply curve. (Supply
curves in stepped format are used in linear programming models like IPM.) See Figure 7-7 for a stepped
version of the supply curve example shown in Figure 7-6. Here each step represents an individual mine,
the width of the step reflects the mine's production, and its height shows the cost of production.
Figure 7-6 Illustration of Final Step in Developing a Cumulative Coal Supply Curve
New oi
Existing?
Mii»ef»
TypeF
Cost[»
Production P
Cum
Productiof*
E
B
U
$ 20
4
4
E
H
u
$ 23
4
8
E
G
u
$ 25
5
13
E
1
u
1 27
3
16
N
F
s
$ 28
2.5
18.5
E
E
s
$ 29
2
20.5
N
A
s
$ 30
2
22.5
N
C
s
$ 32
1
23.5
N
J
s
i 35
0.25
23.75
N
0
s
$ 36
0.5
24.25
Smooth Supply Curve
$40
<5? $30
i 520
©
° 110
5 1 0 15 20 25
Cumulative Production (Tons)
7-17
-------
Figure 7-7 Example Coal Supply Curve in Stepped Format
MINE NAME —
New or ExistincQ
PRODUCTION AMOUNT
G I F E A C J D
13 16 18.5 20.5 22.5 24 25 25.5
13
23
23
23
25
25
25
25
25
27
27
27
28
28
28
29
29
30
3D
32
35
7.2.9 EPA Platform v6 Assumptions and Outlooks for Major Supply Basins
Powder River Basin (PRB)
The PRB is somewhat unique to other U.S. coal basins in that producers are able to adjust production
volumes relatively easily. That said, the decisions on production volumes are largely based on the
market conditions, namely the price. For instance, in a low-demand environment producers tend to
moderate production volumes to maintain attractive prices, and choose to ramp up production when
prices are higher. The evolution of costs in the PRB will be strongly correlated to the rate at which
producers ramp up production at existing mines, which as indicated will depend on market conditions.
Wood Mackenzie anticipates productivity at most existing PRB mining operations to decline at very
modest rates over the forecast horizon, with increasing strip ratios at least partly offset by improved usage
of labor and capital. As most PRB mines are progressing downward, the ratios of overburden to coal
(strip ratios) will increase in the future. The productivity of new mines will be quite low during the early
stages of their life span.
Mining at several locations is steadily proceeding westward toward the Joint Line railroad and, at current
and forecasted levels of production, around 2023 several mines are expected to eventually reach the line.
This event will result in a costly movement across the railroad, requiring significant capital investment and
reduced production as the transition is made. During the move across the Joint Line railroad, strip ratios
will spike, and productivity will fall as new box cuts are created.
7-18
-------
Illinois Basin (ILB)
Production costs in the Illinois basin have been mostly flat with a slight downward trend in recent years as
higher-cost mines close and newer low-cost longwall mines maximize their economies of scale.
Development of these longwalls has been delayed as natural gas prices largely remain below competitive
levels. New developments will be delayed until prices, and demand, recover. In the long-term, the shape
of the ILB supply curve has potential to increase production capacity and decrease costs. However, this
is not due to a lowering of costs at existing mines. Rather it is caused by new mines being coming online
that have lower operating costs than existing mines.
The ILB has vast reserves and potential for large-scale, low-cost mine development. However, a
shrinking customer base will pose a risk to the basin's growth potential as demand could shrink in the
long term.
Central Appalachia (CAPP)
Geologic conditions in the CAPP region are challenging, with thin seams and few underground reserves
amenable to more efficient longwall mining techniques. Costs of production in CAPP rose substantially in
the early 2010s as the region struggled with mining thinner seams depleting reserves. Mining accidents
led to increased inspections, and mine permitting has become increasingly difficult.
In the years prior to 2017, producers cut back production significantly as coal prices plummeted. Many
companies went bankrupt and closed a large proportion of mines. As a result, average costs fell
substantially as high-cost, low-productivity mines were closed. In an effort to retain margins, producers
implemented a variety of tactics at continuing operations to try to keep production costs from continuing to
increase, including shifting more production to lower cost operations and selling lesser quality raw coal to
save on coal preparation/washing costs. In the long term, costs will remain mostly flat as cost optimization
efforts continue within the highly competitive basin.
Northern Appalachia (NAPP)
Similar to CAPP, mining costs in NAPP have remained mostly flat since the closure of high-cost capacity
drove costs downwards. Future mine costs in Northern Appalachia will depend largely on the
development of new reserve areas. However, few thermal projects have been identified - meaning
located at an existing mine or a named project. The remainder are reserves that are available for
development in the region but no engineering or permitting work has begun.
7.3 Coal Transportation
Table 7-25 presents the coal transportation matrix.
Within the United States, steam coal for use in coal-fired power plants is shipped via a variety of
transportation modes, including rail, barge, truck, conveyor belt, and lake/ocean vessel. A given coal-
fired power plant typically has access to only a few of these transportation options and, in some cases,
has access to only a single option. The number of transportation options that a plant has when soliciting
coal deliveries influences transportation rate levels that plant owners are able to negotiate with
transportation providers.
Within the Eastern United States, rail service is provided predominately by two major rail carriers in the
region, Norfolk Southern (NS) and CSX Transportation (CSX). Within the Western United States, rail
service is also provided predominately by two major rail carriers, Burlington Northern Santa Fe (BNSF)
and Union Pacific (UP). Plants in the Midwestern United States may have access to rail service from
BNSF, CSX, NS, UP, the Canadian National (CN), Canadian Pacific (CP), or short-line railroads. Barge,
truck, and vessel service is provided by multiple firms, and conveyor service is only applicable to coal-
fired plants located next to mining operations (e.g., mine-mouth plants).
7-19
-------
Between 2016 (when the coal transportation rate assumptions for EPA Platform v6 November 2018
Reference Case were finalized), and 2020, coal production in the United States declined by
approximately 192 million tons/year, or 26% (from 728 million tons in 2016 to an estimated 536 million
tons in 2020.)77 Approximately 48 gigawatts of coal-fired generating capacity (or about 18% of the total
coal-fired generating capacity in the United States) retired in the period between the end of 2016 and the
end of September 2020.78
Transportation rate levels for most coal movements declined significantly in real terms between 2016 and
2020, as sustained low prices for natural gas and major expansions in renewable generation during this
period reduced the coal volumes used for electric generation further below the already low levels
experienced in 2016. However, since natural gas prices were very low throughout the 2016-2020 period
(averaging $2.65/MMBtu in nominal dollars between January 2016 and November 2020, at Henry Hub).79
the decline in coal transportation rates between 2016 and 2020 was not sufficient to make coal-fired
generation price-competitive with natural gas-fired generation in most areas of the U.S. Instead, the 2020
coal transportation rates shown in this analysis represent strategic decisions by the railroads and other
providers of coal transportation to preserve as much contribution margin as possible on the remaining
coal traffic (while accepting volume declines viewed as largely unavoidable), rather than competing
aggressively for incremental coal volumes. Rail rates for short-distance coal movements to captive plants
either stayed the same or increased in real terms between 2016-2020, as the railroads sought to partially
offset nationwide declines in coal volumes at the small subset of plants where they have the most market
power.
In this market environment, in which the railroads and other providers of coal transportation are generally
seeking to extract the maximum margins from coal traffic which is expected to steadily decline in volume
over the long term, any future arrangements tying coal transportation rates to natural gas pricing would
likely have to be very limited and site-specific (as was already the case in 2016.)
During 2021-2050, rates for most modes of coal transportation are expected to be flat to declining in real
dollars from the 2020 levels, reflecting relatively low levels of expected coal demand throughout the
forecast period used in EPA Platform v6.
The transportation methodology and rates presented below reflect expected long-run equilibrium
transportation rates as of August 2020, when the coal transportation rate assumptions for EPA Platform
v6 were finalized. The forecasted changes in transportation rates during the 2021-2050 forecast period
reflect expected changes in long-term equilibrium transportation rate levels, including the long-term
market dynamics that will drive these pricing levels.
All the transportation rates discussed in this document are expected 2020 rates and are shown in 2019
real dollars.
7.3.1 Coal Transportation Matrix Overview
Description
The general structure of the coal transportation matrix in EPA Platform v6 Post-IRA 2022 Reference Case
is similarto the structure used in EPA platform v6 November 2018 Reference Case. Each of the coal-fired
77 The coal production data cited here is U.S. Energy Information Administration (EIA) data. 2016-2019 data is from
the quarterly coal report released October 2020, is available at https://vwwy.eia.gov/coal/production/quarterlv/. 2020
data is estimated based on a 24.1% decline from 2019 coal production levels for 2020 year-to-date through
12/12/2020, as shown in ElA's Weekly Coal Production data (available at
https://www.eia.gov/coal/production/weeklY/).
78 Data from EIA Electric Power Monthly, February 2017, and November 2020 releases, available at
https://www.eia.gov/electricity/monthly/.
79 EIA data available at: https://www.eia.gov/dnav/ng/hist/rngwhhdm.htm
7-20
-------
power plants included in EPA Platform v6 Post-IRA 2022 Reference Case is individually represented in
the coal transportation matrix. This allows the coal transportation routings, coal transportation distances,
and coal transportation rates associated with each individual coal-fired generating plant to be estimated
on a plant-specific basis. The coal transportation matrix shows the total rate to transport coal from
selected coal supply regions to each individual coal-fired generating plant.
The coal supply regions associated with each coal-fired generating plant in EPA Platform v6 are largely
unchanged from the previous version of EPA Platform v6. The coal supply regions associated with each
coal-fired power plant are the coal supply regions which were supplying each plant as of the first half of
2020, have supplied each plant in previous years, or are considered economically and operationally
feasible sources of additional coal supply during the forecast period in EPA Platform v6. A more detailed
discussion of the coal supply regions can be found in previous sections.
Methodology
Each coal supply region and coal-fired power plant is connected via a transportation link, which can
include multiple transportation modes. For each transportation link, cost estimates, in terms of $/ton,
were calculated utilizing mode-based transportation cost factors, analysis of the competitive nature of the
moves, and overall distance that the coal type must move over each applicable mode. An example of the
calculation methodology for movements including multiple transportation modes is shown in Figure 7-8.
Figure 7-8 Calculation of Multi-Mode Transportation Costs (Example)
Rail Cost ($/ton) =
I Rate (mills/ton-mile) x Rail
Mileage
Transloading
Cost ($/ton)
Barge Cost ($/ton) =
Loading Cost ($/ton) + Barge Mill Rate
(mills/ton-mile) x Barge Mileage
Calculation of Coal Transportation Distances
Definition of applicable supply/demand regions
Coal-fired power plants are linked to coal supply regions based on historical coal deliveries, as well as
based on the potential for new coal supplies to serve each coal-fired generating plant going forward. A
generating plant will usually have transportation links with more than one supply region, depending on the
various coal types that can be physically delivered and burned at that particular plant. On average, each
coal-fired generating plant represented in IPM is linked with about eight coal supply regions. Some plants
may have more than the average number of transportation links, and some may have fewer, depending
on the location of each plant, the transportation modes available to deliver coal to each plant, the boiler
design and emissions control technologies associated with each plant, and other factors that affect the
types of coal that can be burned at each plant.
For mine-mouth plants (plants for which the current coal supply is delivered from a single nearby mine,
generally by conveyor belt or using truck transportation) that are 200 MW or larger, Hellerworx has
estimated the cost of constructing facilities that would allow rail delivery of alternative coal supplies, and
the transportation rates associated with the delivery of alternative coal supplies. This includes the
construction of rail spurs (between one and nine miles in length depending on the proximity of each plant
to existing railroad lines) to connect each plant with existing railroad lines.
7-21
-------
Transportation Links for Existing Coal-Fired Plants
Transportation routings from particular coal supply regions to particular coal-fired power plants were
developed based on third-party software80 and other industry knowledge available to Hellerworx. Origins
for each coal supply region were based on significant mines or other significant delivery points within the
supply region, and the destination points were plant-specific for each coal-fired generating plant
represented in IPM. For routes utilizing multiple modes (e.g., rail-to-barge, truck-to-rail, etc.), distances
were developed separately for each transportation mode.
Transportation Links for New Coal-Fired Plants
Representative coal transportation costs for new coal-fired power plants not yet under construction (i.e.,
coal transportation costs for a new coal-fired power plant modeled by IPM) were estimated by selecting
an existing coal-fired power plant within each IPM Region whose coal supply alternatives, and coal
transportation costs, were considered representative of the coal supply alternatives and coal
transportation costs that would likely be faced by new coal-fired power plants within that same IPM
Region. In cases where there are no existing coal plants within a particular IPM Region, the coal supply
alternatives and coal transportation costs applicable to that IPM Region were estimated using a
methodology similar to that used for the existing coal plants.81 Using this consistent methodology across
all of the IPM regions helps ensure that coal transportation costs for new coal plants are properly
integrated with and assessed fairly vis-a-vis existing coal-fired assets within the IPM modeling structure.
7.3.2 Overview of Rail Rates
Competition within the railroad industry is limited. Two major railroads in the Western U.S. (BNSF and
UP) and two major railroads in the Eastern U.S. (CSX and NS) currently originate most of the U.S. coal
traffic that moves by rail.
As noted earlier in this section, rail rates for most coal movements declined significantly during 2016-
2020, and coal demand for electric generation declined significantly as well. Continued strong
competition from natural gas-fired generation and renewables over the duration of the forecast period
used in EPA Platform v6 is expected to limit future coal demand, and to lead to further real declines in rail
rates over the long term.
The differential between rail rates at captive plants and rates at competitively served plants widened
slightly during 2016-2020, due to flat or increasing rates at the relatively small subset of coal-fired
generating plants where the railroads still have significant market power (short-distance movements to
captive plants).
Since August 2016, the Surface Transportation Board ("STB") has been engaged in a process (STB Ex
Part 665, Sub. No. 2, Expanding Access to Rate Relief) designed to make it easier for small shippers to
obtain rail rate relief from the STB. On September 11, 2019, the Board issued a Notice of Proposed
Rulemaking (NPRM), proposing to adopt Final Offer Rate Review as a rate setting mechanism. This
would be far cheaper and faster than the SAC approach. While designed for small rate cases, it is
obvious that the STB is searching for a means of making rate relief more widely available to shippers.
Whether this will be adopted, and if adopted withstand legal challenge is unknown, but the STB will likely
continue to seek ways to make its regulatory authority feasible for shippers to use. It is also unclear if
shippers would spend much to engage in a risky process to try and reduce rail rates to a coal fired power
plant with limited future prospects.
80 Rail routing and mileage calculations utilize ALK Technologies PC*Miler software.
81 Since the Canadian government has phased out coal-fired generation in Ontario, and in late 2016 announced plans
to phase out coal-fired generation in Alberta by 2030, coal-fired generation was not modeled in the Canadian
provinces where it is not currently used.
7-22
-------
However, it is unlikely, that any new regulatory mechanisms will have widespread impact on coal rates.
Under the legislation that currently governs rail rate relief (the Staggers Act, passed in 1980), the STB is
statutorily prohibited from mandating rates that are less than 180% of long-run variable costs (LRVC).
Very few rail rates for coal are set above this level (with the possible exception of some short-distance
movements to captive plants, which are a small segment of the total coal traffic.) Competition from
natural gas-fired generation has caused many high-cost coal plants to be shut down. Any future
regulations relating to greenhouse gas emissions would also add to coal's costs relative to all other fuel
sources. In summary, the market trends described throughout this analysis are likely to have much
greater impacts on rail rates for coal transportation than any future changes in the regulatory scheme.
All the rail rates discussed below include railcar costs and include fuel surcharges at expected 2020 fuel
price levels. When the rail rate assumptions used in EPA Platform v6 were finalized in August 2020, the
latest Form EIA-923 data that was available for the analysis of historical delivered coal prices and rail
rates was data through May of 2020. Therefore, almost all the data that was relied upon to estimate the
trends in historical rail rates between 2016 and early 2020 reflects rail contracts that would have been
negotiated prior to the beginning of the COVID-19 lockdowns in the United States (i.e., prior to mid-March
2020.) The forward-looking portion of the rail rate analysis (2021-2050) also focused on the expected
long-term trends within the coal and rail industries over the entirety of this 30-year period, rather than on
short-term disruptions related to COVID-19. Thus, neither the 2020 rail rate estimates nor the forecast of
expected long-term trends in rail rates should be biased by any short-term disruptions related to COVID-
19.
Overview of Rail Competition Definitions
Within the transportation matrix, rail rates are classified as being either captive or competitive (see Table
7-12) depending on the ability of a given coal demand region to solicit supplies from multiple suppliers.
Competitive rail rates are further subdivided into high- and low-cost competitive subcategories.
