United States
Em ironmental Protection
Agency	
Office of Air and Radiation
 Washington. DC 20460
EPA-452/R-98-0()3/\
  September 1998
        REGULATORY IMPACT ANALYSIS
         FOR THE NOx SIP CALL, FIP, AND
              SECTION 126 PETITIONS

          Volume 1: Costs and Economic Impacts

-------
 REGULATORY IMPACT ANALYSIS
 FOR THE NOx SIP CALL, FIP, AND
       SECTION 126 PETITIONS

Volume 1:  Costs and Economic Impacts
                   Prepared by

         Office of Air Quality Planning and Standards
             Office of Atmospheric Programs
           U.S. Environmental Protection Agency
                 September 1998
                            U.S. Environmental Protection Agency
                            Region 5, Library (PL-12J)
                            77 West Jackson Boulevard, 12th FbV
                            Chicago, IL  60604*3590

-------
This page intentionally left blank

-------
                                  ACKNOWLEDGMENTS
       This Regulator} Impact Analysis was prepared as a joint effort of the Office of Air Quality Planning
and Standards (OAQPS). and the Office or Atmospheric Programs (OAP) The preparation of this report
was managed b\ Scott Mathias (EPA/OAQPS). Sam Napohtano (EPA/OAP). and Ravi Snvastava
(EPA/OAP) Both offices thank the many individuals and companies that made significant contributions to
this stud\

       OAQPS thanks John Bachmann (EPA/OAQPS). Tammy Croote (EPA/OAQPS). Robin Dennis
(EPA/ORD). Ron Evans (EPA/OAQPS), Richard Ha\nes (USDA), Bryan Hubbell (EPA/OAQPS), Norm
Possiel (EPA/OAQPS). Rosalma Rodriguez (EPA/OAQPS), Allyson Snvik (EPA/Region 6). Eric Slaughter
(Association of National Estuary Programs). Larry Sorrels (EPA/OAQPS). Abt Associates. Inc . ASI, Dyntel.
Mathtech. Inc . Pechan-Avanti Group, and Systems Applications International, Inc

       OAP thanks Keun Culhgan (EPA/OAP). Sarah Dunham (EPA/OAP). Gene Sun (EPA/OAP). Peter
Tsingotis  (EPA/OAP). and ICF Resources. Inc

-------
This page intentionally left blank.

-------
                                    EXECUTIVE SUMMARY

        EPA has finalized the nitrogen oxides (NOx) State implementation plan (SIP) call rule  The ''NOx
SIP call" requires selected eastern States to take actions to reduce emissions of NOx that contribute to
nonattainment of ozone standards in downwind States. For the purposes of this analysis, EPA has modeled
an illustrative State implementation scenario This Regulator}, Impact Analysis (RIA) and associated analyses
are intended to generally inform the public about the potential costs and economic impacts that may result
from this scenario, but specific State actions will ultimately determine the actual costs and benefits of the
NOx SIP call

        At the same time that EPA promulgates the NOx SIP call. EPA is proposing NOx Federal
implementation plans (FIPs) that may be needed if any State fails to comply with the final NOx SIP call
EPA is also proposing a response to Section 126 petitions which were filed by eight northeastern States
asking EPA to address air pollution transported from upwind States Pursuant to Executive Order 12866.
this RIA presents the potential costs and economic impacts of these rulemakings

        The existing 1-hour and ne\\ 8-hour national ambient air quality standards (NAAQS) for ozone set
levels necessary for the protection of human health and the environment  Under the Clean Air Act
Amendments of 1990 (CAAA). attainment of these standards depends on the implementation of State-
specific pollution control strategies contained in SIPs. in conjunction with EPA promulgation of national
controls for some sources  of pollution, to reduce NOx and volatile organic compound (VOC) emissions  The
NOx SIP call creates an effective, efficient and equitable approach for EPA and the States to promote
attainment with the current and ne\\ ozone standards

        In the NOx SIP call. EPA is setting ozone season NOx budgets for States that are in the SIP call
region  In nearly all cases, these budgets \\ill require States to seek lower emissions from their sources to
enable the State to meet its budget le\ el To arm e at what the NOx budgets should be for the States, the
Agenc>  considered altematne le\els of reductions that States could reasonabh require of selected stationary
sources  to reduce their summer NOx emissions in the future The final set of sources that EPA based the
State NOx budgets on  includes large electnciu  generating units, industrial boilers and combustion turbines.
stationary internal combustion engines, and cement manufacturing operations  Table ES-1 lists the major
regulator) altematnes  that EPA considered for each  of the abo\e sectors when it determined State-le\el NOx
emissions budgets in the NOx SIP call  The shaded areas in the table show  the options that EPA selected
based largely on the  Agenc\ "s determination (as explained in the preamble to this rulemaking) that the ozone
season NOx controls for a sector \\ere highK cost-effectn e and could be reasonably implemented in the near
future  For the electricity generating units  and industrial boilers and combustion turbines, the Agency
estimates the costs and emissions changes  based on an emissions cap-and-trade program  For the remaining
sectors.  EPA based its  analysis on States placing direct controls on the units covered

        In this rule, EPA has offered to administer an emissions trading program for the States.   However.
each State is free to join the program, or alternatively set up their own program to meet their NOx budget
Therefore, the actual NOx SIP call costs could vary from those that EPA estimates for the approach on which
it based the NOx budgets
                                             Page ES-1

-------
                                     Table of Contents

TABLE OF CONTENTS     .  .       .   . .        	vn

LIST OF FIGURES    	xi

LIST OF TABLES    	xii

LIST OF ACRONYMS AND ABBREVIATIONS         	xx

EXECUTIVE SUMMARY	       	ES-1

Chapter 1 INTRODUCTION AND BACKGROUND
       1 1  Introduction	          	  1-1
       12  The Clean Air Act    	            ...    1-2
           1 2 1  Ozone Requirements	    .  .      ...           .          1-3
           1 2 2  NOx Control and Ozone Reduction    	                  .    .      1-4
           123  Title IV NOx Requirements	                       1-6
           1 2 4  Nev\ Source Performance Standards    	          	1-7
           125  ReasonabK Available Control Technology Requirements        	      1-8
           126  Northeast Ozone Transport Region  ..          ...                   1-8
       13  Oven ie\\ of the NOx SIP Call Rulemakmg    ...                     	1-9
       14  Relationship Between NOx SIP Call. FIP: and  Section 126 Petitions     ...       1-9
       15  Statement of Need for the NOx SIP Call   .                 ...             1-11
           1 5 1  Statutory Authority and Legislative Requirements                            1-11
           1 5 2  Health and Welfare Effects of NOx Emissions                                1-11
           153  Need for Regulatory Action                   ...                     1-12
       16  Requirements for this Regulator. Impact Anah sis  . .       .                        1-13
           161  Executive Order  12866                   	       .              1-13
           1 6 2  Regulator^, Flexibility Act and Small Business Regulatory Enforcement Fairness Act of
                 1996   '         "      .      .            ..'...             1-13
           163  Unfunded Mandates Reform Act        .  .                                1-14
           164  Paperwork Reduction Act    .                                      .1-15
           165  Executive Order  12898    ....                        .  1-15
           166  Health Risks for Children    	       .     .    .       1-15
       17  Structure and Organization of the Regulatory Impact Analysis                  .      1-16
       1 8  References	       ...    .       .    .     . .   1-18

Chapter 2. REGULATORY ALTERNATIVES
       2 1
       22
Elements Considered in Developing Regulatory Alternatives 	
2 1 1 Tvpe of Control 	
2 1.2 Geographic Scope 	
2 1.3 Potentially Affected Sources 	
2.14 Stringency of Control Level
2 1 5 Effective Dates
2 1 6 Emissions Budget Trading Svstem Design
Definition of Regulatory Alternatives ... ....
	 2-1
	 2-1
	 2-2
	 2-4
2-4
2-5
2-5
	 2-6
                                            V/7

-------
       2 3  References                    .  .           .        ...                .2-8

Chapter 3  PROFILE OF REGULATED ENTITIES
       3.1  Electricity Generating Units   	         	       .     ...      .      3-3
       3 2  Industrial Boilers and Turbines . .        	       ...      .  .    .        . .  3-8
       3 3  Other Stationary Sources     	     . .   .  .     	3-13
       3 4  Other NOX Sources    	         	       ...   3-16
       35  Overview of Baseline Emissions    ....     .        	                  .  3-18
       36  References    	      ...      .    .   3-18

Chapter 4  METHODOLOGY FOR ESTIMATING EMISSIONS. COST, AND ECONOMIC
          IMPACTS FOR THE ELECTRJC POWER INDUSTRY
       4 1  Anahlical Oveme\\	       	     .  .        . .   .  .   4-1
       4 2  Integrated Planning Model Assumptions and Use       ...         ...     	4-2
           4 2 1   Macro Energy and Economic Assumptions     	     .                 4-4
           422   Electric Energy Cost and Performance Assumptions . .                          4-6
           423   Pollution Control Performance and Cost Assumptions    . .       .       .     .4-8
           424   Air Emissions Rates under the Base Case       ....        .        .   4-9
       43  Allowance Allocations and Trading ..        ..          .       ..                  4-10
           431   Purpose of Allowances and Assumptions about Allocations                     4-10
           432   Trading Assumptions  ....      .        . .      .  .                     4-11
       44  Administrative Costs	      . .            ...                      4-11
           441   Administrative Costs to Affected Electricity Generating Units ..     .            4-11
           442   Transaction Costs of Trading Allowances    ..        ...         .         4-12
           443   Admimstratn e Costs to States and Local Governments                       .  4-13
           444   Admimstratn e Costs to the U.S  EPA       .                           .4-13
       4 5  Direct Economic Impacts                            .                           4-14
           451   Costs to Electric Power Producers Relatn e to Revenues                        4-14
           45.2   Assessment of Potential for Passing on Cost Increases    ..       .             4-14
           453   Assessment of Potential for Closures and Additions   .           .             4-15
       4 6 Indirect Economic Impacts                                                      4-15
       47 Limitations of the Analysis  ...         	       ..     .       .      4-15
       4 8 References     .          . .               .....           .             4-16

Chapter 5  METHODOLOGY FOR ESTIMATING EMISSION REDUCTIONS. COSTS AND
          ECONOMIC IMPACTS FOR NON-ELECTRICITY GENERATING UNITS
       5.1  Analytical Overview     	     	       	        .   .   5-1
       5.2 NOx Control Technology  	            	        .       ....   5-2
           5.21   NOx Control Technology for Industrial Boilers and Turbines	          .  5-2
           5.2 2   NOx Control Technology for Other Stationary Sources	      	5-3
       5 3 Control Costs and Cost Effectiveness Methodology	   5-4
       5 4 Administrative Costs for Industrial Boilers and Turbines	  5-6
       5.5 Economic Impacts Analysis	5-6
           5.5.1   Overview of the Economic Impact Analysis Methodology	       5-6
           5 5.2   Data Sources	      5-8
       5 6 Small  Entity Economic Impacts	      . .    .   .5-10
       5.7 References	        	         5-10
                                             viu

-------
Chapter 6  RESULTS OF COST. EMISSIONS. AND ECONOMIC IMPACT ANALYSES FOR THE
          ELECTRIC POWER INDUSTRY
       61   Comparison of Uniform Options to the Initial Base Case     .          	           6-1
            6 1 1   Technology Selection        ....      	     	   6-2
            6 1.2   Emissions  .         	        ...         .      6-3
            6 1 3   Costs    ....       	   6-7
            6 1.4   Cost-Effectiveness   	       6-8
            6 1 5   Initial Base Case Compared to Final Base Case	        	6-11
       6 2  Regional versus Uniform Approach to Trading	      	    6-11
            6 2 1   Comparison of Baseline and Uniform 0.15 Option to the Two-Region
                  Alternate e	6-12
            622   Comparison of Three-Region Alternative to Uniform 0.15 Alternative	6-13
       6 3  Other Program Designs and Sensitivity to Modeling Assumptions  	      ...  6-18
            6 3 1   Costs without Interstate Trading      ...       	      . .    6-18
            632   Banking                       .       ....            ...     .    6-19
            633   Uniform 0 35 Ib/mmBtu Trading Alternative          	         6-20
            634   Sensitn ity to IPM Assumptions              .     	            6-21
       6 4  Direct Economic Impacts                    .  .        	        6-25
            641   Costs Relatne to Electricity Generation and Revenues     .                    6-26
            642   Potential Electricity Price Changes                            .              6-30
            643   Distribution of Cost Impacts Across Generation Types   .          .            6-30
            644   Potential Impacts on Small ElectnciU Generators                              6-32
            645   Potential for Closures and Additions of CapaciU                          .    6-33
       6 5  Indirect Economic  Impacts   .                                                    6-33
            6 5 1   Potential Employment Impacts           .                               .6-33
            652   Potential Impacts on Industrial Users of Electncit}                     .        6-35
            653   Potential Impacts on Households              ...            .           6-38
       66  Admmistratn e Costs                                                           6-40
       6 7  References                             .                .                        6-41

Chapter 7    RESULTS OF COST. EMISSIONS REDUCTIONS. AND ECONOMIC IMPACT
            ANALYSES FOR NON-ELECTRICITY GENERATING UNITS
       7 1   Compliance Costs  and Cost-Effectiveness   .                          .              7-1
            7 1 1   Results for Industrial Boilers and Combustion Turbines           .  .            7-2
            7 1 2   Results for Internal Combustion Engines     . .     	        .  .            7-3
            7 1 3   Results for  Cement Manufacturing (Cement Kilns)                  ...      .  7-5
            7 1 4   Results for Glass Manufacturing	     ....       .  7-7
            7 1 5   Results for Process Heaters      	    7-8
            7 1 6   Results for Commercial and Institutional Incinerators .        	      .   7-9
            717   Summary of Results for Non-Electricity Generating Sources        .    ...       7-11
            7.1.8   Administrative Costs for Non-Electricity Generating Units   	     7-12
       7.2  Potential Economic Impacts  .     	       	   7-13
            7 2 1   Overview of Potentially Affected Sources. Establishments and Firms	7-15
            7 2.2   Results for the Preferred Alternative Combination (60%/$5.000)	  7-17
            7.2.3   Comparison By Alternative    	7-22
       7 3   Small Entity Impacts  .       	       ...         	      .  .  7-23

-------
       7 4  References	                   	      	7-24

Chapter 8  IMPACTS ON GOVERNMENT ENTITIES
       8 1  State Requirements	       	      	8-1
           8 1 1  Planning Requirements    . .    	8-2
           8 1 2  Data Collection	  8-4
       8 2  Administrative Costs Associated with Electricity Generating Units	 8-6
       8.3  Administrative Costs Associated with Other Stationary Sources	     ... 8-8
       8 4  Go\ ernment-Owned Entities    .  .         	8-9
       8.5  References   	  8-10

Chapter 9  INTEGRATED COST: EMISSIONS. AND SMALL ENTITY IMPACTS SUMMARY
       9 1  Emission Reductions .     .  .       .       	     	     9-1
       9 2  Compliance Costs and Cost-Effectiveness    	     	  9-3
           9 2 1  Cost-Effectiveness Comparisons       	     . .  9-6
       9.3  Integrated Potential Small Entity Impacts        	     	  9-7

Appendix    STATE-BY-STATE OZONE SEASON NOx EMISSIONS FOR ELECTRICITY
           GENERATING UNITS BY REGULATORY ALTERNATIVE	A-l

-------
                                         List of Figures

Figure 1-1
       Flowchart of Analytical Steps      	        ...           ..            .. Page 1-16

Figure 2-1
       States Included in EPA's NOx SIP Call     	      Page 2-3

Figure 3-1
       Partitioning of NOx SIP Call Stationary Sources    	  Page 2-3

Figure 4-1
       Integrated Planning Model Regions in the Configuration Used by EPA ...     .    ..      Page 4-3

Figure 6-1
       Ozone Season NOx Emissions in 2007 from the Electric Power Industn for States in
       the SIP Call Region. Uniform Trading Alternatives Compared to the Initial Base Case     Page 6-5

Figure 6-2
       0/one Season NOx Emissions in 2007 from the Electric Power Industn- for States in
       the SIP Call Region  Uniform 0 15 Trading Alternate e Compared to the Initial Base
       Case and the State Budget Component under the 0 15 Ib/mmBtu Limit      	Page 6-6

Figure 6-3
       Comparison of 2007 Ozone Season NOx Emissions in the Initial Base Case with the
       State Component of the NOx Budget Under the T\\o-Region Alternative and the
       Emissions Results from the T\\ o-Region Alternatn e                      .            Page 6-16

Figure 6-4
       Comparison of 2007 O/.one Season NOx Emissions in the Initial Base Case with the
       State Component of the NOx Budget for the  Three-Region Alternatn e and the
       Emissions Results from the Three-Region Alternative       .     .         .            Page 6-17
                                               XI

-------
                                          List of Tables

Table ES-1
        Regulator} Alternatives by Source Category Groupings for the NOx SIP Call            . Page ES-2

Table ES-2
        Estimate of Emissions Reductions. Total Annual Costs, and Cost-Effectiveness in
        2007 of the EPA's Selected Approach to NOx SIP Call   	      Page ES-3

Table 2-1
        Regulatory Alternatives for Electricity Generating Units (ECU)    .        ...      ... Page 2-7

Table 2-2
        Regulatory Alternatives for Non-EGU Sources in the NOx Budget Trading Program
        (Large Industrial Boilers and Combustion Turbines)   	              .  Page 2-7

Table 2-3
        Regulatory Alternate es for Non-EGU Sources NOT in the NOx Budget Trading
        Program    .                   .  .     	       .           . Page 2-8

Table 3-1
        Distribution of Capacities of Potential!)  Affected Electricity Generating Utihh
        Units b\ Type in the Year 2000                              .                        Page 3-4

Table 3-2
        Distribution of Capacities of Potentially Affected Electricity Generating Utility
        Units in the Year 2000                                            .                .  Page 3-5

Table 3-3
        Distribution of Capacities of Affected Electricity Generating Utility Units
        (>25 MW) by State in the Year 2000               .       . .    "      .            .    . Page 3-6

Table 3-4
        Distribution of Capacities of Affected Electncity Generating UtiliU Units b> State
        b> Percentage in the Year 2000  	      . .       	         .  . Page 3-7

Table 3-5
        Distribution of Fossil-Fueled Units Analyzed for Rulemakmg by Initial Base Case
        NOx Emission Rate in the Year 2000   	Page 3-8

Table 3-6
        Number of Fossil-Fuel Fired Industrial Boilers and Combustion Turbines by Industry
        1995 Data   	Page 3-10
                                               xn

-------
Table 3-7
       Number of Large Fossil-Fuel Fired Boilers and Combustion Turbines b\ Fuel
       1995 Data      	        Page 3-11

Table 3-8
       Number of Large Fossil-Fuel Fired Industrial Boilers and Combustion Turbines
       by State 1995 Data  	Page 3-12

Table 3-9
       Number of Large Other Stationary Sources by Industry 1995 Data    	     .      . Page 3-14

Table 3-10
       Number of Large Other Stationary Sources by State 1995 Data	Page 3-15

Table 3-11
       Number of Total and ''No Control" Non-EGU Sources by Industry 1995 Data   .  .     .  Page 3-17

Table 3-12
       O\ en ie\v of 2007 Baseline Ozone Season NOx Emissions in the SIP Call Region   	Page 3-18

Table 4-1
       Key Baseline Assumptions for Electricity Generation    .    .        ....       ... Page 4-5

Table 5-1
       Available NOx Control Technologies for Stationary Industrial Boiler
        and Combustion Turbine Sources  .  .        .      .       ....               Page 5-3

Table 5-2
       Available NOx Control Technologies for Other Non-EGU Stationary Sources    .    .     Page 5-4

Table 5-3
       Equipment Life for Various Non-EGU Control Technologies	      .   Page 5-5

Table 6-1
       Estimated Emission Control Responses for Coal-Fired Steam Units to the NOx
       SIP Call in 2007 (MW Capacity for the SIP Call Region)  	Page 6-2

Table 6-2
       Estimated Emission Control Responses for Oil/Gas-Fired Steam Units to the NOx
       SIP Call in 2007 (MW Capacity for the SIP Call Region)	Page 6-3

Table 6-3
       Estimated Emission Control Responses to the NOx SIP Call in 2007 - Added
       Natural Gas Combined-Cycle (MW Capacity for the SIP Call region)  	Page 6-3
                                              xin

-------
Table 6-4
       Estimated Ozone Season NOx Emissions and Reductions under the Uniform
       Trading Alternatives and the Initial Base Case (1.000 tons)	Page 6-4

Table 6-5
       Incremental Annual Costs for Uniform Alternatives Relative to the Initial Base Case
       (Compliance Costs above Initial Base Case, million 1990S)	Page 6-7

Table 6-6
       Number of Fossil Fuel-Fired Units in IPM Runs Expected to Buy, Sell or Do
       Nothing in the NOx SIP Call Trading Program (0 25, 0 15, and 0 12 Uniform
       Alternatives)      	Page 6-9

Table 6-7
       Summan of Estimated Emission Reductions. Cost, and Cost-Effectiveness
       for the Uniform Alternatives of the NOx SIP Call  Selected Years          . .      .  .   Page 6-10

Table 6-8
       Comparison of Estimated 2007 Incremental Ozone Season NOx Emission Reduction.
       Cost, and Cost-Effecti\ eness for Different Regulator Alternatives     .     .     ... Page 6-11

Table 6-9
       Comparison of Ozone Season NOx Emission Reductions  Uniform 0 15 Trading
       Altername. and the Two-Region Altematne (RegionaliU  1)            ....  Page 6-12

Table 6-10
       Comparison of 2007 Ozone Season NOx Emission Reductions  0 15 State Budgets.
       Uniform 0 15 Trading, and the Three-Region Alternative (Regionahh 2)  .              Page 6-14

Table 6-11
       Comparison of the Differences Between 2007 Ozone Season NOx Emissions for
       0  15 State Budgets.  Three-Region Alternate e (Regionally 2). and Uniform 0 15
       Trading            ....            .   .                    .                   Page 6-15

Table 6-12
       Emission Reductions. Cost, and Cost-Effectiveness with and without Interstate
       Trading in 2007 for the SIP Call Region              	        	Page 6-18
Table 6-13
       Effects of Banking on Estimated Ozone Season NOx Emission Reductions.
       Incremental Cost, and Cost-Effectiveness for the 0.15 Trading Alternative	Page 6-20

Table 6-14
       Comparison of Estimated 2007 Ozone Season NOx Emission Reductions,
       Incremental Cost, and Cost-Effectiveness Between Uniform 0.35 Trading Alternative
       and 0 15 Trading Alternative .        	      	Page 6-21
                                               xiv

-------
Table 6-15
       Effects of Alternative Discount Rate Assumptions on the Emission Reductions and
       Cost in 2007 for the Uniform 0 15 Trading Alternatne        ...      .            Page 6-22

Table 6-16
       Emission Control Choices by 2007 under Lower SNCR Effectiveness 0.15 Trading    .. Page 6-22

Table 6-17
       Ozone Season NOx Emission Reductions. Incremental Cost, and Cost-Effectiveness
       in 2007 Under Lower SNCR Effectiveness 0 15 Trading	Page 6-23

Table 6-18
       Ozone Season NOx Emission Reductions. Incremental Cost, and Cost-Effectiveness
       in 2007 Assuming Shorter Equipment Life.  0 15 Trading        ....      .    .. Page 6-24

Table 6-19
       Effects of Assuming Full NERC Demand/No CCAP Reduction on the Effects of
       the NOx SIP Call in 2007 0 15 Trading	        . Page 6-25

Table 6-20
       Effects of Assuming Retail Competition—More Demand/Less Nuclear Po\\er on the
       Effects of the NOx SIP Call in 2007  0 15 Trading              	Page 6-25

Table 6-21
       Generation Changes  and Costs Compared to Generation in 2007 for Uniform
       Alternatives of Differing  Stringency  ...     ....           .        .    .        . Page 6-26

Table 6-22
       NOx SIP Call Compliance Costs by Alternatn e Compared to Re\ enues from
        Electncit> in 2007 (1990$)  .         ..              	       Page 6-27

Table 6-23
        Potential Net Cost (After Allowance Purchases/Sales) b> Unit Type under 0 15
        Trading Alternative in 2007 (mills/kWh. 1990S)                                     Page 6-32

Table 6-24
        Potential Impact on Emplovment in the Control Technology Sector  0 15 Trading
        (Construction and Installation)	      	     ....   Page 6-34

Table 6-25
        Potential Impact of 0.15 Trading Alternative on Labor Requirements in the Control
        Technology Sector (O&M) in 2007 (Full-Time Equivalent) 	Page 6-34

Table 6-26
        Potential Effects of 0 15  Trading Alternative on Coal Production and Emplovinent
        Demand in 2007         	       	    Page 6-34
                                              xv

-------
Table 6-27
       Potential Effects of 0 15 Trading Alternative on Natural Gas Production and
       Emplo\-ment Demand in 2007    	Page 6-35

Table 6-28
       Summary of Potential Labor Demand Impacts of 0 15 Trading Alternative in 2007	Page 6-35

Table 6-29
       Potential Impacts of Electricity Rate Increases in 2007 on Energy-Intensive
       Industries of 15 Trading Alternative (by Two-Digit SIC Code, 1990S)	Page 6-37

Table 6-30
       Potential Impacts of Electricity Rate Increases in 2007 on Energy -Intensive
       Industries of 15 Trading Alternative (by Four-Digit SIC Code, 1990$)	Page 6-38

Table 6-31
       Potential Impacts of Electricity Rate Increases in 2007 on Households b> Income
       Category of 15 Trading Alternative (1990S)   ..            .  .         ...      .   Page 6-39

Table 6-32
       Potential Administrative Costs for Electricity Generating Units. 2007 (1990$)           Page 6-40

Table 6-33
       Potential Allowance Trading Transaction Costs for Electncit> Generating Units
       for the NOx SIP Call b\ Uniform Alternate         ...      .      .            Page 6-41

Table 6-34
       Potential Total Administrate e Costs to Owners of ElectriciU Generating Units in 2007
       (million 1990$)          .            .       .             	          .   Page 6-41

Table 7-1
       2007 Ozone Season NOx Emission Reductions for Large Industrial Boilers and
       Turbines                              .              .         	        . .   Page 7-2

Table 7-2
       2007 Cost and Cost-Effectiveness Results for Large Industrial Boilers and
       Combustion Turbines    	     	        . .   .      ....   Page 7-3

Table 7-3
       2007 Ozone Season NOx Emission Reductions for Large Stationary 1C Engines3  	Page 7-4

Table 7-4
       2007 Cost and Cost-Effectiveness Results for Large Stationary 1C Engines	Page 7-5

Table 7-5
       2007 Ozone Season NOx Emission Reductions for Large Cement Manufacturing
                                               xvi

-------
       Operations                .           ...        .                  .                 Page 7-5

Table 7-6
       2007 Cost and Cost-Effectiveness Results for Large Cement Manufacturing Operations     Page 7-6

Table 7-7
       2007 Ozone Season NOx Emission Reductions for Large Glass Manufacturing
       Operations     ....        	Page 7-7

Table 7-8
       2007 Cost and Cost-Effectiveness Results for Large Glass Manufacturing Operations ....  Page 7-8

Table 7-9
       2007 Ozone Season NOx Emission Reductions for Large Process Heaters    .  .    .       Page 7-8

Table 7-10
       2007 Cost and Cost-Effectiveness Results for Large Process Heaters    ...       ...  Page 7-9

Table 7-11
       2007 Ozone Season NOx Emission Reductions for Large Commercial and Industrial
       Incinerators                          ...         .          ...          .  Page 7-10

Table 7-12
       2007 Cost and Cost-Effectn eness Results for Large Commercial and Industrial
       Incinerators .         .               ....            	              Page 7-10

Table 7-13
       2007 Ozone Season Emission Reductions and Total Annual Compliance Costs
        for Non-Electncit} Generating Sources Used to Establish State NOx Emissions
        Budgets under the SIP Call                . .           .        	         Page 7-11

Table 7-14
        Control Technologies Selected for Non-Electncit> Generating Sources for Reglulaton Alternatives
        Used to Establish State NOx Emissions Budgets under the SIP Call   .                .  Page 7-12

Table 7-15
        Average Per Source Annual Administrative Costs for Other Stationary Sources in
        2007 (1990 dollars)	Page 7-13

Table 7-16
        Potential Regulatory Option Combinations for Non-Electricity Generating Unit
        Economic Impact Analysis	Page 7-14

Table 7-17
        Number of Firms and Other Entities Potentially Affected, by Source Category	Page 7-15
                                               XVJl

-------
Table 7-18
       Number of Firms Potentially Affected, by Sector and Size	     .       .  Page 7-16

Table 7-19
       Number of Potentially Affected Firms by Firm Costs as a Percentage of
       Sales/Expenditures:  60%/$5.000    	Page 7-17

Table 7-20
       Number of Establishments b\ Costs as a Percentage of Value of
       Shipments/Expenditures and Sector and Firm Size 60%/$5,000    	Page 7-18

Table 7-21
       Number of Establishments By Establishment-Level Costs as a Percentage of Value of
       Shipments/Expenditures and Industry 60%/$5,000	Page 7-19

Table 7-22
       Number of Firms by Firm Costs as a Percentage of Sales/Expenditures and by
       Regulator.- Alternative	    ....        .  .         	   Page 7-22

Table 7-23
       Number of Establishments b\ Establishment-Level Costs as a Percentage of Value of
       Shipments/Expenditures and by Regulators Altematne                  ...       .  Page 7-22

Table 7-24
       Number of Potentially Affected Small Entities by Cost as a Percentage of
       Sales/Expenditures by Regulator. Alternative           .  .       	     . Page 7-23

Table 7-25
       Potential!) Affected Small Entities that Ma\ Incur Compliance Costs of Greater Than
       One Percent of Sales/Revenues for the 60%/$5.000 Regulator. Alternate e              Page 7-24

Table 8-1
       Schedule of Reporting Activities for Each Year During the Period 2003
       through 2008     .                                       	    Page 8-5

Table 8-2
       Administrative Costs Associated with EGUs to States in the SIP Call domain
       in 2007 (1990$)	               	Page 8-7
Table 8-3
       Administrative Costs Associated with EGUs and non-EGUs in the SIP Call
       Region to EPA in 2007 (1990$)	Page 8-8

Table 8-4
       2007 Annual Costs To Potentially Affected Government-Owned ECU NOx
       Emissions Sources  0.15 Trading Regulator. Alternative	       	Page 8-9
                                             xvni

-------
Table 8-5
       2007 Annual Costs To Potentially Affected Government-Owned NOx Emissions
       Sources  60%/$5,000	             	   Page 8-10

Table 9-1
       2007 Ozone Season NOx Emissions and Emission Reductions for Selected
       Combinations of Electricity Generating Units and Other Stationary Source
       Regulator} Alternatives from the Initial Base Case (thousands of NOx Tons)     	Page 9-2

Table 9-2
       2007 Annual NOx SIP Call Compliance Costs for Selected Combinations of
       Electricity Generating Unit and Other Stationary Source Regulatory Alternatives
       (millions of 1990 dollars)    	Page 9-4

Table 9-3
       2007 Ozone Season Average Compliance Cost-Effectiveness for Selected
       Combinations of Electricity Generating Unit and Other Stationary Source
       Regulatory Alternatives (1990 dollars per ton of NOx reduced in the
       ozone season) . .     .       . .     .      ...       ....       	Page 9-5

Table 9-4
       A\erage Cost-Effectiveness of NOx Control Measures Recenth Undertaken
       or Proposed (1990 dollars)	       . .      .            .  Page 9-6

Table 9-5
       Number of Potentially Affected Small Entities for the NOx SIP Call
       Rulemakmg                   .  .                                .                 Page 9-8

Table A-1
        2007 Ozone Season Emissions Estimates for the Electric Power Industn for
        States in the SIP Call Region ....              ...                   .            Page A-2

Table A-2
        Comparison of Electric Power Industn, 2007 Ozone Season Emissions for the
        Initial Base Case, the 0 15 Budget Component, and the 0 15 Trading Option
        for States in the SIP Call Region	          . - Page A-3

Table A-3
        Comparison of State-by-State 2007 Ozone Season Emissions for the Initial Base Case.
        Regional Budgets, and Regional Alternatives	Page A-4
                                                XJX

-------
                          Select List of Acronyms and Abbreviations
ACT - Alternative Control Techniques
AEO - Annual Energy Outlook
AF - Air/Fuel Adjustment
AF RATIO - Air-to-Fuel Ratio
AFS - AIRS Facility System
AIM - Architectural and Industrial Maintenance
AIRS - Aerometnc Information Retrieval System
ANPR - Advanced Notice of Proposed Rulemakmg
beM - Total Atmospheric Light Extinction Coefficient
BACT - Best Available Control Technology
BBL - Barrel
CAA - Clean Air Act
CAAA - Clean Air Act Amendments of 1990
CAPI - Clean Air Power Initiate e
CC - Combined Cycle
CCAP - Climate Change Action Plan
CEM - Continuous Emissions Monitoring
CI - Compression Ignition
CO - Carbon Dioxide
CTGs - Control Technique Guidelines
D&B - Dun & Bradstreet
DoD - Department of Defense
DOE - Department of Energy
DUNS  -Dunn & Bradstreet Numbering System
ECOS - Environmental Council of States
EIA - Energy Information Administration
EIS - Environmental Impact Statement
EGUs - Electncit} Generating Units
EIIP - Emission Imentoiy Improvement Program
EO - Executive Order
EPA - Em ironmental Protection Agency
ESEERCO - Empire State Electric Energy Research Corporation
ETS - Emissions Tracking S\stem
FCM - Fuel Consumption Model
FGR - Flue Gas Rebuming
FHWA - Federal Highway Administration
FINDS - Facihtv Index System
FIP - Federal Implementation Plan
FTE - Full Time Equivalent
g/bhp-hr - Grams per Brake Horsepower Hour
GDP - Gross Domestic Product
GLM - General Linear Model
GSA - General Sen-ices Administration
GW - Gigawatts
HPMS - Highway Performance Monitoring System

-------
hr - Hour
1C - Internal Combustion
ICI - Industnal/Commercial/Institutional
ICR - Information Collection Request
lEc - Industrial Ecolog}
IGCC - Integrated Gasification Combined Cycle
IMPROVE - Interagenc} Monitoring for Protection of Visual Environments
IPM - Integrated Planning Model
IPPs - Independent Po\\er Producers
IR - Ignition Timing Retardation
ISCST3 - Industrial Source Complex Short Term
Kg/ha - Kilograms Per Hectare
knr - Square Kilometer
kWh - Kilowatt Hour
LAER - Lowest Achievable Emissions Rates
Ib - Pound
L-E - low emission
LNB - Lo\\-NOx Burners
mills/kWh - Mills Per Kilowatt Hour
MM4 - Mesoscale Model, \ersion 4
mmBtu - Millions of British Thermal Units
Mm- Megameter
MOU - Memorandum of Understanding
MW - Megawatts
MWe - Megawatt of ElectnciU
MWh - Mega\\ att Hours
NAA - Nonattainment Area
NAAQS - National Ambient Air Qualih Standards
NAPAP - National Acid Precipitation Assessment Program
NATS - National Allowance Tracking S\stem
NEEDS - National Electric Energ> Data System
NEPA - National Em ironmental Protection Act
NERC - North American Electric Reliability Council
NET - National Emission Trends
NH3 - Ammonia
NLEV - National Lo\\ Emission Vehicle
NOAA - National Oceanic and Atmospheric Administration
NOx - Oxides of Nitrogen
NPR - Notice of Proposed Rulemakmg
NPS - Non-Point Source
NSPS - New Source Performance Standards
NSR - New Source Review
03 - Ozone
OMB - Office of Management and Budget
OMS - Office of Mobile Sources
O&M - Operation and Maintenance
OTAG - Ozone Transport Assessment Group
OT and WI - Oxygen Trim and Water Injection
                                             XX!