Competition levels are affected both by the ability to take delivery of coal supplies from multiple rail
carriers, the use of multiple rail carriers to deliver coal from a single source (e.g., BNSF/UP transfer to
NS/CSX for PRB coal moving east), or the option to take delivery of coal via alternative transportation
modes (e.g., barge, truck, or vessel).
Table 7-12 Rail Competition Definitions
Competition Type
Definition
Captive
Demand source can only access coal supplies through a single provider; demand
source has limited power when negotiating rates with railroads.
High-Cost Competitive
Demand source has some, albeit still limited, negotiating power with rail providers;
definition typically applies to demand sources that have the option of taking delivery
from either of the two major railroads in the region.
Low-Cost Competitive
Demand source has a strong position when negotiating with railroads; typically, these
demand sources also have the option of taking coal supplies via modes other than rail
(e.g., barge, truck, or lake/ocean vessel).
Rail Rates
As previously discussed, rail rates are subdivided into three competitive categories: captive, high-cost
competitive, and low-cost competitive. Moves are further subdivided based on the distance that the coal
supply must move over rail lines: <200 miles, 200-299 miles, 300-399 miles, 400-649 miles, and 650+
miles. Within the Western U.S., mileages are only subdivided into two categories (<300 miles and 300+
miles), given the longer distances that these coal supplies typically move.
Initial rate level assumptions were determined based on an analysis of recent rate movements, current
rate levels in relation to maximum limits prescribed by the STB, expected coal demand, diesel prices,
recent capital expenditures by railroads, and projected productivity improvements. In general, shorter
7-23
-------
moves result in higher applicable rail rates due to the lesser distance over which fixed costs can be
spread. As previously discussed, rail rates reflect anticipated 2020 costs in 2019 real dollars.
Rates Applicable to Eastern Moves
Rail movements within the Eastern U.S. are handled predominately by the region's two major carriers, NS
and CSX. Some short movements are handled by a variety of short-line railroads. Most plants in the
Eastern U.S. are served solely by a single railroad (i.e., they are captive plants). The practical effect of
this is that CSX and NS do not compete aggressively at the limited number of plants that have access to
both major railroads, and the rates for high-cost competitive plants tend to be similar to the rates for
captive plants. Table 7-13 presents the 2020 eastern rail rates.
Table 7-13 Assumed Eastern Rail Rates for 2020 (2019 mills/ton-mile)
Mileage Block
Captive
High-Cost Competitive
Low-Cost Competitive
< 200
122
122
104
200-299
71
71
60
300-399
57
57
48
400-649
53
53
45
650+
33
33
28
Prior to the EPA Platform v6 November 2018 Reference Case update in 2016, CSX introduced a new
structure for some of its rail contracts that includes both fixed and variable components. This was an
attempt to help coal-fired generating plants located on the CSX system compete more effectively with
natural gas-fired generation, by offering the generators the opportunity to include only the variable cost
component in their dispatching costs.
However, many larger generators (whose systems included both CSX-served plants, and plants served
by NS or other transportation providers) felt that this contracting structure might tend to favor CSX-served
plants at the expense of other plants on their own systems, and/or unnecessarily complicate dispatching.
Therefore, use of the contracting structure that includes fixed and variable rail rate components was
discontinued in EPA Platform v6 Post-IRA 2022 Reference Case. This change will have a very limited
effect on the IPM modeling for coal-fired generating plants, since this contracting structure was
experimental and was only used at a limited number of plants in EPA Platform v6 November 2018
Reference Case.
Rates Applicable to Midwestern Moves
Plants in the Midwestern U.S. may be served by BNSF, CN, CP, CSX, NS, UP or short-line railroads.
However, the rail network in the Midwestern U.S. is very complex, and most plants are served by only one
of these railroads. The Midwestern U.S. also includes a higher proportion of barge-served and truck-
served plants than is the case in the Eastern or Western U.S. Table 7-14 depicts 2020 rail rates in the
Midwest.
Table 7-14 Assumed Midwestern Rail Rates for 2020 (2019 mills/ton-mile)
Mileage Block
Captive
High-Cost Competitive
Low-Cost Competitive
< 200
122
122
104
200-299
80
80
68
300-399
57
57
48
400-649
57
57
48
650+
33
33
28
7-24
-------
Rates Applicable to Western Moves
Rail moves within the Western U.S. are handled predominately by BNSF and UP. Rates for Western coal
shipments from the PRB are forecast separately from rates for Western coal shipments from regions
other than the PRB. This reflects the fact that in many cases coal shipments from the PRB are subject to
competition between BNSF and UP, while rail movements of Western coal from regions other than the
PRB consist primarily of Colorado and Utah coal shipments that originate on UP, and New Mexico coal
shipments that originate on BNSF. PRB coal shipments also typically involve longer trains moving over
longer average distances than coal shipments from the other Western U.S. coal supply regions, which
means these shipments typically have lower costs per ton-mile than non-PRB coal shipments. In the
west, there are enough plants that have access to both BNSF and UP or a neutral carrier that the western
railroads are concerned with losing coal volume to the competing railroad and therefore offer more of a
rate discount to plants that can access both railroads (e.g., high-cost competitive).
Prior to the EPA Platform v6 November 2018 Reference Case update in 2016, BNSF offered temporary
spot rail rate discounts to a few selected generating plants using PRB coal to improve the utilization of
these plants during periods of unusually lower natural gas prices. However, since Hellerworx believes
that these discounts were only offered experimentally and temporarily to a few captive generating plants
using PRB coal in the Gulf Coast region, they were not modeled in EPA Platform v6 November 2018
Reference Case. The sustained low prices for natural gas during 2016-2020 appear to have made both
BNSF and UP even more reluctant to tie their rail rates to natural gas prices as of 2020 than they were in
2016. Therefore, the rail rate discounts related to natural gas pricing were also not modeled in EPA
Platform v6 Post-IRA 2022 Reference Case.
Over the forecast period, coal volumes are likely to continue to decline significantly from the 2020 levels
in most forecast scenarios. Therefore, other commodities such as intermodal traffic and oil which have
greater growth potential than coal are likely to become even more important strategically to the railroads
in the future than they are in 2020, and the railroads are expected to be generally unwilling to offer large
discounts from their base rates to compete for incremental coal volumes throughout the forecast period.
Non-PRB Coal Moves
The assumed non-PRB western rail rates for 2020 are shown in Table 7-15.
Table 7-15 Assumed Non-PRB Western Rail Rates for 2020 (2019 mills/ton-mile)
Mileage Block
Captive
High-Cost Competitive
Low-Cost Competitive
< 300
69
32
32
300+
40
28
28
The assumed PRB western rail rates for 2020 are available in Table 7-16.
PRB Moves Confined to BNSF/UP Rail Lines
Table 7-16 Assumed PRB Western Rail Rates for 2020 (2019 mills/ton-mile)
Mileage Block
Captive
High-Cost Competitive
Low-Cost Competitive
< 300
46
19
19
300+
21
15
15
PRB Moves Transferring to Eastern Railroads
For PRB coal moving west-to-east, the coal transportation matrix assumes that the applicable low-cost
competitive assumption is applied to the BNSF/UP portion of the rail mileage, and an assumption of either
$2.30 per ton or 28 mills per ton-mile (whichever is higher) is applied to the portion of the movement that
7-25
-------
occurs on railroads other than BNSF and UP. (The $2.30 per ton assumption is a minimum rate for short-
distance movements of PRB coal on Eastern railroads.)
7.3.3 Truck Rates
Truck rates include loading and transport components, and all trucking flows are considered competitive
because highway access is open to any trucking firm. The truck rates shown in Table 7-17 are expected
2020 rate levels, in 2019 dollars. The lower truck rates in EPA Platform v6 Post-IRA 2022 Reference
Case (as compared to EPA Platform v6 November 2018), reflect the fact that the actual change in diesel
fuel prices between 2016 and 2020 was significantly lower than was forecast in 2016.
Table 7-17 Assumed Truck Rates for 2020
Market
Loading Cost (2019 $/ton)
Transport (2019 mills/ton-mile)
All Markets
1.00
100
7.3.4 Barge and Lake Vessel Rates
As with truck rates, barge rates include loading and transport components, and all flows are considered
competitive because river access is open to all barge firms. The transportation matrix subdivides barge
moves into three categories, which are based on the direction of the movement (upstream vs.
downstream) and the size of barges that can be utilized on a given river. As with the other types of
transportation rates forecast in this analysis, the barge rate levels shown in Table 7-18 are expected 2020
rate levels, stated in 2019 dollars.
Table 7-18 Assumed Barge Rates for 2020
Type of Barge Movement
Loading Cost
(2019 $/ton)
Transport
(2019 mills/ton-mile)
Upper Mississippi River, and Downstream on the Ohio River System
3.80
12.2
Upstream on the Ohio River System
3.50
11.8
Lower Mississippi River
2.75
10.3
Notes:
1. The Upper Mississippi River is the portion of the Mississippi River north of St. Louis.
2. The Ohio River System includes the Ohio, Big Sandy, Kanawha, Allegheny, and Monongahela Rivers.
3. The Lower Mississippi River is the portion of the Mississippi River south of St. Louis.
Rates for transportation of coal by lake vessel on the Great Lakes were forecast on a plant-specific basis,
considering the lake vessel distances applicable to each movement, the expected backhaul economics
applicable to each movement (if any), and the expected changes in labor costs and fuel and steel prices
over the long-term.
7.3.5 Transportation Rates for Imported Coal
Transportation rates for imported coal reflect expectations regarding the long-term equilibrium level for
ocean vessel rates, considering expected long-run equilibrium levels for labor, fuel, and equipment costs.
In EPA Platform v6, it is assumed that imported coal is likely to be used only at plants that can receive
this coal by direct water delivery (i.e., via ocean vessel or barge delivery to the plant). The assumption is
based on an assessment of recent transportation market dynamics, which suggests that railroads are
unlikely to quote rail rates that will allow imported coal to be cost-competitive at rail-served plants.
Moreover, import rates are higher for the Alabama and Florida plants than for New England plants
because many of the Alabama and Florida plants are barge-served (which requires the coal to be
transloaded from ocean vessel to barge at an ocean terminal, and then moved by barge to the plant),
7-26
-------
whereas most of the New England plants can take imported coal directly by vessel. The assumed costs
are summarized in Table 7-25.
7.3.6 Other Transportation Costs
In addition to the transportation rates already discussed, the transportation matrix assumes various other
rates that are applied on a case-by-case basis, depending on the logistical nature of a move. These
charges apply when coal must be moved between different transportation modes (e.g., rail-to-barge or
truck-to-barge) - see Table 7-19.
Table 7-19 Assumed Other Transportation Rates for 2020
Type of Transportation
Rate (2019 $/ton)
Rail-to-Barge Transfer
2.00
Rail-to-Vessel Transfer
2.50
Truck-to-Barge Transfer
2.00
Rail Switching Charge for Short line
2.50
Conveyor
1.00
7.3.7 Long-Term Escalation of Transportation Rates
Overview of Market Drivers
According to data published by the Association of American Railroads (AAR), labor costs accounted for
about 33% of the rail industry's operating costs in 2018, and fuel accounted for an additional 16%. The
remaining 51% of the rail industry's costs relate primarily to locomotive and railcar ownership and
maintenance, and track construction and maintenance.
The performance of various cost indices for the railroad industry over the past four years (1Q2016-
1Q2020) is summarized in Figure 7-9. Since the lockdowns related to COVID-19 in the U.S. began on
March 16, 2020, the historical performance of the rail cost indices was assessed based largely on "pre-
COVID" data. This analysis period was selected in order to focus the analysis on the expected longer-
term performance of the rail cost indices during the majority of the 2021-2050 forecast period, and avoid
excessive bias toward the near-term economic disruptions related to COVID-19.
As shown in Figure 7-9, the RCAF82 Unadjusted for Productivity (RCAF-U), which tracks operating
expenses for the rail industry, increased at an annualized rate of 1.8% per year in nominal terms during
1Q2016-1Q2020. Since overall inflation (as measured by the GDP Chained Price Index increased by an
average of 1,9%/year during the same period, the railroad industry's operating costs decreased by an
average of 0.1%/year in real terms during 1Q2016-1Q2020.
As shown by the All-Inclusive Index Less Fuel (All-LF), the railroad industry's overall input costs excluding
fuel (e.g., labor and equipment costs) decreased by an average of 0.7%/year in real terms during
1Q2016-1Q2020. The railroad industry's labor costs decreased by an average of 0.4%/year in in real
terms during the same period.
82 The Rail Cost Adjustment Factor (RCAF) refers to several indices created for regulatory purposes by the STB, calculated by the
AAR, and submitted to the STB for approval. The indices are intended to serve as measures of the rate of inflation in rail inputs.
The meaning of various RCAF acronyms that appear in this section can be found in the insert in Figure 7-9.
7-27
-------
Since the railroads' labor force is largely unionized, Hellerworx considers the real decline in labor costs
during 1Q2016-1Q2020 to be an unusual event, and expects that, on average over the forecast period
used in EPA Platform v6, the rail industry's labor costs are likely to be flat in real terms.
However, since the volume of coal used for electric generation (and thus the volume of coal transported
by the rail industry) is expected to continue to decline significantly during the forecast period in most
forecast scenarios, there will likely be a long-term surplus of the rail equipment used for coal
transportation. Thus, the rail industry's equipment costs are expected to continue to decline in real terms,
by an average of 0.5% per year during the forecast period used in EPA Platform v6.
Figure 7-9 Rail Cost Indices Performance (1Q2016-1Q2020)
1.400
1.300 -
1.200 -
Performance of Rail Cost Indices, 1Q2016-1Q2020
(annualized nominal % change)
RCAF
Unadjusted
for
Productivity
(RCAF-U)
RCAF
Adjusted for
Productivity
(RCAF-U)
All-Inclusive
Index Less
Fuel (All-LF)
Labor
Component
of RCAF-U
PPI for
Industrial
Commodities
Less Fuel
(PPMCLF)
PPI for Rail
Equipment
(PPI-RE)
GDP
Chained
Price Index
1.8%
1.5%
1.2%
1.5%
1.9%
0.0%
1.9%
to 1.100
1.000
~ 0.900
'« 0.800
0.700
RCAF-U — —-RCAF-A
All-LF ~ RCAF-U Labor Comp.
¦ GDP-Chained Price Index PPI-ICLF PPI-RE
The other major transportation modes used to ship coal (barge and truck) have cost drivers broadly
similar to those for rail transportation (labor costs, fuel costs, and equipment costs). However, a
significant difference in cost drivers between the transportation modes relates to the relative weighting of
fuel costs for the different transportation modes. Estimates as shown in Figure 7-10 show that, at 201883
fuel prices, fuel costs accounted for about 16% of long-run marginal costs for the rail industry, 35% of
long-run marginal costs for barges, and 50% of long-run marginal costs for trucks
83 2018 was used as the reference point for fuel prices in this analysis because, at the time the coal transportation
rate assumptions used in EPA Platform v6 Summer Reference 2021 were finalized in August 2020, the latest
analysis of railroad operating expenses available from the AAR contained 2018 data.
7-28
-------
Figure 7-10 Long-Run Marginal Cost Breakdown by Transportation Mode
100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0% J ^ T ^ T
Rail Barge Truck
¦ Fuel ~ Other
7.3.8 Market Drivers Moving Forward
Diesel Fuel Prices
ICF's forecast of long-term equilibrium prices for diesel fuel used in the transportation sector (see Table
7-20) shows expected prices ranging from about $2.39/gallon in 2020 to about $2.88/gallon in 2050 (2019
real dollars). This represents an average annual real increase in diesel fuel prices of about 0.6%/year
during 2020-2050. The coal transportation rate forecast for EPA Platform v6 Post-IRA 2022 Reference
Case assumes that this average rate of increase in diesel fuel prices will apply over EPA's entire forecast
period.
This is a significantly lower rate of increase in diesel fuel prices than the average real increase of
2.0%/yearthat was assumed in EPA Platform v6 November 2018 Reference Case, based on the latest
forecast that was available from the U.S. Energy Information Administration as of mid-2016 (Annual
Energy Outlook 2016, Reference Case forecast for the price of diesel fuel used in the transportation
sector.)
7-29
-------
Table 7-20 EIA AEO Diesel Fuel Forecast, 2020-2050
Year
Rate (2019 $/gallon)
2020
2.39
2025
2.50
2030
2.79
2035
2.98
2040
2.95
2045
2.94
2050
2.88
Annualized % Change, 2021-2050
0.6%
Source: EIA
Labor Costs
As noted, labor costs for the rail industry are expected to increase at approximately the same rate as
overall inflation (flat in real terms), on average over the forecast period. Labor costs in the barge and
truck industries are also expected to increase at approximately the same rate as overall inflation, on
average over the forecast period used in EPA Platform v6.
Productivity Gains
The most recent data which was available from the AAR at the time the coal transportation rate
assumptions used in EPA Platform v6 were finalized in August 2020 (covering 2014-2018) show that rail
industry productivity increased at an annualized rate of approximately 1.0% per year during this period.
Since coal-fired generation is expected to continue to face strong competition from natural gas-fired
generation and renewables during the forecast period used in EPA Platform v6 (which will significantly
limit coal demand), approximately half of the railroad industry's expected productivity gains (0.5% per
year) are forecast to be passed through to coal shippers.
The potential for significant productivity gains in the trucking industry is relatively limited since truckload
sizes, operating speeds, and truck driver hours are all regulated by law. Although it is possible that
increasing use of electric vehicles may reduce trucking costs to some degree at some point during the
forecast period used in EPA Platform v6, both the timing and the magnitude of this change are very
difficult to quantify. Therefore, the potential impact of increasing use of electric vehicles has not been
included in the modeling of coal trucking rates for EPA Platform v6 Post-IRA 2022 Reference Case.
Although increased lock outages and the associated congestion on the inland waterway system as the
river infrastructure ages may reduce the rate of future productivity gains in the barge industry, limited
productivity gains are expected to occur, and these productivity gains are expected to be largely passed
through to shippers since the barge industry is highly competitive.
Long-Term Escalation of Coal Transportation Rates
Based on the foregoing discussion, rail rates are expected to decline at an average rate of 0.7% per year
in real terms during the 2021-2050 forecast period used in EPA Platform v6. Over the same period,
barge and lake vessel rates are expected to decrease at an average rate of 0.3% per year, which
includes some pass-through of productivity gains in those highly competitive industries. Truck rates are
expected to increase at an average rate of 0.3%/year during 2021 -2050, largely due to increases in fuel
costs. Rates for conveyor transportation and transloading services are expected to be flat in real terms,
on average over the forecast period used in EPA Platform v6.
7-30
-------
The basis for these forecasts is summarized in Table 7-21.
Table 7-21 Summary of Expected Escalation for Coal Transportation Rates, 2020-2050
Mode
Component
Component
Real Escalation
Productivity Gains
Real Escalation
Weighting
Before Productivity
Adjustment (%/year)
Passed Through to
Shippers (%/year)
After Productivity
Adjustment (%/year)
Rail
Fuel
Labor
Equipment
16%
33%
51%
0.60%
0.0%
-0.5%
Total
100%
-0.2%
0.5%
-0.7%
Barge &
Fuel
35%
0.6%
Vessel
Labor &
Equip.
65%
0.0%
Total
100%
0.2%
0.5%
-0.3%
Truck
Fuel
Labor &
Equip.
50%
50%
0.6%
0.0%
Total
100%
0.3%
0.0%
0.3%
Conveyor
Total
0.0%
0.0%
0.0%
Transloading
Total
0.0%
0.0%
0.0%
Terminals
7.3.9 Other Considerations
Estimated Construction Costs for Railcar Unloaders and Rail Spurs at Mine-Mouth Plants
To allow mine-mouth generating plants (i.e., coal-fired generating plants which take all of their current
coal supply from a single nearby mine) to access additional types of coal, the costs of constructing
facilities that would allow rail delivery of coal was estimated for almost all of the mine-mouth generating
plants with total capacity of 200 MW or more.
The facilities needed for rail delivery of coal to generating plants of this relatively large size were assumed
to be: a) a rotary dump railcar unloader capable of handling unit train coal shipments, which is estimated
to cost about $25 million installed (in 2019$). b) at least three miles of loop track, which would allow for
one trainload of coal to be unloaded, and a second trainload of coal to simultaneously be parked on the
plant site preparatory to unloading, and c) at least one mile of additional rail spur track to connect the
trackage on the plant site with the nearest railroad main line. Since construction costs for rail trackage
capable of handling coal trains is estimated at about $3 million per mile (in 2019$), the minimum
investment required to construct the facilities needed for rail delivery of coal was estimated at $37 million.
In some cases, the length of the rail spur required to reach the nearest main line (which was estimated on
a plant-specific basis) is considerably longer than one mile. In cases where a rail spur longer than one
mile was required to reach the main line, the cost of the additional trackage was estimated using the
same construction cost of $3 million per mile (2019$) referenced earlier.
The total cost of the facilities required for rail delivery of coal was converted to an annualized basis based
on the assumption that, for capital recovery estimation purposes, each plant's average coal burn during
the forecast period used in EPA Platform v6 should be discounted to 50% of the 2019 historical level84,
and a capital recovery factor of 10.58%.
84 This is intended to represent aa plausible estimate of the average coal burn that might occur at coal-fired
generating plants that remain operational for a significant portion of the 2021-2050 forecast period used in EPA
Platform v6, across a range of different forecasting scenarios.
7-31
-------
The cost of transporting additional types of coal to each mine-mouth generating plant was then calculated
using the same methodology described earlier in this section, and added to the annualized cost for the
rail delivery facilities, to arrive at an estimated all-in cost for delivering additional types of coal to the mine-
mouth plants.
7.4 Coal Exports, Imports, and Non-Electric Sectors Demand
The coal supply curves used in EPA Platform v6 represent the total steam coal supply in the United
States. While the U.S. power sector is the largest consumer of thermal coal - roughly 95% of U.S.
thermal coal consumption in 2019 was used in electricity generation - non-electricity demand must also
be taken into consideration in IPM modeling to determine the market-clearing price.85 Furthermore, some
coal mined within the U.S. is exported out of the domestic market, and some foreign coal is imported for
use in electricity generation, and these changes in the coal supply must be detailed in the modeling of the
coal supply available to coal power plants. The projections for imports, exports, and non-electric sector
coal demand are based on ElA's AEO 2020.
In EPA Platform v6, coal exports and coal-serving residential, commercial, and industrial demand are
designed to correspond as closely as possible to the projections in AEO 2020 both in terms of the coal
supply regions and coal grades that meet this demand. The projections used exclude exports to Canada,
as the Canadian market is modeled endogenously within IPM. First, the subset of coal supply regions
and coal grades in EPA Platform v6 are identified that are contained in or overlap geographically with
those in EIA Coal Market Module (CMM) supply regions and coal grades that are projected as serving
exports and non-electric sector demand in AEO 2020. Next, coal for exports and non-electricity demand
are constrained by CMM supply region and coal grade to meet the levels projected in AEO 2020. These
levels are shown in Table 7-22, Table 7-23 and Table 7-24.
Table 7-22 Coal Exports in v6 (Million Short Tons)
Name
2023
2025
2028
2030
2035
2040
2045
2050
Central Appalachia - Bituminous Low Sulfur
3.91
3.99
3.99
4.01
3.42
3.3
2.39
2.03
Central Appalachia - Bituminous Medium Sulfur
1.32
1.32
1.35
1.32
2.7
2.87
4.22
4.68
East Interior - Bituminous Medium Sulfur
6.07
8.06
4.2
4.2
4.23
0
0
0
Northern Appalachia - Bituminous High Sulfur
7.58
3.9
2.91
2.34
1.47
0.75
11.14
11.14
Northern Appalachia - Bituminous Medium Sulfur
1.3
5.64
6.73
7.36
8.57
9.27
1.34
1.3
Rocky Mountain - Bituminous Low Sulfur
6.2
6.07
5.9
5.8
5.59
5.43
5.31
5.21
Western Montana Subbituminous Medium Sulfur
6.09
7.39
10.31
10.3
8.57
7.82
7.98
8.7
WY PRB - Subbituminous Low Sulfur
5.69
4.55
2.08
2.2
3.87
4.59
4.44
3.74
IPM then endogenously determines which IPM coal supply region(s) and coal grade(s) will be selected to
meet the required export or non-electric sector coal demand as part of the cost-minimization coal market
equilibrium. Since there are more coal supply regions and coal grades in EPA Platform v6 than in AEO
2020, the specific regions and coal grades that serve export and non-electric sector demand are not pre-
specified but modeled.
Table 7-23 Residential, Commercial, and Industrial Demand in v6 (Million Short Tons)
Name
2023
2025
2028
2030
2035
2040
2045
2050
Arizona/New Mexico - Bituminous Low Sulfur
0.11
0.11
0.11
0.11
0.11
0.11
0.11
0.12
Arizona/New Mexico - Subbituminous Medium Sulfur
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
Central Appalachia - Bituminous Low Sulfur
1.37
1.39
1.39
1.38
1.26
1.23
1.2
1.19
Central Appalachia - Bituminous Medium Sulfur
4.74
4.84
4.85
4.8
4.41
4.27
4.19
4.12
Dakota Lignite - Lignite Medium Sulfur
4.43
4.55
4.62
4.62
4.42
4.32
4.25
4.19
East Interior - Bituminous High Sulfur
4.39
4.51
4.57
4.57
4.38
4.29
4.22
4.17
East Interior - Bituminous Medium Sulfur
0.34
0.34
0.34
0.34
0.31
0.3
0.29
0.28
85 https://vwwy.eia.gov/coal/annual/pdf/acr.pdf
7-32
-------
Name
2023
2025
2028
2030
2035
2040
2045
2050
Northern Appalachia - Bituminous High Sulfur
0.4
0.41
0.42
0.42
0.4
0.39
0.39
0.38
Northern Appalachia - Bituminous Medium Sulfur
1
1.05
1.05
1.04
0.97
0.96
0.92
0.9
Rocky Mountain - Bituminous Low Sulfur
4.94
5.08
5.13
5.12
4.86
4.81
4.83
4.89
Southern Appalachia - Bituminous Low Sulfur
0.12
0.12
0.12
0.12
0.11
0.11
0.11
0.1
Southern Appalachia - Bituminous Medium Sulfur
0.73
0.75
0.74
0.73
0.66
0.63
0.62
0.6
West Interior - Bituminous High Sulfur
0.28
0.28
0.28
0.28
0.26
0.25
0.25
0.24
Western Montana - Subbituminous Low Sulfur
1.46
1.5
1.52
1.52
1.46
1.43
1.41
1.39
Western Wyoming - Subbituminous Low Sulfur
0.71
0.73
0.74
0.74
0.71
0.71
0.73
0.75
Western Wyoming - Subbituminous Medium Sulfur
0.88
0.91
0.92
0.92
0.88
0.89
0.91
0.94
WY PRB - Subbituminous Low Sulfur
2.41
2.48
2.51
2.51
2.4
2.35
2.31
2.28
Imported coal86 is only available to 19 coal facilities, which are eligible to receive imported coal. These
facilities, which may receive imported coal, along with the cost of transporting this coal to the demand
regions, are in Table 7-25. The total U.S. imports of steam coal are limited to AEO 2020 projections as
shown in Table 7-24.
Table 7-24 Coal Import Limits in v6 (Million Short Tons)
2023
2025
2028
2030
2035
2040
2045
2050
Annual Coal Imports Cap
0.74
0.42
0.53
0.01
0.01
0.01
0.01
0.01
86 Imported coal is assumed to have a SO2 emission factor of 1.1 Ibs/MMBtu, a mercury emission factor of 7.74
Ibs/TBtu, and a HCI emission factor of 0.018 Ibs/MMBtu.
7-33
-------
Attachment 7-1 Mining Cost Estimation Methodology and Assumptions
Labor Costs
Productivity and labor cost rates are utilized to estimate the total labor cost associated with the mining
operation. The estimate excludes labor involved in any coal processing / preparation plant.
Labor productivity is used to calculate mine labor and salaries by applying an average cost per employee
hour to the labor productivity figure reported by MSHA or estimated based on comparable mines.
Labor cost rates are estimated based on employment data reported to MSHA. MSHA data provides
employment numbers, employee hours worked, and tons of coal produced. These data are combined
with labor rate estimates from various sources such as union contracts, census data and other sources
such as state employment websites to determine a cost per ton for mine labor. Hourly labor costs vary
between United Mine Workers of America (UMWA) and non-union mines and include benefits and payroll
taxes. Employees assigned to preparation plants, surface activities, and offices are excluded from this
category and are accounted for under coal washing costs and mine overhead.
Surface Mining
The prime (raw coal) strip ratio and overburden volume is estimated on a year-by-year basis. Estimates
are entered of the amount of overburden 87 moved each year, split by method to allow for different unit
mining costs. The unit rate cost for each method excludes any drill and blast costs, and labor costs, as
these are accounted for separately. Drill and blast costs are estimated as an average cost per volume of
prime overburden. If applicable, dragline re-handle is estimated separately, and a summation gives the
total overburden moved.
• The different overburden removal methods are:
• Dragline - the estimated volume of prime overburden moved
• Dragline re-handle - the estimated volume of any re-handled overburden
• Truck and shovel - including excavators.
• Other - examples would be dozer push, front end loader, or cast blasting. If overburden is moved
by cast blasting the unit rate is taken to be zero as the cost is already included in the drill and
blast estimate.
• Surface mining costs also include the cost of coal mining estimated on a raw ton basis.
Underground Mining
Raw coal production is split by type into either continuous miner or longwall. Cost estimates can be input
either on a unit rate or a fixed dollar amount, as the cost structure of underground mining generally has a
large, fixed component from year to year. Costs are divided into:
• Longwall
• Continuous miner
• Underground services
Underground services costs cover categories such as ventilation, conveyor transport, gas drainage, and
secondary roof support etc.
87 Overburden refers to the surface soil and rock that must be removed to uncover the coal.
7-34
-------
Mine Site Other
This covers any mine site costs that are outside the direct production process. Examples are ongoing
rehabilitation/reclamation, security, and community development costs.
Raw Haul
Costs for transporting raw coal from the mining location to the raw coal stockpile at the coal preparation
plant or rail load out. A distance and a unit rate allow for an increasing cost over time if required.
List of tables that are uploaded directly to the web:
Table 7-25 Coal Transportation Matrix in EPA Platform v6 Post-IRA 2022 Reference Case
Table 7-26 Coal Supply Curves in EPA Platform v6 Post-IRA 2022 Reference Case
7-35
-------
8. Natural Gas
This chapter discusses the representation of and assumptions for natural gas. The chapter starts with a
brief synopsis of ICF's Gas Market Model (GMM), the primary tool used for generating the natural gas
supply curves. This is followed by discussion of the approach taken to translate GMM results to IPM
inputs for the EPA's Platform v6 Post-IRA 2022 Reference Case (EPA Platform v6). Lastly, brief
descriptions of modeling methodologies and data used in GMM are presented.
Natural gas supply curves and seasonal basis differentials are key inputs to IPM and are developed using
GMM. GMM and IPM are iterated in tandem to develop a forecast of Henry Hub gas price and total
power sector gas demand that informs the derivation of the supply curves. The approach is described as
follows:
• IPM takes the natural gas supply curves, which are developed based on GMM outputs and
specified as a function of Henry Hub prices.
• For each year, delivered price adders and three sets of seasonal natural gas transportation
differentials (summer, winter, and winter shoulder) are added to the supply curves to generate the
final delivered curves by IPM region.
• IPM projects the power sector's demand for natural gas. The projected demand is then matched
with the supply curve to find the market-clearing price.
• IPM's linear programming formulation takes into consideration the gas supply curves, as well as
competing fuels such as coal, and detailed power plant modeling in determining electric market
equilibrium conditions.
Like IPM, GMM is a large-scale linear programming model that incorporates a detailed representation of
gas supply characteristics, demand characteristics, and an integrating pipeline transportation model to
develop forecasts of gas supply, demand, prices, and flows. GMM is a full supply/demand equilibrium
model of the North American gas market. The model solves for monthly natural gas prices throughout
North America, given different supply/demand conditions, the assumptions for which are specified by
each scenario.
On the supply side, prices from GMM are determined by production and storage price curves that reflect
prices as a function of production and storage utilization. Prices are also influenced by "pipeline discount"
curves, which reflect the change in basis or the marginal value of gas transmission as a function of load
factor. On the demand side, prices are represented by a curve that captures the fuel-switching behavior
of end-users at different price levels. The model balances supply and demand at all nodes in the model
at the market clearing prices. Figure 8-1 shows the supply side of the calculation in GMM, and Figure 8-2
shows the interaction of IPM and GMM.
8-1
-------
Figure 8-1 GMM Gas Quantity and Price Response
Production And
Storage Gas Price
Pipeline
Value
Gas
Price
Deliverability
100%
, Inelastic ,
Demand
Distillate \
Switching \
Residual Oil Switching
Production
Production
And Storage
Only Includes Storage
During The Withdrawal Season
Natural Gas Modeling
Pipeline Load Factor
Gas
Transmission
Quantity Consumed
Gas
Demand
Includes Storage
During The Injection Season
Figure 8-2 IPM/GMM Interaction
Power Sector Modeling
Natural Gas Supply Curves
Power Sector Gas Demand
Integrated
Planning
Model flPM)
Activities ,
f
Assumption Updates
Scenario Analysis
f
Activities
i
3
r
r
Assumption Updates
Scenario Analysis
¦i
r
Products
Natural Gas Projections
Natural Gas Supply Curves
Products
Environmental Policy Assessments
Inputs for Air Quality Modeling
8-2
-------
8.1 GMM
GMM is designed to perform comprehensive assessments of the entire North American gas flow pattern.