-------
OTC - Ozone Transport Commission
OTR - Ozone Transport Region
PM - Paniculate Matter
ppm - Parts Per Million
PRA - Paperwork Reduction Act of 1995
PSD - Prevention of Significant Deterioration
RJA - Regulatory Impact Analysis
RACT- Reasonably Available Control Technology
RFA - Regulatory Flexibility Act
RVP - Reid Vapor Pressure
SBA - Small Business Administration
SBREFA - Small Business Regulator, Enforcement Fairness Act of 1996
SCR - Selectn e Catalytic Reduction
SI - Spark Ignition
SIC - Standard Industrial Classification
SNCR - Selective Non-Catalytic Reduction
SNPR - Supplemental Notice of Proposed Rulemakmg
SO: -  Sulfur Dioxide
SUBS - Statistics of U S  Businesses
tpd  - Tons Per Da>
tp\  - Tons Per Year
TRJ - Toxics Release Inventory
TV A - Tennessee Valle\  Authority
UAM-V - Urban Airshed Model - Variable Scale
UMRA - Unfunded Mandates Reform Act
USDA - United States Department of Agriculture
VMT  - Vehicle Miles Tra\ eled
VOCs - Volatile Organic Compounds
VOS - Value of Shipments

-------
                                             Table ES-1
                         Regulators Alternatives by Source Category Groupings
                                        for the NOx SIP Call
Electricity Generating Units
(EGUs)-Emissions Budgets
Based on a NOx Limit of:
025 Ib/mmBtu"
0 20 Ib/mmBtu
0 1 5 Ib/mmBtu in Northeast.
0 20 Ib/mmBtu in Midwest &
Southeast
0 12 Ib/mmBtu in Northeast.
0 15 Ib/mmBtu in Midwest.
0 20 Ib/mmBtu in Southeast
0.15 Ib/mmBtu
0 12 Ib/mmBtu
Non-Electricity Generating Units (non-EGUs)
Industrial Boilers and
Combustion Turbines— Emissions
Budgets Based on a Reduction
from Uncontrolled Levels of:
40%
50%
60%
70%
All Other Stationary Sources-
Emissions Budgets Based on
Highest Ozone Season NOx
Reduction Achievable without a
Source Paying More Than:
$l,500/ton
$2,000/ton
$3.000/ton
$4,000/ton
S5,000/ton
 1 See Chapter 1 for a breakdown of the States co\ered in the NOx SIP call
Emission Reductions, Costs, and Cost-effectiveness

        Table ES-2 summarizes EPA estimates of the emission reductions, costs, and cost-effectiveness for
the regulator, approach that EPA selected as the basis for the NOx SIP Call's NOx budgets  (Please note
Since these estimates \\ere calculated EPA has fine-tuned its estimates of the NOx budgets and an addendum
to this executne summary pro\ ides revised emission reduction, cost and cost-effectiveness information)
Overall. 82% of the emission reductions expected under this regulatory alternative are expected to come from
the electric  power industr\, at an average ozone season cost-effectiveness of $ 1,468 per ton  The table
indicates the estimates of direct control costs for sources including costs associated with emissions
monitoring and reporting The table also indicates the total administrative costs to State governments and
EPA  In EPA's analysis to support this rule, the Agency has shown that for the electric power industry, the
largest source of emissions for  which it considered controls, a single trading program across the SIP call
region can provide a similar reduction to what direct command-and-control requirements would accomplish,
but do the job at lower cost For this reason, the Agency is encouraging States to participate in the trading
program that it plans to administer
                                              Page ES-2

-------
                                              Table ES-2
                          Estimate of Emission Reductions, Total Annual Costs,
              and Cost-Effectiveness in 2007 of the EPA's Selected Approach to NOx SIP Call
Sector
Electricity Generating Units a
Industrial Boilers and Turbines b
Internal Combustion Engines c
Cement Manufacturing c
Administrate Costs for EGUs
Administrative Costs to States and
EPA
Total
Ozone Season
NOx Emission
Reductions
(1,000 tons)
938
104
83
16

1,141
Total Annual Cost
(millions 1990S)
$1,378
$153
$100
$24
$6
$2
SI, 660"
Average Ozone Season
Cost-Effectiveness
(S per ozone season ton)
$1,468
$1,467
$1,215
$1,458

' Does not include additional monitoring costs (see later rou)
b Includes additional monitoring and other administrative costs associated with participating in the XOx emissions trading program
' Includes additional monitoring and other administrative costs associated with the SIP call rule
d Numbers do not add due to rounding
Economic Impacts

        EPA considered \\hat the economic impacts could be. if States implemented the regulatory approach
that EPA used to calculate the NOx SIP call budgets for electricity generating units  Electricity prices could
potentialK- rise in the NOx SIP call region by as much as 1.6 percent in 2007, if the power industry is pricing
its  po\ver on the basis of marginal costs in a fully competitive environment  The price increase will be less, if
these assumptions regarding the nature of the competitive environment do not hold  There will be more ne\\
electric generation capacity built in response to the rule than will retire early (there will be little generation
capacity that closes)  On net. EPA expects this NOx SIP call to create more new jobs (from pollution control
operations and increased natural gas use) than it reduces (due to a small decline in forecasted coal demand)

        The analysis of non-EGU sources indicates that fewer than 5% of potentially affected firms
experience costs in excess of 1% of revenues, and just over 2% of potentialK affected firms experience costs
in excess of 3% of revenues   EPA also examined the potential affect of the NOx SIP call on small entities
that meet the Small Business Administration's definition of "small." The Agency adopted several ways of
minimizing potential impacts on small entities for the final NOx SIP call rulemaking  Of the nearly 1,200
small entities (both EGU and non-EGU) in the NOx SIP call region that have large NOx emissions sources,
only 150 are potentialK- affected by the SIP call rule, and only 41 have potential compliance costs in excess of
1% of total revenues  EPA expects States to use these results to help them design control strategies that will
reduce or eliminate adverse impacts on small entities
                                               Page ES-3

-------
Limitations

        Evaluating the potential costs of a rulemaking provides one framework for policy makers and the
public to assess pohc> alternatives  Not all the potential costs can be captured in any analysis   However.
EPA is generally able to estimate reasonably well the costs of pollution controls based on today's control
technology and assess the important impacts when it has sufficient information for its analysis.  EPA
compiled through the OTAG process and from many other sources sufficient information for this
rulemaking  There are. ho\\e\er, important limitations in the RIA analysis.

        There are some data limitations in some aspects of the RIA. despite the Agency's extensive efforts to
compile information for this rulemaking. While they exist. EPA believes that it has used the models and
assumptions that are made to conduct its analysis in a reasonable way based on the available evidence, but
this should be kept in mind when reviewing various aspects of the RIA's results

        Another factor that adds to the uncertainty of the results is the potential for pollution control
innovations that can occur over time   It is impossible to estimate how much of an impact, if am. ne\\
technologies that are just now emerging may ha\e in lowering the compliance costs for the NOx SIP call.
\\hich goes into effect in 2003  We can onK recognize their possible influence

        There is the uncertainty regarding future costs that exists due to the flexibility that occurs under the
emissions cap-and-trade program that EPA is encouraging the States  to set up.  The analysis that EPA has
done to date has been fairly consen atn'e in considering  the electric power industn,' and large industrial boilers
and combustion turbines operating separately under their own trading programs  In reality, they should enter
the same trading pool and there should be greater efficiency and lower costs that result

        Qualitative and more detailed discussions of the above and other uncertainties and limitations are
included in the analysis Where information and data exists, quantitatne characterizations of these
uncertainties are included  However, data limitations prevent an o\erall quantitative estimate of the
uncertainty associated with final estimates Nevertheless, the reader should keep all of these uncertainties and
limitations in mind \\hen reviewing and interpreting the  results
Addendum to Executive Summary

        In response to comments. EPA has revised the State NOx budgets that it set for the electric po\\er
industn on the basis of 15 Ibs/mmBtus in the final days of the rulemaking process  The SIP call region
budget was lowered from 564 thousand tons of NOx during the ozone season to 544 thousand tons of NOx
The Agency also decided to create a "compliance supplement pool" for use in 2003  and allow banking with
flow controls in the trading program that EPA is encouraging States to undertake

        For the adjustment of the NOx budget to 544 thousand tons for the electricity generating units, the
Agency estimates that there will be a reduction of ozone season NOx emissions by 958 thousand tons in 2007
at an annual cost of $1,440 million  This is an average cost-effectiveness of $1.503 per ton of NOx
reductions during the ozone season  The total ozone season NOx emission reductions from the NOx SIP call
if the States implement the program the way EPA used to set the budget is 1.161 thousand tons
                                             Page ES-4

-------
       An adjustment to the emissions inventory for the non-EGU sources was also made as a result of
public comments  The reanaly sis following these emission inventory adjustments indicated only minor
changes in the costs
                                            Page ES-5

-------
                       Chapter 1.  INTRODUCTION AND BACKGROUND
1.1      Introduction
        This document presents the cost and economic impact estimates for a Regulator. Impact Analysis of
the final NOx SIP call rule, which addresses regional transport issues related to ozone attainment1.  This rule
requires certain States to take action to reduce emissions of nitrogen oxides (NOx) that contribute to
nonattainment of ozone standards in downwind States  This RIA also satisfies the analytical requirements for
the proposed NOx Federal Implementation Plan (FIP) and Clean Air Act (CAA) section 126 petition actions
The proposed FIP may be needed if any State fails to revise its SIP to comply with the final NOx SIP call
The proposed action under CAA section 126 responds to petitions filed with EPA by eight Northeastern
States requesting that EPA provide relief from emissions sources in several upwind States that ma\ be
contributing to ozone nonattainment in the petitioning States"

        The Clean Air Act (CAA) requires States to demonstrate attainment of the National Ambient Air
QuahU  Standards (NAAQS) for ozone  Mam  States have found it difficult to demonstrate attainment of the
ozone NAAQS due to the widespread regional transport of ozone and its precursors, NOx and volatile
organic  compounds (VOCs)  The Ozone Transport Assessment Group (OTAG) was established in 1995 to
undertake an assessment of the regional transport problem in the Eastern half of the United States  OTAG
\\as a collaboratne process among 37 affected States, the District of Columbia, the U S Environmental
Protection Agency (EPA), and interested members of the public, including environmental groups and industry
representatn es

        OTAG concluded that regional reductions in NOx emissions are needed to reduce the transport of
ozone and its precursors  OTAG recommended that major sources of NOx emissions (utility and other
stationan sources) be controlled under State NOx budgets, and also recommended development of an
emissions trading program

        After a revie\\ of OTAG's analysis, findings, and recommendations. EPA proposed a rule to limit
summer season NOx emissions in a group of States that the Agency believes are significant contributors to
ozone in do\\mvmd areas 3  In a November 7. 1997 Notice of Proposed Rulemakmg  (NPR). EPA made a
determination that transport of ozone from certain States in the OTAG region4 makes a significant
contribution to nonattainment. or interferes \\ith the maintenance of attainment, with the ozone NAAQS in
downwind States (FR 1997a)  EPA proposed a summer season NOx budget (in tons of NOx) for each of
these States These States will be required to amend their State Implementation Plans (SIPs) through a call-in
procedure established in  Section 110 of the Clean Air Act Amendments of 1990 (CAAA) In a May 1998
        1 Ground level (or troposphenc) ozone is an air pollutant that forms when its two primary components, oxides
of nitrogen and volatile organic compounds , combine in the presence of certain meteorological conditions

        2 Unless necessary to provide specific emphasis, the term "NOx SIP call'" will be used  (rather than "FIP" or
''section 126 petitions'") throughout this report when referring to the regulator}' framework that is analyzed and reported
in this RIA  See section 1 4 for additional detail on the analytical relationship between these three regulator}' actions

        3 NOx emissions reductions were proposed for 22 States and the District of Columbia

        * The OTAG region consists of 37 States east of 104° W  longitude

                                              Page 1-1

-------
Supplemental Notice of Proposed Rulemakmg (SNPR). EPA made technical corrections to the State NOx
budgets, and developed a proposed trading rule to provide for emissions trading (FR 1998a)  The SNPR also
included an anahsis of the air quality impacts of the proposed rule The State NOx emissions budgets.
trading rule, and related provisions are no\\ being promulgated as a final rule

       A technical background support document prepared for the November 7. 1997 NPR estimated costs
and emissions reductions associated with an assumed strategy that States might take to achieving the
proposed budgets (EPA.  1997a)  These analyses were updated to reflect technical corrections to the
population of sources and grovsth estimates on which the State-specific budgets were based and assess the
effects of the proposed trading system, in an analysis supporting the April 1998 SNPR (EPA 1998a).

       This document provides  the cost and economic impact estimates for the Regulatory Impact Analysis
(RIA) of the final rule  This analysis expands  and updates the previous anaKses. to reflect the provisions of
the final rule and to provide analysis of the potential economic impacts as well as the cost and emissions
impacts associated  with the rule

       The remaining sections of this chapter address the following topics

        1.2     Relevant requirements of the Clean Air Act.

        1.3     Oven lew of the NOx SIP call rulemakmg.

        1  4     Relationship between the NOx SIP call. FIP. and  section 126  actions.

        1  5     Statement of need for the NOx SIP call.

        1  6     Administrative requirements addressed b> this RIA.

        1  7     Structure of the  RIA and organization of this document, and

        1 8     References for Chapter 1


1.2     The Clean Air Act

        The 1970 Clean  Air Act Amendments required EPA to issue, periodically re\iew, and. if necessary
revise, NAAQS for ubiquitous air pollutants (Sections 108 and 109).  States are required to submit SIPs to
attain those NAAQS. and Section 110 of the CAA lists minimum requirements that  SIPs must meet
Congress  anticipated that all areas would attain  the NAAQS by 1975   In 1977. the CAA was amended to
provide additional time for areas to reach the NAAQS  and included the requirement  that States reach the
NAAQS for ozone by  1982 or 1987.  In addition, the 1977 amendments included provisions that required
SIPs to consider  adverse  downwind effects and allowed downwind States to petition for tighter controls on
upwind States that contribute to their NAAQS nonattainment status

        In 1990. the Clean Air Act was again amended  This section outlines  requirements of the 1990
Clean Air Act Amendments (CAAA) related to  NOx reductions and the NOx SIP call. The discussion
includes the ozone  and NOx requirements and a review of the guidelines for new or advanced air emissions
control technologies

                                             Page 1-2

-------
1.2.1    Ozone Requirements

        The CAAA included provisions designed to address the continued nonattainment of the existing
ozone NAAQS. specified requirements that would apply if EPA revised the existing standard, and addressed
transport of air pollutants across State boundaries

        In 1991 and 1992. areas not in attainment with the 1-hour ozone NAAQS were placed in one of five
classifications, based on the degree of nonattainment  Requirements for moving toward attainment, including
definitions of "major source" for VOCs and NOx. attainment dates and new source offset ratios, were
established for each of the five classifications  Within an area known as the Northeast Ozone Transport
Region (OTR). all sources emitting 50 tons or more of ozone forming pollutants a year are defined as ''major
sources." regardless of their current attainment classification  Certain emissions limits apply to major
sources, and e\ en more stringent requirements apply for ne\\ major sources in nonattainment areas

        Since passage of the 1990 CAAA. EPA has revised the NAAQS for ozone EPA is required to
re\ iew the NAAQS at least every five years to determine whether, based on new research, revisions to the
standards are necessan to continue to protect human health and the environment As a result of the most
recent review. EPA revised the NAAQS for both particulate matter and ozone The previous ambient air
qualm standard for ozone \\as 0 12 ppm based on 1-hour averaging of monitoring results  The revised
standard \\as set at 0 08 ppm based on an 8- hour averaging period  The 1-hr standard remains in effect until
EPA determines that a gnen area has air quality meeting its 1-hour standard  This is necessan to ensure
continued progress in those areas and a smooth transition between the two standards

        On JuK 16. 1997. President Clinton issued a directive to EPA on the implementation strategy for the
new ozone and particulate NAAQS The goal of the implementation strategy is to provide flexible, common-
sense, and cost-effectn e means for communities and businesses to comply with the ne\\ standard  The EPA
has issued proposed guidance for public comment on implementation of the revised standards (August 24.
1998. 6? FR 45060)  Additional guidance will be proposed m October 1998  The August and October
guidance will be combined and issued as one document in December 1998   The implementation strateg>
includes

        Endorsement of a Regional Approach Citing EPA's work with the OTAG. the implementation
        strategy notes that ozone needs to be addressed as a regional problem  The Directn e indicates that.
        based on OTAG recommendations. EPA will propose a rule to provide a flexible, common-sense.
        and cost effectn e means for communities and businesses to comply with the new standards  The
        strategy states that EPA  will  encourage and assist the States to develop a regional emissions cap-and-
        trade system, modeled on the current acid rain program, as a way to achieve reduction in NOx
        emissions at lower cost

        Transitional Classifications- Areas that attain the 1-hour standard but that do not attain the new 8-
        hour standard will be eligible for a specific "transitional" classification, if they participate in a
        regional strategy and/or submit early plans addressing the new  standard  EPA will revise its rules for
        new source review (NSR) and conformity so that States will be able to comply with the new
        standards with only minor revisions to the existing programs in such transitional areas   Areas which
        will achieve attainment as a result of the regional strategy need not implement any additional local
        controls. Areas that will not achieve the 8-hour standard even with the regional strategy are eligible
                                             Page 1-3

-------
       for transitional status if the\ submit re\ ised SIPs in the > ear 2000 demonstrating attainment of the 8-
       hour standard on the same schedule as the regional transport requirements.

       Cost-Effective Implementation Strategies EPA will encourage States to design strategies for both
       the PM and ozone standards that focus on getting low cost reductions and that limit the cost of
       control to under $10.000 per ton for all sources. EPA will encourage market-based strategies to
       lower the cost of attainment and stimulate technology innovation.

The NOx SIP call, therefore, plays an important role in the implementation strategy  for the new ozone
NAAQS. by instituting a regional strategy that will encourage cost-effective attainment of the nev\ standard
1.2.2    NOx Control and Ozone Reduction

        To address the CAAA provisions regarding continued nonattarnment of the existing ozone NAAQS.
EPA's post-1994 attainment strategy guidance for the 1-hour ozone standard called for continued emissions
reductions within ozone nonattamment areas together with a national assessment of the ozone transport
phenomenon  Recognizing that no individual slate or jurisdiction can effectively assess or resohe all of the
issues rele\ ant to ozone transport, the Environmental Council of States (ECOS) formed a national \\ork
group to address ozone pollution 5  OTAG was established to assist states east of tbe Mississippi River to
attain federal  ozone standards and to develop regional strategies to address regional transport problems 6 The
multi-state, multi-stakeholder OTAG process included input from State and local governments, industry
environmental groups, and the Federal government.  The stated goal of OTAG was to

        Identify  and recommend a strategy to reduce transported ozone and its precursors which, in
        combination with other measures, will enable attainment and maintenance of the national ambient
        ozone standard in the OTAG  region A number of criteria will be used to select the strategy
        including, but not limited to. cost effectiveness, feasibility, and impacts on ozone levels (OTAG.
        1995)

        OTAG's work included development of a comprehensive base-year (1990) emissions m\ enton for
use in all OTAG analyses  The inventory contained information provided by the States  and re\ ie\\ed b\
OTAG for point, area, and mobile sources  State-specific growth  factors-were used so project emissions  for
the \ears 1999 and 2007. which represent the CAAA attainment dates for certain nomattainmem  areas
Baseline 2007 emissions v>ere also adjusted to reflect the effect of various controls required under existing
regulatory programs or expected from future programs

        OTAG then conducted modeling of NOx and ozone across the OTAG regkan for several scenarios
using geographic and atmospheric models
        5 ECOS is a national organization of environmental commissioners with members from the 50 Stales and
 territories

        6 Information on OTAG and copies of documents produces by the group can be accessed on-line at
 http //\v\vu epa gov/ttn/otag

                                              Page 1-4

-------
        Strategy Modeling  OTAG Strategy Modeling \vas done in several phases, and included anaK sis of
more than 25 emission control strategies  OTAG found that domain wide emissions of NOx in the 2007
baseline are approximate!} 12 percent lo\\er than 1990 and emissions of VOC are approximately 20 percent
lo\\er  Thus, existing CAA programs are expected to produce a reduction in ozone concentrations in mam
nonattainment areas  However, the analysis showed that some areas currently in nonattainment will likely
remain so in the future and that new 8-hour nonattainment and/or maintenance problem areas may develop as
a result of economic growth in some areas

        Geographic Modeling  OTAG conducted geographic modeling to isolate the effects of NOx
reductions on specific subregions  Among other results. OTAG found that a regional strategy focusing on
NOx reductions across a broad portion of the region will help mitigate the ozone problem in mam  areas of
the East Further, a regional NOx emissions reduction strategy coupled with local NOx and/or VOC
reductions ma\ be needed to achieve attainment and maintenance of the NAAQS in the region,

        This anahses conducted b\ OTAG (OTAG 1997). as well as EPA's analyses in support of the nev\
ozone NAAQS (EPA  1997b). showed the important role that reducing NOx emissions plays in the reduction
of ozone le\ els The extensive air qualm  modeling performed by OTAG indicated that both ozone and NOx
can be transported long distances, up to 500 miles While reductions in either NOx and VOCs ma\ reduce
ozone in localized urban areas, only NOx reductions would result in lower ozone levels across the region
The OTAG anaK ses showed a correlation between the magnitude and location of NOx reductions  and the
magnitude of reductions in ozone levels in downwind areas  OTAG. therefore, reached the following
conclusion

        Regional NOx reductions are effective in producing ozone benefits, the more NOx reduced, the
        greater the benefit Ozone benefits are greatest where emission reductions are made and dimmish
        \\ ith distance  Ele\ ated and lew  le\ el NOx reductions are both eft'ecm e (OTAG 1997, pp 51 -52)

        Based on the e\idence of the relationship between NOx emissions and regional ozone levels. OTAG
recommended that a range of NOx controls be applied in certain areas of the OTAG region A wide \ aricn of
sources are responsible for NOx emissions, including electnciU  generating units, other (non-utihh)
stationary  sources, area sources, non-road mobile sources, and highwa\  vehicle sources OTAG did not
suggest am one "right" approach to reducing major source NOx emissions However. OTAG developed a
number of specific recommendations for EPA pertinent to the NOx SIP call, including the following "

        OTAG-related controls should be implemented in the "fine grid" states 8

        The range of utility NOx controls should fall between Clean Air Act controls and the less stringent of
        85% reduction from the 1990 rate (Ib/mmBtu) or 0 15 Ibs  of NOx /mmBtu summer heat input

        The stringency of controls for individual large non-utility point sources should be established in a
        manner equitably with utility controls, and RACT should be considered for individual medium non-
         Summaries of the OTAG findings and recommendations are provided in OTAG 1997

        8 The fine grid states include those modeled using UAM-V at a grid resolution of 12 km2 All other areas
constitute the coarse grid which is modeled at a grid resolution of 36 km2. Coarse grid states are Florida. Louisiana.
Texas. Arkansas. Oklahoma. Kansas. Nebraska, North Dakota, South Dakota, and Minnesota


                                             Page 1-5

-------
        utility point sources where appropriate 9 OTAG recommended that EPA calculate statewide NOx
        tonnage budgets based on a specified relationship between control levels for coal-fired power plants
        and control targets (emission reduction percentages) for large and medium non-utilit} point sources

        OTAG stated that market-based approaches are recognized as having a number of benefits in relation
        to traditional command and control regulations, and that States have the option to select market
        systems that best suit their needs They described two basic approaches that States might use to
        implement NOx emissions market systems, and recommended that a joint State/EPA Workgroup be
        formed to develop design features and implementation provisions for market systems that could be
        selected by the States

OTAG also made recommendations that EPA develop and adopt a variety of specific national regulations that
were assumed for the modeling to result in reduced emissions of VOCs and/or NOx. and to reach closure on
the Tier 2 Motor Vehicle Study

        The recommendations resulting from the extensive anal} sis and air quality modeling conducted by
OTAG have played a major role in the design of the NOx SIP call
1.2.3   Title IV NOx Requirements

        Title IV of the CAAA requires annual reductions in NOx emissions  The Acid Rain NOx Program
under Title IV incorporates a two-phased strateg> to reduce NOx emissions In the first phase, starting
January 1.  1996. some Group 1 boilers (i e . dry bottom \\all-fired boilers and tangential!) fired boilers) are
required to comply with  specific NOx emission limitations10  In the second phase, starting January 1. 2000.
the remaining Group 1 boilers must comply \\ith more stringent NOx emission limits " Further. Group 2
boilers (i e . wet bottom  wall-fired boilers, cyclones, boilers using cell-burner technology. and \ertically fired
boilers) must comply with recently established emission limits 1:

        Compliance results for 1996 show that, from 1990 to 1996. the Phase I affected population's
average NOx emission rate declined by 40 percent O\erall NOx emission reductions between 1990 and
1996 for the affected boilers totaled about 340;000 tons, i e a reduction of 33 percent (EPA. 1997c)  In
Phase II. about 1  17 million tons per > ear of NOx reductions are projected to result from the Acid Ram NOx
Program requirements (EPA.  1996)
        9  OTAG provided specific definitions of large and medium point sources, for purposes of their
recommendations

        10 The affected dry-bottom wall-fired boilers must meet a limitation of 0 50 Ibs of NOx per inmBtu averaged
over the year, and tangentially fired boilers must achieve a limitation of 0 45 Ibs of NOx per mmBtu. again averaged
over the year (FR 1995)
        " Annual averages of 0 46 Ib/mmBtu for dry-bottom wall-fired boilers and 0 40 Ib/mmBtu for tangentially
fired boilers

        I: The limits are 0.68 Ib/mmBtu for cell burners, 0 86 Ib/mmBtu for cyclones greater than 155 MWe. 0 84
Ib/mmBtu for \\et bottom boilers greater than 65 MWe. and 0 80 Ib/mmBtu for vertically fired boilers (FR 1996a)
                                              Page 1-6

-------
        In developing State budgets for the NOx SIP call. EPA considered the NOx reductions committed to
by Title IV NOx Program requirements
1.2.4   New Source Performance Standards

        The EPA is under court order to promulgate a new source performance standard (NSPS) on fossil-
fuel-fired utilih and industrial boilers in September 1998, and subpart GG of Part 60 regulates NOx
emissions from combustion turbines  The final standards revise the NOx emission limits for steam
generating units in subpart Da (Electric Utility Steam Generating Units) and subpart Db
(Industnal-Commercial-Institutional Steam Generating Units).  Only those electricity generating units and
industrial steam generating units for which construction, modification, or reconstruction is commenced after
Juh  9. 1997 would be affected by these revisions.

        The NOx emission limit in the final rule for new subpart Da units is 201 nanograms per joule (ng/J)
[ 1 6  lb/mega\\att-hour (MWh)] gross energy output regardless of fuel type  For existing sources that become
subject to subpart Da through modification or reconstruction, the NOx emission limit is 0 15 Ib/milhon Btu
heat  input  For subpart Db units, the  NOx emission limit being proposed is 87 ng/J (0 20 Ib/milhon Btu)
heat  input from the combustion of am gaseous fuel, liquid fuel, or solid fuel. ho\\ever. for lo\\ heat release
rate units firing natural gas or distillate oil. the current NOx emission limit of 43 ng/J (0 10 Ib/milhon Btu)
heat  input is unchanged

        In de\ eloping the State budgets for the NOx SIP call. EPA considered the potential NOx reductions
attributable  to this NSPS
1.2.5   Reasonably Available Control Technology Requirements

        In the 1977 amendments to the CAA Congress required that all SIPs for nonattamment areas contain
reasonabh a\ ailable control measures (RACM) or reasonabh available control technology (RACT)  In the
1990 Amendments to the Act. Congress created RACT requirements specificalh for ozone nonattamment
areas under the 1-hour standard (see subpart 2 of part D of title I)  Since 1977, EPA has defined RACT for
ozone as the lowest emission limitation that a particular source is capable of meeting by the application of
control technology that is reasonably available considering technological and economic feasibility  The EPA
historically has interpreted the RACT requirement in ozone nonattamment areas to apply independent of a
State's ability to demonstrate that an area will attain the ozone standard, with certain exceptions

        In the ozone-specific RACT requirement enacted in 1990, States were required to correct all existing
deficiencies in RACT rules in marginal nonattamment areas to ensure the rules were adopted consistently  on
a national basis  In addition, all nonattamment areas classified moderate and above were required to adopt
RACT for each source category  for which EPA issued a Control Techniques Guideline (CTG) Over the
years, EPA has issued CTG documents to assist the States in determining RACT for VOCs. Each CTG
contains information on available air pollution control techniques and provides a "presumptive norm" for
RACT for a specific source category  Finally, RACT for controlling NOx was also required in certain
nonattamment areas classified moderate and above

        In developing implementation guidance for the revised 8-hour NAAQS, EPA is addressing the
RACM/RACT requirement under subpart I of part D of title I, rather than subpart 2  The EPA has proposed

                                            Page 1 -7

-------
implementation guidance for the revised ozone NAAQS which addresses several issues, including
RACM/RACT  The proposed policy states that "For the 8-hour ozone NAAQS. if the [nonattainment] area
is able to demonstrate attainment of the standard as expeditiously as practicable with emission control
measures in the SIP. then RACM/RACT will be met and additional measures would not be required as being
reasonably available " (August 24,  1998, 63 FR 45060)  The policy will be finalized by December 31, 1998.
1.2.6    Northeast Ozone Transport Region

        Section 184 of the CAAA delineated a multi state ozone transport region (OTR) in the Northeast and
required specific additional NOx and VOC controls for all areas in this region (not only nonattainment areas).
 Section 184 also established the Ozone Transport Commission (OTC) for the purpose of assessing the
degree of ozone transport in the OTR and recommending strategies to mitigate the interstate transport of
pollution  The OTR consists of the States of Connecticut, Delaware, Maine, Man land. Massachusetts, New
Hampshire, New Jersey. New York, Pennsylvania, Rhode Island. Vermont, parts of northern Virginia, and the
District of Columbia  The OTC was first convened in  1991, and began analysis and evaluation of ozone
reduction strategies for the region. They concluded that regional reductions of NOx emissions are
particularly important in reducing ozone The OTR States confirmed that they would implement RACT on
major stationary sources of NOx. and agreed to a phased approach for additional controls, beyond RACT. for
power plants and other large fuel combustion sources

        This agreement, known as the OTC Memorandum of Understanding (MOU) for stationary source
NOx controls \\as approved  on September 27, 1994 All OTC States, except Virginia, are signatories to the
OTC NOx MOU  The OTC NOx MOU establishes an emissions trading system to reduce the costs of
compliance \\ith the control requirements

        In de\ eloping State budgets for the NOx SIP call. EPA considered the NOx reductions committed to
b> the OTR states in the OTC NOx MOU. along with the OTAG recommendations discussed above
1.3     Overview of the NOx SIP Call Rulemaking

        EPA relied extensively on the OTAG analyses and recommendations in developing the NOx SIP call
As recommended b\ OTAG. the rule establishes ozone season'3 NOx emission budgets for 22 States and the
District of Columbia M  The 23 jurisdictions will be required to amend their SIPs by the year 2000. to allocate
emissions control requirements among sources and to develop compliance programs for each affected source
category to ensure that the NOx budget is met   These compliance programs should include, necessary
pollution control measures; monitoring, reporting, and accounting procedures to ensure source emissions are
not exceeding the State's NOx budget; and enforcement requirements
        13 The ozone season for this rule is the penod May 1 - September 30

        14 The States covered by the rule include Alabama, Connecticut, Delaware, Georgia, Illinois, Indiana.
Kentucky, Maryland, Massachusetts, Michigan, Missouri, New Jersey, New York, North Carolina, Ohio. Penns\h ania.
Rhode Island. South Carolina, Tennessee, Virginia. West Virginia, and Wisconsin

                                             Page 1-8

-------
        Consistent \vith OTAG's recommendation that NOx emissions reductions be achieved pnmariK
from large stationary sources in a trading program, EPA is encouraging States to consider additional controls
on electricity generating units and other large stationary sources as a strategy for meeting statewide budgets
State budgets were developed using assumptions consistent with such a strategy. The budget for each State
was developed for components of major source categories. For non-road and highway vehicle sources.
budgets are based on estimates of the effectiveness in each State of national measures that EPA is taking to
control emissions from mobile sources  For electricity generating units and other stationary sources, the
budgets are based on applying further reasonable controls  A major factor in determining controls is the cost-
effectiveness of control measures

         EPA also followed OTAG's recommendation in urging States to consider implementing market
based s> stems to reduce the costs of complying with the new limits on NOx emissions EPA is encouraging
the States and the District of Columbia to join a trading program administrated by EPA. which is reflected in
a model NOx Budget Trading Rule This trading system would place a collective cap on NOx emissions from
electricity generating units and other large boilers and combustion turbines, and provide for trading of
allowances similar to the CAAA Title IV S0: Allowance Trading Program already in place

        Chapter 2 of this report describes a number of regulatory alternatives that EPA considered in the
development of this final rule
1.4     Relationship Between NOx SIP Call, FIP, and Section 126 Petitions

        In conjunction with promulgating the NOx SIP call. EPA has begun efforts to respond to petitions
filed b> eight northeastern States (FR 1998b)  These petitions were filed under section 126 of the CAA.
\\hich authorizes States to petition EPA to address air pollution transported from upwind States  The
petitions request that EPA make a finding that NOx emissions from certain major stationary sources
significantly contribute to ozone nonattamment problems in the petitioning States If EPA makes such a
finding, the Agenc> would be authorized to establish Federal emissions limits for these sources   The
petitions recommend control levels for EPA to consider  In an April 30. 1998 Advanced Notice of Proposed
Rulemaking (ANPR) (63 FR 24058). EPA presented a schedule for taking actions on the petitions, made a
preliminary identification of upwind sources that may significantly contribute to 1-hour and 8-hour ozone
nonattamment problems in the petitioning States (using information developed for the NOx SIP call NPR).
and requested comment on legal and pohc\ issues raised by section 126 of the CAA. In responding to the
section 126 petitions. EPA intends to be consistent with the approaches taken in the NOx SIP call

        At the same time that EPA promulgates the NOx SIP call rule, EPA is proposing NOx Federal
Implementation Plans (FIPs) that may be needed if any State fails to comply with the final NOx SIP call rule
The FIP requirements are intended to be consistent with the approaches taken in the final NOx SIP call,
including a proposed federal NOx Budget Trading Program for electric utility sources and other large
industrial boilers and combustion turbines

        Since the final NOx SIP call and the proposed FIP and section 126 petition actions are generally
consistent in the manner in which they assess affected emissions sources. EPA is preparing only a single RJA
for all three actions  Even though the facts of the analysis contained in this report do not differ significantly
for any of the three actions, the results have slightly different interpretations  In the case of the final NOx SIP
call, the results in this report are illustrative of potential costs and economic impacts that may result from the
                                              Page 1-9

-------
SIP call  The NOx SIP call itself does not directh impose regulatory requirements on emissions sources
Instead, the SIP call requires States to develop strategies to meet the State NOx budgets contained in the final
NOx SIP call rule  States have discretion on which emissions sources to control to realize the required
reductions  It is EPA's position that for the final NOx SIP call the analytical requirements associated with
the Regulatory Flexibility Act (RFA)and the Unfunded Mandates Reform Act (UMRA) do not apply (see
sections 1 6.2 and 1.6 3 below)  Nonetheless, EPA has performed analyses consistent with the analytical
requirements in the RFA and UMRA, and summarized the results of these analyses in the RIA