It is a large-scale, dynamic linear program that models economic decision-making to minimize the overall
cost of meeting natural gas demand. GMM is reliable and efficient in analyzing the broad range of natural
gas market issues. Figure 8-3 presents the geographic coverage of GMM.
Figure 8-3 Geographic Coverage of GMM
GMM Pipeline Network
um («c
Copy right 2020, ICF
Important features of GMM are described below.
Natural Gas Market Prices in GMM are determined by the marginal (or incremental) value of natural gas
at 121 regional market centers. The regional market centers are also referred to as nodes. Prices are "at
the margin", not "average." Marginal prices do not translate directly into pipeline or utility revenues.
Prices represent "market center" prices as opposed to delivered prices. Gas prices are determined by the
balance of supply and demand in a regional marketplace. Supply is determined considering both
availability of natural gas deliverability at the wellhead, the transportation capacity and cost to deliver gas
to market centers.
Natural gas prices are determined from spot gas price curves that yield price as a function of deliverability
utilization: Curves reflect price for gas delivered into the transmission system (including gathering cost).
Gas storage withdrawal price curves are added to the production price curves during the withdrawal
season. Pipeline value curves are then added to yield a total supply curve for a node. The intersection of
the supply curve and the demand curve (including net storage injections) yields the marginal price at a
node. Price is set by the demand curve when all available supply is utilized.
8-3
-------
Demand is modeled for residential, commercial, industrial, and power sectors for each of the 121 nodes.
GMM solves for gas demand across different sectors, given economic growth, weather, and the level of
price competition between gas and oil. Econometric equations define demand by sector. The industrial
and power sectors incorporate fuel competition, dispatch decisions, new power plant builds, economic
growth, and weather. GMM solves the power generation dispatch on a regional basis to determine the
amount of gas used in power generation, which is allocated along with end-use gas demand to model
nodes. GMM iterates with IPM to better capture electric sector demand for natural gas.
Transportation is modeled by over 530 transportation links between the nodes, balancing seasonal,
sectoral, and regional demand and prices, including pipeline tariffs and capacity allocation. Node
structure was developed to reflect points of change or influence on the pipeline system. These points
include major demand and supply centers, pipeline hubs and market centers, and points of divergence in
pipeline corridors.
Pipeline capacity expansions address the physical constraints of transporting gas from supply regions to
demand regions. They therefore contribute to determining the supply curves and seasonal basis. For the
near-term, pipeline capacity expansions are input to GMM based on identifiable, near-term development
plans and ICF's market assessment. For the longer term, new "generic" pipeline capacity is added in
GMM when the market value of the added capacity exceeds its cost. Generic pipeline capacity in the
model can be added starting 2025 and is deployed in response to expected growth in natural gas
markets.
ICF includes projects that satisfy certain criteria in its analysis. The criteria are listed below.
• First Criteria: The project is already under construction; OR...
• Second Criteria: The project has the necessary approvals to proceed from FERC and other
relevant regulatory proceedings; OR...
• Third Criteria: The project has been filed with FERC and has the necessary firm shipper
commitments; OR...
• Fourth Criteria: The project has been filed with FERC and does not have the necessary shipper
commitments, but does appear to have sufficient market support; OR...
• Fifth Criteria: The project has NOT yet been filed with FERC but appears to have sufficient
market support.
For the fourth and fifth criteria, ICF typically considers supply growth directly upstream of the project,
market growth for markets that are relevant to the project's delivery point/s, and basis differentials that
exceed the per unit cost of pipeline expansion as indicators of market support. If the indicators are all
positive, ICF will add the project as a "generic" project and size it based on the level of market support. In
the case in which there are multiple generic projects for a single GMM link, the generic projects will be
sized in aggregate based on the total level of market support for expansion of the link. Generic projects
are classified as such until one of the first three criteria are satisfied.
For certain markets like New York, New Jersey, and New England, ICF looks closely at regulatory support
for the project which could override the criteria above in determining the pipeline additions in GMM. For
example, if a project like Northeast Supply Enhancement Project (NESE) has been denied water permits
even though it has broad market support, ICF does not include it in its base case.
Pipeline cost assumptions used in GMM have been derived by considering data from Oil and Gas
Journal (OGJ) surveys of pipeline projects. Using regression analysis of the OGJ data across years, we
estimated an average U.S. pipeline cost of $228,000 per inch-mile for 2019 (in 2019 dollars) for large gas
transmission pipelines. The pipeline cost for future years is kept flat in real terms post 2019. Regional
cost multipliers have also been derived from OGJ data as the pipeline costs vary by region. Cost
8-4
-------
multipliers can be different across regions; for example, costs are relatively high in the Northeast where
projects have been very difficult and time consuming to construct.
Supply is modeled by using node-level natural gas deliverability or supply capability, including import and
export levels while accounting for gas storage injections and withdrawals at different gas prices. Total
supply in the United States comes from three sources: production from natural gas fields located in the
lower 48 states, Canadian imports, Alaska, and LNG imports/exports. Natural gas production activity is
represented in 82 of the 121 model nodes where historical production has occurred, or where future
production appears likely.
Natural Gas Storage activity is represented for 24 United States and two Canadian storage regions, with
activity allocated to individual nodes based on historical field level storage capacity. Regional differences
in the physical and market characteristics of storage are captured in the storage injection and withdrawal
relationships separately estimated for each region.
Net monthly withdrawals are calculated from a "storage supply curve" that reflects the level of withdrawals
relative to gas prices. The curve has been fit to actual historical data. Net monthly injections are
calculated from econometrically fit relationships that consider working gas levels, gas prices, and weather
(i.e., cooling degree days). The level of gas storage withdrawals and injections are calculated within the
supply and demand balance algorithm based on working gas levels, gas prices, and extraction/injection
rates and costs.
Storage levels have an impact on GMM's seasonal basis differentials, which are an important component
in constructing the gas supply curves and/or basis differentials that are then input into IPM. The arbitrage
value of storage is driven by the seasonal difference in the supply-area gas prices plus the seasonal
difference in pipeline transportation value. Storage expansions (or increased utilization of existing
storage) decreases seasonal basis differentials in the region surrounding the storage facilities.
8.2 Translating GMM Results to IPM Natural Gas Supply Curves88
In this section, we describe GMM results underlying the natural gas supply curves for EPA Platform v6. A
typical GMM run generates the following outputs:
• Natural gas prices
• Natural gas production by region
• Natural gas consumption by region and sector
Table 8-1 summarizes the supply/demand balance and Henry Hub price for a GMM run reflecting natural
gas assumptions as of end of 2021. The GMM run underlies the natural gas supply curves. The regional
breakout in the demand/supply data is by census region and the mapping to the state and GMM nodes is
provided in Figure 8-4 and Figure 8-5. Table 8-8 provides additional results.
Table 8-1 Supply/Demand Balance and Henry Hub Price for a GMM Run Underlying the Natural
Gas Supply Curves in v6
Demand (Bcf per year)
2028
2030
2035
2040
2045
2050
New England
831
852
778
774
788
775
Mid-Atlantic
4,538
4,658
4,331
3,681
3,627
3,684
East North Central
5,267
5,386
5,044
4,920
4,805
4,746
West North Central
2,007
1,983
1,809
1,779
1,758
1,745
South Atlantic
4,910
5,064
4,869
4,422
3,997
3,675
88 The GMM results presented in this section are illustrative and consistent with a draft version of the EPA Platform
v6. GMM was not rerun for a final calibration with EPA Platform v6 using IPM.
8-5
-------
Demand (Bcf per year)
2028
2030
2035
2040
2045
2050
East South Central
2,234
2,237
1,968
1,794
1,830
1,784
West South Central
7,663
7,421
6,778
6,513
6,415
6,287
Mountain
2,272
2,380
2,426
2,177
2,321
2,174
Pacific (contiguous)
2.897
2,700
2,551
2,319
2,168
2,153
Alaska
284
281
274
274
274
274
Total L-48
32,620
32,680
30,554
28,379
27,709
27,023
Total United States
32,904
32,961
30,827
28,652
27,983
27,297
Exports/Imports (Bcf per year)
Net LNG Exports from US
5,648
5,949
6,058
6,043
6,039
6,111
Net Pipeline Exports to Mexico
2,981
2,981
2,945
2,909
2,887
2,874
Net Pipeline Imports from Canada
1,331
1,438
1,094
816
1,029
969
Supply (Bcf per year)
New England
0
0
0
0
0
0
Mid-Atlantic
8.481
8,283
7,176
6,435
6,192
6,004
East North Central
2,906
2,913
2,827
2,778
2,774
2,755
West North Central
1,261
1,240
1,184
1,148
1,136
1,133
South Atlantic
2,506
2,492
2,365
2,280
2,244
2,205
East South Central
579
497
338
169
131
106
West South Central
20,556
20,546
20,672
19,922
19,546
19,404
Mountain
4,459
4,346
3,889
3,543
3,377
3,237
Pacific (contiguous)
157
157
158
152
144
138
Alaska
286
273
284
281
289
288
Total L-48
40,904
40,475
38,610
36,427
35,544
34,981
Total United States
41,190
40,748
38,894
36,708
35,832
35,269
2028
2030
2035
2040
2045
2050
Henry Hub, 2G19$/MMBtu
2.90
2.53
2.02
2.15
2.37
2.33
New
England
Mountain
West North
Central
East South
Central
South
Atlantic
West South
Central
Atlantic
Offshore
California
Offshore
Elba
Island
LNG
Juarez
Reynosa
Copyright 2018, ICF
8-6
-------
Figure 8-5 Supply Region Definition
Canada
West Coast
Rocky Mountain
Midcontinent
Northeast
Southwest
Northern
Nodes
Everett
< LNG
Atlantic
OUshore
California
Otlshore
Juarez
Lake
Charles LNG
Reynosa
Copyright 2018, ICF
8.2.1 Supply Curves for EPA Platform v6
Henry Hub is a pipeline interchange hub in Louisiana Gulf Coast near Erath, LA, where eight interstate
and three intrastate pipelines interconnect. Liquidity at this point is very high and it serves as the primary
point of exchange for the New York Mercantile Exchange (NYMEX) active natural gas futures markets.
Henry Hub prices are considered as a proxy for U.S. natural gas prices. Natural gas from the Gulf moves
through the Henry Hub onto long-haul interstate pipelines serving demand centers. Due to the
importance and significance of Henry Hub, GMM generated supply curves are specified at Henry Hub
prices.
For IPM modeling, GMM generates a price forecast over a time horizon and a set of time dependent
price/supply curves based on that price path for each year in the forecast. For each year, the mid-point
price of the supply curve is set equal to the solved Henry Hub price from GMM, and the mid-point volume
is set equal to the solved gas consumption for the power sector from GMM. Each supply curve's elasticity
is set equal to the effective price-elasticity for gas supply in that year. In this manner, even while GMM
has itself projected particular levels of gas supply and consumption (and corresponding market-clearing
prices) overtime, the information included in those projections is input into IPM in the form of gas supply
curves that enable IPM to solve for levels of power sector gas consumption and resulting gas prices that
respect a least-cost power production future. The power generation gas use by model region from IPM
run outputs are used as inputs in GMM to generate a new set of supply curves and basis which are used
by IPM as inputs for the next iteration. This iteration process is repeated until the power generation gas
use from IPM and GMM converge.
The final resulting supply curves developed for years 2028, 2030, 2035, 2040, 2045, and 2050 are shown
in Figure 8-6 and Table 8-10. Over time, gas supply becomes more price elastic because producers
8-7
-------
have more time to respond to the market changes. In the longer term, resource depletion tends to offset
elasticity making the curves slightly less elastic than they are between 2028 and 2030.
Figure 8-6 Supply Curves for 2028, 2030, 2035, 2040, 2045, and 2050
«/»
a
o
CN|
0)
o
$7.00
$6.00
$5.00
$4.00
$3.00
$2.00
$1.00
$0.00
*
4
2028
2030
¦2035
¦2040
>2045
¦2050
5 6 7 8 9 10 11 12 13 14 15 16 17 18 192021 222324 25
Quantity (TBtu)
If fewer than 10 run periods are defined, ignore the "0" values in legend.
The static national supply curves used for EPA Platform v6 are robust for typical scenario analysis,
although EPA reevaluates price dynamics in scenarios to ensure that IPM and GMM are iterated in cases
where the regional natural gas demand in the power sector is expected to be significantly different from
the reference case.
8.2.2 Basis
Basis is the difference in gas price in a given market from the widely used Henry Hub reference price.
Basis reflects the price in a given market based on demand, available supply, and the cost of transporting
gas to that location. A negative basis value represents that the gas price in that area is lower than the
Henry Hub price. Basis between two nodes in GMM is the difference in prices between the two nodes.
The GMM utilizes its network of 121 nodes that comprises 530 gas pipeline corridors to assess the basis
between two desired nodes. The pipeline corridors between nodes are represented by pipeline links and
can be characterized by their maximum capacity. Each of the links has an associated discount curve
(derived from GMM natural gas transportation module), which represents the marginal value of gas
transmission on that pipeline segment as a function of the pipeline's load factor. The basis value is
calculated by using the supply/demand balance in two nodes along with the resulting prices in each node
and the cost of transporting gas between the two nodes as determined by the discount curve on that link.
The discount curve is a function of the pipeline tariffs and the load factor. The discount curves are
continuously calibrated to accurately reflect historical basis values. Their parameters can be adjusted to
account for regulatory changes that can affect pipeline values.
8-8
-------
The GMM solves for basis monthly. Basis pressure (i.e., spiking basis) will generally occur when average
monthly load factors rise to above 80%. Since many U.S. markets are winter peaking, the higher basis
typically occurs in the winter months when gas use and pipeline utilization are highest. The IPM relies on
seasonal basis that reflects averages of the monthly basis values solved for in the GMM for three
seasons. IPM uses the gas supply curves and regional price relationships (differentials) on a seasonal
basis overtime as inputs, based on GMM-projected future of gas supply/demand. While EPA Platform v6
has the flexibility to re-determine the relationship of power sector gas demand to supply and to
accordingly find different gas price futures, EPA Platform v6 will maintain the future (basis differential)
price relationship between Henry Hub and each regional location in a national supply picture as originally
determined by these GMM projections. Table 8-9 provides the full set of seasonal basis differentials at the
IPM region level.
8.2.3 Delivered Price Adders
As stated in Section 8.1, GMM prices are market center prices and not delivered prices. To estimate
delivered prices at a power plant, an adder is applied to the seasonal basis from GMM. The delivered
price adder is calculated for each state by comparing the GMM historical prices with historical delivered
gas prices to electric power plants based on EIA-176 data. The delivered price adders implemented in
EPA Platform v6 are shown in Table 8-2.
Table 8-2 Delivered Price Adders
State
Adder (2019$/MMBtu)
State
Adder (2019$/MMBtu)
Alabama
0.01
Nebraska
0.54
Arizona
0.03
Nevada
0.15
Arkansas
0.14
New Hampshire
-
California
0.22
New Jersey
0.20
Colorado
0.19
New Mexico
0.03
Connecticut
0.05
New York
0.20
Delaware
0.01
North Carolina
0.31
Florida
0.02
North Dakota
0.04
Georgia
0.00
Ohio
0.04
Idaho
0.06
Oklahoma
0.02
Illinois
0.15
Oregon
0.01
Indiana
0.13
Pennsylvania
0.04
Iowa
0.14
Rhode Island
0.00
Kansas
0.15
South Carolina
0.15
Kentucky
0.17
South Dakota
0.01
Louisiana
0.04
Tennessee
0.03
Maine
0.03
Texas
0.22
Maryland
0.16
Utah
0.12
Massachusetts
0.03
Virginia
0.07
Michigan
0.16
Washington
0.11
Minnesota
0.40
West Virginia
0.14
Mississippi
0.03
Wisconsin
0.17
Missouri
0.12
Wyoming
0.11
Montana
0.45
Canada
0.15
8.3 GMM Assumptions
This section describes the key GMM assumptions and data used for EPA Platform v6.
8.3.1 GMM Resources Data and Reservoir Description
This section describes the approach used in GMM and documents the changes to the resource data and
reservoir characterization work conducted for EPA Platform v6.
8-9
-------
U.S. Resources and Reserves
This section describes the U.S. resource data sources and methodology used in GMM for EPA Platform
v6.
Current U.S. and Canada gas production is from over 470 trillion cubic feet (Tcf) of proven gas reserves.
ICF assumes that the U.S. and Canada natural gas resource base totals roughly 4,000 Tcf of unproved
plus discovered but undeveloped gas resource. This can supply the U.S. and Canada gas markets for
over 100 years (at current consumption levels). Shale gas accounts for over 50 percent of remaining
recoverable gas resources. No significant restrictions on well permitting and fracturing are assumed
beyond restrictions that are currently in place.
Data sources: Conventional resource base assessment is based on data from the U.S. Geological Survey
(USGS), Minerals Management Service (MMS), and Canadian Gas Potential Committee (CGPC) using
ICF's Hydrocarbon Supply Model (HSM).
In the area of unconventional gas, ICF has worked for many years with the Gas Research Institute
(GRI)/Gas Technology Institute (GTI) to develop a database of tight gas, coalbed methane, and Devonian
Shale reservoirs in the U.S. and Canada. Along with USGS assessments of continuous plays, the
database was used to help develop the HSM's "cells", which represent resources in a specific geographic
area, characterizing the unconventional resource in each basin, historical unconventional reserves
estimates and typical decline curves. ICF has built up a database on gas compositions in the United
States and has merged that data with production data to allow the analysis of net versus raw gas
production.
Resources are divided into three general categories: new fields/new pools, field appreciation, and
unconventional gas. The methodology for resource characterization and economic evaluation differs for
each.
New Fields
Conventional new discoveries are characterized by size class. For the United States, the number of
fields within a size class is broken down into oil fields, high permeability gas fields, and low permeability
gas fields based on the expected occurrence of each type of field within the region and interval being
modeled. The fields are characterized further as having a hydrocarbon make-up containing a certain
percent each of crude oil, dry natural gas, and natural gas liquids. In Canada, fields are oil, sweet non-
associated gas, or sour non-associated gas.
The methodology uses a modified "Arps-Roberts" equation to estimate the rate at which new fields are
discovered. The fundamental theory behind the find-rate methodology is that the probability of finding a
field is proportional to the field's size as measured by its areal extent, which is highly correlated to the
field's level of reserves. For this reason, larger fields tend to be found earlier in the discovery process
than smaller fields. The new equation developed by ICF accurately tracks discovery rates for mid- to
small-size fields. Since these are the only fields left to be discovered in many mature areas, the more
accurate find-rate representation is an important component in analyzing the economics of exploration
activity in these areas.
An economic evaluation is made in the model each year for potential new field exploration programs
using a standard discounted after-tax discounted cash flow (DCF) analysis. This DCF analysis takes into
account how many fields of each type are expected to be found and the economics of developing each.
The economic decision to develop a field is made using "sunk cost" economics where the discovery cost
is ignored, and only time-forward development costs and production revenues are considered. However,
the model's decision to begin an exploration program includes all exploration and development costs.
8-10
-------
Field Appreciation
Field appreciation refers to potential resources that can be proved from already discovered fields. These
inventories are referred to as appreciation, growth-to-known or "probables." The inventories of probables
are increased due to expected future appreciation due to many factors that include higher recovery
percentages of the gas in-place resulting from infill drilling and application of improved technology and
experience gained in the course of developing and operating the field.
Unconventional Gas
The ICF assessment method for shale gas is a "bottom-up" approach that first generates estimates of
unrisked and risked gas-in-place (GIP) from maps of depth, thickness, organic content, and thermal
maturity. Then, ICF uses a different model to estimate well recoveries and production profiles. Unrisked
GIP is the amount of original gas-in-place determined to be present based upon geological factors—
without risk reductions. "Risked GIP" includes a factor to reduce the total gas volume based on proximity
to existing production and geologic factors such as net thickness (e.g., remote areas, thinner areas, and
areas of high thermal maturity have higher risk). ICF calibrates expected well recoveries with specific
geological settings to actual well recoveries by using a rigorous method of analysis of historical well data.
To estimate the contributions of changing technologies ICF employs the "learning curve" concept used in
several industries. The "learning curve" describes the aggregate influence of learning and new
technologies as having a certain percent effect on a key productivity measure (for example cost per unit
of output or feet drilled per rig per day) for each doubling of cumulative output volume or other measure of
industry/technology maturity. The learning curve shows that advances are rapid (measured as percent
improvement per period of time) in the early stages when industries or technologies are immature and
that those advances decline through time as the industry or technology matures. We find the learning
curve effect is roughly 20 percent per doubling of cumulative wells.
Major Unconventional Natural Gas Categories
Definition of Unconventional Gas: Quantities of natural gas that occur in continuous, widespread
accumulations in low quality reservoir rocks (including low permeability or tight gas, coalbed methane,
and shale gas), that are produced through wellbores but require advanced technologies or procedures
for economic production.
Tight Gas is defined as natural gas from gas-bearing sandstones or carbonates with an in-situ
permeability (flow rate capability) to gas of less than 0.1 millidarcy. Many tight gas sands have in situ
permeability as low as 0.001 millidarcy. Wells are typically vertical or directional and require artificial
stimulation.
Coalbed Methane is defined as natural gas produced from coal seams. The coal acts as both the
source and reservoir for the methane. Wells are typically vertical but can be horizontal. Some coals are
wet and require water removal to produce the gas, while others are dry.
Shale Gas is defined as natural gas from shale formations. The shale acts as both the source and
reservoir for the methane. Older shale gas wells were vertical while more recent wells are primarily
horizontal with artificial stimulation. Only shale formations with certain characteristics will produce gas.
Shale Oil with Associated Gas is defined as associated gas from oil shale in horizontal drilling plays
such as the Bakken in the Williston Basin. The gas is produced through boreholes along with the oil.
8-11
-------
Upstream Cost and Technology Factors
In ICF's methodology, supply technology advancements effects are represented in three categories:
• Improved exploratory success rates
• Cost reductions of platform, drilling, and other components
• Improved recovery per well
These factors are included in the model by region and type of gas and represent several dozen actual
model parameters. ICF's database contains base year cost for wells, platforms, operations and
maintenance, and other relevant cost items.
8.3.2 Oil Prices
Natural gas prices and LNG export levels are forecasted by taking oil prices into account.
ICF uses the Refiner Acquisition Cost of Crude Oil (RACC) price as an oil price input to GMM. The
RACC price is a term commonly used in discussing crude oil. It is the cost of crude oil to the refiner,
including transportation and fees. ICF's crude oil price forecast uses futures prices for 2022 and a blend
of futures and our fundamental forecast for 2022-2025. ICF expects an equilibrium marginal production
cost of ~$57/bbl (in 2019$) by 2035 and stays flat beyond 2035 in real terms. The residual oil price
averages between 70 and 100 percent of the RACC price on a dollar per Btu basis. This is the price used
to determine switching in the industrial sector. Table 8-3 shows the RACC price assumption for EPA
Platform v6.
Table 8-3 Refiners' Acquisition Cost of Crude (RACC)
Year
Annual Average Price in 2019$/bbl
2028
53.9
2030
54.7
2035
56.9
2040
56.9
2045
56.9
2050
56.9
8.3.3 Gas Production
Current United States and Canada gas production is from over 470 trillion cubic feet (Tcf) of proven gas
reserves. ICF assumes that the United States and Canada natural gas resource base totals roughly
4,000 Tcf of unproved plus discovered but undeveloped gas resource. This can supply the U.S. and
Canada gas markets for over 100 years (at current consumption levels). Shale gas accounts for over 50
percent of remaining recoverable gas resources. No significant restrictions on well permitting and
fracturing are assumed beyond restrictions that are currently in place.
To estimate the contributions of changing technologies ICF employs the "learning curve" concept used in
several industries. The "learning curve" describes the aggregate influence of learning and new
technologies as having a certain percent effect on a key productivity measure (for example cost per unit
of output or feet drilled per rig per day) for each doubling of cumulative output volume or other measure of
industry/technology maturity. The learning curve shows that advances are rapid (measured as percent
improvement per period of time) in the early stages when industries or technologies are immature and
that those advances decline through time as the industry or technology matures. The learning curve
effect is roughly 20 percent per doubling of cumulative wells.
In ICF's methodology, supply technology advancements effects are represented in three categories:
8-12
-------
• Improved exploratory success rates
• Cost reductions of platform, drilling, and other components
• Improved recovery per well
These factors are included in the model by region and type of gas and represent several dozen actual
model parameters. ICF's database contains base year cost for wells, platforms, operations and
maintenance, and other relevant cost items. Table 8-4 shows the ICF's United States and Canada dry
gas production by source and run year for EPA Platform v6.
Table 8-4 United States and Canada Projected Dry Gas Production by Source (Bcfd)
Year
Conventional Onshore
Coalbed Methane
Tight
Offshore
Shale
Total
2028
13.0
2.5
7.4
2.4
108.9
134.1
2030
12.0
2.3
6.9
2.1
111.0
134.4
2035
10.3
1.8
5.7
1.8
109.2
128.8
2040
9.2
1.4
4.8
1.4
105.4
122.2
2045
8.8
1.2
4.4
1.4
105.0
120.9
2050
8.7
1.1
4.2
1.3
104.4
119.7
8.3.4 Demand Assumptions
Gas demand is calculated by sets of algorithms and equations for each sector and region. Recent data
from DOE/EIA and Statistics Canada have been considered in the calibration of the model. ICF performs
market reconnaissance and data analysis each month to support the GMM calibration. GMM models
natural gas demand in four end-use sectors: residential, commercial, industrial, and power generation.
Residential/Commercial gas demand calculated from regional equations fit econometrically to weather,
economic growth, and price elasticity.
Industrial gas demand is based on a detailed breakout of industrial activity by census region and
includes ten industry sectors, focusing on gas-intensive industries.
Power generation demand in the GMM is modeled for 13 dispatch regions as shown in Figure 8-7 for the
contiguous United States. All the power sector inputs in GMM are changed to be consistent with IPM
results overtime. Most importantly, the total gas use regionally is benchmarked against IPM's gas use.
Pipeline fuel consumption is a function of the fuel rate and the volume of gas moved on each pipeline
corridor. Pipeline gas use is estimated as a percent of natural gas throughput for each link in the pipeline
network.
Lease & Plant gas use is forecasted based on historical percentages of the dry gas produced at each
node. Regional factors determine the share of lease & plant gas use for each supply region.
There are four key drivers for natural gas demand in GMM. They are:
i) Macroeconomic parameters: From Q2 2023 forward, ICF assumes U.S. GDP grows at 2.1%
per year, and Canada GDP grows at 2.0% per year.89
ii) Electric Demand Growth: Electric demand growth rate is assumed to be 0.74% per year
consistent with EPA Platform v6.
89 The U.S. Congressional Budget Office assumes an average annual GDP growth rate of 1.9% between 2022 and
2032 in their July 2022 Long Term Budget Outlook, while the 2022 U.S. Energy Information Administration Annual
Energy Outlook used an average annual GDP growth rate of 2.2% between 2023 and 2050.
8-13
-------
iii) Demographics: Projected demographic trends are consistent with trends over the past 20 years.
U.S. population growth averages about 1% per year throughout our projection.
iv) Weather: Future weather is assumed consistent with regional and monthly average heating and
cooling degree days (HDD/CDD) over the past 20 years (2002 through 2021).
Figure 8-7 GMM Power Generation Gas Demand Regions
Table 8-5 shows the ICF's United States and Canada natural gas demand by sector and run year for EPA
Platform v6.
Table 8-5 GMM United States and Canada Gas Demand Projection (Bcfd)
Year
Reside
ntial
Comme
rcial
Industri
al
Other
Non-
Power
Power
2028
16.2
10.9
29.7
10.5
56.8
37.8
2030
16.2
10.8
30.1
10.6
57.1
38.0
2035
16.2
10.9
31.0
10.2
58.1
31.7
2040
16.3
10.9
31.3
9.7
58.5
25.8
2045
16.5
11.2
31.5
9.7
59.2
24.0
2050
16.8
11.5
31.6
9.6
59.9
22.0
Note: "Other" includes pipeline fuel and lease & plant.
8.3.5 LNG Exports and Pipeline Exports to Mexico
Existing and Potential Liquefied Natural Gas (LNG) Terminals
Based on current global LNG market conditions, ICF assumes that the twelve U.S. LNG terminals
currently under construction are completed and expanded in future. Those terminals are Sabine Pass,
Freeport, Cove Point, Cameron, Corpus Christi, Elba Island, Golden Pass, LNG Canada, Woodfibre,
Calcasieu Pass, Costa Azul, and Driftwood LNG. ICF projects the U.S. LNG export capacity to reach
13.5 Bcfd by end of 2022. By 2028, ICF projects U.S. LNG export capacity will be 19.5 billion cubic feet
per day (Bcfd). ICF assumes an additional 1.7 Bcfd of export capacity will come online in the U.S.
8-14
-------
between 2028 and 2045. The U.S. and Canadian LNG export terminal capacity utilization is projected to
average about 87% between 2028 through 2045. ICF assumes that two LNG export facilities will be built
in British Columbia: Woodfibre LNG and LNG Canada.
Table 8-6 LNG Export Volumes and Capacity (Bcfd)
Year
US Gulf
US East
US West
British Columbia
Capacity Online
Coast
Coast
Coast
(Annual Average)
2028
13.2
1.0
0.0
1.9
19.5
2030
13.9
1.0
0.0
3.4
21.2
2035
14.2
1.0
0.0
3.4
21.2
2040
14.1
1.0
0.0
3.4
21.2
2045
14.1
1.0
0.0
3.4
21.2
2050
14.3
1.0
0.0
3.5
21.2
Pipeline Exports to Mexico
Mexico's demand for natural gas will continue to increase between 2020 and 2030 due to Mexico's
expansion of its domestic pipeline infrastructure, increased power generation gas demand, and lower
domestic production. Since 2015, Mexico's imports of U.S. gas have undergone a 124% increase,
reaching 6.4 Bcfd in 2022. ICF projects that exports will reach 8.2 Bcfd by 2030. ICF assumes the first
phase of the Costa Azul LNG export facility will be built in Mexico, further increasing pipeline exports to
Mexico from the United States.
Table 8-7 U.S. Pipeline Exports to Mexico (Bcfd)
Year
California
West Texas/ New
Mexico
Arizona
South
Texas
2028
0.5
2.0
0.6
5.1
2030
0.5
2.1
0.6
5.0
2035
0.5
2.4
0.7
4.4
2040
0.5
2.4
0.7
4.4
2045
0.5
2.4
0.7
4.4
2050
0.5
2.4
0.7
4.3
List of tables that are uploaded directly to the web:
Table 8-8 EIA Style Gas Report for EPA Platform v6 Post-IRA 2022 Reference Case
Table 8-9 Natural Gas Basis for EPA Platform v6 Post-IRA 2022 Reference Case
Table 8-10 Natural Gas Supply Curves for EPA Platform v6 Post-IRA 2022 Reference Case
8-15
-------
9. Other Fuels and Fuel Emission Factor Assumptions
Besides coal (Chapter 7) and natural gas (Chapter 8), EPA Platform v6 Post-IRA 2022 Reference Case
(EPA Platform v6) also includes assumptions for residual fuel oil, distillate fuel oil, biomass, nuclear, and
waste fuels. This chapter describes the assumptions pertaining to characteristics, market structures, and
prices of these other fuels. As reported in previous chapters, natural gas is represented by an exogenous
supply curve along with a basis differential approach informed by a resource fundamentals model. Coal
is represented by a robust set of supply curves and a detailed representation of the associated coal
transport network. Together they are designed to capture the intricacies of the resource base and market
for these fuels which accounted for about 62% of U.S. electric generation in 2019.90 As with coal, the
price and quantity of biomass combusted is determined by balancing supply and demand using a set of
geographically differentiated supply curves. In contrast, fuel oil, nuclear, waste fuel, and hydrogen prices
are exogenously determined and input to IPM during model set-up as constant price points that apply to
all levels of supply. The following treats each of these remaining fuels in turn and concludes with a
discussion of the emission factors for all the fuels represented in EPA Platform v6.
9.1 Fuel Oil
Two petroleum derived fuels are included in EPA Platform v6. Distillate fuel oil is distilled from crude oil,
and residual fuel oil is a residue of the distillation process. The fuel oil prices are based on the AEO 2020
reference case projection and a long-term crude oil projection of 70 $/barrel and are shown in Table 9-1.
They are regionally differentiated according to the National Energy Modeling System (NEMS) regions
used in the AEO 2020. These prices are mapped to their corresponding IPM regions for use in EPA
Platform v6.