        Ho\\ever. the FIPs, if needed, and the section 126 petition responses will directly impose regulators
requirements on emissions sources EPA is proposing to regulate sources under the FIP and section 126
petition actions w ith strategies that are modeled in this RIA  In these cases the results presented in the RIA
reflect potential outcomes from direct federal regulation, and. depending on the  outcome of the final actions.
ha\ e a higher probability of reflecting the actual outcome of the rules   For the proposed FIP and section 126
actions, it is the EPA's position that the analytical requirements of the  RFA and UMRA do apply
Accordingly. EPA has performed the required analv ses for these proposals and  summarized the results of
these analyses in the RIA.  These analyses will be updated as necessary as part of any final actions EPA takes
on the FIP and section 126 petitions

        The proposed section 126 actions will potentially affect only a subset of the sources potential!}
affected b\ the broader NOx SIP call  Sources in Georgia. South Carolina, and Wisconsin are not affected b\
the proposed section 126 rule  Therefore, the costs and economic impacts associated with the proposed
section 126 rule are hkeK to be smaller than for the final NOx SIP call. Since Georgia. South Carolina, and
Wisconsin are affected under the final NOx SIP call, and would be subject to a  final FIP if the> fail to compK
with the proMsions of the final NOx SIP call. EPA did not see the need to separately address the potential!)
smaller costs and economic impacts for the proposed section  126 rule
 1.5     Statement of Need for the NOx SIP Call

        The following sections discuss the statutory  authority and legislative requirements of the NOx SIP
 call, health and welfare effects of NOx emissions, and the basis for the regulatory actions of the NOx SIP
 call
 1.5.1   Statutory Authority and Legislative Requirements

        Section 110(a)(2)(D) provides that a SIP must contain provisions preventing its sources from
 contributing significantly to nonattamment or interfering with maintenance of the NAAQS in a downwind
 State  This section applies to all pollutants covered by NAAQS and all areas regardless of their attainment
 designation  Section 110(k)(5) authorizes EPA to find that a SIP is substantially inadequate to meet any
 CAA requirement, as well as being inadequate to mitigate interstate transport as described in Sections 184
 and 176A Such a finding would require States to submit a SIP revision to correct the inadequacy within a
 specified period of time.
                                              Page 1-10

-------
1.5.2   Health and Welfare Effects of NOx Emissions''

        NOx emissions contribute to the formation of ozone during the summer season  Ozone is a major
component of smog and is harmful to both human health and the environment  Research has shown the
following health effects of ozone

        Exposure to ambient ozone concentrations has been linked to increased hospital admissions for
        respirator, ailments, such as asthma Repeated exposure to ozone can make people more susceptible
        to respirators  infection and lung inflammation, and can aggravate preexisting respiratory diseases

        Children are at risk for the effects of ozone because they are active outside during the summer
        months when  ozone levels are at their highest Adults who are outdoors and moderately active during
        the summer months are also at risk  These individuals can experience a reduction in lung function
        and increased respirators symptoms, such as chest pain and cough, when exposed to relatively low
        ozone levels during periods of moderate exertion

        Long-term exposures to ozone can cause repeated inflammation of the lung, impairment of lung
        defense mechanisms, and irreversible changes in lung structure, which could lead to premature aging
        of the lungs and/or  chronic respiratory illnesses such as emphysema and chronic bronchitis

        Twenty -one peer reviewed epidemiology studies recently published suggest a possible association
        between ozone exposure and mortality

Ozone has also been shown to ad\erseh affect \egetation. including reductions in agricultural and
commercial forest yields,  reduced growth and decreased sun i\ ability of tree seedlings, and increased tree and
plant susceptibility to disease, pests and other environmental stresses

        NOx emissions also contribute to fine particle matter formation (PM)  Exposure to airborne PM has
a \\ide range of ad\erse health effects  The  ke> health effects associated with PM include  1) premature
mortality. 2) aggravation  of respiratory and cardiovascular disease (as indicated by increased hospital
admissions and emergency room visits, school absences, work loss da\s. and restricted actruty days). 3)
changes in lung function and increased respiratory s\Tnptoms. 4) changes to lung tissues and  structure. 5)
altered respiratory defense mechanisms, and 6) chronic bronchitis  Most of these effects have been
consistent!) associated with ambient PM concentrations, which have been used as a measure  of population
exposure, in a number of community epidemiological studies  Although mechanisms by \\hich particles
cause effects ha\ e not  been elucidated, there is general agreement that the cardio-respiratory s\ stem is the
major target of PM effects  Paniculate matter also is associated with welfare effects, which include visibility
impairment, soiling, and materials damage

        Based on its revie\\ of the scientific  evidence. EPA established standards for PM2 5 and retained the
standards for PMU,  The EPA revised the secondary  (welfare-based) PM NAAQS by making them identical
to the primary standards

        Finally. NOx emissions contribute to a wide range of health and environmental problems
independent of their contribution to ozone or PM formation. Among these problems are acid deposition,
        1? A comprehensive discussion of health and environmental issues related to NOx appears in EPA, 1997d

                                             Page 1-11

-------
nitrates in the drinking water, and nutrient loading in waterways, particular!) in sensitive coastal estuaries
where air deposition is a major portion of nitrogen loadings
1.5.3   Need for Regulatory Action

        The existing and revised ambient air quality standards for ozone set levels necessary for the
protection of human health and the environment.  Under the CAA. attainment of these standards depends on
the implementation of State-specific pollution control strategies contained in SIPs to reduce NOx and volatile
organic compound emissions, in conjunction with EPA promulgation of national controls for some sources of
pollution

        It is clear that, even with planned national measures in place, several States cannot bring existing
nonattainment areas into compliance with the current ozone standard, or avoid the application of very costly
local control measures, unless the transport of ozone from other upwind areas is reduced  Furthermore, many
States AM!! find it hard, if not impossible, to avoid nonattainment with the re\ised ozone NAAQS. or come
into attainment with it in the future, unless mitigation of the ozone transport problem occurs This dilemma
has raised concerns over the fairness of downwind areas having to cope with the pollution coming from areas
up\\md The current regulatory framework requires States to develop SIPs that demonstrate air quality
improvements sufficient to reach specific attainment levels. States have no  control over neighboring States'
actions, and may be unable to meet their air quality goals due to pollutants transported across State lines
The contribution of upwind sources outside of nonattainment areas creates a dilemma for States seeking to
reach air quality goals

        States could develop local ozone mitigation strategies to address the impact of transported ozone
Hov\e\ er. local efforts could lead to undesirable outcomes  Some States might develop SIPs that do not
achie\ e compliance in some serious and severe ozone nonattainment areas, because the States would deem
local measures needed to achieve attainment as too dracoman

        The NOx SIP call is designed to mitigate these problems through a coordinated Federal and State
effort to address regional ozone transport  This rule will  create a more effective, efficient and equitable
approach for EPA and the States to promote attainment with the current and new ozone NAAQS
1.6     Requirements for this Regulatory Impact Analysis

        This section describes various legislative and executive requirements that govern the anahtical
requirements for Federal rulemakmgs. and describes how each analytical requirement is addressed in this
RIA
1.6.1   Executive Order 12866

        Executive Order 12866, "Regulatory Planning and Review" (FR, 1993), requires EPA to provide
the Office of Information and Regulatory Affairs of the Office of Management and Budget with an
assessment of the costs and benefits of significant regulator}' actions  A ''significant regulatory action" is
defined as "any regulatory action that is likely to result in a rule that may
                                              Page 1-12

-------
        Ha\e an annual effect on the economy of $100 million or more or adverseK affect in a material way
        the econoim. a sector of the economy, productivity, competition, jobs, the environment, public health
        or safety. or State, local, or tribal governments or communities.

•       Create a serious inconsistency or otherwise interfere with an action taken or planned by another
        agency.

•       Materially alter the budgetary impact of entitlements, grants, user fees, or loan programs or the rights
        and obligations of recipients thereof, or

        Raise novel legal or policy issues arising out of legal mandates, the President's priorities, or the
        principles set forth in the Executive Order"' (FR. 1993)

For any such regulator, action, the Agency must provide a statement of the need  for the proposed action.
must examine alternative approaches, and must estimate social benefits and costs

        EPA  has determined that the NOx SIP call is a significant regulatory action because its effect on the
econorm is expected to exceed SI00 million per \ ear  This RJA provides the cost and economic impact
information required b> E 0  12866 for a significant  regulatory action  A separate document provides the
benefits information required b\ the Executive Order
1.6.2    Regulatory Flexibility Act and Small Business Regulatory Enforcement Fairness Act of 1996

        The Regulatory Flexibility Act (RFA) of 1980 (PL 96-354) requires that agencies conduct a
screening analysis to determine whether a regulation will have a significant impact on a substantial number of
small entities, including small businesses, governments and organizations  If a regulation will have such an
impact, agencies must prepare a Regulator} Flexibility Analysis, and comply with a number of procedural
requirements to solicit and consider flexible regulator, options that mmimi/e adverse economic impacts on
small entities  The RFA's anahtical and procedural requirements were strengthened  by the Small Business
Regulator,- Enforcement Fairness Act (SBREFA) of 1996

        For reasons explained more fully in the Federal Register notice for the final NOx SIP call, it is EPA's
position that the RFA as amended b>  SBREFA does not apply to the final NOx SIP call, because the rule
does not impose direct requirements on emissions sources   States will ultimately decide what emissions
limits are imposed for specific sources  However, the EPA has determined that the RFA as amended by
SBREFA does apply to both the proposed FIP and section 126 actions.  Therefore. EPA has examined the
potential for small entity impacts to provide pohc\ makers and States with additional decision information.

        The RFA and SBREFA require use of definitions of "small entities", including small businesses.
governments and non-profits, published by the Small Business Administration (SBA)16 Screening analyses
of economic impacts presented in Chapters 6, 7, and 9 of this RIA examine potential  impacts on small
entities
        16 Where appropriate, agencies can propose and justift alternative definitions of "small entirv " This RIA relies
on the SBA definitions
                                             Page 1-13

-------
1.6.3   Unfunded Mandates Reform Act

       The Unfunded Mandates Reform Act (UMRA) of 1995 (PL 104-4) was enacted to focus attention on
federal mandates that require other governments and private parties to expend resources without federal
funding, to ensure that Congress considers those costs before imposing mandates, and to encourage federal
financial assistance for intergovernmental mandates. The Act establishes a number of procedural
requirements The Congressional Budget Office is required to inform Congressional committees about the
presence of federal mandates in legislation, and must estimate the total direct costs of mandates in a bill in
any of the first five years of a mandate, if the total exceeds $50 million for intergovernmental mandates and
$100 million for private-sector mandates

       Section 202 of UMRA directs agencies to provide a qualitative and quantitative assessment of the
anticipated costs and benefits of a Federal mandate that results in annual expenditures of $100 million or
more  The assessment should include costs and benefits to State, local, and tribal governments and the
private sector, and identify any disproportionate budgetary impacts  Section 205 of the Act requires agencies
to identify and consider alternatives, including the least costly, most cost-effective, or least burdensome
alternative that achieves the objectives of the rule

       For reasons explained more fully in the Federal Register notice for the NOx SIP call, it is EPA's
position that section 202 of UMRA does not apply to the final NOx SIP call, because the annual estimated
costs of possible SIP submittals by States is less than $100 million and no Federal or pmate sector mandates
are direct!} imposed  Ho\se\er, EPA has determined that UMRA does apply to both the proposed FIP  and
proposed section 126 rules  Chapter 8 of this RJA presents a summary of analyses of the potential impacts of
the NOx  SIP call on State and local governments, to support compliance with Section 202 of UMRA   This
analysis includes administrative requirements of State and local governments associated with revising SIPs
and collecting and reporting data to EPA  It also includes the compliance and administrative costs to
emissions sources owned b\ go\ eminent entities
 1.6.4   Paperwork Reduction Act

        The Paperwork Reduction Act of 1995 (PRA) requires Federal agencies to be responsible and
 publicly accountable for reducing the burden of Federal paperwork on the public  EPA has submitted an
 Information Collection Request (ICR) to the Office of Management and Budget (OMB) in compliance with
 the PRA The ICR explains the need for additional information collection requirements and provides
 respondent burden estimates for additional paperwork requirements to State and local governments
 associated with the NOx SIP call
 1.6.5   Executive Order 12898

        Executive Order 12898. "Federal Actions to Address Environmental Justice in Minority Populations
 and Lo\\-Income Populations."  requires federal agencies to consider the impact of programs, policies, and
 activities on minority populations and low-income populations. Disproportionate adverse impacts on these
 populations should be avoided.  According to EPA guidance, agencies are to assess whether minority or lo\\-
 income populations face risk or a rate of exposure to hazards that is significant (as defined by the National
                                              Page 1-14

-------
Environmental Pohc> Act) and that "appreciably exceeds or is Iikeh to appreciably exceed the risk or rate to
the general population or other appropriate comparison group." (EPA, 1996b) This guidance outlines EPA's
Environmental Justice Strategy and discusses environmental justice issues, concerns, and goals identified by
EPA and environmental justice ad\ocates in relation to regulatory  actions  The NOx SIP call is expected to
provide health and \\elfare benefits to eastern U.S. populations, regardless of race or income
1.6.6    Health Risks for Children

        Executive Order 13045, "Protection of Children from Environmental Health Risks and Safety
Risks." directs Federal agencies developing health and safety standards to include an evaluation of the health
and safety effects of the regulations on children  Regulatory actions covered under the Executive Order
include rulemakings that are economical!} significant under Executive Order 12866. and that concern an
environmental health risk or safety  risk that the agency has reason to believe may disproportionately affect
children  EPA has developed internal guidelines for implementing the E.O 13045 (EPA, 1998b)

        The NOx SIP call is a '"significant economic action." because the annual costs are expected to
exceed $ 100 million  Both NOx and ozone formed by NOx  are known to affect the health of children and
other sensitive populations, \\hich were addressed in the de\ elopment of the new ozone NAAQS However.
the NOx SIP call is not expected to have a disproportionate impact on children
1.7     Structure and Organization of the Regulatory Impact Analysis

        The potential costs and economic impacts and benefits have been estimated for this rulemaking  The
flo\\ chart in Figure 1-1 summarizes the anahtical steps taken in de\ eloping the these estimates
                                             Page 1-15

-------
                                            Figure 1-1
                                  Flowchart of Analytical Steps
                                       Model Air Quality
                                                I
                                    Select Control Strategies
                                                                        \

                                                                       BENEFITS
         Estimate Control Cost
Estimate Post-Control Air Quality
                    I
               1
   Estimate Small Business and Other
           Economic Impacts
     Estimate Human Health
       and Welfare Effects
                                                                           1
                                                               Estimate Monetized Value of
                                                                Health and Welfare Effects
        The assessment of costs, economic impacts, and benefits consists of multiple analytical components.
dependent upon emissions and air quality modeling  In order to estimate baseline air quahn in the year 2007.
emission imentones are developed for 1995 and then projected to 2007, based upon estimated national
growth in industry earnings and other factors Current CAAA-mandated controls (e g . Title I reasonabh
available control measures. Title II mobile source controls. Title III air toxics controls. Title IV acid ram
sulfur dioxide (SO;) controls) are applied to these emissions to take account of emission reductions that
should be achieved in 2007 as a result of implementation of the current PM  and ozone requirements  These
2007 CAA emissions in turn are input to several air quaht> models that relate emission sources to area-
specific pollutant concentrations  This modeled air quality is used as the base against which several
alternatue control options are measured and cost estimates developed  Given the estimated costs of the
alternative regulator} control  options, the potential economic impacts of these estimated costs on  potentialk
affected industry sectors is subsequently analyzed

        The RJA analyses have been constructed such that costs are estimated incremental to those derived
from the effects of implementing the CAAA in the year 2007  These analyses provide a "snapshot" of
potential costs of this rulemaking in the context of implementation of CAA requirements between now and
2007 and the air quality effects that derive from economic and population growth

        States have discretion in how they achieve their NOx budgets, and different States may choose
different strategies The RIA must, therefore, be based on assumptions about how the States will choose to
implement the NOx SIP call requirements  Consistent with EPA's recommendation that States focus on
                                            Page 1-16

-------
major stationary sources, this RIA assumes that States impose additional controls — incremental to those
alreadx required by other national programs that address NOx emissions — onh for major stationary sources.
and that States implement the cap-and-trade system for electricity generating units and industrial boiler and
turbine sources This assumption is illustratn e of one cost-effective approach States could take to meeting
the NOx SIP call budgets  States may choose different allocations of controls across major stationary sources
than assumed here, or may choose to impose additional controls on area or mobile sources as well  Costs and
economic impacts would differ from those estimated in this RIA to the extent that  States" compliance
strategies differ from the RIA assumptions

        Anal} sis of costs, changes in emissions, and economic impacts is conducted separately for two
groups of sources electncit} generating units and other stationary sources  The Integrated Planning Model
(IPM) allov\s analysis of trading and industry-level adjustments for electricity generating unit sources Other
stationary sources are analyzed separate!}. using assumptions about baseline conditions and control costs that
are generally consistent \Mth the IPM  modeling assumptions used for electricity generating units

        Predicted changes in emissions due to the additional controls for electncit} generating units and other
stationary sources are then combined  to estimate changes in air quality and to calculate the benefits of the
NOx SIP call  These air quality and benefits  analyses are discussed in a separate document

        The remainder of the RIA is organized in the following chapters and appendices

        Chapter 2 presents a discussion of the regulator) alternatives considered by EPA for this rulemakmg.

•       Chapter 3 characterizes the regulated communit}. including the electric power industry. large
        industrial boilers and combustion turbines throughout industry, and other  sources of NOx emissions
        throughout industry.

        Chapter 4 describes the methodolog} used to estimate costs, emissions reductions and economic
        impacts (including  small entity impacts) for the electric power industry.

•       Chapter 5 describes the methodology for estimating costs, emissions reductions and economic
        impacts (including  small entit} impacts) for other stationary sources.

        Chapter 6 presents the results of the analysis of costs, emission reductions, and economic impacts
        under each regulatory option  for the electric power mdustn,.

•       Chapter 7 presents the estimated costs, emission reductions, and economic impacts for other
        stationary sources.

•       Chapter 8 analyzes impacts on State  and Federal governments, as implementers of the regulatory
        program and as  owners of affected sources of NOx emissions, to provide the analyses  required by
        UMRAiand

•       Chapter 9 presents an integrated cost and small entity impacts summary associated with the NOx SIP
        call
                                              Page 1-17

-------
Where appropriate, each chapter includes a discussion of limitations of the analysis A series of appendices
follo\\ Chapter 9 and provide more detailed descriptions of specific methodologies and results
1.8    References

Federal Register 1993 Executive Order 12866. Regulatory Planning and Review Vol. 58. October 4,
1993. pg 51735

Federal Register. 1995. Acid Rain Program, Nitrogen Oxides Emission Reduction Program Vol  60, No
71. April 13, 1995

Federal Register, 1996 Acid Rain Program: Nitrogen Oxides Emission Reduction Program Final Rule
Vol  61, No 245, December 19. 1996

Federal Register. 1997a Finding of Significant Contribution and Rulemaking for Certain States in the
Ozone Transport Assessment Group Region for Purposes of Reducing Regional Transport of Ozone,
Proposed Rule  Vol 62. No 216, November 7. 1997,pg 60318.

Federal Register. 1997b Memorandum of July 16. 1997 from William J. Clinton. President of the United
States, to the Administrator of the Em ironmental Protection Agenc>   Subject  "Implementation of the
Revised Air Quality Standards for Ozone and Paniculate Matter " Vol 62, July 18. 1997. pg 38421

Federal Register. 1997c National Ambient Air Quality Standards for Ozone and Paniculate Matter,
Nonce of Proposed Rulemaking Vol 62. No  241. December 13. 1997

Federal Register. 1998a  Supplemental Notice for the Finding of Significant Contribution and Rulemaking
for Certain States in the Ozone Transport Assessment Group Region for Purposes of Reducing Regional
'Transport of Ozone  Vol  63. No 90. Ma\  11. 1998

Federal Register. 1998b Finding of Significant Contribution and Rulemaking on Section 126 Petitions for
Purposes of Reducing Interstate Ozone Transport  Vol 63, No 83. April 30. 1998

 Ozone Transport Assessment Group. 1995   Pohc>  Paper approved  by the Polic> Group on December 4.
 1995

Ozone Transport Assessment Group, 1997.  Executive Report 1997

U S  Environmental Protection Agency, 1996a  Fact Sheet on Nitrogen Oxides Emission Reduction Program
Office of Air and Radiation. Washington, DC.. December 1996.

U.S. Environmental Protection Agency .1996b  Guidance for Incorporating EnvironmentalJustice
Concerns in EPA 'sNEPA Compliance Analyses (Review Draft). Office of Federal Activities, Washington,
DC., July 12, 1996.

U.S. Environmental Protection Agency. 1997a  Proposed Ozone Transport Rulemaking Regulatory
Analysis  Office of Air and Radiation. Washington. D C., September 1997.
                                            Page 1-18

-------
U S Em ironmental Protection Agencx.  1997b Regulatory Impact Analysis for the Paniculate Matter and
Ozone National Ambient Air Quality Standards and Proposed Regional Haze Rule. Office of Air Quaht>
Planning and Standards. Research Triangle Park, NC July 1997

US Environmental Protection Agency,  1997c / 996 Compliance Report  EPA Report 430/R-97-025,
Office of Air and Radiation. Washington D C . June  1997

U S Environmental Protection Agency,  1997d Nitrogen Oxides Impacts on Public Health and the
Environment. EPA Report 452/R-97-003. Office of Air and Radiation, Washington, D C . August 1997

US Environmental Protection Agency,  1997e EPA Interim Guidance for Implementing the Small Business
Regulatory Enforcement Fairness Act and Related Provisions of the Regulatory Flexibility Act Februan 5.
1997

U S Enuronmental Protection Agencv  1998a Supplemental Ozone Transport Rulemaking Regulatory
Analysis Office of Air and Radiation. Washington. D C . April 7. 1998

U S Em ironmental Protection Agenc\.  1998b Memorandum from Tro\ ato and Kelly to Assistant
Administrators  Subject "Implementation of Executive Order 13045.  Protection of Children from
Em ironmental Health and SafeU Risks" April 21. 1998
                                           Page 1-19

-------
                           Chapter 2. REGULATORY ALTERNATIVES
        This chapter explains the various regulator}, alternatn es considered in this analysis Section 2 1
provides background on the elements that differentiate the options that were considered, and Section 2 2
defines the options that were analyzed and compared.
2.1     Elements Considered in Developing Regulatory Alternatives

        EPA's NOx SIP call sets summer NOx emissions budgets for eastern States that the Agency has
found  significantly contribute to the nonattamment b\ other States of the pre-existing ozone standard (1-
hour) and \\ill contribute in the future to nonattainment b\ other States with the revised ozone standard (8-
hour)  EPA relied heavily on its estimation of the NOx reductions that the electric power industry and other
stationary sources could provide cost-effectively in setting the State budgets   Other factors, such as the
feasibility of implementing controls in a reasonable time frame, also influenced the Agency's final decisions
To estimate the cost-effectiveness of controls for various sources, the Agency considered several ways that
controls could be implemented in the SIP call region  However. States can place controls on their sources of
NOx emissions different!)  than the approach that EPA used in the budget setting process, if they can show
that  control strateg)  will provide the same level of NOx reduction in the SIP call region

        This section describes the elements that make up the \ anous regulatory alternatives considered for
this  anal) sis  The regulator) alternatives described in Section 2 2 represent various combinations of these
elements  Some elements of the rule remain the same for all the options considered Other elements are
considered in varying combinations,  including stringency of controls, geographic  scope, affected sources and
design of the trading system  For all options analyzed, the timing of regulator) requirements was also
considered, as this issue is critical in terms of feasibility of compliance and attainment of both the pre-
existing ant the re\ ised ozone standard
2.1.1    Type of Control

        EPA had to decide on the types of regulator) approaches that the Agenc\ wanted States to consider
in their efforts to lo\\er NOx emissions from various source categories  EPA used those approaches in
estimating the cost-effectneness of ozone season NOx controls at \anous levels for different Upes of
sources  OTAG recommended that the Agency consider controls that allo\\ for emissions trading, rather
than traditional command-and-control regulation  OTAG's anahsis of trading programs had shown that
there could be considerable savings from this hpe of approach for the electric po\\er industry (OTAG.
1997)

        EPA also demonstrated the potential savings from a NOx emissions trading program that could
result in its regulator)  analysis for the proposed NOx SIP call (EPA,  1997a)  That analysis showed that in
2005 a command-and-control program for the electric power industry- would cost about 30 percent more than
a trading program m the NOx SIP call region For that reason, the Agency has focused heavily on developing
regulator) approaches that States can use collectively that are based on allowance-based NOx emissions
trading   It was also clear from OTAG analysis and  EPA's own work that further savings and flexibility
could be gained from allowing banking as part of a trading program  EPA's regulator)- analysis over the last
year has also considered banking options for inclusion in the Model Trading Rule for States  (EPA. 1997b)

                                              Page 2-1

-------
2.1.2    Geographic Scope

        After considering OTAG's recommendations and other relevant information. EPA identified 22
States plus the District of Columbia (i.e.. 23 jurisdictions) as significantly contributing to nonattainment with.
or interfering with maintenance of, air quality standards in a downwind State  The SIP call region is shown in
Figure 2-1 and consists of Alabama. Connecticut. Delaware. District of Columbia. Georgia. Illinois. Indiana.
Kentucky. Massachusetts. Man land. Michigan. Missouri. North Carolina. New Jersey. New York. Ohio.
PennsN Ivama. Rhode Island, South Carolina. Tennessee. Virginia. West Virginia, and Wisconsin

        The final rule reflects State NOx budgets that are developed using the same region-wide stringency
targets and region-wide analyses of cost-effectiveness for all 23 jurisdictions  EPA also considered dividing
the SIP call region into two or three subregions in an effort to make a distinction among the States that ma\
contribute the most to the ozone transport problem and those where the wind patterns may be less likely to
affect air quaht>  in the other States The SIP call region was divided into two regions—Northeast and
Southeast, or into three regions-Northeast. Midwest, and Southeast  Different le\ els of stringency are then
applied in the different regions, as described  below

        The two region area consists of Connecticut. Delaware. District of Columbia. Massachusetts.
Maryland. New Jersey. New York. Ohio. Penns\Kama. Rhode Island. Virginia, and West Virginia m the
Northeast, and Alabama. Georgia. Illinois. Indiana. Kentucky, Michigan. Missouri. North Carolina. South
Carolina. Tennessee, and Wisconsin in the Southeast

        The three region area consists of Connecticut. Delaware. District of Columbia. Man land.
Massachusetts. New Jerse>. New York. Pennsylvania, and Rhode Island in the Northeast. Illinois, Indiana.
Kentucky Michigan. Missouri. Ohio. Virginia. West Virginia, and Wisconsin in the Midwest, and Alabama.
Georgia. North Carolina. South Carolina, and Tennessee in the Southeast
                                              Page 2-2

-------

-------
2.1.3    Potentially-Affected Sources

        EPA has developed State budgets based on the effects of additional controls (beyond those alread\
required by the CAAA-related or reflected in existing SIPs) only for major stationary sources of NOx
emissions  These sources include (1) electricity generating utility boilers. (2) industrial, commercial and
institutional boilers. (3) combustion turbines. (4) reciprocating internal combustion engines. (4) cement
manufacturing operations, and (5) other industrial processes that emit NOx  Only existing or planned
CAAA-related controls are considered in calculating budgets for other sectors (area and mobile sources) that
contribute to NOx emissions  States ultimately have discretion to determine which sources to regulate to
achieve the budget level

        Analysis of costs and economic impacts in this R1A are based on a range of assumptions about which
major stationary sources will actually be targeted for additional controls by the States in implementing the
NOx SIP call  The primary assumption in this analysis is that  States will allocate NOx emissions reduction
requirements to the largest electricity generating utility boilers, industrial, commercial and institutional
boilers, combustion turbines: cement manufacturing units, and internal combustion engines Refer to
Chapter 3 for further discussion of the assumptions regarding options for regulatory  co\erage

        Large electricity generating units are defined as those generating more than 25 megawatts (MW)
Large industrial boilers, combustion turbines, reciprocating internal combustion engines, and other industrial
NOx sources are those capable of firing greater than 250 mmBtu/hour. or that emit greater than or equal to
one ton of NOx per summer da>.
2.1.4   Stringency of Control Level

        In order to develop a cost-effective NOx reduction strategy as a basis for establishing State budgets.
EPA considered various emission reduction levels for the affected sources for the summer ozone season
defined as May 1 through September 30.  For the electricity generating units (EGUs). EPA considered
emissions budgets based on emission limits of 0.12 Ib/mmBtu. 0 15 Ib/mmBtu. 0.20 Ib/mmBtu. and 0 25
Ib/mmBtu '' For the large industrial boilers and combustion turbines. EPA considered a uniform percent
emission reduction from uncontrolled projected 2007 emission levels ranging from 40 percent to 70 percent
For the remaining large nonutihty sources. EPA  considered source category -specific control levels
corresponding to average ozone season cost-effectiveness cut-offs ranging from $1.500/ton to S5.000/ton

        Taking into consideration the emission reductions and associated costs projected under each of the
above scenarios. EPA identified cost-effective NOx reduction strategies Based on the reduced emissions
achieved by this strategy. EPA then established State-specific budgets for ozone season NOx emissions
Alternative budgets are calculated for the different stringency levels considered for EGUs.  The details of
NOx budget development can be found in the budget technical support document (EPA. 1998b)
        1 Limits for each electricity generating unit are expressed as a specific NOx limit of pounds of NOx per mmBtu
 of summer heat input projected for 2007, the year which was the focus of OTAG's analysis (the >ear for which air
 qualm- modeling was done)

                                               Page 2-4

-------
 2.1.5   Effective Dates

        States subject to the NOx SIP call must submit revised SIPs by September 1999  The affected
 sources in the States must implement NOx controls by May 2003. and EPA will assess how each State's SIP
 has actualh performed in 2007.
 2.1.6   Emissions Budget Trading System Design

        To allow for use of the most cost-effective emission reduction alternatives, an emissions budget
 trading program is an optional component of the NOx SIP call. Each of the States subject to the NOx SIP
 call are encouraged to participate in this model NOx Budget Trading Program and thereby provide a
 mechanism for sources to achieve cost-effective NOx reductions.  The trading unit is a NOx Allowance, equal
 to one ton of emitted NOx  Details of the trading program are described in the Federal Register notice
 accompanying the final rule.

        Under the NOx Budget Trading Program, each of the participating States would determine how its
 seasonal State trading program budget is allocated among its sources  Each source would be gi\en a certain
 quantit) of NOx allowances   If a source's actual NOx emissions exceed its allocated NOx allowances, the
 source ma\ purchase additional allowances  Comersely. if a source's actual NOx emissions are below  its
 allocated NOx allowances, then it ma\ sell the additional NOx allowances  Such a program creates a
 competitue market for  NOx allowances that encourages use of the most efficient means for reducing NOx
 emissions

        For purposes of this analysis, trading may occur among any of the sources within the entire SIP call
 region or w ithin each of the subregions  If subregions are de\ eloped for the SIP call region, only mtra-
 regional (within the region) trading would be allowed

        Banking would allow  sources that do not use all of their NOx allowances for a given year to save
 them for later use  If banking is allowed, however, mechanisms such as flow controls can be put in place to
 limit the le\e! of exceedance of the emissions cap Flow controls restrict the use of the banked NOx
 allowances by restricting their use at certain times or within  certain areas. For example, a restriction may be
 placed on the banked allowances that allows only a set amount to be used during a defined time period

        For this RIA. EPA anah zed a variety of trading options, and trading with banking only for the  15
 trading option, where banking begins after the start of the program in 2003. Banking of "early" reductions
 was not modeled for the 0 15 option  because earlier IPM analysis suggested that owners of electricity
 generating units would want to use it to a very limited degree to lower the costs of future compliance (EPA.
 1997). The following considerations were part of the 1997 analysis

 •      Beginning in 2003 (and each year thereafter), the fossil fuel-fired electricity generating units over 25
       MW in the SIP call region are assumed to hold NOx allowances during the summer ozone  season
       equal to 489 thousand tons

     .  Electricity generating  units could trade allowances without restrictions or bank them for later use or
       sale to another generation unit Trading could occur within the entire SIP call region

•      Analysis with and without flow controls

                                             Page 2-5

-------
EPA's analysis in 1997 \\as conducted using the 1996 version of the Integrated Planning Model (IPM)  This
model is described in EPA. 1996 EPA's analysis shows that on strict economic grounds, (i e . under
minimization of the total direct operating costs over the simulation period) limited banking was forecasted b\
the IPM based on the scenarios described above  However, EPA believes that some banking, which the IPM
could not estimate, should occur when some power plants overcontrol their NOx emissions in order to bank
allowances for use in years in which units experience utilization greater than forecasted  More discussion of
this issue can be  found in Chapter 6
2.2     Definition of Regulatory Alternatives

        Based on the elements described above. EPA defined specific options for this analysis  This section
pro\ ides descriptions of the regulator, alternatives considered by EPA for each key NOx stationary  source
segment These regulatory alternatives are summarized in Tables 2-1  (for electricity generating units). Table
2-2 (for industrial boilers and combustion turbines), and Table 2-3 (for other large stationary sources )  The
NOx Budget Trading Program described in the final rule covers electricity generating units, industrial boilers.
and combustion turbines  The NOx Budget Trading Program described in the final rule does not mitialh
include other large stationary sources. ho\\ever. provisions are made for inclusion of these sources should
States, or the sources themselves, express a desire to participate

        Costs and economic impacts for each of these source segments are evaluated separately using
different anahsis techniques  The electnciU  generating units are evaluated using the latest version of IPM
(IPM98). \\hich optimizes nationwide delivery of electricity subject to the SIP call emissions constraint  The
industrial boilers and combustion turbines are evaluated using a least-cost analysis designed to reach the
specified emissions budget level at the lowest possible overall cost  Finally, all other stationary sources are
evaluated at the individual source category le\el  The goal of the cost analyses is to determine what levels of
reductions are highK  cost-effective for each segment

        At the same time that EPA considered each of the alternatives in  each of the tables in setting up State
budgets for ozone season NOx emissions, the Agenc> rexiewed the public comments that it recehed in the
Spring and Summer of 1998  The led EPA to reestimate what the NOx emissions budget for the electric
power industry would be in the 0 15 Trading case, and what the budget would be for the 60%/$5.000
combination of non-electricity generating source alternatives  These alternatives are the basis for EPA's final
approach to setting the budget  The NOx budget for the electric power industry was lowered from 564
thousand ozone season tons to 544 thousand ozone season tons of NOx  The NOx budget for the other large
stationary sources was increased from 558 thousand to 559 thousand ozone season tons of NOx  An
addendum to the RIA presents the final cost results and contains a more thorough explanation of how the
emissions cap will work
                                              Page 2-6

-------
                                                       Table 2-1
                          Regulatory Alternatives for Electricity Generating Units (EGU)
Name of
Alternative
0 25 Trading
0 20 Trading
0 1 5 Trading
0 12 Trading
Regionahn-1
Regionalm-2
State Emissions Budgets
Based on Ozone Season
NOx Limit of:
0 25 Ib/mmBtu
0 20 Ib/mmBtu
0 1 5 Ib/mmBtu
0 12 Ib/mmBtu
Northeast8 0 1 5 Ib/mmBtu
Southeast-Mid\\estb 0 20
Ib/mmBtu
Northeast' 0 1 2 Ib/mmBtu
Mid\\estd 0 15 Ib/mmBtu
Southeast6 0 20 Ib/mmBtu
NOx Emissions Budget
(Cap) in Ozone Season
(1,000 tons)
940
751
564
453
222
454
100
297
189
Scope of
Emissions Trading
Region \vide
Region wide
Region wide
Region wide
Intra-regional only
Intra-regional onh
' Northeast Connecticut. Delaware. District of Columbia. Massachusetts. Maryland. New Jersey, New York. Ohio. Pennsylvania. Rhode Island.
Virginia and \\ "esl Virginia
b Southeast-Midwest  Alabama. Georgia. Illinois. Indiana Kentucky. Michigan. Missouri. North Carolina. South Carolina Tennessee and Wisconsin
' Northeast Connecticut. Delaware District of Columbia Maryland. Massachusetts New Jerse\ New York. Pennsylvania, and Rhode Island
"Midwest  Illinois Indiana. Kentucky. Michigan. Missouri  Ohio.  Virginia. West Virginia, and \\ isconsin
1 Southeast -Vlabama  Georeia North Carolina. South Carolina, and Tennessee
                                                       Table 2-2
               Regulator)  Alternati'ves for Non-EGU Sources in the NOx Budget Trading Program
                                (Large Industrial Boilers and Combustion Turbines)
Name of
Alternate e
40% Control
50% Control
60% Control
70% Control
State Emissions Budgets
Based on Ozone Season NOx Reduction of:
40% from
uncontrolled 2007 baseline
50% from
uncontrolled 2007 baseline
60% from
uncontrolled 2007 baseline
70% from
uncontrolled 2007 baseline
Scope of
Emissions Trading
Region v\ ide
Region wide
Region \\ ide
Region wide
                                                        Page 2-7

-------
                                           Table 2-3
         Regulatory Alternatives for Non-EGU Sources NOT in the NOx Budget Trading Program"
Name of
Alternative
$1.500/ton
$2.000/ton
$3,000/ton
$4.000/ton
$5,000/ton
State Emissions Budgets Based on Source
Category-Specific Evaluation of the Highest
Reduction Achievable for Less Than:
$l,500/ton
$2,000/ton
$3,000/ton
$4,000/ton
$5.000/ton
Scope of
Emissions Trading
Not applicable
Not applicable
Not applicable
Not applicable
Not applicable
'Eg. industrial manufacturing processes
2.3
References
OTAG. 1997a  Draft of Costs of NOx Control Strategies on Electric Power Generation Using the
Integrated Planning Model  For incorporation into the OTAG Final Report, June 1997.