Table 9-1 Fuel Oil Prices by NEMS Region in v6
Residual Fuel Oil Prices (2019$/MMBtu)
AEO NEMS Region
2023
2025
2028
2030
2035
2040
2045
2050
TRE
10.13
11.01
11.55
12.14
12.20
12.71
12.68
12.58
FRCC
8.51
9.39
9.93
10.52
10.58
11.09
11.05
10.96
MISW
9.30
9.46
9.56
10.42
10.48
11.01
10.78
10.50
MISC
3.12
4.00
4.54
5.13
5.19
5.70
5.66
5.57
MISE
5.64
6.52
7.06
7.65
7.71
8.22
8.18
8.09
MISS
10.02
10.90
11.44
12.03
12.09
12.60
12.57
12.47
ISNE
9.92
10.80
11.34
11.93
11.99
12.50
12.47
12.37
NYCW
11.89
12.76
13.31
13.90
13.96
14.46
14.43
14.34
NYUP
10.43
9.97
10.52
11.30
11.36
11.87
11.84
11.74
PJME
9.73
9.56
10.11
10.89
11.05
11.56
11.53
11.43
PJMW
5.58
6.46
7.00
7.59
7.65
8.16
8.13
8.03
PJMC
6.63
7.50
8.05
8.64
8.70
9.20
9.17
9.08
PJMD
9.73
9.56
10.10
10.69
10.75
11.25
11.22
11.13
SRCA
6.65
7.52
8.07
8.66
8.72
9.22
9.19
9.10
SRSE
5.58
6.46
7.00
7.59
7.65
8.16
8.13
8.03
SRCE
6.65
7.52
8.07
8.66
8.72
9.22
9.19
9.10
SPPS
10.13
11.01
11.55
12.14
12.20
12.71
12.68
12.58
SPPC
6.63
7.50
8.05
8.64
8.70
9.20
9.17
9.08
SPPN
6.63
7.50
8.05
8.64
8.70
9.20
9.17
9.08
SRSG
8.31
9.19
9.73
10.32
10.38
10.89
10.86
10.76
CANO
10.37
11.25
11.79
12.38
12.44
12.94
12.91
12.82
CASO
10.37
11.25
11.79
12.38
12.44
12.94
12.91
12.82
90 EIA. Detailed EIA-923 monthly and annual survey data back to 1990. Available at
https://vvww.eia.qov/electricitv/data. php#qeneration
9-1
-------
Residual Fuel Oil Prices (2019$/MMBtu)
AEO NEMS Region
2023
2025
2028
2030
2035
2040
2045
2050
NWPP
7.67
9.32
9.87
10.73
10.62
10.80
10.54
10.45
RMRG
4.90
5.78
6.32
6.91
6.97
7.47
7.44
7.35
BASN
12.06
12.54
13.19
13.94
13.95
13.90
12.88
12.83
Distillate Fuel Oil Prices (2019$/MMBtu)
NEMS Region
2023
2025
2028
2030
2035
2040
2045
2050
TRE
16.62
15.46
16.55
17.06
17.23
17.11
17.20
17.04
FRCC
18.63
18.09
19.13
19.54
19.72
19.58
19.69
19.51
MISW
15.63
14.01
15.03
15.47
15.70
15.59
15.69
15.57
MISC
15.62
14.07
15.10
15.54
15.77
15.66
15.75
15.64
MISE
15.54
13.97
15.00
15.44
15.67
15.55
15.65
15.53
MISS
16.62
15.46
16.55
17.06
17.23
17.11
17.19
17.04
ISNE
17.01
15.99
17.03
17.44
17.61
17.47
17.59
17.41
NYCW
20.02
19.82
20.85
21.27
21.44
21.30
21.42
21.24
NYUP
20.02
19.82
20.85
21.27
21.44
21.30
21.42
21.36
PJME
19.68
19.34
20.38
20.79
20.97
20.86
20.94
20.78
PJMW
17.18
16.25
17.30
17.79
18.16
18.21
18.35
18.20
PJMC
15.54
13.97
15.00
15.44
15.67
15.55
15.65
15.53
PJMD
18.63
18.09
19.13
19.54
19.72
19.58
19.69
19.51
SRCA
18.63
18.09
19.13
19.54
19.72
19.58
19.69
19.51
SRSE
17.83
17.07
17.71
18.11
18.33
18.20
18.28
18.15
SRCE
16.61
15.42
16.46
16.91
17.12
17.01
17.12
16.99
SPPS
16.62
15.46
16.55
17.06
17.23
17.11
17.20
17.04
SPPC
15.66
14.02
15.05
15.49
15.72
15.61
15.70
15.59
SPPN
15.66
14.02
15.05
15.49
15.72
15.61
15.70
15.59
SRSG
19.32
18.85
19.95
20.41
20.59
20.44
20.56
20.43
CANO
18.80
17.86
18.94
19.43
19.61
19.47
19.60
19.46
CASO
18.80
17.86
18.94
19.43
19.61
19.47
19.60
19.46
NWPP
18.82
17.98
19.07
19.59
19.82
19.50
19.63
19.48
RMRG
19.36
18.91
19.99
20.45
20.63
20.49
20.62
20.48
BASN
19.36
18.91
19.99
20.45
20.63
20.49
20.62
20.48
9.2 Biomass Fuel
Biomass is offered as a fuel for existing dedicated biomass power plants and potential (new) biomass
direct fired boilers. In addition to its use as the prime mover fuel for these plants, it is also offered for co-
firing to those coal-fired power plants that have co-fired biomass in the recent past. Section 5.3 provides
further details of these selected plants.
EPA Platform v6 uses biomass supply curves based on those in the Department of Energy's 2016 Billion-
Ton Report (DOE Report). Biomass supply curves at the IPM region and state level are generated by
aggregating county-level supply curves from the DOE Report. Power plants demand biomass from the
supply curve corresponding to the IPM region and state in which they are located. No inter-region trading
of biomass is allowed. Each biomass supply curve depicts the price-quantity relationship for biomass and
varies overtime. There is a separate curve for each model run year. The supply component of the curve
represents the aggregate supply in each region of agricultural residues, forestry residues, energy crops,
waste, and trees. The price component of the curve includes transportation costs of $15 per dry ton.
The supply curves represent the IPM region and state-specific delivered biomass fuel cost at the plant
gate. A storage cost of $20 per dry ton is added to each step of the agricultural residue supply curves to
9-2
-------
reflect the limited agricultural growing season.91 The biomass supply curves are summarized in Table
9-4. The biomass prices are derived endogenously based on the aggregate power sector demand for
biomass in each IPM region and state. The results are unique market-clearing prices for each IPM region
and state. All plants using biomass from that IPM region and state face the same market-clearing price.
9.3 Nuclear Fuel
The AEO 2020 price for nuclear fuel is used as the nuclear fuel price assumption for 2021 -2050 in EPA
Platform v6. The 2028, 2030, 2035, 2040, 2045, and 2050 prices are 0.69, 0.69, 0.70, 0.71, 0.72, and
0.73 2019 $/MMBtu, respectively.
9.4 Waste Fuels
The waste fuels include waste coal, petroleum coke, fossil waste, non-fossil waste, tires, and municipal
solid waste (MSW). Table 9-2 describes the characteristics of these fuels, the extent to which they are
represented in NEEDS, and the assumptions pertaining to their use and pricing. Furthermore, the fuels
are provided to only existing and planned-committed generating units. Potential (new) generating units
that the model builds are not given the option to burn these fuels. In IPM model output, tires, MSW, and
non-fossil waste are included under existing non-fossil other plant type, while waste coal and petroleum
coke are included under coal plant type.
Table 9-2 Waste Fuels in v6
Modeled
Fuel in
NEEDS
Number
of Units
in NEEDS
Total
Capacity in
NEEDS
Description
Supply and Cost
Modeled
By
Assumed
Price
Waste
Coal
20
1,420 MW
"Usable material that is a byproduct of previous coal
processing operations. Waste coal is usually composed of
mixed coal, soil, and rock (mine waste). Most waste coal is
burned as-is in unconventional fluidized-bed combustors.
For some uses, waste coal may be partially cleaned by
removing some extraneous noncombustible constituents.
Examples of waste coal include fine coal, coal obtained
from a refuse bank or slurry dam, anthracite culm,
bituminous gob, and lignite waste."
https://vwwv,eia,aov/tools/alossarv/index,php?id=W
Supply
Curve
Based on
AEO 2020
AEO 2020
Petroleum
Coke
11
518 MW
A residual product, high in carbon content and low in
hydrogen, from the cracking process used in crude oil
refining.
Price
Point
$49.80/Ton
Fossil
Waste
62
1,382 MW
Waste products of petroleum or natural gas including blast
furnace and coke oven gas. They do not include petroleum
coke or waste coal which are specified separately among
the modeled fuels.
Price
Point
0
Non-
Fossil
Waste
223
2,299 MW
Non-fossil waste products that do not qualify as biomass.
These include waste products of liquid and gaseous
renewable fuels (e.g., red, and black liquor from pulping
processes and digester gases from wastewater treatment).
They do not include urban wood waste which is included in
biomass.
Price
Point
0
Tires
2
52 MW
Discarded vehicle tires.
Price
Point
0
91 http://wvwy.extension.iastate.edu/aqdm/crops/pdf/a1-22.pdf,
http://vwwy.rand.org/content/dam/rand/pubs/technical reports/2011/RAND TR876.pdf
9-3
-------
Modeled
Fuel in
NEEDS
Number
of Units
in NEEDS
Total
Capacity in
NEEDS
Description
Supply and Cost
Modeled
By
Assumed
Price
Municipal
Solid
Waste
150
1,935 MW
Residential solid waste and some nonhazardous
commercial, institutional, and industrial wastes.
https://www.eia.gov/tools/glossarv/index. php?id=M
Price
Point
0
9.5 Hydrogen Fuel
The price of hydrogen is assumed to be 1 $/kg or 7.40 $/MMBtu.
9.6 Fuel Emission Factors
Table 9-3 brings together all the fuel emission factor assumptions implemented in EPA Platform v6. For
sulfur dioxide, chlorine, and mercury in coal, where emission factors vary widely based on the rank,
grade, and supply source of the coal, cross references are given to tables that provide more detailed
treatment of the topic. Nitrogen oxides (NOx) are not included in Table 9-3 because NOx emissions are a
factor of the combustion process and are not primarily fuel based.
Table 9-3 Fuel Emission Factor Assumptions in v6
Fuel Type
Carbon Dioxide
(Ibs/MMBtu)
Sulfur Dioxide
(Ibs/MMBtu)
Mercury
(Ibs/TBtu)
HCI (Ibs/MMBtu)
Coal
Bituminous
202.8-212.9
0.67-7.78
2.80 -34.71
0.015-0.214
Subbituminous
209.2-215.7
0.52-2.15
2.03-8.65
0.007-0.014
Lignite
212.6-219.3
1.51 -5.67
7.53 -30.23
0.011 -0.036
Natural Gas
117.08
0
0.00014
0
Fuel Oil
Distillate
161.39
0
0.48
0
Residual
173.91
1.04
0.48
0
Biomass
195
0.08
0.57
0
Waste Fuels
Waste Coal
204.7
7.78
53.9
0.0921
Petroleum Coke
225.1
7.70
2.66
0.0213
Fossil Waste
321.0
0.08
0
0
Non-Fossil Waste
0
0
0
0
Tires
189.5
1.65
3.58
0.06
Municipal Solid Waste
91.9
0.35
71.85
0
Note:
Table 7-4 has coal emission factor on a coal supply region level.
List of tables that are uploaded directly to the web:
Table 9-4 Biomass Supply Curves for EPA Platform v6 Post-IRA 2022 Reference Case
9-4
-------
10. Financial Assumptions
10.1 Introduction and Summary
This chapter presents the financial assumptions used in the EPA Platform v6 Post-IRA 2022 Reference
Case (EPA Platform v6). EPA Platform v6 models a diverse set of generation and emission control
technologies, each of which requires financing92, and incorporates updates to reflect The Tax Cuts and
Jobs Act of 2017.93 The capital charge rate converts the capital cost for each investment into a stream of
levelized annual payments that ensures recovery of all costs associated with a capital investment
including recovery of and return on invested capital and income taxes. The discount rate is used to
convert all dollars to present values and IPM minimizes the present value of annual system costs. The
discount rate is set equal to the weighted average costs of capital. Describing the methodological
approach to quantifying the discount and capital charge rates in the EPA Platform v6 is the primary
purpose of this chapter.
10.2 Introduction to Risk
The cost of capital is the level of return investors expect to receive for alternative investments of
comparable risk. Investors will only provide capital if the return on the investment is equal to or greater
than the return available to them for alternative investments of comparable risk. Accordingly, the long-run
average return required to secure investment resources is proportional to risk. There are several
dimensions to risk that are relevant to power sector operations, including:
• Market Structure -The risk of an investment in the power sector is heavily dependent on
whether the wholesale power market is regulated or deregulated. The risks are higher in a
deregulated market compared to a traditionally regulated utility market. Slightly more than half of
U.S. generation capacity is deregulated (operated by Independent Power Producers (IPPs), or
'merchants').94 IPPs often sell power into spot markets supplemented by near-term hedges. In
contrast, regulated plants sell primarily to franchised customers at regulated rates, an
arrangement that significantly mitigates uncertainty, and therefore risk.95
• Technology - The selection of new technology investment options is partially driven by the risk
profile of these technology investments. For instance, in a deregulated merchant market an
investment in a peaking combustion turbine is likely to be much riskier than an investment in a
combined cycle unit. This is because a combustion turbine operates as a peaking unit and can
generate revenues only in times of high demand, or via capacity payments, while a combined
cycle unit is able to generate revenues over a much larger number of hours in a year from the
energy markets as well as via capacity payments. An investor in a combined cycle unit,
therefore, would require a lower return due to a more diversified stream of revenue, and receive a
lower risk premium than an investor in a combustion turbine, all else equal.
92 The capital charge rates discussed here apply to new (potential) units and environmental retrofits that IPM selects.
The capital cost of existing and planned/committed generating units (also referred to as 'firm'), and the emission
controls already on these units are considered sunk costs and are not represented in the model.
93 The Tax Cuts and Jobs Act of 2017, Pub.L. 115-97.
94 According to EIA Form 860 2019, the current capacity mix is 58% utility and 42% merchant by MW.
95 There is a potential third category of risk, where IPPs enter into long-term (e.g., ten years or longer), known-price
contracts with credit worthy counterparties (e.g., traditionally regulated utilities). With a guaranteed, longer-term
price, the risk profile of this segment of the IPP fleet is similar enough to be treated as regulated plants.
10-1
-------
• Leverage - There are financial risks related to the extent of leverage. Reliance on debt over
equity in financing a project increases the risk of insolvency. This dynamic applies to all
industries, power included.96
• Financing Structure - Lastly, there are also financing structure risks (e.g., corporate vs. project
financing), also referred to as non-recourse financing. There are no clear risk implications from
the structure alone, but rather this element interacts with other dimensions of risks making
considerations of leverage, technology, and market structure more important.
• Systemic - Systemic risk is when financial performance correlates with overall market and
macro-economic conditions such that investment returns are poor when market and economic
conditions are poor, and vice versa. For example, if investors are less likely to earn recovery of
and on investments during recessions, then these risks are systemic, and increase required
expected rates of return. This emphasis on correlated market risk is based on the Capital Asset
Pricing Model (CAPM), which is used to produce key financial assumptions for EPA Platform v6.
Other risks are handled in the cash flows and are treated as non-correlated with the market.
10.2.1 Deregulation - Market Structure Risks
As noted, the power sector in North America can be divided into the traditional regulated sector (also
known as cost of service or utility sector) and deregulated merchant sector (also known as competitive,
merchant, deregulated,97 or IPP sector).
Traditional Regulated
The traditional regulated market structure is typical of the vertically integrated utilities whose investments
are approved through a regulatory process and the investment is provided a regulated rate of return,
provided the utility's investments are deemed prudent. In this form of market structure, returns include
the return of the original investment plus a return on invested capital that are administratively determined.
Returns are affected by market conditions due to regulatory lag and other imperfections in the process,
but overall regulated investments are less exposed to the market than deregulated investments, all else
equal.
Deregulated Merchant
In a deregulated merchant market structure, investments bear a greater degree of market risk, as the
price at which they can sell electricity is dependent on what the short-term commodity and financial hedge
markets will bear. Return on investment in this form of market structure is not only dependent on the
state of the economy, but also on commodity prices, capital investment cycles, and remaining price-
related regulation (e.g., FERC price caps on capacity prices). The capital investment cycle can create a
boom-and-bust cycle, which imparts risk or uncertainty in the sector that can be highly correlated with
overall macro-economic trends. The operating cash flows from investments in this sector are more
volatile as compared to the traditional regulated sector, and hence, carry more business or market risk.98
Overall, there is ample supporting evidence for the theoretical claim that deregulated investments are
more risky than utility investments. For example:
96 We use the terms debt and leverage interchangeably.
97 Wholesale generators cannot be economically unregulated; they can be Exempt Wholesale Generator ("EWG")
subject to FERC jurisdiction. The moniker of deregulated is used to convey greater market risk relative to regulated
utility plants.
98 In this documentation, the terms merchant financing, deregulated, IPP, non-utility and merchant refer to this type of
market structure.
10-2
-------
• All three large publicly traded IPPs" are rated as sub-investment grade100 while all utilities are
investment grade.
• All major IPPs have gone bankrupt over the last 20 years.101
• Estimates of beta, a measure of risk using CAPM, leverage, debt costs, and weighted average
cost of capital, consistently produce higher risk for deregulated power plants.
10.3 Federal Income Tax Law Changes
EPA Platform v6 incorporates updates to reflect The Tax Cuts and Jobs Act of 2017. The four most
significant changes in the federal corporate income tax code are:
• Rate - The corporate tax rate is lowered 14 percentage points from 35%102 to 21%; the 21% rate
is in place starting in 2018 and remains in place indefinitely; the lower tax rate decreases capital
charges in all periods and all sectors, all else held equal. When state income taxes are included,
the average rate decreases 13.1 percentage points, from 39.2% to 26.1%. This applies to both
sectors, utility and IPP.
• Depreciation - The new tax law expands near-term bonus depreciation (also referred to as
expensing) for the IPP sector only until 2027; the utility sector is unaffected.
• Interest Expense - The new law lowers tax deductibility of interest expense for the IPP sector,
which continues indefinitely; the utility sector is unaffected.
• Net Operating Losses - The new law limits the use of Net Operating Losses (NOL) to offset
taxable income. This applies to all sectors, utility and IPP.
Other important features of the new tax law include:
• Annual Variation of Provisions - The legislation specifies permanent changes (tax rate and
NOL usage limit) applying to both sectors, utility and IPP. The legislation also applies temporary
changes that vary year-by-year through to 2027 (depreciation and tax deductibility of interest)
(See Table 10-1) applying to the IPP sector only. This creates different capital charge rates for
each year through 2027. We calculate these parameters for IPM run years 2023, 2025, and 2030
and thereafter. This set covers a wide range of financing conditions even though we do not
estimate every year.
99 Dynegy Inc. Calpine Corp. and NRG Energy Inc are the three IPP's whose ratings were B2, Ba3 and Ba3 in 2016.
100 Below minimum investment grade.
101 Dynegy, Calpine, and NRG were bankrupt - i.e., the three large public IPPs were bankrupt. Also, Mirant (major
IPP), Boston Generating (IPP), EFH (utility with large IPP component), and FES (utility with large IPP component)
have been or are bankrupt.
102 The average state income tax rate is 6.45 percent. State income tax is deductible, and hence, the combined rate
is 26.1% (26.1 =21 +(1-0.21 )*6.45).
10-3
-------
Table 10-1 Summary Tax Changes
Parameter
Previous
2023103
2025
2030 and Later
Marginal Tax Rate -
Federal
35
21
21
21
Maximum NOL (Net
Operating Loss)
Carry Forward
Usage
No limit. All losses
in excess of
income are carried
forward and
usable
immediately.
Carry Forward
cannot exceed
80% of Taxable
Income
Carry Forward
cannot exceed
80% of Taxable
Income
Carry Forward
cannot exceed 80%
of Taxable Income
Tax Deductibility of
Interest Expense
100%104
30% of EBIT;
Utilities MACRS
30% of EBIT;
Utilities MACRS
30% of EBIT;
Utilities MACRS
Bonus
Depreciation105
Q106
IPP 80%107;
Utilities 0%
IPP 40%108;
Utilities 0%
0
• Utilities Versus IPPs - As noted, the legislation treats utilities and IPPs differently. The new tax
code exempts utilities from changes in tax deductibility of interest and accelerated depreciation.
The financing assumptions used in IPM modeling are a blend (weighted average) of the utility and
IPP average. The weighting is 60% utility and 40% IPP, and hence, the greatest weight is on the
least affected sector. This partly mitigates the impacts of the changes.
• Capital Charge Rates - We calculate the capital charge rates for utilities and IPPs, and then
take the weighted average of the resulting capital charge rates. As a result of the legislation,
combined with the IPM model's ability to vary capital charge rates by run year, the blended
average is calculated for specific run years.
• Discount Rates - The discount rate equals the weighted average after tax cost of capital
(WACC) and is affected by the change in the corporate income tax rate only. The discount rate is
invariant overtime, sectors, and technologies. Therefore, the calculation methodology for
discount rate used in IPM is unchanged.
10.4 Calculation of the Financial Discount Rate
10.4.1 Introduction to Discount Rate Calculations
A discount rate is used to translate future cash flows into current dollars by considering factors such as
expected inflation and the ability to earn interest, which make one dollar tomorrow worth less than one
103 IPM run years in the near term are 2023, 2025, and 2028.
104 No limit except losses in excess of income can be carried forward. The losses were limited to first few years.
105 Referred to as expensing. If depreciation exceeds income in first year, it can be carried forward to succeeding
years up to 80% of EBITDA.
106 Bonus depreciation was available but only in the period before IPM runs, and only for new equipment.
107 For thermal power plants coming online in 2023, the 100% would apply only to costs incurred through end of
2022. We are hence assuming practically all capital costs are incurred prior to 2023.
108 Remaining basis depreciated at MACRS schedule.
10-4
-------
dollar today. The discount rate allows intertemporal trade-offs and represents the risk adjusted time value
of money.109
The discount rate adopted for modeling investment behavior should reflect the time preference of money
or the rate at which investors are willing to sacrifice present consumption for future consumption. The
return on private investment represents the opportunity cost of money and is commonly used as an
appropriate approximation of a discount rate.110
The real discount rate for all expenditures (capital, fuel, variable operations and maintenance, and fixed
operations and maintenance costs) in the EPA Platform v6 is 3.76%.111
10.4.2 Summary of Results
The tables below present a summary of the key financial assumption for the EPA Platform v6. A
description of these values and the attendant methodological approaches follow throughout the chapter.
Table 10-2 Financial Assumptions for Utility and Merchant Cases
EPA Platform v6 - Utility WACC using daily beta for 2016-2020
Parameters
Value
Risk-free rate
2.73 %112
Market premium
7.15 %113
Equity size premium
-0.01 %114
Levered beta115
0.72
Debt/total value116
0.58
Cost of debt
3.50 %117
109 The discount rate is the inverse of compound interest or return rate; the existence of interest, especially compound
interest creates an opportunity cost for not having dollars immediately available. Thus, future dollars need to be
discounted to be comparable to immediately available dollars.
110 For a perspective on the legal basis for utilities having the right to have the opportunity to earn such returns under
certain conditions such as prudent operations, see Bluefield Water Works and Improvement Co. v Public Service
Comm'n 262 US 679, 692 (1923). See also Federal Power Comm'n versus Hope Natural Gas Co., 320 US 591, 603
(1944).
111 This rate is based on the weighted average aftertax cost of capital (WACC), which reflects two weightings. First,
it reflects an assumption that 60% of the investments are made by a regulated utility and 40% are made by a
merchant investor (also referred to as a hybrid). Second, it assumes a mix of plant types - 55% renewable and 45%
gas thermal. This weighting reflects the profile of builds over 2015-2019 of renewable and natural gas-fired units. The
financial data used to estimate this rate is primarily from 2016-2020. The EPA Base Case v6 uses 2019 (2019$) as
its real dollar baseline and assumes 1.76% general inflation. Hence, the nominal discount rate is 5.59%.
112 Represents 10-year historical average (2011- June 2020) on a 20-year treasury bond. See discussion of risk-free
rate and market premium. The 5-year average (2016-June 2020) on a 20-year T bond is 2.45%. The 5-year (2016-
June 2020) and 10-year (2011-June 2020) averages for the 30-year bond are 2.66% and 2.99% respectively.
113 Represents the long horizon expected equity risk premium based on differences between S&P 500 total returns
and long-term government bond income returns from 1926-2020 (Duff and Phelps 2020).
114 Size Premiums according to size groupings taken from Duff & Phelps 2020 Valuation. Equity Size Premium is
based on weighted average of each company's Equity Size Premium, weighted by each company's Market
capitalization level.
115 Levered betas were calculated using 5 years (2016-June 2020) and in a sensitivity case discussed separately
later 10 years (2011-June 2020) of historical stock price data. Daily returns were used in the current analysis. In the
previous case, weekly returns for 5 years (2016-2020) were used.
116 Debt/total value ratio is the simple average of net debt to equity ratio for the past 5 years.
117 Cost of debt is based on 5-year (2016-June 2020) weighted average of debt yields for 18 utilities. The weights
assigned are equity share of each utility.
10-5
-------
EPA Platform v6 - Utility WACC using daily beta for 2016-2020
Debt beta
0.00
Unlevered beta118
0.36
Target debt/total value119
0.50
Relevered beta
0.62
Cost of equity (with size premium)120
7.17 %
WACC
4.88 %
EPA Platform v6 - Merchant WACC using 55% Target Debt
Parameters
Value
Risk-free rate
2.73 %
Market premium
7.15 %
Equity size premium
0.89 %121
Levered beta122
1.04
Debt/total value123
0.64
Cost of debt124
6.27 %
Debt beta125
0.00
Unlevered beta126
0.45
Target debt/ total value127
0.55
Relevered beta
0.86
Cost of equity (with size premium)128
9.74%
WACC
6.65%
Table 10-3 Weighted Average Cost of Capital in v6
Utility
Share
Utility
WACC
Merchant
Share
Merchant
WACC
Weighted
Average
Nominal
Inflation
Weighted
Average Real
WACC
WACC
60%
4.88%
40%
6.65%
5.59%
1.76%
3.76%
118 Calculated using Hamada equation.
119 Target debt/total value for utility case is based on historical 5 years of average D/E for utilities
120 Cost of Equity represents the simple average cost of equity derived from Risk-Free Rate, Market Premium,
Relevered Beta, and Target D/E value.
121 Size Premiums according to size groupings taken from Duff & Phelps 2020 Valuation Handbook. Equity Size
Premium is based on weighted average of each company's Equity Size Premium, weighted by each company's
equity capitalization level.
122 Levered betas were calculated using five years (2016-June 2020) of historical stock price data. Weekly returns
were used in the analysis.
123 Debt/total value for merchant case is calculated as simple average of the 5-year total debt to total value for each
IPP.
124 Cost of debt is based on historical 5-year weighted average of yields to maturity on outstanding debt.
125 Debt Beta was previously used as Dynegy was in the process of bankruptcy.
126 Calculated using Hamada equation. In merchant case, it was modified slightly to include the riskiness of debt.
127 The capitalization structure (debt to equity (D/E)) for merchant financings is assumed to be 55/45.
128 Cost of Equity (ROE) represents the simple average cost of equity. In the Merchant ROE, the decrease reflects
primarily the lower beta.
10-6
-------
10.5 Discount Rate Components
The discount rate is a function of the following parameters:
• Capital structure (share of equity and debt)
• Post-tax cost of debt
• Post-tax cost of eq u ity
The WACC is used as the discount rate and is calculated as follows:129
WACC = [Share of Equity * Cost of Equity]
+ [Share of Preferred Stock * Cost of Preferred Stock]
+ [Share of Debt *After Tax Cost of Debt]
The methodology relies on debt and equity (common stock) because preferred stock is generally a small
share of capital structures, especially in the IPP sector. Its intermediate status between debt and equity
in terms of access to cash flow also tends not to change the weighted average.130 Typically, net cash
flows are used to fund senior debt before subordinated debt, and all debt before equity. Therefore, the
risk of equity is higher than debt, and the rates of return reflect this relationship. Notwithstanding,
consistent with our use of utility debt that has recourse to the corporation rather than individual assets, we
use IPP debt that has recourse to the corporation rather than individual assets because the data are more
robust.
10.6 Market Structure: Utility-Merchant Financing Ratio
With two distinct market structures, EPA Platform v6 establishes appropriate weights for regulated and
deregulated financial assumptions to produce a single, hybrid set of utility capital charge rates for new
units. The EPA Platform v6 uses a weighting of 60:40, regulated to deregulated, based on recent
capacity addition shares by market type (see Table 10-4).131
Table 10-4 Share of Annual Thermal Capacity Additions by Market
Entity
2015
2016
2017
2018
2019
Total
Regulated
61%
81%
51%
52%
63%
61%
Merchant
39%
19%
49%
48%
37%
39%
10.7 Capital Structure: Debt-Equity Share
10.7.1 Introduction and Shares for Utilities and IPPs
The second step in calculating the discount rate is the determination of the capital structure, specifically
the debt to equity (D/E) or debt to value (D/V) ratio for utility and merchant investments.132 This is
129 Sometimes abbreviated as ATWACC. The pretax WACC is higher due to the inclusion of income taxes. Income
taxes are included in the capital charges. All references are to the after-tax WACC unless indicated.
130 Debt generally has first call on cash flows and equity has a residual access.
131 In contrast to new units, existing coal units can be classified as belonging to a merchant or regulated market
structure. Hence, for retrofit investments, the EPA Platform v6 assumption is that coal plants owned by a utility get
purely utility financing parameters coal plants owned by merchant companies get purely merchant financing
parameters.
132 A project's capital structure is the appropriate debt capacity given a certain level of equity, commonly represented
as "D/E." The debt is the sum of all interest bearing short- and long-term liabilities, while equity is the amount that the
project sponsors inject as equity capital.
10-7
-------
calculated by determining the total market value of the company, and the market value of its debt and
equity. The market value of the company is the sum of the market value of its debt and equity. We also
determined the capital structure for the various technology types.
The target capitalization structure for utilities was assumed to be 50:50. This was based on the
capitalization over the 2016 to 2020 period. The capitalization structure for merchant financings is
assumed to be 55/45, reflecting the greater risk inherent to this market.133
10.7.2 Utility and Merchant
For utility financing, the empirical evidence suggests that utility rate of return is based on an average
return to the entire rate base. Thus, EPA Platform v6 assumes that the required returns for regulated
utilities are independent of technology. In contrast, the merchant debt capacity is based on market risk
and varies by technology.
10.7.3 Merchant by Technology
Assigning merchant technology risk is difficult because there is a lack of publicly traded securities that
provide an empirical basis for differentiating between the risks, and hence, financing parameters for
different activities.134 Nevertheless, we assigned merchant technology market risk as follows:
• Combined Cycles - The capitalization structure for merchant financing of combined cycles is
assumed to be 55/45.
• Peaking Units - A peaking unit such as a combustion turbine is estimated to have a capital
structure of 40/60. Peaking units have a less diverse, and therefore, more risky revenue stream.
• Coal Units - A new coal unit is estimated to have a capital structure of 40/60, reflecting higher
risk than a combined cycle unit. This is reflected in a lack of proposed new builds, decreases in
coal dispatch, financial assessments by other entities such as EIA and NREL indicating greater
risk, and greater levels of environmental regulatory risk.
• Fossil Units - New, non-peaking fossil fuel-fired plants face additional risks associated with a
potential cost on future CO2 emissions, which the EIA handles by increasing the cost of debt and
equity for new coal plants.135 EPA Platform v6 extends this treatment of risk to new combined
cycle plants.
• Nuclear Units — A new nuclear unit is estimated to have a capital structure of 40/60. There is
high risk associated with a new IPP nuclear unit. This is supported by: (1) the financial
challenges facing existing nuclear units, (2) the very limited recent new nuclear construction, (3)
statements by financial institutions, and (4) the lack of ownership of nuclear power plants by pure
play IPP companies. Of the three pure play companies only one has partial ownership of a single
133 The U.S. wide average authorized rate of return on equity, authorized return on rate base, and authorized equity
ratio during the 5 years (2012-2016) for 146 utility companies was 9.93%, 7.64%, and 50.22% respectively.
According to S&P Global Market Intelligence, the authorized ROE approved for the first half of 2020 was 9.55%.
Similarly, S&P Global Market Intelligence give an average authorized ROE of 9.64% in 2019, 9.59% in 2018, 9.63%
for 2017, and 9.60% in 2016. In contrast, they state the average earned ROE to be 9.75% for the 12 months ended
during the second quarter of 2020, 10.21% in 2019, 10.34% in 2018, 10.00% in 2017.
134 There were only three major IPP companies with traded equity. This is insufficient to conduct statistical analysis.
135 ElA's Annual Energy Outlook 2021; the capital charge rates shown for Supercritical Pulverized Coal without
Carbon Capture include a 3% adder to the cost of debt and equity. See The Electricity Market Module of the National
Energy Modeling System: Model Documentation 2020 (p. 108),
https://www.eia. qov/outlooks/aeo/nems/documentation/electricitv/pdf/m068f2020).pdf
10-8
-------
nuclear power plant. With this one exception, only utilities and affiliates of utilities own nuclear
units.