OTAG. 1997b  Trading and Incentives Work Group Report - Draft of OTAG Final Report  Chapter 7.
June 1997

U S  Em ironmental Protection Agency. 1996  Analyzing Electric Power Generation under the CAAA
Office of Air and Radiation. Washington. D C . JuK  1996

U S. Em ironmental Protection Agency. 1997a Proposed Ozone Transport Rulemaking Regulatory
Analysis  Office of Air and Radiation. Washington.  D C . September 1997

U S  Environmental Protection Agencv 1997b Model NOx Cap and Trade Rule Workshop Working
Papers  Office of Atmospheric Programs. Washington. D C . December 1997

US  Em ironmental Protection Agenc>. 1998a Analyzing Electric Power Generation under the CA4.4
Office of Air and Radiation. Washington. D C . March 1998

U S  Environmental Protection Agenc>, 1998b Development of Modeling Inventory and Budgets for
Regional NOx SIP Call  Office of Air Quality Planning and Standards. Research Triangle Park. September
1998
                                            Page 2-8

-------
                       Chapter 3.  PROFILE OF REGULATED ENTITIES


        This chapter describes the sources potentialK affected by the NOx SIP call  Profiles of the sizes,
types, locations, and NOx emissions characteristics of potentially affected electricity generating units, large
industrial boilers and combustion turbines, and other stationary sources are presented  The OTAG  1990 data
base \vas the starting point for development of the inventory of sources considered in this report, and many
updates to that database have been made by EPA (EPA. 1998)  For the purpose of setting State emissions
budgets under the NOx SIP call. EPA did not choose to apply new controls to all of the source t>pes profiled
in this chapter  Under the NOx SIP call States are free to choose \\hich sources they will control in order to
achieve the NOx budget specified in the NOx  SIP call

        Figure 3-1  illustrates how EPA has partitioned the universe of stationary sources for the analyses
presented in this  RJA  EPA considered a number of factors in determining which sources to control for the
purposes of establishing state emissions budgets  First, as indicated in the proposed NOx SIP call.  EPA has
not assumed additional controls for  sources defined as small  In addition. EPA has determined that  it would
be inefficient to establish reductions for several non-EGU point source categories that  emit a small  amount  of
NOx relatn e to total point source NOx emissions These sources, found in 24 different source categories.
comprise about 11 percent of total baseline large source non-EGU emissions, and about 6 percent of total
baseline non-EGU emissions Further, for a number of sources. EPA \\as not able to identify an applicable
control measure  This group of sources is di\ erse and not subject to categorizing as part of the categories set
out b\ EPA. and  total emissions are lo\\ for this group The Agency has  determined that the effort needed to
collect adequate information on those sources  (about 6.000 small and 258 large) \\ould be time consuming.
uncertain,  and potentialK affect less than five  percent of total non-EGU baseline point source emissions, and
therefore has not assumed any additional control for these sources for the purposes of setting State budgets
Also. EPA has determined that municipal waste combustors should not be required to reduce emissions
be>ond those already required b>  the maximum achievable control technology (MACT) rules for NOx
required under section 111 and  129  of the CAA  FmalK. the Agenc\ is not assigning emissions reductions in
this rulemakme for industrial boilers that are not fossil-fuel fired in order to be consistent with the treatment
of fossil-fuel fired electncitv generating units in the NOx Budget Trading Program

        For the remaining  source categories. EPA is basing emissions control  decisions on the relative
a\ erage cost-effectn eness  of achieving NOx reductions during the ozone season  The methodology used to
anaK/e the cost and a\ erage cost-effectn eness of alternatn e control le\els is discussed in Chapters 4 and 5
The anah sis and results supporting  these decisions are found in Chapters 6 and 7
                                              Page 3-1

-------
                 Figure 3-1
Partitioning of NOx SIP Call Stationary Sources


Small


-.('ill's
MW





Steam
Illl


•.k'Urir
il)
MW





Von-I'lilily :-i\IW
a Noii-utihl\
detuM.iloi'. (Nl'ds)
11 ( "HCIlCl.ltcMS
                  Page 1-2

-------
3.1      Electricity Generating Units

        In  1990. approximately 2 8 trillion kilowatt hours (kWh) of electricity were generated in the United
States and \\ere used in roughly equal proportions by industry, commercial establishments, and households
B\ 2005. EPA projects this total to increase to about 3 6 trillion kWh '  Most of this electricity, almost 70
percent, is generated at fossil-fuel-fired power plants, with coal accounting for most of the fossil fuel used in
these plants

        More than 95 percent of the nation's generating capacity is owned by the electric utilities  Although
utilities are general!} granted monopolies for their service territories, the rates that the utilities may charge are
regulated by the authorities that grant the monopoly (known as a "franchise") Rates for investor-owned
utilities have theoretical!} been  set high enough to cover all reasonably incurred  costs, including capital
investments,  and to provide an allowance for a reasonable rate of return on imested capital  This
arrangement has insulated, to a  large extent, most large producers of electricity from some of the effects of
the market  as \\ell as from regulatory costs A changing regulatory and economic environment, however, is
eroding this insulation  In the future, utilities  are expected to be much less able to pass on their emission
control costs

        A significant portion of the nation's electricity generating industry is in  the region affected b\ the
N0\ SIP call  EPA estimates that 2.014 units will be operating in this region in the year 2000  In addition to
electric utility power units that produce only electricity, this number includes units owned by independent
power producers (IPPs)  This number also includes units that co-generate electricity and steam (co-
generators), whether owned b> utilities or IPPs  Table 3-1 presents the number of fossil-fueled units by
capacity range and t>pe (i e . coal, oil/gas steam, combined cycle, combustion turbine)   Approximate!) 64
percent of the affected fossil-fueled  electric utilit}  units have capacities that are less than or equal to 100
megawatt (MW)  Less than one percent are greater than 1.000 M\V  Table 3-2  presents the distribution of
these units  as a percentage b} type within each capacity range  About 41 percent of these units are coal
powered, providing approximate!} 72 percent of fossil-fueled capacit}  Approximate!} 45 percent are
combustion turbines, which provide about 10  percent of the capacit} of all these units  The table indicates
that coal powered units make up the majority of the capacit} of all units
        1 EPA's generation requirement projections are based on an extension of the electric demand forecast of the
North American Electric Reliabilit} Council, adjusted for the impact of the Climate Change Action Plan

                                               Page 3-3

-------
                                             Table 3-1
                          Distribution of Capacities of Potentially Affected
                            Electricity Generating Utility Units by Type
                                          in the Year 2000

Boiler
Capacit}
0-25 MW
>25-100MW
> 100-200
MW
>200-400
MW
>400-600
MW
>600-800
MW'"
>800-
1000MW
> 1000MW
Total
Coal/Steam
#of
Units
51
213
238

151

93

50

16

12
824
Capacity
(MW)"
789
14.441
34.834

42.785

48.605

34.550

13.831

14.802
204.635
Combined
Cycle
#of
Units
15
62
20

8

0

2

0

2
109
Capacity
(MW7)"
256
3.685
2.910

2.065

0

1.327

0

2.093
12.337
Combustion
Turbine
#of
Units
569
304
32

6

1

0

0

0
912
Capacity
(MW)'
4,844
17,439
4.200

1.611

500

0

0

0
28.594
Oil/Gas Steam
#of
Units
33
38
42

22

15

11

7

0
168
Capacity
(MW)
312
2,661
5.815

7.488

7.479

7.274

5.994

0
37.023
Total
a of
Units
668
617
332

187

109

63

23

14
2.013
CapaciU
(MW)"
6:201
38.227
47.760

53.949

56.584

43.150

19.824

16.895
282.589
Source  1PM data 1CF Resources
                                               Page 3-4

-------
                                            Table 3-2
                         Distribution of Capacities of Potentially Affected
                                Electricity Generating Utility Units
                                         in the Year 2000

Unit
Capacity
(MW)'
0-25
>25-100
>100-200
>200-400
>400-600
>600-800
>800-
1000
> 1000
Total
Coal/Steam
%of
Units in
Capacity
Range
8
35
72
81
85
79
70
86
41
%of
Capacity
in
Capacity
RanRe
13
38
73
79
86
80
70
88
72
Combined Cycle
%of
Units in
Capacity
Range
2
10
6
4
0
3
0
14
5
%of
CapaciU
in
Capacity
Range
4
10
6
4
0
~\
^
0
12
4
Combustion
Turbine
%of
Units in
Capacity
Range
85
49
10
3
1
0
0
0
45
%of
Capacity
in
Capacity
Range
78
46
9
•^
1
0
0
0
10
Oil/Gas Steam
%of
Units in
Capacity
Range
5
6
13
12
14
17
30
0
8
%of
Capacity
in
Capacit}
Range
5
7
12
14
13
17
30
0
13
Total
%of
Units in
Capacity
Range
100
100
100
100
100
100
100
100
100
%of
Capacity
in
Capacit}
Range
100
100
100
100
100
100
100
100
100
Source IPM data. ICF Resources
        Table 3-3 sho\\s the geographic distribution and the total capacih of the affected electncits
generating units by type (coal, combined cycle, combustion turbine, and oil/gas steam) among the States in
the SIP call region  Table 3-4 presents the same information in percentage terms  All States except Rhode
Island and the District of Columbia ha\e coal-powered units and, for many States. coal-po\\ered units make
up the majonh of the capacity of all units  The District of Columbia and West Virginia do not have am
combustion turbine units  Further, mam States do not have combined-cycle units

        Table 3-5 shows the distribution of electricity generating units by type and by NOx emission rate.
The rates in this table are based upon initial base case controls, which include existing Title IV controls.
Reasonabh Available Control Technology requirements, New Source Performance Standards (NSPS) for
nev> and recenth -built power  plants, and implementation of Phase I of the Ozone Transport Commission
Memorandum of Understanding (MOU)  As shown in Table 3-5. over half of all the units analyzed fall in the
range of 0 to 0.2 Ibs NOx/mmBtu  These units provide about one quarter of capacity in the SIP call region
More than half of the capacity emits more than 0 4 Ibs of NOx/mmBtu  The table also shows that a
significant majority of combined cycle, combustion turbine, and oil/gas steam units, in both number and
capacitv fall in the lower ranges of NOx emission rates
                                             Page 3-5

-------
                                                Table 3-3
                                   Distribution of Capacities of Affected
                          Electricity Generating Utility Units (>25 MW) by State
                                            in the Year 2000

State
Alabama
Connecticut
)ela\\ are
)C
Georgia
hnois
ndiana

-------
                                                 Table 3-4
       Distribution of Capacities of Affected ElectricitJ Generating Utility Units by State by Percentage
                                              in the Year 2000

State
Alabama
Connecticut
Delaware
Dist of Columbia
Georgia
Illinois
Indiana
Kentucky
Man land
Massachusetts
Michigan
Missouri
Ne\\ Jerse\
Ne\\ York
North Carolina
Ohio
Penns\hania
Rhode Island
South Carolina
Tennessee
Virginia
West Virginia
Wisconsin
Coal/Steam
% of all
Units in
each
State
672
7 I
333
00
460
484
70 1
775
250
137
43 3
23 5
98
136
51 6
608
548
00
427
48 1
330
1000
35 1
%
Capacity
in each
State
(MW)
895
163
460
00
845
809
953
90 3
537
23 3
729
81 2
197
164
862
929
"85
00
81 2
846
533
1000
759
Combined Cycle
% of all
Units in
each
State
34
36
95
00
00
00
00
00
54
178
1 3
06
250
152
147
20
52
00
00
00
82
00
00
%
Capacity
in each
State
(MW)
20
1 4
133
00
00
00
00
00
49
150
68
07
263
143
33
08
1 9
00
00
00
177
00
00
Combustion
Turbine
% of all
Units in
each
State
293
393
38 1
00
473
385
278
21 1
554
52 1
478
749
467
495
326
31 8
289
1 00 0
573
51 9
526
00
649
%
Capacity
in each
State
(MW)
85
87
144
00
135
2 2
46
90
149
10 1
3 7
r "
32 0
120
102
54
5 1
1 00 0
188
154
120
00
24 1
Oil/Gas Steam
% of all
Units in
each
State
00
500
19 1
1000
68
13 1
20
1 4
143
164
77
1 1
185
21 7
1 1
54
11 1
00
00
00
62
00
00
%
Capacin
in each
State
(MW)
00
735
264
1000
1 9
168
0 1
07
265
51 6
166
05
220
573
03
09
145
00
00
00
170
00
00
Total
% of all
Units
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
%of
Capacin
' 100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
Source IPM data. ICF Resources
                                                  Page 3-7

-------
                                              Table 3-5
                       Distribution of Fossil-Fueled Units Analyzed for Rulemaking
                                by Initial Base Case NOx Emission Rate
                                           in the Year 2000
Initial Base Case
Emission
(Ibs NO\/mmBtu)
0000-0 100
0 101-0200
0201-0300
030] -0400
0401-0500
0501-0600
0601-0700
0701-0800
0 80 1 -0 900
0901-1 000
>1 000
Total
Coal/Steam
#of
Units
4
56
21
184
374
16
40
21
82
13
13
824
CapaciU
(MW)'
353
4.797
5.503
44.203
102,190
999
22.488
2.907
19.598
887
779
204.635
Combined Cycle
#of
Units
16
90
0
0
3
0
0
0
0
0
0
109
Capacity
(MW)"
2,040
10,162
0
0
135
0
0
0
0
0
0
12.33-7
Combustion
Turbine
#of
Units
118
783
3
2
3
1
1
0
0
0
1
912
Capacity
(MW)"
5.472
22.736
136
75
90
2
62
0
0
0
21
28.594
Oil/Gas Steam
#of
Units
12
85
59
8
0
4
0
0
0
0
0
168
Capacity
(MW)'
1,810
25,522
9.302
350
0
39
0
0
0
0
0
37.023
Total
#of
Units
150
1,014
83
194
380
21
41
21
82
13
14
2.013
Capacity
(MW)'
9,675
63.217
14,941
44,628
102,415
1.040
22.550
2,907
19,598
887
800
282.589
Source  IPM data. ICF Resources
3.2     Industrial Boilers and Gas Turbines

        This section provides information on industrial boilers and combustion turbines, including the types
of fuels they use and their emissions, as well as a description of the industries that own them  Sources are
classified b\ unit design capacity  Only large sources (as defined in Chapter 2) are assumed to be potentially
affected b\ controls that ma\ be implemented by  a State to meet its NOx budget le\ el

The following types of sources are described in this section

•       Industrial, Commercial, and Institutional Boilers - Industrial/commercial/ institutional (ICI)
        boilers include steam and hot water generators with heat input capacities from 0 4 to 1.500
        mmBtu/hr  These boilers are used in a range of applications, from commercial space heating to
        process steam generation, in all major industrial sectors  Although coal. oil.  and natural gas are the
        primary fuels, many ICI boilers also burn a variety of industrial, municipal, and agricultural waste
        fuels

•       Stationary Combustion Turbines - Turbines  are used in electric power generators, in gas pipeline
        pump and compressor drives, and in various process industries   This section includes  turbines other
        than those used for electricity generation The primary fuels used are natural  gas and distillate oil.
        although residual fuel oil is used in a few applications
                                               Page 3-8

-------
        Industrial boilers are owned and operated b\ a wide variety of industries, from traditional
manufacturing to sen ice industries like medical care and education  Thirty of the two-digit "major groups"
in the Standard Industrial Classification (SIC) system include establishments with industrial boilers The
industries with the most industrial boilers are: chemicals, paper, petroleum, and primary metals  Table 3-6
shows the industry distribution of large industrial boilers and turbines in the 22 States plus D.C

        Table 3-7 presents a breakdown of industrial, commercial, and institutional boilers by primary fuel
t>pe  (All ICI boilers are referred to as industrial boilers in this RIA)  Natural gas  fired boilers account for
the largest percentage of industrial boilers, with 36 percent of all boilers  Coal and oil industrial boilers make
up the rest of the industrial boilers, with 30 percent and 12 percent of total boilers respectiveh
Approximate!)  22 percent of the industrial boilers are listed as "other "  Other industrial boilers include wood
and wood wastes, pulping liquor, and waste gases  During the mid 1980s there \\as a trend towards use of
dual-fuel boilers, where the preferred configuration was a natural gas system with a fuel oil back up

        Finalh. Table  3-8 shows the distribution of large fossil-fuel fired industrial boilers and combustion
turbines b\ state  For two industrial boilers, the data was not  a\ ailable to match a state with the source
                                               Page 3-9

-------
                                                  Table 3-6
            Number of Fossil-Fuel Fired Industrial Boilers and Combustion Turbines by Industry"
                                                  1995 Data
SIC
10
20
21
22
24
25
26
27
28
29
30
32
33
34
35
36
37
38
39
49
51
72
"9
80
89
Industry
Metal mining
Food and kindred products mfgr
1 obacco products mfgr
Textile mill products
Lumber & wood products. exc furniture
Furniture & fixtures
Paper and allied products
Printing & publishing
Chemicals & allied products
Petroleum refining and related industries
Rubber & plastics products
Stone. cla\. glass & concrete products
Pnman metal industries
Fabricated metal products, exc machinen & trans equip
Industrial & commercial machinen & computer equip
Electronic & other elec equip . exc computer equip
Transportation equipment
Measuring mst . photo, med & opt goods, clocks
Miscellaneous manufacturing industries
Electric, gas. and sanitan sen ices
Wholesale Trade - nondurable goods
Personal semces
Amusement and recreation sen-ices
Health sen ices
Miscellaneous semces
Federal Government
Other Government
Colleges/Universities
Total
Number of
Boilers
1
50
8
11
1
5
153
3
187
45
10
5
138
5
11
5
21
4
7
34
1
1
3
6
1
18
7
14
755
Number of
Turbines
0
0
0
0
0
0
4
0
2
2
0
0
0
0
0
0
1
0
0
43
0
0
0
0
0
0
0
0
52
Source Pechan-Avanti Group
a Excludes 90 large non-fossil fiael fired boilers and turbines
                                                   Page 3-10

-------
                                  Table 3-7
Number of Large Fossil-Fuel Fired Boilers and Combustion Turbines by Fuel*
                                  1995 Data
Source Type - Fuel Type
ICI Boilers - Coal/Wall
ICI Boilers - Coal/FBC
ICI Boilers - Coal/Stoker
ICI Boilers - Coal/Cyclone
ICI Boilers - Residual Oil
ICI Boilers - Distillate Oil
ICI Boilers - Natural Gas
ICI Boilers - Process Gas
ICI Boilers - Coke
ICI Boilers - LPG
Total ICI Boilers
Combustion Turbines - Oil
Combustion Turbines - Natural Gas
Combustion Turbines - Jet Fuel
Total Combustion Turbines
Total ICI Boilers and
Combustion Turbines
Number of Sources
166
6
69
8
75
26
307
86
10
2
755
22
28
i
52
807
     Source Pechan-A\anti Group
    * Excludes 90 large non-fossil fuel fired boilers and turbines
                                   Page 3-11

-------
                                                 Table 3-8
          Number of Large Fossil-Fuel Fired Industrial Boilers and Combustion Turbines b> State*
                                                 1995 Data
State
Alabama
Connecticut
Delaware
District of Columbia
Georgia
Illinois
Indiana
Kentucky
Maryland
Massachusetts
Michigan
Mis>oun
Ne\\ Jerse\
Neu York
North Carolina
Ohio
Penns\hania
Rhode Island
South Carolina

J ennes.^ee
Virginia
West Virginia
Wisconsin
Unknown
Total
Number of
Industrial Boilers
51
12
6
5
15
64
60
22
7
11
38
11
64
48
23
88
39
0
36

49
37
30
36
2
755
Number of
Combustion Turbines
1
2
0
0
0
7
2
1
0
0
5
0
3
2^
0
0
4
0
1


0
0
1
0
52
Source Pechan-Avanti Group
* Excludes 90 large non-fossil fuel fired boilers and turbines
                                                 Page 3-12

-------
3.3     Other Stationary Sources

Other ma]or stationary sources of NOx emissions include the following

•       Stationary Internal Combustion Engines - These units generate electric power, pump gas or other
        fluids, or compress air for pneumatic machinery The primary nonutilitv application of internal
        combustion (1C) engines is in the natural gas industry to power compressors used for pipeline
        transportation, field gathering (collecting gas from wells), underground storage, and in gas
        processing plants  Reciprocating engines are separated into three design classes  2-cycle (stroke)
        lean burn. 4-stroke lean burn, and 4-stroke rich bum  Each of these have design differences that
        affect both baseline emissions as well as the potential for emissions control

•       Cement Manufacturing Operations - There are  four types of kilns that produce cement, long wet.
        long dn. kilns \\ith a preheater, and kilns with a precalcmer  Long wet kilns use a production
        process where the raw materials  are suspended in \\ater to form a slurry  Long dn kilns and kilns
        with a preheater or a precalcmer use a dn production process, \\herem raw cement materials are
        dried to a powder  Each of these types of kilns have design differences that affect both baseline
        emissions as \\ell as the potential for emissions control

•       Other Industrial Processes - Some industrial processes emit NOx  Examples include furnaces at
        iron and steel mills, glass furnaces, process heaters at chemical plants and petroleum  refineries.
        nitric acid plants, and adipic acid plants

Tables 3-9 and 3-10 sho\\ the industn and state distribution, respectively, of the various types of large
stationan sources, other than electnciU generating units and industrial boilers and combustion turbines
Note these figures do not include non-EGU sources for \\hich EPA was unable to identify  applicable control
technologies (see section 3 4)
                                              Page 3-13

-------
                                                        Table 3-9
                               Number of Large Other Stationary Sources by Industry
                                                        1995 Data
SIC
14
20
22
24
25
26
28
29
30
32
33
35
-* ~t
J ;
45
49
51
80
87
89
Industry
Non-metal, non-fuel mming/quaming
Food and kindred products mfgr
Textile mill products
Lumber & \\ood products, exc furniture
Furniture & fixtures
Paper and allied products
Chemicals & allied products
Petroleum refining and related industries
Rubber & plastics products
Stone, clay, glass & concrete products
Pnmarv metal industries
Industrial & commercial machmen & computer equip
Transportation equipment
Transportation b\ air
Electric, gas. and samtan sen ices a
\\Tiolesale Trade - nondurable good;>
Health sen ices
Engineering, accounting, research, mgmt. & related
sncs
Miscellaneous sen ices
Federal Go\ eminent
Other Gin eminent
Colleges/Una ersities
Total
1C
Engines
7
0
0
0
1
0
0
1
1
0
0
0
0
0
286
1
1
0
0
5
2
0
305
Cement
Manu-
facturing
0
0
0
0
0
0
0
0
0
58
0
0
0
0
0
0
0
0
0
0
0
0
58
Other
Indus-
trial
Sources'"
1
3
6
12
0
202
39
21
2
45
72
1
7
1
83
1
5
3
1
4
17
0
526
Total
8
*>
^
6
12
1
202
39
22
3
103
72
1
7
1
369
T
6
3
i
9
19
0
889
Source Pechan-Avanti Group
1 \on-EGL"s classified within SIC 4911 are included in this estimate  These are co-generation units that supply less than 50"! oof their generated
power to the electric power grid
b Includes 90- large non-fossil fuel fired industrial boilers and combustion turbines
                                                        Page 3-14

-------
                                                   Table 3-10
                              Number of Large Other Stationary Sources by State
                                                   1995 Data
State
Alabama
Connecticut
Delaware
District of Columbia
Georgia
Illinois
Indiana
Kentucky
Maryland
Massachusetts
Michigan
Missouri
Ne\\ Jerse\
Xe\\ York
North Carolina
Ohio
Perm s\ Kama
Rhode Island
South Carolina
Tennessee
Virginia
West Virginia
Wisconsin
Total
1C
Engines
36
0
0
0
2
30
42
2
3
0
47
6
0
0
6
12
9
0
11
48
9
35
7
305
Cement
Manufacturing
5
0
0
0
3
2
8
1
9
0
5
7
0
3
0
5
2
0
0
-\
3
2
3
0
58
Other Industrial
Sources"
88
7
6
0
60
31
41
14
16
15
48
16
4
22
28
19
13
0
48
34
31
21
23
526
Total
129
7
6
0
65
63
91
17
28
15
100
29
4
25
34
36
24
0
59
85
42
59
30
889
Source Pechan-Avanti Group
1 Includes 90 large non-fossil fueled fired industnal boilers and combustion turbines
                                                     Page 3-15

-------
3.4     Other NOx Sources

        A vaneh of sources of NOx emissions are not addressed in this R1A because it is assumed that the>
will not be subject to ne\\ controls in the States' rexised SIPs  Among the other stationary sources not
addressed in this RJA are all small stationary sources (e g  small industrial boilers), sources for which control
information could not be identified, and sources already receiving highly cost-effective NOx controls under
other rulemakmgs (e g . municipal waste combustors affected by the 1994 MACT standard)  Other sources
of NOx emissions which are assumed not to be subject to new controls include

•       Area Sources - Area sources are small point sources that include open burning and small
        commercial, industrial and residential fuel combustion sources

•       Mobile Sources - This category is divided into highway vehicles and nonroad sources Highway
        vehicle sources include cars, trucks, buses, and motorcycles with gas and diesel highway engines
        Nonroad sources include commercial marine engines, small engines such as lawn and garden
        equipment, and larger engines such as those used in construction equipment and locomotives

        EPA did not identify additional controls beyond those in the 2007 baseline case for the area, mobile
and nonroad source categories at less than $2.000 per ton. nor did EPA receive comments during the public
comment period suggesting that such feasible, cost-effective controls should be implemented  Therefore.
EPA did not calculate additional emissions budget decreases nor propose rules for these source categories

        There is a large, diverse set of non-EGU sources that meet the large source size definition for which
EPA was not able to identify applicable control technologies  Table 3-11 indicates the industries in which
these sources operate, as well as the total number of non-EGU sources (i e . both those sources for which
control technology is known and not known) in each industn
                                             Page 3-16

-------
                                                    Table 3-11
                       Number of Total and "No Control" Non-EGU Sources by Industry
                                                    1995 Data
SIC
10
13
14
20
22
24
25
26
28
29
30
32
i -»
^ J*
34
35
37
38
45
49
51
80
8"
89
Industry
Metal mining
Oil and gas extraction
Non-metal, non-fuel mining/quarrying
Food and kindred products mfgr
Textile Mill Products
Lumber & \\ood products, exc furniture
Furniture & fixtures
Paper and allied products
Chemicals & allied products
Petroleum refining and related industries
Rubber & plastics products
Stone, clav. glass & concrete products
Pnman. metal industries
Fabricated metal products, exc machinery & trans equip
Industrial & commercial machinery & computer equip
Transportation equipment
Measuring inst . photo, med & opt goods, clocks
Transportation b\ air
Electric, gas. and sanitary sen ices3
Wholesale Trade - nondurable goods
Health services
Engineering, accounting, research, mgmt. & related sncs
Miscellaneous services
Federal Go\ emment
Other Government
Colleges/Universities
Total
No Control
Sources
4
1
1
4
0
2
0
65
34
6
0
7
49
2
1
65
1
1
7
2
1
1
1
2
0
1
258
Total Sources
4
1
9
7
6
14
1
267
73
28
3
110
111
2
2
72
1
2
3^6
3
7
4
2
11
21
1
1.138
Source Pechan-Avanti Group
" Xon-EGL"s classified uithm SIC 4911 are included in this estimate  These are co-generation units that suppK less than 50% of their generated
power to the electric power grid
                                                     Page 3-17

-------
3.5     Overview of Baseline Emissions

        Table 3-12 provides an overview of the contribution of various NOx sources to total baseline NOx
emissions in the 22 States and D C.  This table shows that large sources that are potentially subject to new
requirements under the SIP call (including electricity generating units, industrial boilers, combustion turbines.
internal combustion engines, and cement manufacturing operations) account for approximately 43 percent of
the total projected baseline emissions in these States
                                               Table 3-12
               Overview of 2007 Baseline Ozone Season NOx Emissions in the SIP Call Region
Source Category
ElectnciU Generating Units
Industrial Boilers'1
Combustion Turbines
Internal Combustion Engines
Cement Manufacturing
Other non-EGU Source Categories
"No Control" Sources'
Area/Mobile/Nonroad Sources
Total
Baseline Ozone Season NOx Emissions
Large
Units
1.497,061
203.883
5.809
92.424
42.701
101.964
34.832
na
na
Small
Units
4,714
139.569
4.926
54.885
13.868
41.268
32.845
1,981.845
na
Total
1.501,775
343,452
10.735
147.309
56.569
143.232
67,677
1.981,845
4,252,59-4
Percent of
Total Baseline
Ozone Season
NOx
Emissions"
35%
8°/o
0 3%
3%
1%
3%
2%
4"%
100.0%
 Source Id". Pechan-A\anti Group and SN'PR
 na ~ not estimated or not applicable
 ' Due to rounding percentages do not add to exacth 100%
 b Includes baseline emissions for the 90 large non-fossil fuel fired industrial boilers that are not affected in this rule
 cNon-EGU units for which EPA was not able to identifv control measures
 3.6
References
 Abt Associates. Inc., 1998  Non-Electnaty Generating Unit Economic Impact Analysis for the NOx SIP
 Call  Prepared for the U S  Environmental Protection Agency, Office of Air Quality Planning and Standards,
 September 1998

 U.S. Environmental Protection Agency, 1998  Development of Modeling Inventory and Budgets for
 Regional NOx SIP Call.  Office of Air Quality Planning and Standards. Research Triangle Park, NC,
 September 1998
                                                Page 3-18

-------
             Chapter 4.  METHODOLOGY FOR ESTIMATING EMISSIONS, COSTS,
             AND ECONOMIC IMPACTS FOR THE ELECTRIC POWER INDUSTRY
        This chapter presents the methodology for estimating the costs, emission reductions, and impacts of
the NOx SIP call for the electric power industry The chapter is divided into eight sections, beginning with an
analytical o\erue\v in Section 4 1  Section 4 2 discusses the use  of the Integrated Planning Model (IPM) for
the analysis, including  assumptions about the baseline and about technologies for power generation and
emission control  Allowance allocation and trading issues are presented in Section 4 3. and the estimation of
administrative costs is discussed in Section 4 4  These discussions are followed by Sections 4.5 and 4 6.
\\hich outline the analysis of potential direct and indirect economic impacts  Limitations of the analysis are
presented in Section 4.7. and references are presented in Section 4 8 of the chapter.
4.1     Analytical Overview

        The basic approach to estimating the potential effects of the NOx SIP call on electricity producers is
to project their actions in the absence of the rule, project their actions if they were subject to the rule, and then
compare the two sets of actions Subtracting the total costs of generating electricity in the absence of the rule
from the total costs under the rule, for example, yields the total costs of the rule itself if States and sources
follow the implementation approach modeled in this report  Similarly, subtracting estimated emissions.
generation, and capacity yield the effects of the rule in these three areas

        The scope of these analyses is wide both geographically and in terms of time  While the focus of the
rule is on the 23 jurisdictions affected by the NOx SIP call, the analysis projects the actions of utilities  (and
non-utility generators) in all 48 contiguous States in order to capture effects that can spill out of one region
into neighboring areas  Rather than examining only a snapshot in time, the analysis co\ers a period starting
in 2001 and running out to 2025  Examining the industry o\er mam years makes it possible to take mam
important dynamic effects into account  For example, the effects of efficiency gams over time and the choice
between capital-intense e control measures and measures that increase operating costs can be investigated b>
projecting utility response o\ er a long analytical period  In addition, the effects of allow ing the banking of
emission reductions can be analyzed only in a dynamic framework

        The actions of electricity generators over time are projected using the IPM. which is a detailed
computer model of the electric power industry  IPM is designed to find the most efficient (that is. the least-
cost) way  to satisfy- the demand for electricity under a series of limitations or constraints   The constraints
under which IPM "produces"' electricity can include a limit on tons of NOx emissions during the summer, and
it is by setting this constraint that the effects of the  NOx SIP call can be modeled  Running IPM without a
limit on tons of NOx emissions produces a picture of the baseline situation in which the NOx SIP call is not
in effect  Rerunning IPM after adding a constraint that limits emissions in the SIP call region to a specified
number of summer tons (e.g , 564,000, under the 0 15 option) shows what the industry would  do to comply
with the NOx SIP call while keeping its costs as low as possible Additional runs  with different sets of
constraints are conducted to assess other options, while additional runs with different assumptions make it
possible to test the sensitivity of the results  More detail on how IPM operates is provided in Section 4 2
below and in Analyzing Electric Power Generation Under the  CAM, Office of Air and  Radiation, U S
Em ironmental Protection Agency.  March 1998  This information and the model runs conducted for the
analysis can also be found at an EPA website with the address:  http //wwvv epa gov/capi
                                              Page 4-1

-------
        The IPM runs for the baseline and the various options constitute the heart of the analysis  Before the
results can be presented, however, additional analyses must be conducted to interpret these runs  For
example, in some cases it is necessary to aggregate the detailed results into totals b\ State and region, or to
divide the cost changes by emission changes to estimate cost effectiveness   In addition, tracing the potential
economic impacts of changes in costs and electricity prices be\ ond the electricity generating industry is
outside of the scope of IPM. and must be done using standard techniques of economic impact assessment
(discussed in Sections 4 5 and 4 6)
4.2     IPM Assumptions and Use

        EPA uses IPM to evaluate the emissions and potential cost impacts expected to result from the
requirements of the NOx SIP call on the electric power industry, based on EPA's illustrative implementation
scenario IPM has been used for over ten years by electric utilities, trade associations, and government
agencies both in the U S and abroad to address a \\ide range of electric power market issues  The
applications ha\ e included capacity planning, environmental polic\ and compliance planning, wholesale price
forecasting, and asset \ aluation  EPA has used IPM extensively for environmental policy and regulaton
analysis In particular. EPA has used IPM to analyze NOx emission policy and regulations as part of the
Clean Air Po\\er Initiative (CAPI) in 1996. as an anaKsis tool for the Regulaton Impact Analysis of the
National Ambient Air Quality Standards (NAAQS) for ozone and particulates in 1997. and as a tool to
analyze alternative trading and banking programs during the OTAG process in  1996 and 1997  IPM was
also used for the regulaton analysis of the NPR

        IPM has undergone extensive review and validation over this ten-year period   In April 1996. EPA
requested participants in the CAPI process to comment on the Agencv 's new approach to forecasting electric
pov\er generation and selected air emissions EPA received many helpful comments and made a series of
changes in  its methodology and assumptions based on commenters' recommendations  Most recently. IPM
and EPA's modeling assumptions were reviewed as part of the OTAG process  Again, changes were made to
the methodolog> and assumptions based on commenters" recommendations

        The ^ ersion of IPM used by EPA (IPM98) represents the U S electric power market in 21 regions, as
depicted in Figure 4-1  These regions correspond in most cases to the regions and sub-regions used by the
North American Electric Reliability Council (NERC)  IPM models the electricity demand, generation.
transmission, and distribution within each region as well as the transmission grid that  connects the regions

        The model includes existing utility power plants as well as independent power producers and
cogeneration facilities that sell firm capacity into the wholesale market  Data on the existing boiler and
generator population, which consists of close to 8.000 records, are maintained in EPA's National Electric
Energy  Data System (NEEDS).  In order to make the modeling more time and cost efficient,  the individual
boiler and generator data are aggregated into "model" plants  EPA's application of the model has focused
heavily on  understanding the future operations of coal-fired units, which will have the greatest air emissions
among the  fossil-fired units  The operation of other types of non-fossil fuel-fired generation capacity.
including nuclear and renewables, are also simulated but at a higher degree of aggregation
                                              Page 4-2

-------
                                            Figure 4-1
               Integrated Planning Model Regions in the Configuration Used by EPA
                                                                                     FRCC
       Working with these existing model plants and representations of alternative new power plant options.
IPM determines the least-cost means for supph ing electricity demand while limiting air emissions to remain
below specified policy limits  Multiple air emissions policies can be modeled simultaneousk  For example.
IPM is used in this study to simulate compliance with existing CAAA Title IV SO: emission requirements as
well as actions that EPA has considered for controlling the ozone season NOx emissions in the States covered
b> the NOx SIP call  While determining the least-cost solution. IPM also determines the optimal compliance
strategy for each model plant A wide range of compliance options are evaluated, including the following

•      Fuel Switching - For example, switching from high sulfur coal to low sulfur coal.

•      Repowermg - For example, repovvering an existing coal plant to a gas combined-cycle plant

•      Pollution Control Retrofit - For example, installing selective catahtic reduction (SCR), selective
       non-catalytic reduction (SNCR). or gas rebum (to reduce NOx emissions), or flue gas
       desulfurization (to control SO: emissions).

•      Economic Retirement - For example, retiring an oil or gas steam plant
                                             Page 4-3

-------
•       Dispatch Adjustments - For example, running high-NOx cyclone units less often, and lo\\ NO\
        combined-ex cle plants more often

        IPM provides estimates of air emission changes, incremental electric po\\er generation costs, changes
in fuel use. and other potential impacts for each air pollution policy analyzed

        The model is not limited in scope to facilities owned by electric utilities, but also includes
independent po\\er producers (IPP) that provide electricity to the power grid on a firm-contract basis, as well
as IPP facilities larger than 25 megawatts that provide power on a non-firm basis

        IPM simultaneously models over an extended time period, and reports results for selected years In
addition to reporting for 2003. which is the year that the regulatory approach would begin, these analyses also
pro\ide results for 2001. 2005.  2007, and 2010

        In using IPM to analyze NOx emission policy over the past two years. EPA has developed a set of
data and assumptions that reflect the best available information on the electnciK market and operating
factors  These data and assumptions can be grouped into the following four categories

•       Macro Energy and  Economic  Assumptions - These assumptions are related primarily to
        electricity demand projections, fuel prices, po\\er plant availability heat rates, lifetimes, and capacity.
        factors  Also included in this category are discount rate and year dollar assumptions

•       Electric Technology Cost and  Performance - These assumptions are related to electric technolog>
        cost and performance for existing and new plants, as well  as for existing plant refurbishment and
        repo\\ermg

•       Pollution Control Performance and Costs ~ These assumptions pnmanK co\er the performance
        and unit costs of pollution control technologies for NOx and S0:

•       Air Emissions Rates under the Base Case - These assumptions cover current EPA and State
        requirements that will affect emission levels from \ anous  facilities The focus has been on SO; and
        NOx controls

Each  of these sets of data  and assumptions are briefly discussed below  More detail can be found in EPA's
March 1998 report entitled Analyzing Electric Power Generation under the CAAA
4.2.1   Macro Energy and Economic Assumptions

        In developing the analysis for the NOx SIP call, EPA makes assumptions about major macro energy
and economic factors, as shown in Table 4-1  See Appendix No. 2 of EPA's March 1998 report Analyzing
Electric Power Generation under the CAAA for details on most of the macro energy and economic factors

        In this study, IPM's cost outputs are converted from real 1997 dollars to real  1990 dollars to be
consistent with the cost analyses prepared for the proposed NOx SIP call and the Agency's recently published
Regulatory Impact Analysis of the National Ambient Air Quality Standards for ozone and particulates  The
factor used for this purpose is 0 83. which corresponds to the gross domestic product implicit price deflator
index published by the Bureau of Economic Analysis

                                              Page 4-4

-------
                                               Table 4-1
                          Ke> Baseline Assumptions for Electricity Generation
                      Factor
             Assumption
Discount Rate (percent per year)
Com ersion Factor from 1997 to 1990 Dollars
                 083
Electricity Demand drouth Rate (percent per \ear)a
            1997-2000= 1 6
            2001-2010= 1 8
            >2010     =13
Reductions due to Climate Change Action Plan
(Billion k\\h) for sensitivity anahsis in section 634
              2001 =100
              2003 = 164
              2005 = 228
              2007 = 293
              2010 = 389
             >2019 = 608
Po\\er Plant Lifetimes
Fossil Steam  =

Nuclear      =
Turbines      =
65 \ears if 2 50 MW
45 \ears if < 50 MW
40 year license length
3CM ears
U S Nuclear CapaciU (giga\\atts)
              2001 =93
              2003 = 90
              2005 = 87
              2007 = 86
              2010 = 81
              2020 = 50
Nuclear Capacin Factors (percent)
              2001 =80
              2003 = 80
              2005 = 80
              2007 = 82
              2010 = 81
              2020 = 83
                                               Page 4-5

-------
                                         Table -4 -1 (continued)
                           Key Baseline Assumptions for Electricity Generation
Factor
World Oil Prices (1997$ per BBL)
Wellhead Natural Gas Price (1 997$ per mmBtu)b
Coal Steam Power Plant Availability (percent)
Existing Pouer Plant Heat Rates
Coal Mining Productn itv Increases
(percent change per year)
A\erage Delnered Coal Pncesb
(percent change per \ ear 2001-2010)
Assumption
2001 = 1920
2003= 1990
2005 = 20 50
2007 = 20 80
2010 = 21 20
2020 = 22 40
2001 = 1 90
2003 = 1 95
2005 = 2 00
2007 = 2 00
2010 = 200
1995 = 82
2000 = 83 5
2005/10/20 = 85
No change o\er time
1995-1999 = 3 1
2000-2004 = 28
2005-2009 = 24
2010-2014 = 2 1
2015-2025 = 2 1
-20
 1   Does not include an\ adjustment for potential improsements related to the Climate Change Action Plan
 b   Based on recent ICF anahses using updated coal mining products it\ and suppK for coal, and technology and suppK assumptions for gas  Note
    that the natural gas prices are not an assumption in the model, hut are a forecast of the model


4.2.2   Electric Energy Cost and Performance Assumptions

        In order to simulate the electric power market under baseline conditions and for each of the
regulators  options, assumptions are made on the cost and performance of new po\\er plants as v\ell as for
repowering existing power plants  These characterizations of new po\\er plant cost and performance are  used
in IPM to determine the least cost means for meeting projected future electricity requirements subject to the
baseline emission restrictions and the NOx emission limits specified for each regulator}, option

        Power plant cost and performance assumptions are developed for the following new conventional and
unconventional power plant types

•       New Conventional Power Plants

                Conventional Pulverized Coal.
                Advanced Coal (Integrated Gasification Combined Cycle - IGCC).
                                                Page 4-6

-------
              Combined Cycle.
              Combustion Turbine, and
              Nuclear

•      New Rene\\ able/Nontraditional Options

              Biomass IGCC.
              Solar Photovoltaics.
              Solar Thermal.
              Geothermal: and
              Wind

       Cost and performance projections are developed for 2001. 2003: 2005. 2007. and 2010 in order to
capture changes in technology over time  In general the year 2001 estimates reflect generation technology
that is close to or  identical to existing technology, and the later year estimates reflect advancements in costs
and performance  The Agency relies heavily on \\ork that the Energy Information Administration did in
support of the most recent Annual Energy Outlooks (AE097  and AEO98)  EIA had its approach
peer-reviewed during its development

       In addition to the AEO. ke\ data sources used to develop these assumptions are as follows

       EPRL TAG Technical Assessment Guide. Electricity Supply - 1993. EPRI
               TR-102276-V1R7. June 1993.

       SERI. The Potential of Renewable Energy  An Interlaboratory White Paper.
               SERI/TP-260-3674. March 1990. and

       TVA. Integrated Resource Plan Environmental Impact Statement. Volume Two.
               Technical Documents. JuK  1995

In addition to these assumptions on new power plants. EPA also develops assumptions  on the cost and
performance of repowermg existing power plants  The following three types of repo\\ enng options are
considered

 •      Repo\\ermg Coal Steam to Integrated Gasification Combined-Cycle,

 •      Repowermg Coal Steam to Gas Combined-Cycle, and

 •      Repowermg Oil/Gas Steam to Gas Combined-cycle

 The key sources of data for this section are the repowermg studies conducted by Bechtel Corporation, the
 TVA Integrated Resource Plan EIS. and the EIA life extension report

       For more details on the assumptions made about the cost and performance of new power plants and
 repowermg of existing power plants, see Appendix No. 3 of EPA's March 1998 report Analyzing Electric
Power Generation under the CAAA
                                            Page 4-7

-------
4.2.3    Pollution Control Performance and Cost Assumptions

        EPA develops pollution control cost and performance estimates for the following options

•       Coal-Fired Steam Electric Generating Units

               Combustion Controls,
               Selective CataMic Reduction,
               Selective Non-Catalytic Reduction, and
               Natural Gas Rebum

•       Oil and Gas-Fired Steam Generating Units

               Selective Catalytic Reduction; and
               Selectne Non-Catalytic Reduction

        EPA also develops cost and performance estimates for combining SCR or SNCR with coal plant
scrubbers With these options, the IPM can determine if in some instances, it is optimal to place a scrubber
and SCR or SNCR to reduce S0: emissions and NOx emissions from a given plant simultaneousK  In
determining the least cost means for comph ing with a NOx regulator  policy, the model can choose from
among these pollution control options and change the dispatch of model plants   For example, the model in
some cases can reduce the utilization of high NOx  emitting units and increase the utilization of low NOx
emitting units

        In addition to including the pollution control cost and performance estimates described above. IPM
also takes into account the cost and performance of combustion controls installed beyond those resulting from
implementation of Title IV and Title I (Reasonable A\ailable Control Technologies - RACT) requirements
Note that the Title IV NOx program permits an owner/operator to comph  with the requirements by averaging
the NOx emissions from some units \\ithin the owner/operator system \\ith emissions from other units also
within the same s> stem  This emissions averaging permits an owner/operator to install controls on units that
are cost-effecti\ e to control and average emissions from these units with emissions from units that are less
cost-effectn e to control  EPA accounts for the cost of combustion controls beyond those needed for Title IV
compliance m the following manner  (1) EPA identifies the units that either are (Phase I units) or are hkeh to
(Phase II units) average their emissions with other  controlled units, and (2) EPA reasons that these
uncontrolled units, for the purposes of this proposed rulemakmg. \\ill install the least expensive controls,  that
is. combustion controls, where requirements beyond Title IV are imposed on them  These units can further
reduce their emissions by installing SCR. SNCR. or gas reburn,  as described above  Additionally, using
continuous emissions monitoring (CEM) data, EPA found that some sources with  a common owner or
operator, that could average their emissions under Title IV. consistently emitted well below (20 percent or
more) their Title IV mandated levels   For the purposes of analyses in this  report, such sources are assumed to
emit at their actual CEM-measured levels, not their applicable Title IV Standard

        These performance and pollution control cost assumptions for NOx are based on the following
sources:

        U S Environmental Protection Agency. Regulatory Impact Analysis of NOx Regulations,
               October 1996
                                              Page 4-8

-------
        Bcchtel Po\\er Corporation. Cos; Estimates for NOx Control Technologies Final Report.
               Februar> 1996

        Bechtel Po\\er Corporation, Draft Technical Study on the Use of Gas Reburn to Control
               NOx at Coal-fired Electric Generating Units, June 1996

        Acurex Environmental Corporation. Phase II NOx Controls for NESCAUM and MARAMA
               Region. 1995

        For more details on the assumptions made about pollution control cost and performance see
Appendix No 5 of EPA's March 1998 report. Analyzing Electric Power Generation under the CAAA.
4.2.4   Air Emissions Rates under the Base Case

        Assumptions about the other environmental rules that will be in effect with or without the NOx SIP
call constitute a vital aspect of the baseline because even existing environmental initiatives will lead to NOx
reductions in the future If the reductions that are projected to take place under these initiatives are not
accounted for. the effects of the  NOx SIP call in capping NOx emissions will be o\ erestimated

        Three sets of regulations affecting NOx emissions in the baseline are taken into account in this
analysis  First. EPA factors in regulations under Title 1 of the Clean Air Act. including RACT requirements
for existing sources. EPA's New Source Performance Standards, and controls based on Best Available
Control Technology (BACT) and Lowest Achie\able Emissions Rates (LAER) that \\ould be in effect  for
new sources The analysis also  accounts for the NOx reductions from utility units under Phases I and II of
Title IV's Acid Rain Program, which set rate limitations for most coal-fired generators greater than 25  MW
of capacm

        Finally, the control program agreed upon b\ the Ozone Transport Commission for the Ozone
Transport Region is assumed to go forward in the baseline  The OTC's Memorandum of Understanding
em isions three progressively more stringent control requirements for sources in the OTR  Phase I. Phase II.
and Phase III Though EPA anticipates that all three of these phases will e\ entually be implemented in the
baseline, cases including Phase I alone (i e . RACT controls in place in the OTR) are examined in some of the
baseline anah ses  This baseline, which is referred to as the Initial Base Case, is the primary' basis for
comparison in this RJA Comparisons of the options to a Final Base Case, which assumes that Phase II and
Phase III of the OTC's MOU will also go into effect, are also made in the RJA  Because Phases II and  III are
estimated to cut NOx from electric  generators, any comparison of an option to the Initial Base Case will
appear to be more effective (and more costly) than a comparison of that option to the Final Base Case  In
considering the effects of the OTC's MOU, this anah sis covers only the NOx controls in the SIP call region
In considering the effects of the  OTC's MOU, this anah sis covers only the NOx controls in the SIP call
region  Thus NOx controls resulting from the OTC MOU in Maine. Vermont, and New Hampshire are not
included
                                             Page 4-9

-------
4.3     Allowance Allocations and Trading

        For the purposes of this analysis, the NOx SIP call is assumed to be implemented through an
emissions trading program  The trading program works by allocating the limited rights to emit NOx in
limited quantities during the summer, and allowing sources to choose the extent to which the\ will reduce
emissions or purchase these limited rights or "allowances " For man}- aspects of the analysis, the initial
distribution of the allowances is not important: the only relevant fact regarding the allowances is their total
volume (segmented by region for the regional options), which determines the number of tons of NOx
reductions required The reasons for this separation of the allowance allocations from the rest of the anah sis.
and the circumstances under which the allocation does become important, are described in Section 43.1
Assumptions about the trading of allowances are presented in Section 4.3.2
4.3.1    Purpose of Allowances and Assumptions about Allocations

        IPM works b> finding the least-cost method for producing electric power for the mdustn as a whole.
assuming the entire industry in the area of the NOx SIP call is subject to an overall cap on ozone season NOx
emissions  The model places pollution controls or makes dispatch changes to electricity generating units that
lead to the achievement of emission reductions at the lo\\est cost  As a result, some firms" power plants are
projected to be tighth controlled, at significant cost, while other firms" plants ha\e no controls beyond those
assumed in the baseline

        Realistically this pattern would not be seen unless some system existed to give incentives to the
firms with the most cost-effective control possibilities to bear the greatest part of the control burden  The
NOx  SIP call envisions that these incentives will take the form of compensation for allowances, which must
be purchased b>  the firms that elect to under-control their plants'  emissions  Firms are assumed to either bu\
or sell allowances depending on their own costs of control in comparison to the market price of allowances
As the price reacts to changes in demands and supplies of allowances, the market will help ensure that the
costs  of incremental reductions of NOx are the same for  all participants

        Projecting how the NOx emissions cap will be divided  initial!)  among firms through awards of
allowances is not important for estimating the total costs of the NOx SIP call or the control methods that will
be used if it can be assumed that the allowance market will be efficient   If the market is efficient, the onk
effect of allocating more allowances to a given firm will be that a firm will be able to sell more allowances
after controlling emissions to an efficient degree. Experience with the SO: allowance market under Title IV
demonstrated that these markets can function efficiently, with significant trading volumes  and minimal
transaction costs (U S EPA. Forthcoming-Fall 1998)

        The initial distribution of the allowances is, however, very important in assessing potential impacts
and trading patterns  If. for example, allowances are distributed in proportion to baseline  NOx emissions.
owners of coal plants would be able to sell many more allowances than if allowances are distributed in
proportion to baseline generation

        For this analysis, it is assumed that States will divide allowances only among affected sources, m
proportion to their 1995 or  1996 fuel input, allowing both for growth in capacity and growth in electricity
output to 2007.  These assumptions are somewhat simplified versions of the allowance distribution system
recommended by EPA for State consideration, which provides for shifts m allowance distributions over time
in response to changing capacity use, unit closures, and new builds.  These simplifications should have little

                                             Page 4-10

-------
effect on the anaKsis. which reflects the approach EPA recommends if a State allocates on the basis of input
Ho\ve\er. if the State allocates on an output basis or differs a great deal on the input approach from EPA's
recommendation, the outcome could be much different
4.3.2   Trading Assumptions

        As noted above. IPM calculates costs and emissions reduction choices as though there are no
constraints on the transfer of allowances from source to source Implicitly, then, the analysis assumes a
completely efficient, fnctionless market for allowances More realistically, there will be some transaction
costs incurred when allowances are transferred  Based on the experience with the Title IV S02 allowance
trading program, the cost of allowance transactions has been assumed to be 1 5 percent of the value of the
transaction (U S EPA, Forthcoming-Fall 1998)  The basis for this estimate is  explained in Section 442
The effects of transaction costs of the magnitude estimated for the NOx SIP call on the total costs of the rule
and the distribution of control efforts is likely to be negligible  because many transferred allowances are
between units within company systems, not between power companies, and most allowances are not
transferred at all Also, transaction costs are expected to be a  very small percentage of the value of those
allowances that are transferred
4.4     Administrative Costs

        Electric utilities. State and local air quality regulatory agencies, and EPA will incur administrate e
costs in addition to the costs of complying with the NOx SIP call  The primary basis for determining the
amount of these administrate e costs is supporting data from EPA's Information Collection Request (ICR)
(EPA. 1998d) for the proposed Regional NOx Federal Implementation Plan (FIP)  Even though this ICR is
conducted for the FIP. EPA assumes that the unit costs \\ould be identical for the NOx SIP call, though there
could be changes in who bears the costs  All of the admimstratn e costs are annualized and presented in 1990
dollars for the \ ear 2007
4.4.1    Administrative Costs to Affected Electric Generating Units

        The o\\ners or operators of affected electricity generating units will incur administrative costs
associated with the following actn ities

•       Monitoring emissions.

•       Certifying compliance.

•       Modifying permits, and

•       Trading allowances.

        Electricity generating units will be required to have in place monitoring equipment to measure their
NOx emissions  This is already required under Title IV for units covered by the S02 allowance program  In
addition. the\ will be required to submit a monitoring plan to the State or local agenc\ for review and
appro\ al  On a regular basis, they must also submit a report certifying compliance The sources will also be

                                             Page 4-11

-------
required to obtain air permits and. before beginning construction of the control technologies, the facilities
may need a construction permit The operating permit will also need revision to incorporate the revised
emission limitations  These administrative costs are based on supporting data from EPA's ICR
(EPA, 1998d)  In general, the administrative costs are equal to the unit costs multiplied by the number of
affected electricity generating units

        Utilities will also incur transaction costs in trading allowances between companies. The
methodology for estimating transaction costs is discussed next
4.4.2   Transaction Costs of Trading Allowances

        IPM calculates costs and emission reduction choices as though there are no constraints on the
transfer of allowances from source to source. Implicit!}, then, the analysis assumes a completely efficient.
fnctionless market for allowances  More realistically, there will be some transaction costs incurred when
allowances are transferred  The transaction costs include the costs to gather information on the market.
search for allowances, make bids and offers, negotiate terms and conditions of allowance transfer agreements.
and ensure that the allowances transfer  Many companies will hire an emissions trading broker to perform
these sen ices, but the company will also incur other costs related to the decision-making process as well as
costs of legal counsel

        For this analysis of the NOx market. EPA assumes that total transaction costs are approximate!} 1 5
percent of the \ alue of the allowances traded In real it}. the percentage may van.' depending upon the
quantity of allowances traded, the familiarity of the traders with the market,  and the overall maturity of the
market While total S02 transaction costs have declined over time, recent evidence suggests total S0:
transaction costs range from one to two percent (U S EPA. Forthcoming-Fall 1998)'  These total transaction
cost estimates are for both buyer and seller combined and include brokerage fees and internal decision-
making costs  The decline in total transaction costs is believed to be attributable to impro\ ed market
maturity and trading famihant} This analysis assumes an average total transaction cost of 1 5 percent for
the NOx market, thus accounting for market \ anation o\ er time  Here we are onl\ counting inter-utilit}
trading costs There will also be ultra-utility trading, but no costs are assigned to these trades

        The total value of allowances traded between companies under each option is equal to the product of
the number of allow ances and the price of an allowance estimated for each option   For this anal} sis.  EPA
projects the \olume of allowances traded between units and between utilities and the value of each allowance
EPA estimates the total volume of allowance transactions for the 0.15 trading option by comparing IPM
projections for emissions for each unit to an estimate of the allowances that the unit would receive under the
NOx SIP call  Allowance allocations are assumed to be made based on the baseline fuel inputs of the units
The difference between each unit's emissions and the allocated allowances is assumed to be equal to the
amount of each allowance transaction, either the quantity acquired or transferred to another.  Then, the total
quantity is found by summing all allowances acquired across all units. To determine the number of
transactions occurring between companies, the emissions for each unit are summed on a utilit} -by-utility
basis and compared to the sum of the allowances allocated to each unit on utihty-by-utility basis The
difference of the two is equal to the total inter-utility transactions
         These numbers are based on preliminary estimates and may change in the final report

                                              Page 4-12

-------
        The number of transactions will be greater or fewer for each regulatory option based on the
differences bemeen the marginal cost functions at the various control levels as well as the shapes of these
cost functions  Transaction volumes are calculated for the 0.12. 0.15, and 0 20 trading options  The price of
allowances is assumed to be approximately equal to the marginal cost of NOx reductions, as estimated by
IPM

        Transaction costs were derived based on trading between firms operating utilities only  Transaction
costs for firms operating non-EGU sources were not estimated  Since the transaction costs are only a very
small component of total annual compliance costs of the NOx SIP Call  for affected utilities, and will likely be
a ver> small component of total annual compliance costs for affected non-EGUsources. the inclusion of
non-EGU  transaction costs should not significantly change the compliance costs estimate of the SIP Call
4.4.3    Administrative Costs to States and Local Governments

        The following administrative costs will be incurred b\ State and local air quality regulator} agencies

•       Certifying monitoring plans.

•       Monitoring compliance by conducting audits, and

•       Re\ ie\\mg and appro\ ing permit modification applications

        State or local agencies will need to certify monitoring plans prepared by affected sources  States and
local agencies will also monitor the compliance of electric generating units by reviewing the emissions data
and conducting occasional audits  It is assumed that States and local  agencies will audit 10 percent of the
electricity generating units  These agencies will also be responsible for reviewing and approving permit
applications for both construction permits (except in Prevention of Significant Deterioration (PSD) areas that
have not been delegated the authority to implement the program) and operating permits  For this analysis, the
administrative costs associated with permitting are allocated to only the States and local agencies, because the
costs associated with PSD permits re\ ie\\ed and appro\ed by the U S EPA are assumed to be minimal. The
administrative costs attributable to the State and local go\ ernments are calculated by multiplying the unit
costs and the number of affected units based on supporting data from EPA's ICR (EPA. 1998d)


4.4.4    Administrative Costs to the U.S. EPA

        EPA's primary administrative costs are associated with upgrading the allowance tracking system.
administrating the allowance tracking system, and collecting the NOx emissions monitoring data  This effort
is incremental to current EPA collection of NOx emissions data from acid rain units and certain OTC units
that already provide emissions data   EPA recently modified the allowance and emissions tracking systems for
the Ozone Transport Commission's NOx Budget Program; therefore, these systems will require minimal
upgrades to expand to the SIP call region  These administrative costs incurred by EPA are equal to the unit
costs multiplied by the number of affected electricity generating units and are based on supporting data from
EPA's ICR  (EPA.'l998d)
                                             Page 4-13

-------
4.5     Direct Economic Impacts

        The Agency anaKzed the potential direct impacts of the rule on the electric power producers.


4.5.1    Potential Costs to Electric Power Producers Relative to Revenues

        Costs of the NOx SIP call for this illustrative implementation approach are compared to the revenues
of electric power producers at two levels industry -wide and to small entities Industry-wide comparisons are
made b> expressing the potential costs of the rule on a per-kilowatt hour basis, using IPM outputs on the
generation of electricity from fossil fuel, and then comparing this increase in unit costs to average rates per
kilowatt hour  Data are obtained from EIA Form 861 to find revenues per kilowatt hour for utilities in the
SIP call region

        Potential costs to small entities are estimated in more detail Specific data on small utility revenues
is collected from EIA Form 861. and total revenues for small non-utility owners is obtained from Dun &
Bradstreet  These revenue estimates are compared to cost estimates for individual units owned by each small
entity  The cost estimates take into account the least-cost means of compliance, including the option of
allowance purchases, as discussed belov.

        The IPM analysis focuses on estimating industry -wide costs and emission reductions Where it is
necessary to derive rough estimates of potential costs for particular firms, the first step is to find the projected
emission control choices  made for each of the firm's units in the cost-minimizing solution The  cost
functions built into IPM can then be used to calculate the fixed and \ anable control costs for each unit,  and
these costs are summed across all of the units owned by the firm to yield total control costs

         The next step in  estimating potential costs to a given firm is to incorporate purchases or sales of
allowances  Estimates of the total summer NOx emissions from the firm's plants are compared to the firm's
allocation of allowances (based on an allocation of the SIP-call-region-\\ide cap in proportion to baseline fuel
use)  This comparison gives the net purchases or sales of allowances, •uhich must then be multiplied by an
estimated allowance price  Allowance price estimates are based on EPA estimates of the incremental costs
per ton of reducing NOx  It should be noted that these potential firm-by-firm impacts do not take account of
changes in dispatching, and can therefore o\ erstate net economic impacts on the potentially affected entities
Allowance transaction costs are considered  too small to be considered in this analysis


 4.5.2    Assessment of Potential for Passing on Cost  Increases

         An assessment of the potential effects of the NOx SIP Call on electricity prices is conducted using
 estimates of changes in marginal cost combined with judgment on the effects of power industry  restructuring
 on the competitiveness of the market. Potential changes in marginal costs of generation, weighted by demand
 segment (e.g., peak load  or base load), are assumed to be passed on to consumers, under the simplifying
 assumption of perfectly inelastic demand for electricity and a competitive market  This simplified estimate is
 used to provide an upper bound estimate of potential price increases. EPA recognizes that there will be price
 elasticity of demand effects, such that the quantity demanded will respond to price changes in both the short
 and long run. This elasticity' of demand will limit the increase in prices  It is recognized. ho\\e\ er, that these
 assumptions may not hold at all times and in all States
                                               Page 4-14

-------
4.5.3   Assessment of Potential for Closures and Additions

        The chance that power plants might close as a result of the NOx SIP call is assessed using IPM,
v> hich determines whether it is more cost-effective to control the emissions from a given plant, buy
allowances and continue to operate  it. or close it down  IPM is also used to project capacity additions, based
on the costs of building new capacity m comparison to its value
4.6     Indirect Economic Impacts

        The economic effects of the NOx SIP call can be transmitted through market interactions to entities
that are not directly affected by its provisions.  These potential indirect effects are analyzed using estimated
changes in electncih  rates, emission control technology, and fuel use.  The effect of rising electricity rates is
assessed b> multiplying the per-kilowatt-hour increase by the number of kilowatts used by typical
manufacturers or consumers, and comparing this increased cost to revenues or incomes.  Data are obtained
from the Census of Manufacturers. Bureau of the Census, and surve\s by the U S Energy Information
Administration Potential cost increases are assessed both in terms of nationwide averages and sensitive
subgroups (including energy-intensive industries and  low-income households)

        Potential impacts on emplo\ment in the industries providing fuel and pollution control equipment are
assessed by measuring changes in fuel and control equipment purchases in combination with projected labor
products ity in these industries
4.7     Limitations of the Analysis

        This analysis incorporates a fine-grained representation of the behavior of a large number of
industrial entities, it covers both a long period of time and a \\ide geographical area  As with am similar
attempt to project the future in detail, it is subject to limitations and uncertainties. Thus, several factors could
lead to cost and emissions impacts abo\ e or belo\\ the reported impacts   Those factors include the following

•       Speed of Deregulation - EPA has assumed that electric utility deregulation will continue to move
        ahead at a stead\ pace  The Agency has also assumed that deregulation will affect the electricity
        market in specific ways including lower cost of transmission, higher coal plant availability,  and lo\\er
        reserve margins Should deregulation occur more quickly or more slowly than assumed, or  affect the
        electricity system in different ways, the estimated costs and emissions impacts for these regulator*
        options may differ

•       Pollution Control Costs and Performance - EPA has used estimates of pollution control costs and
        performance that reflect the current state-of-the-art However, technological progress stimulated by
        competition could lead to improvements in the performance and cost of pollution control technolog\
        in the future For this reason, the Agency's estimates of future cost impacts for the regulatory-
        options considered could be overstated

•       Regulatory Program Implementation - EPA has assumed that the regulatory program resulting
        from the NOx SIP call will be implemented smoothly and at specific points in time
                                             Page 4-15

-------
       Data Limitations - EPA has constructed a database for this anahsis that consists of information on
       virtually even boiler and generator in the U S  The Agency has assembled the best information on
       each boiler and generator that is publicly available  Inevitably, when working with information on
       such a large number of facilities, some units may not be represented correctly  Improvements to the
       database could lead to changes in estimates of emissions and potential cost impacts for the regulatory
       options analyzed
4.8    References

Acurex Environmental Corporation, 1995.  Phase II NOx Controls for NESCAUM and MARAMA Region

Bechtel Po\\er Corporation. 1994  Electric Utility Repowenng Assessment, Final Report, July 1993-
February 1994

Bechtel Power Corporation. 1996a  Cost Estimates for NOx Control Technologies, Final Report February
1996

Bechtel Power Corporation. 1996b. Draft Technical Study on the  Use of Gas Reburn to Control NOx at
Coal-fired Electric Generating Units June. 1996

Energ> Information Administration. 1996  Annual Energy Outlook 1997  December 1996

Energy Information Administration. 1997  Annual Energy Outlook 1998  December 1997

Energy Information Agency  Life extension report

Energy Information Administration. Form 861

EPR1. 1993. TAG Technical Assessment Guide. Electricity Supply - 1993. EPRI TR-102276-V1R7  June
1993

SERI. 1990 The Potential of Renewable Energy An Inlerlaboratory White Paper, SERI/TP-260-3674
March 1990

TVA. 1995  Integrated Resource Plan Environmental Impact Statement. Volume Two. Technical
Documents  July 1995

US Environmental Protection Agency, 1996 Regulatory Impact Analysis of NOx Regulations. Office of
Atmospheric Programs. Acid Rain Division, Washington.  D.C , October 1996

U.S. Environmental Protection Agency, 1998a.  Analyzing Electric Power Generation under the CAA4
Office of Air and Radiation, Washington, D.C., March 1998.