• Renewable Units — A new merchant renewable unit is estimated to have a capital structure of
65/35. This is the highest debt share among the major classes of generation options, and
therefore, the lowest cost of capital. This is in part because renewables have access to a third
source of financing in tax equity. Tax equity receives the tax benefits such as ITC, PTC, losses
available to defray income tax, overtime by making a payment upfront. These benefits are not
transferable to other companies. There is a risk that the tax credits may become less valuable
overtime (e.g., the company providing the tax equity does not have sufficient taxable income), or
the project may not perform and have inadequate operations to generate expected PTC volumes.
This risk is less than typical equity, since the tax credits value is not subject to as much variation
as regular equity. These projects are also easier to hedge because they have zero variable
costs, and hence, the annual volume of output is less uncertain, all else equal, and often receive
support via power purchase agreements and renewable energy credits. Limits of relying on even
greater debt include the scheduled lowering of the PTC and ITC overtime, and the potential for
performance problems.
Table 10-5 Capital Structure Assumptions in v6
Technology
Utility
Merchant
Combustion Turbine
50/50
40/60
Combined Cycle
50/50
55/45
Coal & Nuclear
50/50
40/60
Renewables
50/50
65/35
Retrofits
50/50
40/60
10.8 Cost of Debt
The third step in calculating the discount rate is to assess the cost of debt.136 The utility and merchant
cost of debt is assumed the same across all technologies.
Table 10-6 Nominal Debt Rates in v6
Technology
Utility
Merchant
Combustion Turbine
3.50%
6.27%
Combined Cycle
3.50%
6.27%
Coal & Nuclear
3.50%
6.27%
Renewables
3.50%
6.27%
Retrofits
3.50%
6.27%
10.8.1 Merchant Cost of Debt
The cost of debt for the merchant sector was estimated to be 6.27%. It is calculated by taking a 5-year
(2016-2020) weighted average of debt yields from existing company debt with eight or more years to
maturity. The weights assigned to each company debt yields were based on that company's market
capitalization. During the most recent 5 years (2016-2020), none of the existing long-term debt exceeded
twelve years to maturity, hence above average yields are based on debt with maturity between eight and
twelve years.
136 Measured as yield to maturity.
10-9
-------
10.8.2 Utility Cost of Debt
The cost of debt for the utility sector was estimated to be 3.5%. It is calculated based on the 5-year
(2016-2020) average of a set of 18 investment grade utilities weighted by enterprise value (see
Table 10-7).
Table 10-7 Utilities Used to Calculate Cost of Debt
Name
Ameren Corp
American Electric Power Co Inc
Cleco Corporate Holdings LLC
CMS Energy Corp
Empire District Electric Co/The
MGE Energy Inc
Vectren Corp
Evergy Kansas Central Inc
WEC Energy Group Inc
CH Energy Group Inc
Consolidated Edison Inc
Eversource Energy
Southern Co/The
Avista Corp
IDACORP Inc
Pinnacle West Capital Corp
PNM Resources Inc
Xcel Energy Inc
10.9 Return on Equity (ROE)
10.9.1 Introduction and Beta
The final step in calculating the discount rate is the calculation of the required rate of return on equity
(ROE). The ROE is calculated using the formula:
ROE = risk free rate + beta x equity risk premium + size premium
The formula is the key finding of the CAPM and reflects that a premium on return is required as
investment risk increases, and that premium is proportional to the systemic risk of the investment.137
Systemic risk is measured by the impact of market returns on the investment's returns and is measured
by beta.138
There are several additional aspects of estimating beta:
137 The financial literature on CAPM originally did not emphasize the size premium (also referred to as the liquidity
premium). It emerged from later findings that the estimated required return was too low for small stocks (i.e., with low
equity value).
138 Beta is the covariance of market and the stock's returns divided by the variance of the market's return.
10-10
-------
• Time Period - The most common practice is to use five years of historical returns to estimate
beta.
• Returns - Daily returns are commonly used to estimate beta except for illiquidly traded stocks
when weekly returns are used to avoid underestimating beta. The utility estimates presented use
daily data and the IPP estimates used weekly estimates.
• Unlevered Betas - It is useful to estimate unlevered betas that eliminate the effects of leverage.
This facilitates comparison across investments with different leverage levels and allows
recalculation to account forgoing forward changes in leverage levels. This recalculation involves
a technique known as the Hamada139 equation.
• Debt Betas - When a company is facing financial distress, the debt can become the new equity
as part of corporate reorganization under the federal bankruptcy code. Hence, during the
bankruptcy period, the debt trades like equity. There is a technique to adjust the beta by
calculating a debt beta. This technique is employed because in past analyses (e.g., 2012-2016),
IPP companies were bankrupt.
10.9.2 Risk-Free Rate and Equity Risk Premium
The risk-free rate of return and equity risk premium are market parameters and are not company-specific.
They also determine the average market-wide level of returns on equity. Therefore, the average return of
the market equals the sum of the risk-free rate of return and equity risk premium.
The EPA estimate is based on the approach of using long-term averages for both the risk-free rate and
the market risk premium. This avoids using or giving large weight to the currently depressed risk-free
interest rates.
In the current analysis, EPA used the 10-Year Risk-Free rate of 2.73%, based on the 10-year (2011 —
2020) average of U.S. Treasury 20-year bond rates. Additionally, the Duff and Phelps Long-Term (1926-
2020) Market Premium of 7.15% was adopted in this analysis. Thus, the total of the risk-free rate and the
market premium is 9.88%. As noted, this sum equals the expected return of the market (i.e., the beta is
one).
10.9.3 Beta
Utility betas average 0.72 during the 2016 to 2020 period on a levered basis (see Table 10-8). This
estimate is based on daily returns.
Table 10-8 Estimated Annual Levered Beta for S15ELUT Utility Index Based on Daily Returns140
Year
Levered Beta
2016-2020
0.72
139 In corporate finance, Hamada's equation is used to separate the financial risk of a levered firm from its business
risk.
140 S15ELUT Index comprises of 20 utilities. They are: American Electric Power Co Inc, ALLETE Inc, Duke Energy
Corp, Eversource Energy, Entergy Corp, Evergy Inc, Edison International, Exelon Corp, FirstEnergy Corp, Hawaiian
Electric Industries Inc, IDACORP Inc, Alliant Energy Corp, NextEra Energy Inc, OGE Energy Corp, Pinnacle West
Capital Corp, PNM Resources Inc, PPL Corp, Southern Co/The, and Xcel Energy Inc. We have excluded NRG as it is
an IPP Company.
10-11
-------
IPP levered betas average 1.04 based on weekly returns from 2016-June 2020. After decreasing
leverage for IPPs from 64% to 55%, the relevered beta was 0.86. The unlevered betas (i.e., betas without
debt impacts) of utilities is 0.33, and of IPPs is 0.45.141
10.9.4 Equity Size Premium
It is observed that the long-run returns of smaller, less liquidly traded companies have higher returns than
predicted using the market risk premium. Therefore, an equity size of liquidity premium is added. Based
on the 2020 Duff and Phelps Valuation Handbook there was a significant equity size premium for IPPs of
0.89% and a minimal premium for utilities at -0.01%.
10.9.5 Nominal ROEs
Utility
The utility ROE is 7.17% in nominal terms. The utility ROE is the single most influential parameter in the
estimate of the discount rate because of the 60% weight given to utilities compared to IPPs, and the
decrease in interest rates due to the tax shield on debt (debt interest payments are tax deductible).
The estimated utility ROE in EPA Platform v6 is lower than what state and federal commissions have
awarded the shareholder-owned electric utilities recently.142 In some cases, commissions use a different
approach or assumptions.143 Regardless of methodology, the trend overtime is to lower returns and this
is a long-term analysis focused on cost of capital for future investments that can occur 25 years or more
in the future. Thus, it could be that returns are trending toward this level and that sufficient capital can be
attracted in the future at these lower rates. Another possible explanation is that while the utilities are
allowed to earn higher returns, actual earnings will be overtime lower than allowed and closer to the
required utility ROE estimated here.
IPP
The nominal ROE for IPPs is 9.74%. The IPP required ROE is sensitive to the amount of debt and the
analysis assumes future delevering. Specifically, the IPP ROE assumes 55% debt rather than 64% debt,
which is the 2016-2020 average.
141 Unlevered betas are lower than levered betas. Levered beta is directly measured from the company's stock
returns with no adjustment made for the debt financing undertaken by the company. The leveraged beta of the
market equals one.
142 Based on Bloomberg data, the average authorized ROEs for nine Utility Companies (Southern Company,
American Electric Power Co, WEC Energy, CMS Energy, Cleco Corp, Allete Inc., Black Hills Corp, and NextEra
Energy) was 9.86% in 2019. This was less than the average earned ROE according to S&P Global Intelligence of
10.21 % in 2019, and slightly higher than their average authorized ROE of 9.64%.
143 Some regulatory commissions use what is known as the dividend growth model. This model assumes that the
current market price of a company's stock is equal to the discounted value of all expected future cash flows. In this
approach, the time period is assumed to be infinite, and the discount rate is a function of the share price, earnings per
share and estimated future growth in dividends. The challenge with using this approach is estimating future growth in
earnings. Commissions rely on stock analyst forecasts of future growth rates for dividends. In other cases,
commissions may allow for other parameters such as flotation costs (costs of issuing stock). We did not use this
approach because it is less commonly used. There also appears to be a tendency of allowed rates of return as a
group to be too low during periods with high financial costs and too high during periods of low financing costs. This
may be to ensure comparability with similar utility companies. There is also a literature that indicates that as betas
deviate from 1, the CAPM returns are too low and too high. We did not address these issues directly in part because
the results were comparable to other results, with the exception of being lower than allowed returns.
10-12
-------
10.9.6 WACC/Discount Rate
The WACCs are 4.88% in nominal terms for utilities and 6.65% in nominal terms for IPPs (see Table
10-3). Using a 60:40 utility/merchant weighting, the weighted average WACC under utility financing and
merchant financing is a 5.59% WACC. The real hybrid WACC is 3.76%.
10.10 Calculation of Capital Charge Rate
10.10.1 Introduction to Capital Charge Rate Calculations
The capital charge rate is used to convert the capital cost into a stream of levelized annual payments that
ensures capital recovery of an investment. The number of payments is equal to the book life of the unit or
the years of its book life included in the planning horizon (whichever is shorter). Table 10-9 to
Table 10-11 presents the capital charge rates by technology type used in EPA Platform v6. As discussed
in section 10.3, the changes to the Tax Code have caused capital charge rates to vary by run year,
therefore the tables below show the rates for the individual run years through 2030. Capital charge rates
are a function of underlying discount rate, book and debt life, taxes and insurance costs, and depreciation
schedule.
Table 10-9 Real Capital Charge Rate - Blended (%)144 in v6
New Investment Technology Capital Hybrid (60/40 Utility/Merchant)
2023
2025
2028 and Beyond
Environmental Retrofits - Utility Owned
10.58%
10.58%
10.58%
Environmental Retrofits - Merchant Owned
12.66%
12.70%
12.99%
Advanced Combined Cycle
8.29%
8.30%
8.39%
Advanced Combustion Turbine
8.64%
8.63%
8.69%
Ultra-Supercritical Pulverized Coal
7.92%
7.93%
8.01%
Nuclear without Production Tax Credit
7.90%
7.89%
7.94%
Biomass
7.66%
7.65%
7.65%
Wind, Solar and Geothermal
8.15%
8.15%
8.15%
Wind, Landfill Gas, Solar, and Geothermal without Property Tax and Insurance
7.00%
6.99%
6.99%
Landfill Gas
8.14%
8.14%
8.18%
Hydro
7.66%
7.67%
7.75%
Energy Storage
10.94%
10.93%
10.94%
Energy Storage without Property Tax and Insurance
9.79%
9.78%
9.80%
Table 10-10 Real Capital Charge Rate - IPP (%)
New Investment Technology Capital (IPP)
2023
2025
2028 and Beyond
Environmental Retrofits - Merchant Owned
12.66%
12.70%
12.99%
Advanced Combined Cycle
9.43%
9.46%
9.70%
Advanced Combustion Turbine
10.08%
10.05%
10.19%
Ultra-Supercritical Pulverized Coal
9.42%
9.43%
9.64%
Nuclear without Production Tax Credit
9.41%
9.38%
9.49%
Biomass
8.73%
8.72%
8.71%
144 Capital charge rates were adjusted for expected inflation and represent real rates. The expected inflation rate
used to convert future nominal to constant real dollars is 1.76%. The future inflation rate of 1.76% is based on an
assessment of implied inflation from an analysis of yields on 10-year U.S. Treasury securities and U.S. Treasury
Inflation Protected Securities (TIPS) over a period of 5 years (2016-2020).
10-13
-------
New Investment Technology Capital (IPP)
2023
2025
2028 and Beyond
Wind, Solar and Geothermal
9.14%
9.12%
9.12%
Wind, Landfill Gas, Solar, and Geothermal without Property Tax and Insurance
7.99%
7.97%
7.97%
Landfill Gas
9.15%
9.15%
9.28%
Hydro
10.61%
10.67%
11.01%
Energy Storage
11.77%
11.74%
11.77%
Energy Storage without Property Tax and Insurance
10.62%
10.58%
10.63%
Table 10-11 Real Capital Charge Rate - Utility (%)
New Investment Technology Capital Utility
2023
2025
2028 and Beyond
Environmental Retrofits - Utility Owned
10.58%
10.58%
10.58%
Advanced Combined Cycle
7.52%
7.52%
7.52%
Advanced Combustion Turbine
7.69%
7.69%
7.69%
Ultra-Supercritical Pulverized Coal
6.93%
6.93%
6.93%
Nuclear without Production Tax Credit
6.90%
6.90%
6.90%
Biomass
6.94%
6.94%
6.94%
Wind, Landfill Gas, Solar, and Geothermal
7.50%
7.50%
7.50%
Wind, Landfill Gas, Solar, and Geothermal without Property Tax and Insurance
6.35%
6.35%
6.35%
Landfill Gas
7.46%
7.46%
7.46%
Hydro
7.01%
7.01%
7.01%
Energy Storage
10.38%
10.38%
10.38%
Energy Storage without Property Tax and Insurance
9.24%
9.24%
9.24%
10.10.2 Capital Charge Rate Components
The capital charge rate is a function of the following parameters:
• Capital structure (debt/equity shares of an investment)
• Pre-tax debt rate
• Debt life
• Post-tax return on equity
• Other costs such as property taxes and insurance
• State and federal corporate income taxes
• Depreciation schedule
• Book life
Table 10-12 presents a summary of various assumed book lives, debt lives, and the years over which the
investment is fully depreciated. The EPA Base Case v6 assumes a book life of 15 years for retrofits. This
assumption is made to account for recent trends in financing of retrofit types of investments.
Table 10-12 Book Life, Debt Life, and Depreciation Schedules in v6
Technology
Book Life
(Years)
Debt Life
(Years)
U.S. MACRS Depreciation
Schedule (Years)
Combined Cycle
30
20
20
Combustion Turbine
30
15
15
Coal Steam and IGCC
40
20
20
10-14
-------
Technology
Book Life
(Years)
Debt Life
(Years)
U.S. MACRS Depreciation
Schedule (Years)
Nuclear
40
20
15
Solar, Geothermal, and Wind
30
20
5
Landfill Gas
30
20
15
Biomass
40
20
7
Hydro
40
20
20
Batteries
15
15
7
Environmental Retrofits
15
15
15
Depreciation Schedule
For the utility sector, the U.S. MACRS depreciation schedules were obtained from IRS Publication 946
that lists the schedules based on asset classes.145 146 The document specifies a 5-year depreciation
schedule for wind energy projects and 20 years for electric utility steam production plants. These exclude
combustion turbines and nuclear power plants, which each have a separate listing of 15 years. As a result
of the tax code changes, the merchant sector is allowed to depreciate assets on an accelerated schedule
through 2027. Accelerated depreciation is allowed starting in 2018 with 100% depreciation and phases
out at 20% annual between 2023 and 2027.
Taxation and Insurance Costs
The maximum U.S. corporate income tax rate is 21%.147 State taxes vary but the weighted average state
corporate marginal income tax rate is 6.45%. This yields a net effective corporate income tax rate of
26.1%.
U.S. state property taxes are approximately 0.9%, based on a national average basis. This is based on
extensive primary and secondary research conducted by EPA using property tax rates obtained from
various state agencies.
Insurance costs are approximately 0.3% on a national average basis.
145 MACRS refers to the Modified Accelerated Cost Recovery System, issued after the release of the Tax Reform Act
of 1986.
146 IRS Publication 946, "How to Depreciate Property," Table B-2, Class Lives and Recovery Periods.
147 Internal Revenue Service, Publication 542.
10-15
------- |