U.S. Environmental Protection Agency. 1998c.  Transaction Costs Associated with Emissions Trading.
Prepared by ICF Incorporated for the Climate Polity and Programs Division. Forthcoming, Fall 1998.
                                            Page 4-16

-------
US  Environmental Protection Agency. 1998b  Feasibility of'Installing NOx Control Technologies by May
2003 Offic of Atmospheric Programs, Acid Ram Division. Washington. D C., July 1998

US  Environmental Protection Agency. 1998d  Supporting Analysis for the Information Collection Request
for the Federal Implementation Plan Office of Air and Radiation. Washington. D C . September 1998
                                            Page 4-17

-------
   Chapter 5. METHODOLOGY FOR ESTIMATING EMISSION REDUCTIONS, COSTS AND
            ECONOMIC IMPACTS FOR NON-ELECTRICITY GENERATING UNITS
        This chapter describes the methodologies used to estimated emission reductions, costs and economic
impacts for stationary sources other than electricity generating units (EGUs) that are potentially affected by
the NO\ SIP call  Section 5 1 provides an analytical overview for the methodologies described in this
chapter  Section 5 2 describes the available NOx control technologies for two sets of sources  industrial
boilers and turbines, and other stationary sources '  Section 5.3 presents information on control costs and
describes the cost-effectiveness methodolog}, and Section 5 4 discusses administrative costs associated with
the NOx SIP call  Section 5.5 provides an oven lew of the economic impact analysis methodology and data
sources used to conduct such an analysis, and Section 5 6 discusses the methodology for estimating small
entrn impacts  Fmalh.  Section 5 7 provides references for the chapter
5.1     Analytical Overview

        The basic approach to estimating the potential effects of the NOx SIP call to other stationary sources
is to project their actions in the absence of the rule, project their actions if they were subject to the rule, and
then compare the two sets of actions  The actions of other stationary sources in the absence of the rule is
referred to as the 2007 CAAA baseline, or 2007 base case  Total annual compliance costs and NOx
emissions changes  are estimated incremental to the base case

        The geographic scope of these analyses is the 23 jurisdictions affected b\ the NOx SIP call  The
anah ses pro\ ide results for 2007. the >ear in  which all required emissions reduction strategies are to be fulh
implemented All results are presented in 1990 dollars

        The potential emission reductions and control costs to other stationary sources affected b> the NOx
SIP call are estimated using a model that  is primarily based on data and assumptions from Alternate e
Control Technolog} (ACT) documents prepared by EPA for mam of the industries in this source categon
that are potentially affected b\ the rule The costs for SNCR and SCR control applications to industrial
boilers are dcmed  from a separate stud>  that  also serves as the basis for the cost estimates used in the
Integrated Planning Model (IPM) for utiht\ boilers  For sources identified in the NOx budget trading
program (industrial boilers and combustion turbines), this model estimates emission reductions and control
costs for 2007 using a least-cost approach applied across the entire SIP call region   For source not in the
trading program (e g .stationan, 1C engines and cement manufacturing operations) the model applies control
measures at individual emissions  units based on a cost ceiling calculated in terms of average cost-
effectiveness The least cost  approach used for the trading sources provides a proxy for State-level emissions
trading programs free of transactions costs  The approach for sources outside the trading program provides
estimates of the costs for meeting each State's emissions budget under a command-and-control scenario

        The least-cost analyses are performed for 4 different regulator}' alternatives, and the command-and-
control analyses are performed for 5 different regulators' alternatives  More detail on the control technologies
        1 Other stationan sources refers to a large variety of non-electricity generating source types  This analysis
limits itself to t\\o source types  stationan reciprocating internal combustion engines, and cement manufacturing
operations (including coal-fired cement kilns)

                                              Page 5-1

-------
used to provide emission reductions are in Section 5.2 below and in Ozone Transport RulemakingNon-
Electricity Generating Unit Cost Analysis. September. 1998, and more detail on the control cost model
operates is provided in Section 5 3 below and in the cost report previously mentioned

        Monitoring and administrative (record keeping and reporting) costs are also estimated for potentially
affected other stationary sources and are added to the control costs to provide an estimate of total compliance
costs More detail on ho\\ these costs are estimated is provided in Section 5 4 below  and in Support for
Revising ICRfor Reporting Requirements for NOx SIP call, September, 1998

        FmalK. the total compliance costs at the source level are aggregated to the establishment or plant
level and further aggregated to the entity level, and are used to estimate the potential economic impacts
associated with the entities potentially affected by the NOx SIP call  These analyses consist of estimating
compliance costs as a percentage of sales or revenues for affected entities (firms or institutions that own
affected other stationary sources).  The Agency also conducted analyses for the set of potentially affected
small entities (using SBA size definitions)  More detail on how these impacts are estimated is pro\ ided in
Section 5  5 and Section 5 6 below and in Non-Electricity Generating Unit Economic Impact Analysis for
the NOx SIP Call. September. 1998
5.2     NOx Control Technology

        This section describes a\ailable technologies for controlling emissions of NOx for industrial.
commercial and institutional (ICI) boilers2 combustion and turbines (Section 52 1) as well as other non-EGU
stationary sources (Section 522)

        In general. lo\\-NOx burners (LNB) is applied as the default control technology for industrial boilers
and turbines due to its possible application to most am industrial burner application (Pechan. 1998)  Other
issues imolved m choosing a control technology  include ease of retrofit and reduction performance  While all
controls presented in this analysis are considered generally technically-feasible for each class of sources.
source-specific cases may exist \\here a control technology is in fact not technical!} -feasible   In their
response to the NOx SIP call. States ma> wish to consider case-specific feasibility when establishing control
requirements
 5.2.1   NOx Control Technology for Industrial Boilers and Turbines

        There are three types of control technologies considered for industrial boilers  selective catahlic
 reduction (SCR), selective non-catal\tic reduction (SNCR). and low-NOx burners  As stated above, the
 default control technology chosen was LNB due to its breadth of application  In some cases, LNB
 accompanied by flue gas reburnmg (FOR) is applicable, such as when fuel-borne NOx emissions are
 expected to be of greater importance than thermal NOx emissions  When circumstances suggest that
 combustion controls do not make sense as a control technology (e g . sintering processes, coke oven batteries)
 SNCR is the appropriate choice
         : The terms 'TCI boiler'" and "industrial boiler" are used interchangeable in this RIA

                                               Page 5-2

-------
        Control technologies applicable to gas turbines include  water injection (WI). steam injection. low-
NOx burners, selectne catalytic reduction (SCR), selective non-catahtic reduction (SNCR). and
combinations of SCR \vith LNB. oxygen trim (OT). \vater injection, or steam injection  Table 5-1 lists the
control technologies a\ ailable for non-EGU industrial boilers and combustion turbines by type of fuel
                                              Table 5-1
                   A\ ailable NOx Control Technologies for Stationary Industrial Boiler
                                   and Combustion Turbine Sources
Source T\pe/Fuel Type
1CI Boilers -
1CI Boilers -
1C I Boilers -
1CI Boiler* -
1CI Boilers -
ICI Boilers -
ICI Boiler.-, -
ICI Boilers -
ICI Boilers -
ICI Boilers -
Combustion
Combustion
Combustion
Coal/Wall
Coal/FBC
Coal /Stoker
Coal/C\ clone
Residual Oil
Distillate Oil
Natural Ga.s
Process Cias
Coke
LPG
1 urbines - Oil
Turbines - Natural Gas
Turbines - Jet Fuel
Available
Control Technolog)
SNCR, LNB, SCR
SNCR - Urea
SNCR
SNCR. Coal Rebum. NCR, SCR
I.NB. SNCR. LNB -i- FOR. SCR
LNB. SNCR, LNB + FGR. SCR
LNB. SNCR. LNB + FGR. OT + Wl. SCR
LNB. LNB - FGR. OT t- WI. SCR
SNCR. LNB. SCR
LNB. SNCR. LNB + FGR. SCR
Water Injection, SCR + Water Iniection
Water Injection. Steam Injection. LNB, SCR + LNB. SCR
+ Steam Injection. SCR -t- Water Injection
Water Iniection. SCR + Water Iniection
Source Pechan-As dnli Group
5.2.2   NOx Control Technology for Other Stationary Sources

        Other stationary sources included in the analysis include reciprocating internal combustion (1C)
engines and cement kilns used in the cement manufacturing process.  In the case of 1C engines, most are "lean
burn" designs which are considered best available control technolog> (BACT) for NOx control  Thus, it is
assumed that no further add-on emissions controls are practical or necessars.  NOx control technology
available for cement kilns includes those a\ ailable to industrial boilers and turbines, namely  LNB. SCR.
SNCR  In addition, mid-kiln firing, ammonia-based SNCR, ignition retard (IR). adjustments of the air/fuel
ratio (AF RATIO), and low emission engines (L-E) can be utilized where appropriate  Table 5-2 lists the
control technologies available for 1C engines and cement  kilns
                                              Page 5-3

-------
                                              Table 5-2
                M ailable NOx Control Technologies for Other Non-EGU Stationary Sources
Source Type/Fuel Type
Internal Combustion Engines - Oil
Internal Combustion Engines - Gas
Internal Combustion Engines - Gas,
Diesel. LPG
Cement Manufacturing - Drv
Cement Manufacturing - Wet
In-Process. Bituimmous Coal. Cement
Kilns
Available
Control Technology
IR, SCR
IR, AF RATIO, AF + IR, L-E (Medium Speed),
L-E (Lo\\ Speed). SCR
IR, SCR
Mid-Kiln Firing, LNB. SNCR - Urea Based, SNCR - Ammonia Based.
SCR
Mid-Kiln Firing, LNB. SCR
SNCR - Urea based
Source  Peehan-Avanti Group
5.3     Control Costs and Cost Effectiveness Methodology

        This section describes the methods used to develop estimates of costs by control technology and b>
source catcgon  This section also describes the approaches used for each group of sources to assign control
technologies to specific sources assumed to be potential!} subject to new controls

        Two t\pes of costs will be incurred in association with the addition of NOx control technologies  a
one-time capital cost for new equipment installation, and increased annual operating and maintenance costs
In general, economies of scale exist for pollution control technologies for both capital costs and operating and
maintenance costs  Thus, the size of the unit to which controls are applied will determine, in part, the cost of
implementing the pollution control(s)

        Control cost estimates by source si/.c are developed using EPA's Alternative Control Techniques
(ACT) Documents for each major source category 3 Additional control cost equations for  SCR and SNCR
are adapted  from  information originally developed for EGU sources for use in the IPM analysis (Pechan.
1998)  All costs are converted from the original source year to 1990 dollars using the GDP price deflator
Capital costs are annuahzed using a seven percent interest rate and an equipment life appropriate for each
control technique, as specified in the relevant  ACT Documents  Table 5-3 lists the equipment life
assumptions used in this  analysis  To take account of the effects of size on costs, cost equations are applied
to estimate costs as a function of boiler design capacit}   Engineering judgement and knowledge of the
affected industries was used to assign control  cost equations to specific source types (Pechan. 1998)
        3 The ACT Documents did not provide total O&M costs for combustion turbines The total was calculated b\
subtracting the annuahzed capital from the total annual cost This may have added some uncertainty, as both capital and
total annual costs \\ere rounded to the nearest thousand dollars in the ACT documents The ACT Document for 1C
engines also contained no O&M cost data Thus, operating and maintenance costs \\ere back calculated for these source'*
as \\ ell
                                               Page 5-4

-------
                                              Table 5-3
                       Equipment Life for Various ISon-EGU Control Technologies
Source Category/Control Technology
ICI Boilers (All fuels)
LNB. LNB + FOR. OT + WI
Coal Reburn. NGR
SCR, SNCR
Combustion Turbines (All fuels)
All Controls
Stationan 1C Engines (All fuels)
All Controls
Cement Manufacturing (Wet & Drv Kilns, and
Bituminous Coal-Fired Kilns)
All Controls
Equipment Life
10
20
20
10
15
15
        Sources ma\ be controlled or uncontrolled in the 2007 baseline  Controlled NOx sources tend to be
those in ozone nonattainment areas, or in the Northeast ozone transport region, that are subject to RACT
regulations  The cost analysis takes into account these baseline controls  Where sources are uncontrolled, all
a\ ailable controls are considered  For controlled sources, only those control altematues that pro\ ide NOx
emission reductions be\ond the baseline le\el of control are considered

        Separate methods are used to determine \\hat controls are applied for trading (industrial boiler and
turbine) and non-trading sources  Ho\\ever. in both cases the allocation of controls across sources is based
on the total annual costs per ton of NOx reduced in the ozone season for different controls 4

        For trading sources, a least-cost analysis is conducted A NOx emissions budget for the collection of
large industrial boilers and turbines is established at different le\els of stringenc\   The least costh controls.
in terms of total annual cost per ozone season ton remo\ed. across the entire set of possible source-control
measure combinations arc selected in order until the required NOx emission budget is achie\ ed  Costs used
in the least-cost modeling  are based on source capacity. if capacif) information is available, and on average
dollar per ton costs if not

        For non-trading sources, a more conventional source-category  specific cost analysis is conducted
Regulatory alternatives that place a limit on the cost per ton of reduction are examined, and the most effecti\e
NOx control technique that has a cost per ton below that limit is selected for each unit  The cost per ton
reduction for these control techniques is then multiplied by the difference between 2007 baseline and
controlled emissions, to estimate total annual costs for each source.
         Total annual cost is the sum of annuahzed capital costs and annual operating and maintenance costs

                                               Page 5-5

-------
5.4     Administrative Costs for Industrial Boilers and Turbines

        In addition to control costs, potentially affected sources could incur administrative costs associated
with the collection and reporting of NOx emissions  Estimates are developed of the administrate e costs for
requirements beyond those that exist in the baseline. The additional requirements include one-time activities
and annual activities

        The one-time activity for an industry source to read and interpret the reporting requirements of the
rule is estimated to be 1 hour for technical staff and 1 hour for managerial staff The effort to re\ ise a Title V
permit to incorporate NOx monitoring requirements of the final rule is estimated to be 0.8 hours of technical
staff (i e . 4 hours annuahzed o\er 5-years)

        Annual activity industry burden items associated with the collection and reporting of data include a
requirement to submit a year-end compliance certification report  The burden associated with this activity is
estimated to be 2 hours of technical staff time, and 0 5 hour of managerial staff time for trading sources  For
non-trading sources, the burden estimate is 8 hours of technical staff time, and 2 hours of managerial staff
time

        Owners of sources that are eligible to participate in the emissions trading program and elect to do so
will incur some administrative costs associated with the trading system  Chapter 4 (Section 4 4) pro\ ides
information on the administrate e costs associated with trading for EGUs  These same costs apph to
industrial boilers and turbines and other stationary sources that participate in the trading program, and these
costs are pro\ ided in Chapter 8 (Section 8 3)
5.5     Economic Impact Analysis

        This section describes the methodology used to estmate economic impacts for estabhshements and
firms that are potentially directly affected b\ the NOx SIP call  These arc distinguished from indirect
impacts, which are impacts on related parties - suppliers (including the pollution control industry).
customers, or competitors of the potential!) directK  affected establishments — that result from the rule
Indirect impacts would also include impacts on local taxpa\ers where sources owned b\ local governments
(e g . schools or municipal combustion units) are subject to increased costs
5.5.1    Overview of the Economic Impact Analysis Methodology5

        Consistent with the analysis of electric power industry sources described in Chapter 4, this analysis
examines the economic impacts of incremental costs incurred by potentially affected sources in the year 2007
No attempt is made to forecast changes in economic conditions between 1995 and 2007. however The
financial characteristics of the establishments and firms affected by the rule are assumed to remain the same
as reported in 1995 (the latest year for which Census data are currently available )  To provide results in units
comparable to the cost analyses prepared for the proposed NOx SIP call, costs are expressed in 1990 dollars
        -"" A more detailed explanation of the methods and results for the economic impact anahsis of non-EC il" source-
can be found in Abt. 1998

                                              Page 5-6

-------
Therefore, the 1995 financial data used to assess economic impacts are adjusted to 1990 dollars using the
o\erall GDP deflator6

        Several industries and other sectors (e g . schools, colleges, hospitals and go\ ernments) are
potentially subject to new controls as a result of the NOx SIP call  States will ultimately decide what control
measures are necessary to meet the NOx emissions budgets stipulated in the SIP call, so the exact sources
that will face new controls is not known at this time  Based on the simulated control scenarios that are used
by EPA to develop the State emissions budgets, the economic impact analysis for non-EGU sources relies on
a screening analysis to focus on the  sectors  that may potentially experience impacts  More detailed analysis
of market-le\el impacts and indirect impacts is needed only if the screening analysis shows that a substantial
number of establishments in any industry might be subject to significant  impacts  The more detailed market-
le\el analysis \\ould assess the distribution  of impacts among subsectors of the affected industry and their
suppliers, customers and competitors

        Potential economic impacts are assessed at both the plant and firm le\ el  Impacts at the plant.
facility or establishment level are re^ ant for assessing the potential for plant closures, and to calculate
aggregate impacts for specific industries   Impacts at the firm-level are evaluated to determine whether small
entities ma>  be significant!) impacted as part of the illustrative implementation scenario, and to determine
\\hether the  combined effect of requirements at multiple establishments owned by the same firm would
impose a significant burden at the firm le\ el

        The screening analysis is based on  calculating the ratio of total annual compliance costs to annual
sales (for businesses) or (for non-profits or go\ ernments) other measures of revenues or receipts  Two
screening thresholds are used one percent and three percent  Where total annual costs represent less than one
percent of annual sales or re\enues.  it is assumed that the rule will not cause significant burdens to the
establishment or firm in question Establishments or firms that are predicted to incur costs of three percent of
sales or revenues or more are assumed to be potential candidates for significant impacts under our illustrate e
implementation scenario Cases where annual costs equal between one and three percent of sales/receipts are
borderline cases  In an industry that operates with low profit margins, costs of this magnitude could represent
an economic burden, \\hile in higher-margin industries this level of costs  \\ould not impose significant
impacts

        The screening  anah sis docs not indicate which establishments or firms will in fact experience
significant economic burdens as a result of the NOx SIP call, for two reasons

•       First, the NOx SIP call does not impose specific requirements on sources, but rather requires States
       to set NOx emissions limits that will achiexe the aggregate NOx emissions budget established for
       each State  States have discretion in ho\\ the\ choose to allocate required reductions across sources
       The actual allocation of reductions  may differ from that assumed in this RIA  In particular. States
        6 Note that the adjusted data represent 1995 economic conditions expressed in 1990 dollars, not 1990
economic conditions

         The terms plant, facilm and establishment are used interchangeably to refer to a single location, which may
include one or more emission sources subject to additional requirements under the NOx SIP call Costs estimated at the
source level are aggregated to the facility level to provide the required inputs for the economic impact analysis  In
addition, a single firm ma\ own multiple plants or establishments Firm-level analysis requires aggregating costs for
multiple establishments o\\ned b\ the same parent firm

                                               Page 5-7

-------
        may choose to impose less stringent limits in those cases where the limits assumed in this analysis
        would impose significant economic burdens

•       Second, the affected firms may be able to recover some of the added costs by increasing prices to
        customers  This outcome is more likely where a substantial number f firms in a given industry
        sector are affected and less likely if only a few firms in an industry sector incur costs 8 A detailed
        market-level analysis  would be required to determine to what extent Finns would be able to recover
        costs through price increases  The screening analysis makes a worst-case assumption about impacts
        on profits — that all costs are borne by the directly -affected firms, arad no costs are recovered
        through price increases

The economic impact screening analysis can therefore be viewed as providing a general indication of the
potential for significant impacts for EPA's illustrative implementation scenarno. rather than a prediction of
specific outcomes  The screening analysis can be used to eliminate establishments and industries which can
safely be assumed not to experience significant impacts and highlight other cases for more detailed
im estigation The results may help States decide how to implement the requirements in ways that limit the
most significant impacts identified in the screening analysis
5.5.2    Data Sources

        The screening analysis relies on Dun & Bradstreet (D&B) data, where available, to determine the
size of individual affected establishments and the entities that own them  D&B DUNS identifiers are
collected for as many of the potentially affected establishments as possible using EPA's FINDS (the Facility
Indexing System) (EPA. 1998) and Toxic Release Inventors- (TRI) (EPA. 1995) databases  A D&B record
for each potentially affected establishment is then accessed to identify the firEB that owns the establishment
(the D&B "ultimate")  The D&B record also pro\ ided estimates of employment at the potentially affected
establishment ("employment here") and employment and sales at the ultimate firm levelb

        The D&B employment data are used for two purposes

•       To classify the firms owning potentially affected establishments as small or large, for those
        establishments in industries for \\hich the SBA small-firm criteria are expressed in numbers of
        employees.

•       To determine the size category for each potentially affected establishment so that the appropriate
        Census economic data can be selected for the establishment-le\ el impacts analysis

The D&B "ultimate" sales data are used to assess the ratio of total annual compliance costs to sales at the
firm level
        8 In the latter case, the affected firms would most likely not be able to raise tiheir prices to recover costs because
of competition from firms that do not incur the added costs

        9 In some cases, sales at the establishment level is also provided by D&B These data often in fact reflect sales
at the firm level or some intermediate le\ el in the firm organization, however, and were not believed to be consistent
enough to be used in the analysis of economic impacts

                                              Page 5-8

-------
        Because reliable sales or revenue data are generally not available for indiv idual establishments, the
economic impact analysis relies on Census data to estimate average SIC establishment-level sales, revenues
and receipts  Census data are reported for industries defined by 4-digit SIC codes  Mam of the 4-digit SICs
are v ery broad and include establishments of varying sizes and characteristics  Census data are also
disaggregated by establishment- and firm-size.  Where establishment employment data are available from
D&B. they are used to select Census financial data for the size group as well as industry appropriate for each
affected establishment

        Where D&B employment data are not available for individual establishments. Census data on the
sales/revenues/receipts for the average establishment and for the average small entity (e g . firm) in each
industry (four-digit SIC) are used to screen for potentially significant impacts

        Total annual compliance costs described in Section 5.3 are before-tax costs, \vhich is in general the
appropriate measure for estimating the total social costs of the rule  To estimate economic impacts, however.
the more relevant costs are after-tax costs  From the potentiallv affected establishment's perspective, the
costs associated \\ith the NOx SIP call are tax-deductible, as are other business expenses  The burden of
these costs is therefore shared by the affected firms  and the U S taxpayer in the form of lost tax re\enues

        Fullv adjusting for the tax consequences of the estimated costs would be complex, given the range of
compliance options invohed and the fact that some  of the affected facilities are not subject to Federal
corporate income taxes (e g  go\ ernment entities or  non-profit hospitals and schools ) The economic impact
analvsis is therefore conducted using before-tax costs, \\hich overstates impacts on establishments for which
these costs are tax-deductible

        For three sectors, additional data sources are used to obtain financial data

•       For establishments owned bv electric utilities (in particular, those in SICs 4911 and 4931). data are
        obtain from the Energy Information Administration (E1A)  The E1A sources pro\ ide both total
        mega\\att hours (MWh) generated and total sales for the parent electric utilities of the potential!)
        affected establishments  The former are used to determine which establishments. 1996. were owned
        bv  small utilities (based on the SBA threshold of 4 million MWh). and the latter is used as the
        measure of firm-lev el sales

•       For colleges and universities, data on rev enues (tuition and fees) are obtained from the National
        Center for Education and Statistics !

•       For government-owned sources, data on rev enues and expenditures are obtained from the Census of
        Governments

        Census data are obtained from the Department of Census' Statistics of U S  Businesses and the
various  1992 Economic Censuses Data on sales (value of shipments, receipts or revenues, depending on the
sector) for the appropriate SIC and size category are divided by the number of establishments or firms,  to
provide the average sales/rev enues/receipts per establishment or firm.
        10 This measure of financial strength is used rather than a broader measure—which includes income from
endowments—to provide a-conservative screen for potential impacts

                                              Page 5-9

-------
5.6     Small Entity Economic Impacts

        A small entities impact analysis is required to compK with RFA requirements, as described in
Chapter 1  The analysis is designed to determine whether EPA can certify that the NOx SIP call will not
impose '"significant impacts on a substantial number of small entities." While the RFA does not apply to this
action, as discussed in Chapter 1. EPA has elected to evaluate the potential impacts of the rule on small
entities, based on assumptions about how the States could implement the requirements

        The screening analysis described in Section 5  5 provides the information needed to assess whether
the NOx SIP call might impose a significant impact on a substantial number of small entities if States were to
directK  adopt the illustrative implementation scenario examined in this RIA  For businesses, the D&B data
on firm-level employment and revenues are compared  with the SBA size standards to determine which
establishments are owned by small entities  Additional data are collected to characterize the size of affected
non-federal go\ eminent, utility, and college and universib,  entities, as described pre\iousl\

        Once it is determined which establishments are small using the SBA definitions, the firm-le\el
screening analysis results are used to screen for potential small entity impacts The results of this screening
analysis for sources other than electricity generating sources are combined with the results  of the small entin
analysis for electric utilities (described in Chapter 4) to provide an assessment of potential small entity
impacts for the rule as a whole  The results of the combined small entity analysis are provided in Chapter 9
5.1     References

Abt Associates. 1998  Non-Electncjty Generating Unit Economic Impact Analysis for the NOx SIP Call
Prepared for the U S  Environmental Protection Agenc\. Office of Air Quaht} Planning and Standards.
September 1998

National Center for Education and Statistics. Integrated Post-Secondary Education Data S\stem. FY 1994-
95. unpublished

Pechan-Avanti Group. 1998  Ozone Transport Rulemaking Non-Electricity Generating Unit Cost Analysis
Prepared for the U S  Em ironmental Protection Agenc\. Office of Air Quality Planning and Standards .
September 1998

Support for Revising ICRfor Reporting Requirements for NOx SIP call  Technical Memorandum Industry
Indirect Burdens for NOx SIP Call September 1998

U.S  Census Bureau. 1992 Economic Censuses
        Census of Agriculture (SICs 02. 07)
        Census of Mineral Industries (SICs 10-14)
        Census of Construction and Housing (SICs 16-17)
        Census of Manufacturers (SICs 20-39)
        Census of Transportation.  Communications and Utilities (SICs 40-49)
        Census of Wholesale Trade (SIC 50-51)
        Census of Retail Trade (SICs 52-59)
        Census of Sen ice Industries  (SICs 70-89)
        Census of Governments (SICs 91-97)

                                             Page 5-10

-------
U S  Census Bureau. 1995  Statistics ofUS Businesses (available from the Small Business Administration
at hup ll\\\\\\ sba go\/ADVO/stats/mt_data html)

US  Department of Energy. 1996  FormElA-861.  Energy Information Administration. 1996

US  Environmental Protection Agency.  1995   Toxics Release Inventory  Office of Information Resources
Management. 1995  (available from the U S EPA at http://\v\vw.epa gov/enviro/tri/tn_overyie\v.html)

US  Em ironmental Protection Agenc\.  1998   Facility Indexing System (FINDS)  Office of Information
Resources Management. September 1998  (a\ ailable from the U S  EPA at
http //\\-\\-\\ epa gov/em iro/finds/fnd_oveme\\ html)
                                           Page 5-11

-------
      Chapter 6. RESULTS OF COST, EMISSIONS, AND ECONOMIC IMPACT ANALYSES
                           FOR THE ELECTRIC POWER INDUSTRY
        This chapter summarizes the potential cost. NOx emission reductions, and economic impacts
associated with the NOx SIP call for electricity generating sources  The results of various regulator)
alternatives are presented and compared with each other and with one of two baselines  Section 6 1
introduces the  annual cost and emission measures and compares four uniform regulator}.' alternatives (based
on NOx emission rates of 0 25. 0 20, 0 15. and 0.12 Ibs/mmBtu) to the Initial Base Case (as well as the Final
Base Case)  Section 6 2 compares the results  under the 015 uniform (that is. NOx-SIP-call-region-wide)
alternative to alternatives that restrict trading to two or three sub-regions, with emission rates that van' by
sub-region   Section 6 3 presents the results of an analysis of alternative program designs and sensitivity
analyses, in which the results for the 0 15 alternative are shown under various alternative assumptions
regarding electnciU demand, equipment life, control measure effectiveness,  and the discount rate  Potential
direct and indirect economic impacts of the rule are discussed in Sections 6 4 and 6 5 of this chapter.
respectn el>   Finalh. Section 6 6 presents administrative costs, and Section 6 7 contains references for the
chapter

        The comparisons between altematnes. comparisons of major program alternatives, and the
sensitivit> anahses are made in reference to the 0 15 Ib/mmBtu alternative because that alternative is the
basis for the final NOx SIP call emissions budgets  For most of the comparisons, results are presented only
for the year 2007 Limiting the presentation to a single year simplifies the exposition, and the similantx in
costs and emission reductions from year to year ensures that little is lost by the simplification  The >ear 2007
is selected in pan because many areas of the affected region are obligated to reach compliance \\ith the one-
hour ozone standard in that \ ear  In addition,  modeling predicts that annual  compliance costs reach their peak
near 2007. so presenting onh that year avoids understating the costs of the rule
6.1     Comparison of Uniform Alternatives to the Initial Base Case

        EPA considered four geographically uniform alternatn es in developing the NOx SIP call, each one
based on a different allowable emissions rate  For example, the 0 15 alternative is based on limiting summer
NOx emissions to 0 15 Ib/mmBtu of fuel heat input during the summer season after allowing for growth in
electnciU demand to 2007  This alternate e imposes a seasonal cap of 564 thousand tons of NOx   The 0 15
alternative pro\ ides the point of comparison for sensitn its and other analyses in this chapter  The Integrated
Planning Model was used to generate predictions of the technology selection, costs, and emissions for
electricity generating units under the various alternatives

        Trading is assumed to be allowed both within and among the 23 jurisdictions in the SIP call region
EPA examines the cost and emission reduction impacts of each of the uniform regulatory alternatives
incremental to the Initial  Base Case level  The Initial Base Case assumes compliance with RACT, BACT.
and NSPS requirements,  as well as Phase I of the Ozone Transport Commission (OTC) Memorandum of
Understanding (MOU). as such it includes all currently applicable Federal or State NOx control measures
The Final Base Case assumes the controls included in the Initial Base Case, as well as Phase II and Phase III
of the OTC MOU  In the Final Base Case, many units have already implemented SNCR and SCR controls to
meet the more stringent requirements of the Final Base Case   This section presents the results of the cost and
emissions reductions analyses and translates those results into measures of cost-effectn-eness for the uniform
                                             Page 6-1

-------
regulator* altematnes  Section 6 1 5 of this chapter contains a comparison of the costs of the 0 15 trading
alternate e in the Initial and Final Base Cases
6.1.1    Technology Selection

        Tables 6-1 and 6-2 present emission control responses for coal and oil/gas fired boilers in the SIP
call region under each of the uniform alternatives  If States choose to follow the implementation scenario that
EPA has modeled. Table 6-3 shows additions to natural gas combined cycle capacity that will occur as part of
compliance  Industry \\ ill increase its use of natural gas over coal to generate power as part of its approach to
compliance  Some coal, oil, or gas-fired boilers will be retrofit with SCR. SNCR, or gas reburn. Others will
have no incremental control technology added beyond the types of controls required under Title IV. BACT
and OTC Phase I/RACT in the Initial Base Case  In addition, some boiler capacity will could close in
response to the wa\  States implement the NO\ SIP call  Not shown in the table are combustion turbines and
combined cycle units, which are not expected to be retrofit with additional controls in response to the NOx
SIP call  IPM anah sis does sho\\. ho\\e\er. that about 2.000 to 4.000 MW of combined cycle capacity
would be added, depending on the alternative  The IPM runs project that almost all of the control technology
retrofits needed to reduce emissions to the cap under the uniform alternatives would come from the coal-fired
boilers, which tend to be both larger and higher m baseline emissions than other types  As alternames
become more stringent. Title IV controls (i e . combustion controls such as lo\%  NOx burners) are augmented
with SNCR and then with SCR (which is capable of greater NOx reduction)  The same general pattern is
seen for the oil/gas- fired boilers, though the percentages of them that are  retrofit with SNCR or SCR are
smaller than for the coal boilers
                                              Table 6-1
                    Estimated Emission Control Responses for Coal-Fired Steam Units
                                      to the NOx SIP Call in 2007
                                 (M\V Capacity for the SIP Call Region)
Emission Control Response
Close Unit
Comply \\ ith BACT
Title IV NOx Controls Onh a
Add SNCR
Add SCR
Add Gas Reburn
0.25 Trading
18
4.158
97.895
93,003
7,208
-
0.20 Trading
16
4.158
40.242
133,240
23.384
1.242
0.15 Trading
113
4,158
4.545
129.690
63,267
509
0.12 Trading
18?
4,158
4,8^9
83.172
109,761
129
 Source ICF anahsis
 ' This row shows the M\V capacit) adding only Title IV \O.\ controls Therefore, the numbers tend to decrease with increases in option strmgenc)
                                               Page 6-2

-------
                                            Table 6-2
               Estimated Emission Control Responses for Oil/Gas-Fired Steam Units
                                    to the NOx SIP Call in 2007
                              (MW Capacity for the SIP Call Region)
Emission Control Response
Close Unit
No Further Controls Be\ond
OTC Phase 1/RACT
Add SNCR
Add SCR
0.25 Trading
182
33.564
447
-
0.20 Trading
181
33.253
759
-
0.15 Trading
201
29.890
4.102
-
0.1 2 Trading
151
23,624
7.008
3.410
Source 1CF anahsis
                                            Table 6-3
                    Estimated Emission Control Responses to the iNOx SIP Call
                          in 2007 - Added Natural Gas Combined-Cycle
                               (MW Capacity for the SIP Call region)
Emission Control Response
Added Capacin of Combined C\clea
0.25 Trading
1 .895
0.20 Trading
1.798
0.15 Trading
2.200
0.12 Trading
4.156
Source !Ct anaKw
1 Aho\e 1e\e! in the Initial Base Case which is 47.308 M\V
6.1.2    Emissions

        The control technologies presented in Tables 6-1 and 6-2. \\hich result from the implementation
scenario modeled b\ EPA. \\ill reduce NOx emissions b\ hundreds of thousands of tons in the SIP call
region   Although the NOx SIP call focuses on ozone season NOx emissions, reductions of NOx emissions
under the uniform alternatnes will occur \ ear-round because some of the control strategies (e g . combustion
controls) function continuous!)  For the 0  15 alternatne in 2007. the annual reductions amount to 1.183
thousand tons o\er the Initial Base Case, \\ith 245 thousand of those tons (21 percent) from outside the ozone
season  Table 6-4 sho\\s the incremental ozone season tons of NOx emitted under each of the uniform
alternate es compared to the Initial Base Case  The rule requires sources to be in compliance starting in May
2003  From that point on, the emissions for electricity generating units are assumed to be capped under the
scenarios modeled by EPA. resulting in 564 thousand tons of NOx per ozone season under the 0.15
alternate e  Initial Base Case emissions continue to increase after this point due to forecasted growth in
electric power generation, while the cap remains constant  As a result, the incremental NOx emission
reductions gro\\ yearh after 2003  Figure  6-1 shows State-by-State emissions results for each of the uniform
trading alternatives, compared to the emissions under the Initial Base Case1.  Figure 6-2 shows  the emissions
for the Initial Base Case, the State budget levels, and the IPM analysis results for the 0  15 trading
         The data used to develop Figure 6-1 is included in Appendix B. Table B-l

                                             Page 6-3

-------
alternate e:  The incremental reduction in ozone season tons under the 0 15 alternate e amounts to about 62
percent of baseline ozone season NOx emissions in the period between 2003 and 2010 — slighth less in the
earl> \ears and slight!} more in the later years. By contrast, the reductions in annual tons are less than 35
percent of baseline annual tons, because emission reductions m the winter months are onh' about 12 percent
of baseline emissions  This disparity stems from the fact that the most widely used control strategies —
SNCR and SCR — can be shut off at the end of each ozone season to limit operating costs  Most
importantly, as Figure 6-2 shows, a uniform trading program can lead to reductions throughout the NOx SIP
call domain that are comparable to what would occur under a command-and-control approach where States
set emission rates for EGUs aimed at hitting each State's NOx budget le\el
                                            Table 6-4
                      Estimated Ozone Season NOx Emissions and Reductions
                  under the Uniform Trading Alternatives and the Initial Base Case
                                           (1,000 tons)
Case/Alternathe
Initial Base Case
025 Trading
(Reduction)
0 20 Trading
(Reduction)
0 1 5 Trading
(Reduction1)
0 12 Trading
(Reduction)
2003 1 2005
1,462
940
(523)
751
(711)
564
(899)
453
(1.009)
1.497
940
(557)
751
(746)
564
(933)
453
(1.043)
2007
1,502
940
(563)
751
(751)
564
(938)
453
(1.049)
2010
1.511
940
(572)
751
(760)
564
(948)
453
(1.058)
Source ICh anaKsis
Numbers do not sum due to rounding
        : The data used to develop Figure 6-2 is included in Appendix B, Table B-2

                                             Page 6-4

-------
                                                                  Figure 6-1
                      O/.onc Season NO\ Emissions in 2007 from the Electric Power Industry for States in the SIP Call Region:
                                         Uniform Trading Alternatives Compared to the Initial Base C'ase
      Initial Base Case

      O 25 Trading Case

      O.2O Trading Case

      O 15 Trading Case

|    |  O 1? Trading Case

Scale:  Ohio Base Case =  163.132  Tons

Source:   K"l; Analysis
                                                                   I'lljiC 6-5

-------
                                                       Figure 6-2
              O/one Season INOx Emissions in 2007 from the Electric Power Industry for States in the SIP Call Region:
  Uniform 0.15 Trading Alternative Compared to the Initial Base Case and the State Budget Component under the 0.15 Ih/mmBtu Limit
       Initial Base Case
      Component of Budget NOx
      for 0 15 Option
      0 15 Option

Scale  Ohio Base Case = 163,132 Tons
                                                       Page 6-6

-------
6.1.3    Costs

        EPA calculated the annual cost of the uniform alternates incremental to the Initial Base Case level3
Table 6-5 presents EPA's estimates of the total annual costs that the electric power industry could incur in the
years 2003. 2005. 2007. and 2010 For each of the uniform alternatives, the incremental cost rises until
2007. but begins to fall in subsequent years  Because of growth in demand for electricity, the fixed cap of
564.000 tons per ozone season becomes progressively tighter over time; as fuel input grows, the fixed
allocation of allowances leads to a tighter and tighter effective limit  This effective tightening tends to dm e
the costs of meeting the emissions cap higher o\ er time  Countering this tendency, on the other hand, is a
reduction over time in the costs of new generation and control technologies, and the possibility of retrofitting
existing plants to function as combined-cycle units  B\ 2010. these improvements begin to dominate, and
incremental costs begin to decline
                                              Table 6-5
         Incremental Annual Costs for Uniform Alternatives Relative to the Initial Base Case
                      (Compliance Costs above Initial Base Case, million 1990S)2
Alternative
0 25 Trading
0 20 Trading
0 1 5 Trading
0 12 Trading
2003
$589
$894
$1.308
$1.766
2005
$628
$935
$1.354
$1.816
2007
$643
$948
$1.378
$1.846
2010
$632
$932
$1.341
$1.757
Source ICh anahsis
1 Compliance costs do not include administrate monitoring or transaction costs which are minor in comparison to the total cost of the rule t nits
co\eredh> Title IV will ha\e monitoring de\ ices in the baseline, which will reduce the incremental monitoring costs  See Section 6 6
        Trading

        Some regulated sources have years of experience with inter-firm and mtra-firm emissions trading  In
the mid-1980s. EPA published the Emissions Trading Policy Statement (51 FR 43831). which allowed
sources to obtain emission credits for use as emission offsets and in bubbles  In 1990. the Clean Air Act
Amendments (CAAA) expanded the potential pool of sources that would need to obtain offsets  The 1990
CAAA was also the ad\ent of the acid ram S0: allowance market, under which the electric power industry
learned to use emissions trading as a compliance strategy  In addition, a variety of market-based programs
have been implemented at the State and local \c\ els  Most recently, the Ozone Transport Commission
adopted a Memorandum of Understanding committing the signatory States to the development and proposal
of a regional NOx emissions cap-and-trade program, similar to the one proposed under the NOx SIP call
Under these emissions trading programs, especially the allowance markets, the affected sources became
familiar with emissions trading markets and the procedure for buying and selling allowances
        3 All cost data and cost-effectiveness calculations are presented in 1990 dollars

                                               Page 6-7

-------
        Sources are assumed to participate in the NOx allowance market to utilize the most cost-effectn e
compliance option and because of experience gained \\ith other trading programs  The potentially affected
sources are expected to trade allowances within their own company and/or with other companies  For
example, based on the IPM results for the 0.15 alternative, 641 units are projected to obtain about 78,000
allowances from the 506 units projected to provide excess allowances.  Only 23 percent of the units (333
units) are expected to use only the allowances allocated to them. Table 6-6 shows the number of expected
trades for individual units, including inter- and mtra-firm transactions, under the 0.25. 0.15, and 0.12 uniform
alternatives  It is notable that a substantial  number of trades could occur under each of the alternatives
examined, with mam units able to generate excess allowances for the use of other units  The number of
allowances traded between and within companies under each of the uniform alternatives varies somewhat
under the 0 12 alternative, approximately 73.000 allowances may be traded, while under the 0 25 alternate e.
approximate!) 118.000 allowances may be  traded  Under the 0.15 alternative, about 37.000 of the 78.000
traded allowances (about half) are projected to be inter-firm trades  Though EPA  has estimated the volumes
of inter-firm allowance trades only for the 0.15 alternative, the number of allowances traded among individual
firms ma\ \ an from alternative to alternative
6.1.4   Cost-Effectiveness

        The average cost-effectiveness of the regulatory alternatives is calculated from the Initial Base Case
le\cl  Cost-effecti\ eness is calculated as the total annual costs of the alternative dnided by ozone season
emission reductions  Table 6-7 shows the emissions change and the annual costs and cost-effectiveness that
the EPA estimates for the potentially affected part of the electric power industry in the years 2003. 2005.
2007. and 2010  As shown in the table, the average costs per ozone season ton of NOx removed under the
0 15 alternative for each of the four years differ slightK. but for each year is less than $1.500 per ton of NOx
removed  The highest cost per ton removed is seen in 2007 The effects of the growth in electnciK demand
and the application of the fixed cap of ozone season tons of NOx on cost, described in the preceding section.
explains the pattern in cost-effectiveness over time Comparing the change in total costs to the change in
emissions, it can be seen that the cost per ozone season ton removed increases   Thus, costs are rising faster
than emission reductions, as more costly measures are pressed into sen-ice on smaller and less-mtensivcK
used units

        The increasing per-ton cost can be seen more clearh by presenting the changes in costs and tons for
each alternative relative to the next-most-stringent alternative,  instead of relative to the base case  This
approach, which shows the incremental per-ton costs of just the additional tons of reductions as the
alternatives grow more stringent, is presented in Table 6-8
                                              Page 6-8

-------
                                                 Table 6-6
                              Number of Fossil Fuel-Fired Units in IPM Runs
                Expected to Buy, Sell, or Do Nothing in the NOx SIP Call Trading Program"
                                (0.25,0.15, and 0.12 Uniform Alternate es)
Fuel T>pe
BUY Allowances
Sell Allowances
Do Nothing
0.25 Trading
Coal
Oil /Gas
Combined Cvcle/ Combustion Turbine
Integrated Gasification/ Combined C\cle
Total
517
9
15
0
541
265
123
207
1
596
0
10
335
0
345
0.15 Trading
Coal
Oil 'Gas
Combined C\clc/ Combustion Turbine
Integrated Gasification' Combined C\cle
Total
507
39
95
0
641
275
91
139
1
506
0
10
323
0
333
0.12 Trading
Coal
Oil/Ga>
Combined C\cle/ Combustion Turbine
integrated Gasification' Combined C\cle
1 otal
462
38
112
0
612
320
92
130
1
543
0
10
315
0
325
1 AJIcmance transfers within companies are included among the purchases and sales though mone\ would not necessanK change hands in these
internal transactions
                                                  Page 6-9

-------
                                                   Table 6-7
                  Summary of Estimated Emission Reductions, Cost, and Cost-Effecti\ eness
                       for the Uniform Alternatives of the NOx SIP Call:  Selected Years
Year/AlternatK e
Reductions in Ozone
Season NOx Emissions
(1,000 tons)
Annual Cost above
Initial Base Case
(million 1990S)
Cost per Ozone Season
Ton of NOx Removed
(1990S/ton)
2003
025 Trading
0.20 Trading
0 1 5 Trading
0 1 2 Trading
523
711
899
1.009
$589
$894
$1,308
$1,766
SI. 127
$1,258
$1,455
$1,750
2005
0 25 Trading
0 20 Trading
0 1 5 Trading
0 1 2 Trading
557
746
933
1.043
$628
$935
$1,354
$1.816
$1.128
$1.254
$1,451
$1.741
2007
0 25 Trading
0 20 Trading
0 1 5 Trading
0 12 Trading
563
751
938
1.049
$643
$948
$1.378
$1.846
$1.143
$1.263
$1.468
$1.760
2010
0 25 Trading
0 20 Trading
0151 rading
0 12 Trading
572
760
948
1.058
$632
$932
$1.341
$1,757
$1.106
$1.226
$1.415
$1.660
Source ICFanaKsis
Because of rounding, the cost-effecti\ eness values do not equal the ratio of the costs to the XOx reductions shown in the table
                                                   Page 6-10

-------
                                             Table 6-8
          Comparison of Estimated 2007 Incremental Ozone Season NOx Emission Reduction, Cost,
                       and Cost-Effectiveness for Different Regulatory Alternatives
Alternate c
0 25 Trading3
0 20 Trading
0 15 Trading
0 1 2 Trading
Reduction Incremental to
Next-Most-Stringent
Alternative
(1,000 ozone season tons)
563
188
187
111
Cost, Incremental to
Next-Most-Stringent
Alternative
(million 1990S)
$643
$305
$430
$468
Incremental Cost-
Effectiveness, Relative to
Next-Most-Stringent
Alternative
(1990S/ozone season ton)
$1.143
$1,618
$2,294
$4.240
Source ICF anahsis
Because of rounding, the cost-effecti\eness values do not equal the ratio of the incremental costs to the NOx reductions shown in the table
1 Compared to the Initial Base Case
6.1.5   Initial Base Case Compared to Final Base Case

        The Initial Base Case assumes compliance with RACT. BACT. and NSPS requirements, as well as
Phase I of the Ozone Transport Commission (OTC) Memorandum of Understanding (MOU)  As such, it is
best suited for estimating the cost of further controls abo\ e the requirements alread\ in place   The Final
Base Case assumes the controls included in the Initial Base Case, as \\ell as Phase II and Phase III of the
OTC MOU  In the Final Base Case, mam units have alread> implemented SNCR and SCR controls to meet
the more stringent requirements of the Final Base Case and \\ill not need to spend as much to meet the NOx
SIP call requirements  For this reason, the cost of the NOx SIP call is lower \\hen compared to the Final Base
Case than to the Initial Base Case  For example, the incremental annual cost of the 0 15 alternatne is S1.250
million when compared to the Final Base Case, but $1.378 million  compared to the Initial Base Case
6.2     Regional versus Uniform Approach to Trading

        The uniform alternate es presented in the preceding section set State-b\ -State caps that are all based
on the same nominal emission rate, and envision unrestricted trading of emission allowances among all SIP
call States  EPA also examined altematn es that attempt to target the emission reductions to the sources that
might have the greatest impact on severe non-attainment areas downwind  These regional trading alternatives
set tighter caps for States closer to the Northeast, and less stringent caps for the Midwest and Southeast
Because a trading system that allowed unrestricted trade of allowances from region to region would
undermine the stratified caps set up by these regional alternatives, interstate trading is limited to States within
the same regions

         The remainder of this section compares the cost and emissions results under the 0.15 trading
alternatne with the two-region (Regionality 1) and three region (Regionally 2) alternatives in turn  The
comparisons place less emphasis on cost per ton of NOx removed, and more emphasis on showing the
differences between alternatnes in terms of emissions by region and the effects of trading  Because the intent
                                             Page 6-11

-------
of the regional alternatives is to cost-effectively reduce ozone levels in severe nonattainment areas, the cost
per ton of NOx remo\ed is of secondary importance compared to the distribution of emission reductions


6.2.1    Comparison of Baseline and Uniform 0.15 Alternative to the Two-Region Alternative
        (Regionally 1)

        The summary of the effects of the two-region alternative (Regionahty 1) relative to the 0.15 trading
alternate e  are shown m Table 6-9   The two-region alternative restricts trading to within two regions, split
approximate!} along Northern and  Southern boundaries 4 Because the two-region alternative reduces the
stringency of the cap (to a nominal 0.20  Ib/mmBtu) in the Region 1 while leaving it constant (at a nominal
0.15 Ib/mmBtu) in Region 2. it results in smaller emission reductions m Region 2 compared to the 0.15
trading alternative  Figure 6-3 shows the State-by-State  emission results for the two-region alternate e
compared to the Initial Base Case s

        The t\\o-region altematne would be less costly than the 0 15  trading alternatne. \\hich is to be
expected gi\ en its lower stringency  The cost of the 0 15 trading alternative compared to the baseline is
SI.3 78 million, while the cost of the t\\o-region alternatne is $1.118 million   Thus, the incremental 1117
thousand tons of ozone season NOx emissions eliminated in 2007 under the uniform 0.15 trading alternatn e
relatne to the t\\o-reeion altematne \\ould cost an additional $260 million
                                              Table 6-9
                         Comparison of Ozone Season NOx Emission Reductions:
             Uniform 0.15 Trading Alternative, and the Two-Region Alternative (RegionaliU  1)
                                             (1,000 tons)
Incremental Measure
(from Baseline)
Total Tons Reduced
Tons reduced. Region 1
(0 1 5 Ib/mmBtu)
Tons Reduced. Region 2
(020 Ib/mmBtu)
Reductions
Relative to Initial
Baseline:
0.15 Trading
938
351
587
Reductions
Relative to Initial
Baseline: T«o-
Region
Alternative
826
340
486
Incremental
Effect of Tw o-
Region
Alternative
-112
-11
-101
Reductions under
Two-Region
Alternate e as a
Percentage of
0.15 Trading
88%
97%
83%
Source 1CF anahsis
        4 The t\\o-region area consists of Connecticut. Delaware, District of Columbia, Massachusetts, Maryland.
New Jersev New York. Ohio, Penns\lvama, Rhode Island, Virginia, and West Virginia in Region 1 (Northeast and
Mid-Atlantic States), and Alabama, Georgia, Illinois. Indiana, Kentucky, Michigan, Missouri, North Carolina. South
Carolina, Tennessee, and Wisconsin in Region 2 (Southeast and Midwest States)

        5 The State Budgets for the t\\o-region alternative used in Figure 6-3 are calculated using the \anablc erms
rates for each SIP call region  The data used to develop Figure  6-3 is included in Appendix B, Table B-3
                                              Page 6-12

-------
6.2.2   Comparison of Three-Region Alternative (Regionality 2) to Uniform 0.15 Alternative

        The three-region alternate e sets a cap based on 0 12 Ib/mmBtu in Region 1.015 Ib/mmBtu in
Region 2. and 0 20 Ib/mmBtu in the Region 3.6 The emission results under the three-region alternative are
similar to those of the 0 15 altematne.  States in the Northeast tend to reduce NOx emissions beyond the
0 15 Ib/mmBtu levels as part of their compliance strategy  Under the 0.15 alternative, these States will
receive credits that they can sell to electricity generating units in the Southeast and Midwest Under the three-
region alternative, the electricity generating units in the Northeast are held to a tighter standard than under the
0 15 alternative, and will not recen e as many credits as under the 0 15 alternative

        Tables 6-10 and 6-11 display the same information about emissions between alternatives, but in
different forms  Table 6-10 shows total reductions, while Table 6-11 shows differences  Both of the first
two columns of Table 6-10 refer to the 0.15 alternative  The first column shows the total NOx reductions (in
thousands of tons per year) assigned to each of three regions, based on the caps gi\en to  States within Region
1 (the Northeast). Region 2 (the Miduest). and Region 3 (the Southeast)  These figures represent the
emission reductions by region that would be provided b\ the 0 15 alternative in the \ear 2007 if there is no
interstate trading, that is. if States are held to their emissions budgets  The second column shows the
projected emissions reductions under the 0 15 alternative,  given the interstate trading projected b\ IPM for
2007  The fact that the total reduction. 938 thousand tons, is the same in both columns reflects the fact that
trading redistributes emission reductions but  does not change the total

        The third column of Table 6-10 shows the tons of reductions by region under the three-region case.
these figures represent the emissions under the different caps specified by EPA  No  interregional trading is
allo\\ed under this altematne  Across the ro\\s of Table 6-10, \\c see that overall reductions are lower under
the regional alternate e. but reductions in the Northeast and Mid\\est are higher than under the 0 15
alternatue The Southeast, \\ith a cap based  on a limit of 0 20 Ib/mmBtu. has emissions of over 37 thousand
tons higher than the 0 15 alternative  Figure  6-4 shows the State-by-State emission results for the three-
region alternate e compared to the Initial Base Case and the expected results if the State Budgets  were set
with the limit of 0 15 Ib/mmBtu

        The effects of trading and differences between alternatnes are also shown in Table 6-11  The first
column shows the difference between emission reductions under the 0 15 trading alternatne and the
reductions that \\ould be required  given the State-by-state budgets if trading is not allowed  This column
shotts that the pattern of trading under the 0  15 trading altematne \\ould tend to reduce emissions in the
Northeast by an additional 14 thousand tons compared to the State budgets, \\hilc allowing the other t\\o
regions to reduce their emissions less than required to meet the State budgets

        The second column of Table 6-11 compares the results of the three-region alternative to 0 15
alternative without trading. The three-region alternative does not reduce total emissions by as much.
emissions would be higher b\ 22 thousand tons in 2007 under the three-region alternative than under the 0.15
        6 The three-region area consists of Connecticut, Delaware, District of Columbia, Maryland, Massachusetts,
New Jersey. New York. Pennsylvania, and Rhode Island in Region 1 (Northeast States), Illinois, Indiana. Kentucky,
Michigan, Missouri. Ohio. Virginia. West Virginia, and Wisconsin in Region 2 (Midwest and Adjacent OTR States).
and Alabama, Georgia, North Carolina, South Carolina, and Tennessee in Region 3 (Southeast States)

         The data used to develop Figure 6-4 is included in Appendix B. Table B-3

                                              Page 6-13

-------
alternative, largeh  as a result of higher emissions in the Southeast  The three-region alternative does reduce
emissions in the Northeast by almost 25 thousand tons more than under the 0 15 alternative before trading.
while providing for smaller reductions m the Midwest and especially the Southeast

        As shown  in the third column of Table 6-10. the effects of the three-region alternative are smaller if
the changes due to  interregional trading in the 0 15 alternative are considered. The three-region alternative
provides about 11 thousand additional tons of reductions in the Northeast compared to the 0 15 alternative
after trading, and about five thousand extra tons of reductions in the Midwest The annual cost of the three-
region alternative compared to the Initial Base Case is $ 1.349 million  The savings for this alternative
compared to the 0  15 alternative would be on the order of $29 million
                                               Table 6-10
                       Comparison of 2007 Ozone Season NOx Emission Reductions:
         0.15 State Budgets, Uniform 0.15 Trading, and the Three-Region Alternate e (Regionally 2)
                                               (1,000 tons)
Incremental Measure
(from Initial Base Case)
Total Tons Reduced
Tons reduced. Region 1
(0 1 2 Ib/mmBtu)
Tons reduced. Region 2
(0 1 5 Ih/mmBtu)
Tons reduced. Region 3
(0 20 Ib/mmBtu)
Reductions under
0.15 State Budgets"
938
117
607
214
Reductions under
Uniform
0.15 Trading
938
131
603
204
Reductions under
Three-Region
Alternative
916
142
607
167
Source 1CI- anaKsis
1  The Stale Budget le\el< were set with a 0 15 Ib mmBtu rate and a ozone season cap of 564.000 tons This column shows Initial Base Case
Emissions - State Budgets for each region Total emission reductions under the Slate Budgets are equal to the emission reductions under the 0 1'
alternate e
                                                Page 6-14

-------
                                                Table 6-11
                Comparison of the Differences Between 2007 Ozone Season NOx Emissions for
          0.15 State Budgets, Three-Region Alternative (Regionality 2), and Uniform 0.15 Trading
                                                (1,000 tons)
Incremental Measure
(from Baseline)
Total Tons Reduced
Tons reduced. Region 1
(0 1 2 Ih/mmBiu)
Tons reduced. Region 2
(0 1 5 Ih/mmBtu)
Tons reduced. Region 3
(0 20 Ib/mmBtu)
Reductions under
0.15 Trading Relative to
State Budgets"
00
138
-44
-93
Reductions under Three-
Region Alternative
Relative to
State Budgets"
-220
247
0
-467
Reductions under Three-
Region Alternative
Relative to
0.15 Trading
-220
109
4 5
-374
Source Id- anahsis
1 The State budgets are listed in the SNPR for the 0 15 alternative and in .Appendix B. Table B-l
                                                 Page 6-15

-------
                                                             Figure 6-3
               Comparison of 2007 Ozone Season NOx Emissions in the Initial Base Case with the State Component of the NOx Budget
                          Under the Two-Region Alternative and the Emissions Results from the Two-Region Alternative
      Initial Base Case
I    I  Component of NOx Budget
'    '  Under the Two-Region Option
i	1  Two-Region Option
Scale  Ohio Base Case=163,132 Tons
Soiiuv:    ICI-Aiuihsis

-------
                                                           Figure 6-4
                      Comparison of 2007 O/onc Season NOx Emissions in the Initial Base Case with the State Component
                of the INOx Budget for the Three-Region Alternative and the Emissions Results from the Three-Region Alternative
^gg   Initial Base Case
|    |   Component of NOx Budget
^_^   Under the Three-Region Option
[    I   Three-Region Option
 Scale: Ohio Base Case=163,132 Tons

 Source:    IC1!7 Analysis
                                                                                                                   DE
                                                                                                                       Rl
                                                                                                              DC
I	I  Region 1

l'Z.?ZI  Region 2

|     |  Region 3
                                                            'ape 6-17

-------
6.3     Other Program Designs and Sensitivity to Modeling Assumptions

        This section compares the major program alternatives considered by EPA in the development of the
NOx SIP call   Section 6 3  1 contains a discussion of costs in the absence of interstate trading.  Section 6.3 2
contains the results of the banking/no banking analysis. Section 633 presents an analysis of a 0.35
Ibs/mmBtu alternative, and Section 634 presents IPM results for the 0 15 trading altematne under varying
assumptions on technology, discount rate, and electricity demand
6.3.1    Costs without Interstate Trading

        Another program option is to set State budgets based on a uniform emission rate and to restrict
trading between sources within State boundaries  Table 6-12 presents the results of the analysis restricting
trading and shows the difference in outcomes under this "no interstate trading" case for the 0 15 Ib/mmBtu
alternate e  The cost increase for a program with 23 regions (where each of the jurisdictions covered b\ the
NOx SIP call would be its own region and allowing trading only within each State or jurisdiction) compared
to the 0 15 alternative is approximately two percent  The cost difference is due to the fact that there are some
differences in control costs within State boundaries, and effective trading can therefore occur within State
boundaries  This estimate of the difference in cost between the 0 15 trading alternative and the alternative
that does not allow interstate trading is dependent on  the assumption that all States will set up trading
programs for electricity generating units within their boundaries  If States  adopt rate-based approaches, the
cost could be expected to be higher, though this possibility was not explicitly modeled for this report s The
distribution  of emissions under the no interstate trading case, which is equal to the State budgets, is shown  in
Table 6-10
                                              Table 6-12
        Emission Reductions, Cost, and Cost-Effectiveness with and without Interstate Trading in 2007
                                        for the SIP Call Region
Incremental Measure
(from Base Case)
NOx Reductions
( 1 .000 ozone season tons)
Cost (million 1990$)
Average Cost per Ton (1 990$)
Region"
Region 1
Region 2
Region 3
Region Total
Region Total
Region Total
0.15 Alternative
with Interstate
Trading
131
603
204
938
$1,378
$1.468
0.15 Alternate e
without
Interstate
Trading
117
607
214
938
$1,407
$1,499
Effects of
Pre>enting
Interstate
Trading
- 14
4
9
0
$29
-
Source ICFanaKsis
' Regions 1, 2. and 3 are defined in Section 6.2 2
        8 The effects of a program that did not allo\\ trading v\ithm States was addressed in EPA September 1 •>;"

                                               Page 6-18

-------
6.3.2    Banking

        Under an allowance based emissions trading program with banking, sources can create reductions
beyond required le\ els in one season, thus freeing up some allowances for use in a later season Each banked
allowance represents one ton less emissions in the current season  Banking is a cross-cutting option, because
it can be used with any of the uniform or regional alternatives. Banking encourages earl) reductions, provides
flexibility. and reduces cost for regulated sources. It also dispels the "use it or lose it" conception concerning
the use of allowances, and accommodates changes in generation activity that may occur in response to
interruptions of power supply from sources that do not emit NOx  On the other hand, banking can create
uncertainty about actual  emissions in a given season

        EPA considered several banking alternatives, including  options with (1) no banking. (2) banking of
emission reductions after the start of the program. (3) banking of "earh" reductions (i e , those that come
before the beginning of the program), (4) and banking from an earlier phase of the program to a later phase

        Banking of "earh" reductions was only modeled for the 0 15 alternative because earlier IPM analysis
suggested that owners of electricity generating units would want to use it to a very limited degree to lower the
costs of future compliance, although not all the important advantages of banking were incorpated into that
anah sis (EPA.  1991 a]  A two-phase banking program was also not  modeled   Table 6-13 presents IPM
results for the 0 15 alternative with and without banking, where banking begins after the start of the program
in 2003

        Banking is most \ aluable for programs in which costs per ton remo\ ed rise over time, which is most
likeK to occur if the effective stringenc\ of the regulations rises o\er time In the case of the  NOx SIP call,
the effectn e stnngena rises only slighth o\ er time as a  result of a fixed emissions cap interacting with a
gro\\mg demand for electricity   Over time, however, anticipated impro\ements in technology (including
combined-c) cle retrofits) will counteract the effects of growth, so that the cost per ton removed is expected to
fal! e\entualh

        Gn en the fact that costs per ton remo\ ed are expected to rise onh slightly at first and then fall, it is
not surprising that little banking \\as predicted by IPM in the 0 15  alternative with banking beginning in
2003 Both costs and the geographic distribution of emission reductions \\ould be almost the same with or
without  a banking program
                                             Page 6-19

-------
                                             Table 6-13
        Effects of Banking on Estimated Ozone Season NOx Emission Reductions, Incremental Cost,
                         and Cost-Effectiveness for the 0.15 Trading Alternative

Reductions of Ozone season NOX
Emissions (1,000 Ions)
Costs. Incremental to Initial Base Case
(million 1990$)
Average Cost/Ton of Ozone season NOx
Removed (1990$)
| 2003
Without Banking
With Banking
Beginning in 2003
Without Banking
With Banking
Beginning in 2003
Without Banking
With Banking
Beginning in 2003
' 899
902
$1,308
$1,326
$1,455
$1,469
2005
933
931
$1,354
$1.348
$1.451
$1.448
2007
938
935
$1.378
$1.358
$1.468
$1.453
2010
948
948
$1.341
$1.338
$1,415
$1.412
Source ICFanaKsis
        This analysis does not consider a banking plan in which emission reductions prior to 2003 could be
used to ease the transition to the NOx SIP call, and help ensure that allowances were available for planning
purposes eark in the compliance period  Under that type of banking program, more tons would be banked
and the sa\ ings and other ad^ antages (especially in terms of reduced uncertainty) would be greater Annual
emissions starting in 2003. hov\e\er. would be higher and less predictable than in programs that did not allo\\
earh emissions to be banked
6.3.3    Uniform 0.35 Ib/mmBtu Trading Alternative

        EPA analyzed the emissions and cost-effectiveness results of imposing a 0 35 Ib/mmBtu standard for
NOx. assuming no implementation of the OTC MOU Phases II and III (i e . the Initial Base Case)  Under
this scenario, no electricity generating units use post-combustion controls; rather, they add onh  Title IV
controls  The o/one season emissions reductions compared to the Initial Base Case m 2007 are an additional
187 thousand tons  B\ comparison, the 0 15 trading alternative provides  ozone season NOx emission
reductions of 938 thousand tons The lo\\er reductions in NOx emissions correspond to lower total and
average costs  Table 6-14 illustrates these results
                                             Page 6-20

-------
                                              Table 6-14
        Comparison of Estimated 2007 Ozone Season NOx Emission Reductions, Incremental Cost, and
         Cost-Effecti>eness Between Uniform 0.35 Trading Alternative and 0.15 Trading Alternative

NOx Reduction
( 1 .000 ozone season tons)
Annual Cost
(million 1990$)
Average Cost-Eftectiveness
( 1 990$/ozone season ton)
0.35 Trading
187
$217
$1.165
0.15 Trading
938
$1,378
$1,468
Incremental Effect
751
$1.161
$303
Source 1C!- anal\Ms
Numbers ma> not sum due to rounding
6.3.4   Sensitivity to IPM Assumptions

        Sensitivity analyses on several ke\ assumptions in the IPM analysis are presented in this section
These anahses. which respond to mam  of the comments received on the proposed NOx SIP call, include the
effects of a higher discount rate, lower SNCR effectn eness. shorter equipment life, and higher growth rates
of demand were developed to assess the robustness of the results outlined above  The sensitiMty analyses are
all evaluated relatne to the uniform 0 15 trading altername

        Effects of Higher Discount Rate

        This sensitn ity analysis  assumes that the after-tax cost of capital is eight percent per annum rather
than the six percent rate that is assumed in the rest of the IPM modeling  Table 6-15 presents cost and
emission reduction differences  between the t\\o scenarios  As shown, the incremental cost of the 0 15 trading
alternatee rises b\ about t\\o percent \\hen a higher discount rate is assumed  Emission reductions are higher
under the higher discount rate assumption, because the higher discount rate  leads operators to delay installing
ne\\. lower-emitting units  The dela\  in introducing ne\\ units leads to higher baseline emissions, and the
need for greater reductions to reach the NOx SIP call cap
                                              Page 6-21

-------
                                              Table 6-15
       Effects of Alternative Discount Rate Assumptions on the Emission Reductions and Cost in 2007
                                for the Uniform 0.15 Trading Alternative

NOx Reduction
( 1 ,000 ozone season tons)
Incremental Annual Cost
(million 1990$)
Average Cost-Effectiveness
(1 990$/ozone season ton)
Expected Discount Rate
(6%)
938
$1,378
$1,468
Higher Discount Rate
(8%)
940
$1,414
$1,504
Incremental Effect
2
$36
$36
Source ICFanaKsis
        Effects of Lower SNCR Effectiveness

        This scenario assumes that, on coal-fired boilers emitting below 0.50 Ib/mmBtu. SNCR will reduce
NOx by 30 percent rather than 40 percent assumed in the rest of the analyses  Table 6-16 presents
technology choices under the expected effectiveness scenario and the lower assumed effectiveness scenario.
as well as the incremental change  Table 6-17 presents cost differences between the two scenarios  As
shown, the incremental cost of the 0 15 trading alternative rises (by about  11 percent) when lower SNCR
effectn eness is assumed  Shifts in the application of technology also take place with lower SNCR
effectiveness  Table 6-17 shows that  less capacity is retrofitted with SNCR when a lower effectiveness is
assumed,  and more capacity is retrofitted with SCR.  EPA examined  whether this potential increase in the
installation of SCR would be feasible  by 2003 and found that it would be feasible (EPA,  1998a)  Also shown
in Table 6-16 is a small increase in capacity of gas combined cycle units  Costs increase in part because of
the reduced cost-effectiveness  of the SNCR units, and the consequent need to substitute more expensive
control measures
                                              Table 6-16
             Emission Control Choices b> 2007 under Lower SNCR Effectiveness: 0.15 Trading
MW of Capacity
SNCR
SCR"
GasCC
Expected SNCR
Effectiveness
(40% reduction from
coal-fired units with
baseline NOx below
0.5 Ib/mmBtu)
133,792
63,267
20,443
Lower SNCR
Effectiveness
(30% reduction from
coal-fired units with
baseline NOx below
0.5 Ib/mmBtu)
84,761
93,638
20.997
Incremental Effect of Lower SNCR
Effectiveness
Total MW
-49,031
30,371
554
Percentage
Change
-36 %
47%
27%
Source 1CF anaKsis
1 In part because the increased need to install SCR is offset by reductions in SN'CR installation, EPA's analysis of the feasibility of installing control
technologies found that the necessary retrofits under this scenario could be accomplished See "Feasibility' of Installing XO\ C'ontrol Technologies b\
Ma\ 2003." U S EPA. July 1998
                                               Page 6-22

-------
                                             Table 6-17
         Ozone Season NOx Emission Reductions, Incremental Cost, and Cost-Effectheness in 2007
                            Under Lower SNCR Effectiveness: 0.15 Trading

NOx Reduction
(. 1 .000 o/one season tons)
Annual Cost
(million 1 990S)
A\ erage Cost-Lffccti\ eness
( 1 990$''o7one season ton)
Expected SNCR
Effectiveness
(40% reduction from
units with baseline NOx
below 0.5 Ib/mmBtu)
938
$1.378
$1.468
Lower SNCR
Effectiveness
(30% reduction from
units with baseline NOx
below 0.5 Ib/mmBtu)
938
$1.526
$1.626
Incremental Effect of
Lower SNCR
Effectiveness
0
$148
$158
Source  ICP anaKsi
        Effects of Shorter Equipment Life

        This scenario assumes that all equipment life is 15 years rather than 20 \ears as assumed in the other
anah ses The changes in the technologies used to comply with the NOx SIP call under this scenario arc
similar to those seen for the higher discount rate scenario  capital-mtensne technologies are used less, with
greater emphasis on dispatching changes  Table 6-18 presents cost differences between the two scenarios
As shown, the incremental cost of the 0 15 trading alternative \\ould rise b\ six percent if equipment life were
shorter than assumed b\  EPA
                                             Page 6-23

-------
                                             Table 6-18
         Ozone Season NOx Emission Reductions, Incremental Cost, and Cost-ErTecti\ eness in 2007
                            Assuming Shorter Equipment Life:  0.15 Trading

NOx Reduction
(1 .000 ozone season tons)
Annual Cost
(million 1990$)
A\erage Cost-Effectiveness
( 1 990S 'ozone season ton)
Expected Life
(20 years)
938
$1,378
$1.468
Lower Assumed Life
(15 years)
941
$1.461
$1.552
Incremental Effect
3
$8?
$84
Source 1CF anaKsis
Xumbers ma\ not sum due to rounding
        Effects of Alternative Demand Scenarios

        Tables 6-19 and 6-20 present the effects on projected cost and cost-effectiveness of changing
electricity demand forecasts, using the 0 15 trading alternatne as a basis for comparison  The first alternate e
scenario, shown in Table 6-19. assumes the full increase in demand projected by NERC. unlike EPA's
baseline assumptions, this alternatne does not allo\\ for reductions related to the Climate Change Action Plan
(CCAP)9  The second alternative scenario, shown  in Table 6-20.  assumes that retail competition (in addition
to the alread> assumed wholesale competition) occurs throughout the country  Three mam quantitatn e
effects of retail competition were modeled electricity price reductions, which will induce increases in
electricity demand,  initiation of time-of-day pricing, and increased retirement of nuclear generation units due
to their inabihu to be competitn e

        Both of these alternatne demand scenarios lead, in the Initial Base Case, to more electnciU
production from fossil-fueled units  Greater projected fossil-fueled production leads, in turn, to higher
emissions in the Initial Base Case, and the need for greater reductions to meet a gn en emissions cap  The
total cost and the  cost-effecm eness are therefore higher under the alternatn e demand scenarios
        9 The CCAP reductions are listed in Chapter 4, Table 4-1

                                              Page 6-24

-------
                                              Table 6-19
               Effects of Assuming Full NERC Demand/No CCAP Reduction on the Effects of
                                the NOx SIP Call in 2007: 0.15 Trading

Projected Demand
Full NERC Demand/
No CCAP Reduction
Increase Due to Altematn e
Assumption
NOx Reduction
(1,000 ozone season tons)
938
996
58
Annual Cost
(million 1990S)
$1.378
$1,502
$124
Cost-Effecth eness
(1990S/ton)
$1,468
$1.508
$40
Source  1CF anaKsis
Numbers ma\ not sum due to rounding
                                              Table 6-20
         Effects of Assuming Retail Competition—More Demand/Less Nuclear Power on the Effects
                               of the NOx SIP Call in 2007: 0.15 Trading

Projected Demand
Retail Competition -- More Demand.
Less Nuclear Po\\ er
Increase Due to Alternatne
Assumption
NOx Reduction (1,000
ozone season tons)
938
996
58
Annual Cost (million
1990S)
$1,378
$1.491
$113
Cost-Effectiveness
(1990S/ton)
$1.468
$1.498
$30
Source  ICF anaKsis
Numbers ma\ not sum due to rounding
6.4     Direct Economic Impacts

        This section presents the results of the cost analyses from the perspective of potential economic
impacts  Direct impacts, which are presented in this section, are those borne b\ the entities that potential!}
incur costs because they are required to reduce emissions  Indirect impacts, on the other hand, fall on entities
that are affected through their interactions with the directly affected entities  They are presented in
Section 6.5

        This section moves from the broadest level of impacts down to more specific assessments  Costs of
the rules relative to all electricity generation are presented first, followed by consideration of the potential
distribution of costs across t>pes of generators   Fmalh. potential impacts on small owners of electricity
generating units are summarized
                                              Page 6-25

-------
6.4.1    Costs Relative to Electricity Generation and Revenues

        Table 6-21 shows the potential impact of the compliance costs of the rule at the broadest level by
comparing them to the total amount of electricity generated annually  Annuahzed costs for the year 2007 are
shown in the third row of the exhibit for four uniform alternatives. These costs are then compared to
electricity generation to show costs in terms of mills (i e., tenths of cents) per kilowatt-hour. Also shown in
the table is the fact that generation in the SIP call region is lower under each of the alternatives than in the
base case  Because power production will be growing over time with or without the NOx SIP  call, producers
in the SIP call region will still generate more power m 2007 than in the current year — the effect of the NOx
SIP call will be to lower the rate of generation growth in the SIP call  region and shift some of it to nearb>
States where power is less  expensive to produce10. Furthermore, because some utilities will own capacity
both inside and outside of the SIP call region, some of the shifts in generation will represent a  shift within
corporations, rather than a shift in output from one group of firms to another
                                               Table 6-21
                      Generation Changes and Costs Compared to Generation in 2007
                             for Uniform Alternatives of Differing Stringency

Total Generation in SIP Call Region (millions MWhrs)
Percent of National Po\\er Generation
Costs Relame to Initial Baseline (billion 1990S)
Cost per Unit
(mills/kWh. 1990S)
Initial
Base
Case
2.075
56%
-
-
0.25
Trading
2.026
55%
064
031
0.20
Trading
2.0263
55%
095
047
0.15
Trading
2.018
55%
I 38
068
0.12
Trading
1,995
54%
1 85
093
Source ICF anahsis and U S Energ\ Information \dmimstration.Annual Energy Outlook 1998 December 1997
1 Generation is lower under the 0 20 altematne than the 0 25 alternatne. the difference is not apparent because of rounding
        These potential costs can be put into perspectne b\ comparing them to the typical revenues recened
b\ electricity suppliers  Table 6-22 sho\\s. for the same alternatives presented in the preceding exhibit, the
incremental per-kilo\\att cost of generation in comparison to an estimate of per-kilowatt-hour revenues
received by utilities and other suppliers in the SIP call region  Revenues, in turn, closely approximate the
total costs of supplying electricity to the end-use customer (including amortization of equipment and a return
on invested capital) "  Table 6-22 shows that the potential costs of the rule are  less than two percent of the
revenues of electricity suppliers for all of the alternatives, and climb above one  percent only for the 0 15 and
        1 ° EPA did not analyze the potential change in electric demand (it is held constant), rather, EPA analyzed the
change m the mix of suppliers that ma\ result from implementation of the SIP call

        1' Table 6-22 shows the costs of the NOx SIP call m 2007 in comparison to 1996 revenues from electricity.
Per-unit revenues can be expected to change over time (as a result of increased competition m the industry). so the
percentage impact of the rule will differ from the impact shown in the table
                                               Page 6-26

-------
0  12 alternatnes  Under traditional cost-of-sen ice regulation, this potential cost increase could be expected
to constitute the price increase as well
                                               Table 6-22
                              NOx SIP Call Compliance Costs by Alternative
                              Compared to Revenues from Electricity in 2007
                                                (1990S)

Average Per-unit Revenues. 23 Jurisdictions. Base Case
(mills/WrD'
Cost per Unit, all Generation in SIP Call Region (mills/kWh)
Cost as a Percentage of Revenues
0.25
Trading
6000
031
0 52%
0.20
Trading
6000
047
0.78%
0.15
Trading
6000
068
1 13%
0.12
Trading
6000
093
1 55%
Source Axerage re\enues per k\Vh calculated b\ 1CF using EIA Form 861 for 1996. other figures calculated b\ 1CF
'The table shoxxs the costs of the NO\ SIP call in 2007 in comparison to current rexenues from electricitx  Per-unit rexenues can be expected to
change oxer time (as a result of increased competition in the industn). so the  percentage impact of the rule xxill differ from the impact shoxxn in the
.-LI.
tabl
        Effects of Cost Changes on Electricity Producers

        Whether potential costs of the magnitude shoun in Tables 6-21  and 6-22 have a significant impact
on electncit) producers depends in part on whether the costs will be accompanied by offsetting price changes
In the past, because the electric power industn,- was tightly regulated, it was reasonable to assume that their
commissions would ha\ e app^ ed (perhaps after a lag) rate increases sufficient to co\ er the costs related to
emission control programs  Alternatively, part of the rule-related increases in operating costs may have been
reflected in fuel adjustment clauses The lack of competition helped to limit the reduction in demand resulting
from a rate increase  Utilities could thereby expect to continue receiving an adequate return on im ested
capital, and concerns about economic impacts were limited to the lag between cost and rate increases, and the
impacts of the rate increases on electricity demand

        More recentK. the restructuring of the industry and the prospect of competition among utilities and
non-utility producers has added uncertainty to the task of projecting  economic impacts on utilities
Competition at the wholesale level may complicate the process of passing on unusually high costs, while
greater uncertainty over the speed of retail deregulation increases the difficulty of projecting the price impact
on  customers  This section discusses some potential effects of the restructuring process on the recovery  of
the costs of the rule by electricity generators

        The potential effects of the NOx SIP call on the electric power industry will depend in part on the
timing of the rule relatn e to the progress of the restructuring process The NOx SIP call is scheduled to go
into effect by mid-2003  Competition at the wholesale level is  already underway and will be fully
implemented prior to 2003  Competition at the retail level is expected to spread widely by 2002. and to be
largel} complete by 2003 to 2005 (in the States where it occurs)  Thus, for most of the period in which the
NOx SIP call will impose costs, the assumptions of freely competitive markets should  apply.  The following
discussion of effects on utilities, therefore, begins from a free-market perspective  Following that discussion,
the effects of possible limits on competition early in the period  are discussed.
                                               Page 6-27

-------
        Effects of Cost Increases Under Competition

        In a world m which electricity prices are set by competitn e market forces at both the wholesale and
the retail level, theory predicts that the interaction of buyers and sellers will ensure that prices reflect the
marginal costs of generation (that is. the incremental costs of producing another unit of output)12 The need to
limit NOx even as electricity output increases means that marginal costs of generation will rise producing
more electricity under the NOx SIP call might require using more reagent for SNCRs as capacity factors
increase, or it might require the use of higher-cost fuels than are used under the Initial Base Case  Assuming
that the demand for electricity is relatively inelastic (which  is a reasonable consen ative assumption for the
short run), the price of electricity is predicted to rise by up to the amount that marginal costs rise u

        Because suppliers  in a competitive market are not required to produce at a loss, this increase in prices
can be expected to be high  enough to at least cover the variable costs  actually  incurred by every producer
Am  kilow att-hours that would have cost more to generate than the revenues they would bring in would not be
produced  The fact that price will be high enough to cover  variable costs, however, does not by itself mean
that there would be no impacts on generators.  Some generating units might have unusually high variable
costs for any given level of output; such units will not be used during time periods  in which the market price
of electricity is too low given their variable costs   Overall, generators in the SIP call region are predicted to
cut output during the ozone season b\ several percent, largeh from power plants that would use control
strategies with high marginal costs (e.g . combinations of SNCR and  allowance purchases)   The owners of
these units \\ould lose the net revenues they would have earned on  those kilowatt-hours of output that they
did not produce, if their State chose to implement the NOx SIP call as modeled b>  EPA

        A more significant issue is that the adjustment of prices in response to changes in marginal costs
does not guarantee that all  increases  m fixed costs will be covered  Fixed costs (such as the capital costs of
installing emission control  systems) do not affect the free market equilibrium unless they are large enough to
result in the retirement of plants as a way to avoid incurring unrecoverable costs   Short of early retirement.
which  IPM does not project for any significant amount of capacity, there is no  necessary connection between
fixed costs and price changes

         Some or all of the fixed costs of the rule can be recouped  if the price increases exceed the a\ eragc
change in variable costs  A relatnely large increase  in marginal costs could occur, for example, if generating
units that are a\ailable for  generating incremental power tend to be those with  high variable costs of control
(e g  . those with SNCR and a need to purchase allowances  for even, additional mmBtu of fuel used) Because
the change in the price of electricity will be determined by the increase in marginal costs for these marginal
units, the price increase could be higher than the average increase in variable costs
        12 The total price of electricity to consumers includes, in addition to the costs of energy, additional costs for
 transmission, distribution, and (where appropriate) charges for peak capacity  Capacity charges were found not to be
 affected significantly under the 0.15 alternative, and transmission and distribution charges are assumed not to change in
 response to the NOx SIP call  Thus, this analysis and discussion focuses on the marginal costs of producing energ\,
 which are expected to rise as a result of the NOx SIP call

        13 A price increase could result in a more significant reduction in the quantity of electricity demanded in the
 long run than m the short  run  A reduction in electricity demand would lower the total costs of the NOx SIP call, and
 limit the size of the price increase, while reducing the revenues received by electricity producers These possible
 impacts have not been analyzed for this report

                                                Page 6-28

-------
        This pattern appears to fit the case of the 0 15 alternative  Among coal-fired boilers, those units
projected to retrofit \\ith SNCR tend not to be used to the maximum extent possible (in part because their
costs of operation are high). B> contrast, the coal-fired boilers retrofit with SCR, which have lo\ver increases
in marginal cost, tend to be run almost continuous!}, and cannot contribute to further increases in output
Thus, increases in electricity output tend to come from the under-utilized units retrofit with SNCR. and the
marginal costs of electricity tend to reflect their high variable costs of control  Based on model results, it
appears that marginal-cost-based prices would rise by more than enough to cover in aggregate both the fixed
and variable costs of the rule

        There is. however, no guarantee or even expectation that the increased revenues will accrue in exact
proportion to increased costs  Rather, owners of units  for which emissions are unusually low in the baseline
(or for which costs of control are low) will tend to  gain when prices rise as a result of the rule Conversely.
owners of high-emitting units that are costh to control will not necessarily recover all of their increased costs
The additional costs might then be borne b\ the owners of those units

        Because of the restructuring process that is accompanying the shift from regulated to free market
conditions, the identities of the owners of the units is somewhat uncertain The traditional, vertically
integrated utiht> provides  all of the functions of generation, transmission, distribution, and marketing
services  Under competition, some of these functions are like!} to be split off  utilities might sell off their
generation capacity and/or turn over their marketing functions to other entities, possibly keeping their
transmission and distribution functions  The owners of the power plants, who will presumably be responsible
for reducing emissions and for holding sufficient allowances, might well be entities other than the utilities that
currenth own them

        Whether this change in ownership also means  that any particular!} high costs under the NOx SIP call
will fall on the ne\\  owners is less clear  To the extent  that unrecoverable costs of compliance can be clearly
predicted before the sale of the power plant, these  potential costs will most likely be considered full} in the
negotiations over the selling price  That is. unusually high-emitting plants will tend to have lower market
values than clean ones  Onh the costs that cannot be foreseen (resulting, perhaps, from unforeseen increases
in allowance prices) \\ill fall on the new owners  A significant portion of the costs of the rule might then be
borne b} the current owners of those generation assets  that will cost the most to control  Because most
utilities own a range of different t}pes of generation capacity, with van-ing baseline NOx rates,  some of the
\ anability in costs can be expected to be canceled  out  high costs for high-emitting generation will be
balanced in mam- cases by plants with low or zero emissions

        Effects of Cost Increases during the Transition to Competition

        As mentioned above, there may be cases in which electricity is sold at regulated retail prices for a
period after the implementation of the NOx SIP call The effects of the regulations will ultimately depend on
the decisions made b\ the utility regulators, and cannot therefore be predicted with assurance. Understanding
the reason for the continued regulation of retail electricity prices might narrow the uncertainty over their
possible effects

        One of the mam reasons that some retail price regulation will persist is to deal with the problem of
"stranded costs" — fixed costs that were incurred  under a regulated environment, for generation assets that
would not be able to cover their full costs in a free market  Regulatory responses to the problem of stranded
costs van- by SIP call region and situation In some cases, the issue of stranded cost recovery has been
separated from the issue of free market rates b}- including on utility bills a separate,  non-by-passable charge

                                               Page 6-29

-------
to cover stranded costs  The rest of the bill would be determined by the market, and could presumabh take
the effects of the NOx SIP call into account

        In other cases, utility regulators ha\ e agreed to fix retail prices for the electricity sold by utilities that
own these assets for a set period,  rather than allowing the generating prices that customers will ultimately see
to float down to free-market levels. If the utility is able to reduce its costs (of generation, or of purchasing
power from other suppliers), while continuing to sell at a fixed rate, it will be able to cover some portion of its
stranded costs  In some cases, the utilities agree to provide a discount from previous regulated rates, so that
consumers can realize some of the advantages of the free market even during the transition.  If the terms of
the agreed-upon rate freeze (including the size of the initial discount) do not take into account the additional
costs associated with the NOx SIP call, the utility might be in the position of absorbing the costs of the rule
for the length of the price freeze   State utility regulators may need to be cognizant of the potential effects of
the NOx SIP call on the need for stranded cost recovers
6.4.2   Potential Electricity Price Changes

        As discussed above, a reasonable basis for projecting price changes in response to cost increases is
the increase in marginal costs of electricity production  Marginal cost changes (that is. cost changes that van
with firm or industry output) are key because they immediately affect the market equilibrium, if demand can
be assumed to be relative!} inelastic, microeconomic theory strongly suggests that almost all of a marginal
cost increase will be quickly translated into increased prices  Because electricity demand has been relativeh
inelastic,  in the short run, the change in marginal costs resulting from the rules is a reasonable upper-bound
estimate of the change in electricity prices in the short run

        If all States implemented the NOx SIP call as illustratn ely modeled b> EPA. annual average
marginal  costs of electricity production could rise b\  1 0 mill/kWh in 2007. or about 1 6 percent of average
revenues  per kWh  Overall, the marginal cost changes in 2007 fall  in the middle of the changes over the
period from 2003 to 2010.

        Prices could rise by these same amounts as an upper bound estimate, to the extent that electnciU
demand is not completeh inelastic, prices would not rise as much If prices were to go up on the order of the
changes in marginal costs, total revenues to the electric po\%er mdustn. in the absence of a change in
generation, would rise by about $ 1 9 billion in 2007, and the net revenues to the industn after the increase in
costs \\ould be on the order of half of a billion dollars.  Estimating economic impacts is  made somewhat more
complex  by the fact that generation is projected to be lower in some years under the NOx SIP call than in the
base case, by 2 8 percent in 2007 for the 0.15 alternative, for example   This decline would require utilities
within the SIP call region to purchase more power from outside the SIP call region (or import it from
generating capacity they own outside the SIP call region), though this increased cost would be largely
balanced out by savings of the  costs of generation
 6.4.3   Distribution of Cost Impacts Across Generation Types

         Impacts on electricity producers will also depend heavily on the characteristics of their power plants
 Utilities with an unusually high percentage of capacity and generation from units that start out with low
 emission rates, or are relative inexpensive to control, will generally have lower costs per kilowatt-hour of
 electricity produced  These utilities might include those that have already been regulated under other

                                               Page 6-30

-------
regulations, and therefore have low baseline emissions  Other utilities may have a preponderance of small
coal-fired boilers with high baseline emissions rates  These utilities may have to install controls to reduce
baseline emissions rates, and then may still have to purchase allowances in order to comply

        Though costs per kilowatt hour can be different for every affected unit, it can be instructive to show
costs for typical units in the most important generator categories.  Table 6-23 presents IPM results for four
types of units  combustion turbines; gas combined cycle; oil/gas-fired boilers: and coal-fired boilers  Within
each type, results are shown for small and large examples, where sizes are selected as the lower and upper
quartiles of the size distribution. respective!)  The population is further subdivided into examples with low
initial NOx rates (at the lower quartile of rates) and high initial NOx rates (at the upper quartile)  Finally, for
the boilers, examples for cases with SCR and SNCR are displayed

        Costs, displayed in mills per kilowatt hours, are calculated by adding the cost of the control
technology (if any) to the number of allowances that could be sold, times an estimate of the value  of the
allowances  For the purposes of exposition, allowance prices are assumed to be $3.000 per ton, on the basis
of the marginal cost of NOx reductions m the 0 15 trading alternative over the analytical period   The number
of allowances that could be sold is calculated by comparing controlled emission rates to an estimate of the
allowances that would be allocated to the unit under the 0 15 alternative, and multiplying by an estimate of
the fuel input to the unit over the ozone season The total net cost of compliance, considering both control
measure costs and allowance costs or revenues, is then divided by estimated annual generation to yield an
a\ erage cost per kWh

        As seen in the table, typical combustion turbines and combmed-c\cle plants realize savings rather
than costs from the  rule (not counting admimstratn e or monitoring costs)  Because their emission rates are
typicall) low even in the baseline (due in some cases to previously installed control devices), they are not
assumed to be retrofitted with additional emission  control devices

        Oil and gas-fired boilers with lo\\ initial rates can experience savings analogous to those  for
combustion turbines and combined cycle units  Oil and gas boilers with high rates can have net costs, with or
without the addition of control technology  If electricity prices rise appreciably (in step with changes in
marginal costs, for example) as a result of the NOx SIP call, some owners of oil and gas-fired boilers would
be better off because their control costs \\ould be lower than the industry-wide increase in marginal costs

        Coal-fired boilers, which provide the majorit)  of fossil generation, can have costs in the range of one
mill per kilowatt hour This cost is comparable to, though somewhat higher than, the average costs for all
generation under the 0.15 alternative That cost increase, of 0 68 mills/kWh. is shown in Table 6-21  As
shown in Table 6-23, costs can be expected to be higher for smaller units  (10-16 mills/kWh) than for  large
units  (04-10 mills/kWh).  Costs will also tend to be higher for those units with higher baseline rates
(including some Group 2 boilers, which were not required to reach low rates under Title IV)

        Analysis of the IPM results also shows changes in capacity factors for some of the typical units  in
response to the NOx SIP call. Units that employ SNCR are most likely to reduce their capacity factors, and
those with low controlled rates (either coal with SCR or gas turbme/CC) are likely to increase their capacity
factors  As discussed above, this pattern leaves the marginal units more likely to have high marginal costs of
generation, because the units with available capacity face additional costs of purchasing reagent and
allowances when they increase generation  The units that reduce their capacity factors will lose the revenues
that would ha\ e accompanied their lost output; on the other hand, they also save their variable costs of
operation for those kilowatt hours

                                             Page 6-31

-------
                                              Table 6-23
  Potential Net Cost (After Allowance Purchases/Sales) by Unit Type under 0.15 Trading Alternative in 2007
                                          (mills/kWh, 1990S)

Unit Type
Combustion
Turbine
Gas Combined
C\ cle
Oil/Gas Fired
Boiler
Coal Fired Boiler
Small Units
Low Initial
NOx Rate
(Unaffected)
-1 3
00
1 0
(SNCR)
High Initial
NOx Rate
(Unaffected)
-03
08
1 6
(SNCR)
Large Units
Low Initial NOx Rate
-1 3
-20
-04
D (SNCR)
09 04
(SCR) (SNCR)
High Initial NOx Rate
-04
-03
05 01
(SNCR')
10 05
(SCR) (SNCR)
Source ICF analysis
6.4.4   Potential Impacts on Small Electricity Generators

        To investigate the possibility that small utilities and other small affected entities could be ad\ erseK
affected by the NOx SIP call. EPA has conducted a screening analysis of small entity impacts   That analysis
re\eals that a relative!} small number of small utilities are potentially affected in this anal} sis. in part because
coverage is limited to units greater than 25 M\V. and m part because small utilities are more common in the
\\estern states that are outside the 23 jurisdictions named in the SIP call  Of almost 900 utilities nationwide
that generate electricity. o\ er 700 are considered small by SBA's definition (of less than 4 billion kWh per
}ear) Fewer than 250 of these small utilities are found in the SIP call region, however, and of these onh
about 190 own fossil-fuel fired units Excluding those utilities that have no units greater than 25 M\V lea\ es
41 small potentially affected utilities

        Though mam  of these small utilities will be affected to a minor degree onh. about half ma}
experience cost increases that are greater that one percent of their electricity-based revenues  under EPA's
illustrative implementation scenario  The small utilities that may be more serious!} affected tend to be those
relying more hea\ ily on coal-fired boilers, especial!} cyclones (which tend to have high uncontrolled
emissions and are not subject to tight controls under Title IV), and those with units whose baseline NOx
emission rates are unusually high  While these utilities constitute almost half of affected small utilities, the}
are less then ten percent of the small utilities that may be affected in the absence of the size cut-off
established by EPA to limit impacts on small sources

        In a search for small non-utility generators. EPA identified approximately 100 affected units that
generate electricity but are not owned by a utility.  The owners of almost all of these units are identified using
a data base of non-utility generators. Data collected on the revenues, SIC codes, and total generation of the
owners or their parent companies are used to divide the owners into small and large entities  Those for whom
data on size are unavailable are assumed to be small in order to estimate a conservative "worst case"
scenario  In all.  73  small non-utility entities  with units greater than 25 MW are analyzed
                                               Page 6-32

-------
        Estimated costs of compliance are calculated under the conservative assumption that all small non-
 utilit)  units comply through the purchase of allowances  This approach \\ould tend to overstate compliance
 costs because it does not consider cases in which emission reductions can be achieved at costs below the
 marginal cost of reductions in the SIP call region  In the 0.15 trading alternative. 12 entities judged to be
 small are projected to face costs in excess of one percent of revenues under EPA's illustrative implementation
 scenario  These 12 entities constitute about 16 percent of the 73 small non-utilities anahzed

        Adding 20 small utilities to 12 small non-utilit} entities yields a total of 32 small entities in the
 trading program with projected costs in excess of one percent of revenues   These 32 entities constitute about
 27 percent of all small affected entities, though it would be an even smaller percentage if a 25 MW cut-off
 was not used
6.4.5   Potential for Closures and Additions of Capacity

        A potential!) important measure of the economic impacts of a rule is the number of potential closures
predicted to result from the rule  Closures occur when the costs of compliance are so high as to make the net
present \ alue of future  operation negatn e, leading to the abandonment of a productn e asset as the least-
costly alternative  New installations of capacity induced by the rule constitute another important measure

        The results of the IPM anahsis sho\\ that some capacity could be shut down or retired early as a
result of the NO\ SIP call, if the  States implemented the SIP call as EPA modeled it  At most. 183 MW of
coal-fired capacih and  151 MW  of oil/gas-fired capacity out of a total of over 230.000 MW, are projected to
close  Thus, all but 0 7 percent of capacih would continue to operate under the NOx SIP call  These
potential closures \\ill be more than offset (in terms of capacity) by an increase in combmed-c\ cle units of
between 1.798 and 4.156 MW (depending on the  alternative)
6.5     Indirect Economic Impacts

        In addition to impacts on the entities that are potentially directly affected b\ the rules, there will be
some impacts on sectors of the econorm that interact with the electricity generating industry  This section
bnefh  examines the potential effects on fuel suppliers, industrial users of electnciU. and households


6.5.1    Potential Employment Impacts

        Emission control devices will ha\ e to be installed as a result of the NOx SIP call  Thus, the rule will
generate an initial demand for workers to install emission control technology and a continuous demand for
workers to operate and maintain the technology  Tables 6-24 and 6-25 present the potential impact on
emplo\-ment in the control technology sector for the 0 15 trading alternative
                                             Page 6-33

-------
                                              Table 6-24
              Potential Impact on Employment in the Control Technology Sector: 0.15 Trading
                                     (Construction and Installation)
Labor Required for Construction and Installation
2000 - 2003
Combustion
Controls
(worker-years)
6,511
SCR
(worker-years)
21,535
SNCR
(worker-years)
14,312
Total
(worker-years)
42,358
Increased Annual
Labor Demand,
assuming construction
and installation take
three years (FTEs)
14.119
Source ICFanahsis
                                              Table 6-25
Potential Impact of 0.15 Trading Alternative on Labor Requirements in the Control Technology Sector (O&M)
                                                in 2007
                                         (Full-Time Equivalent)
SCR
316
SNCR | Total
669
985
Source ICF anahsis
No additional O&M assumed to be required for combustion controls
        The NOx SIP call may also affect demand for labor in the coal and natural gas sectors  Coal
produced is hkeh to decrease while natural gas production is likely to increase  The resulting decrease in the
demand for coal workers (which amounts to less than one percent of total coal mining emplo>Tnent) and
increase in the demand for natural gas workers are presented in Tables 6-26 and 6-27  EPA did not estimate
the potential additional indirect impact on labor demand for  coal transportation workers, but it is expected to
be smaller  the than potential changes in production workers
                                              Table 6-26
      Potential Effects of 0.15 Trading Alternative on Coal Production and Employment Demand in 2007



Eastern U S
Western U S

Total Nationwide
Coal Production,
Initial Base Case
(million tons)
511 7
5137
Change in Coal
Production in
Response to NOx
SIP Call, 0.15
Alternative
(million tons)
-46
-2.6


Labor Hours
(thousands)
-809
-113


Change in Labor
Requirement (FTE)
-392
-44
Source ICF analysis  Assumes growth in output per worker to 11.734 tons yr for eastern miners, 58.433 tons yr for western miners See base
assumptions and EPA. June 1997
                                               Page 6-34

-------
                                              Table 6-27
  Potential Effects of 0.15 Trading Alternative on Natural Gas Production and Employment Demand in 2007
Initial Base
Natural Gas Use
(billions of cubic feet )
1.928
Increase in
Natural Gas Use
(billions of cubic feet )
80
Labor Demand
(FTE per billion cubic feet
per year)
748
Change in Labor
Requirement
(FTE)
600
Source ICF anahsis
        Table 6-28 presents the potential overall impact on labor demand in 2007.  The increase in demand
for control technology construction and installation workers is not taken into account in calculating the net
change in labor demand because these \\orkers are needed only during the construction and installation stage
(the first three years)  As shown, the labor demand in 2007 is likely to increase b\ 1.149 workers as a result
of the NOx SIP call
                                              Table 6-28
             Summary of Potential Labor Demand Impacts of 0.15 Trading Alternathe in 2007
Market Segment
Coal Production - East
Coal Production - West
Natural Gas Production
Emission Control Technolog\ (O&M1 a
Net Change
Change in Labor Requirement (FTE)
-392
-44
600
985
1,149
        Source ICFanaKsis
        ' There will also be a labor requirement equivalent to 14.119 TTLs. per \ear for pollution control astern installation hetueen
        2001 and 2003
6.5.2   Potential Impacts on Industrial Users of Electricity

        The potential costs of the NOx SIP call are expected to be passed along to electnciU users through
rate increases, as discussed in Section 6 4  Whether the rate increases significantly affect industrial users
depends both on the size of the increases and the amount of electricity used, relative to industrial output
Total net electricity use by manufacturing sectors in 1994 was 2.656 trillion Btu, which equals 778 billion
kWh The value of total shipments from the manufacturing sector in 1995 was $3.119 billion (in 1990
dollars), of which $1.485 billion represented value  added (as opposed to the value of purchased materials and
other inputs)  Thus, on average, the manufacturing sector used 0.25 kWh of electricity per dollar of output
(or 0.52 kWh per dollar of value added)   If the price of electricity rises by  1.0 mill/kWh in 2007. which is the
generation-weighted increase in marginal costs, industry would experience a cost increase of 0.25 *
                                              Page 6-35

-------
$10/1000 or a 0 03 cents for each dollar of shipments, or 0 05 cents for each dollar of value added The cost
increases, then, would be a fe\v hundredths of a percent of the value of output on average u

        These calculations refer only to the average manufacturing entity.  Some industries use considerably
more electricit) than others, meaning that the potential impacts could be more serious for some entities.
Table 6-29 shov>s the electricity use, value added, and value of shipments for the six two-digit manufacturing
industries that use the most electricity per dollar of value added Table 6-29 also shows the effect of a 1 0
mill increase in the electricity prices per kWh as a percentage of the two output measures  As seen, even the
costs for the most electricity-intensive two-digit industries is below a quarter of one percent  Table 6-30
shows the electricity demand, value of shipments and the effect of an increase in electricity prices per kWh as
a percentage of value of shipments for the four-digit manufacturing industries that use the most electricity per
dollar of output. A total of nine industries at the four-digit level had costs greater than the industry with the
highest costs at the t\\o-digit level. Even for these few electricity-intensive industries, cost impacts would be
less than one percent
         M ] 997 Statistical Abstract of the United States, U.S Department of Commerce, Bureau of the Censuv 1 ablcs
 1219 and 930, pp  739 and 587

                                               Page 6-36

-------
                                                      Table 6-29
      Potential Impacts of Electricity Rate Increases in 2007 on Energy-Intensive Industries of .15 Trading
                                                     Alternative
                                          (by Two-Digit SIC Code, 1990S)





Industry
Pnman Metals
Petroleum and
Coal Products
Textile Mill
Products
Stone. Cla\. and
Glass Products
Paper and Allied
Products
Chemicals and
Allied Products
All Other
Manufacturing




SIC
Code
33

29

22

32

26

28



Net
Electricity
ITco
(billion
kWh)
144

35

. 33

36

65

152

311
Total
Output of Industry
(billion 90S)

Value
Added
$61

$28

$29

$37

$^0

$ni

$1.091
Value of
Shipments
$156

$131

$70

$66

$150

$315

$2.231
Electricitj Used per
Dollar of Output
(kWh/90S)

Value
Added
237

1 26

1 15

099

093

089

028
Value of
Shipments
092

027

047

055

043

048

0 14
Potential Increase of
1 mill/kWh' as a
Percentage of Output

Value
Added
0 00%

0 13%

0 12%

0 10%

0 09%

0 09%

0 03%
Value of
Shipments
0 09%

0 03%

0 05%

0 05%

0 04%

0 05%

001%
««JKe  !9y~ Statistical Abstract table 930 and 1219. pp 58". 739-743. and ICF calculations One k\Vh assumed to be 3 412 Btu
' Potential increase under full competition in the electric pouer mdustr> uith marginal cost pricing  Actual increase mas be less if full competition
does not occur (under cost of semce pricing, the increase is estimated to be 0 7 mill k\Vh)
                                                      Page 6-37

-------
                                               Table 6-30
     Potential Impacts of Electricity Rate Increases in 2007 on Energy-Intensive Industries of .15 Trading
                                               Alternative
                                     (by Four-Digit SIC Code, 1990S)
Name
Industrial Gases
Electrometallurgical Products,
Except Steel
Industrial Inorganic Chemicals.
NEC
Cement. Hvdraulic
Pnman Smelting and Refining of
Nonferrous Metals
Lime
Paper Mills
Glass Containers
Steel Works, Blast Furnaces
(Including Coke")
SIC Code
2813
3313
2819
3241
3339
3274
2621
3221
3312
Electricity
Demand
(million kWh)
22.816
4,797
42,786
10.789
4,205
1,151
37.503
4.268
45.463
Value of
Shipments
(million
1990S)
$3,051
$1,020
$14,530
$5,125
$2,509
$805
$33.485
$3.825
$45.587
Electricity
Demand per
Dollar of
Output
(UWH/90S)
748
470
294
2 11
1 68
1 43
1 12
1 12
1 00
Potential
Increase of 1
mill/kWh" as
a Percentage
of Value of
Shipment*
075%
0 47%
0 29%
021%
0 17%
0 14%
0 11%
0 11%
0 10%,
Source EI\. 1996 Eleciric Sola, and Revenue. Dec 1997. Bureau of Economic .Analyses. Shipments of M
1 Potential increase under full competition in the electric power industry with marginal cost pricing  Actual r
does not occur (under cost of ser% ice pricing the increase is estimated to he 0 7 mill k\Vh)
anufactunng Industries www bea doc go\
increase ma\ he less if full competition
6.5.3   Potential Impacts on Households

        Impacts on household budgets \\ill be smaller than the percentage increase m electricity prices.
because electricity is onh one component of expenditures  Households used 961 billion k\Vh of electricity in
1993. or an a\erage of almost  10.000 kWh per household " An increase of 1 mill/kWh would add about $10
annually to the average household budget  As median income per household was over $29.500 in 1995 (in
1990 dollars),  the typical increase m electricity cost would take an additional 0.03 percent (that is. a thirtieth
of one percent) from the income of a typical household 16

        The impacts would be higher for households with unusually high electricity demand or unusually lo\\
incomes  Table 6-31 shows typical annual electricity bills for households in different parts of the income
distribution, and the effect that an increase in electricity prices of 1 0 mill would have on their incomes
        151997 Statistical Abstract, Table 929. p 587

        16 1997 Statistical Abstract, Table 719. p 466
                                                Page 6-38

-------
                                               Table 6-31
          Potential Impacts of Electricity Rate Increases in 2007 on Households by Income Category
                                        of .15 Trading Alternative
                                                 (1990S)



Income Categorv
Less than $15, 000
$15.000 to $34.999
$35.000 to $74.999



Assumed Annual
Income
$7,500
$25,000
$55.000



Typical Annual
Electricity Bill
$544
$704
$832


Electricity as a
Percentage of
Income
7.3%
28%
1 5%
Potential
Increase of 1
mill/kWh' as a
Percentage of
Income
0 12%
0 05%
0 03%
Source 799  Siaiisnca! Abstract. Table 720. p 467. and ICF calculations, assuming an average price of electricity of 6 cents per k\Vh
' Potential increase under full competition in the electric power industry with marginal cost pricing Actual increase ma\ be less if full competition
does not occur (under cost of service pricing, the increase is estimated to be 0 7 mill \\Vh)
        Table 6-31 shows that electricity use rises with income, but not in direct proportion  households with
very low incomes spend almost as much as those with substantially higher incomes  Thus, electncih takes a
larger percentage of income from the poor than from the rich Because of the size of the increase in electricity
prices expected to result from the  NOx SIP call is small, the effects on even the least-well-off households
will be much smaller than one percent  Most households are likely to see a net reduction in electnciK rates
over the coming decade as a result of the savings from restructuring, despite the increases from the NOx SIP
call

        Variations in impacts on residential  consumers in different parts the SIP call region are also
estimated under EPA's illustrative implementation scenario  Because per-household electricity use vanes
from State to State, costs impacts by State vary as well  The 1996 median electricity expenditures as a
percentage of household income was 2 1 percent across the entire SIP call region   For six States, electricity
expenditures as a percentage of income was less than  1 5 percent, consumers in these States are likely to have
relati\ ely smaller cost increases  On the other hand, the ratio of electricity expenditures to income was
greater than 3 percent in a total of five States, with the highest ratio reaching 3 9 percent in Tennessee  The
NOx SIP call could have somewhat  larger effects on consumers in these States, because am given increase in
electricity prices will be applied to larger portions of their incomes  Even in these States, though,  the price
increases likely to result from the NOx SIP call will be less than a tenth of one percent of income.1
          1997 Statistical Abstract, Table 722. p 468  El A Electric Sales and Revenue, 1996, Tables 5 and 6.
p  17-18
                                                Page 6-39

-------
6.6     Administrative Costs

        Administrative Costs to Electricity Generating Units

        Administrative costs to operators of electricity generating units are associated with monitoring NOx
emissions, reporting compliance information, permitting, and allowance trading.  Table 6-32 presents the
administrative costs to electricity generating units for emissions monitoring, reporting, and permitting  The
costs presented are incremental to costs in the Initial Base Case.  EPA estimates that the average unit costs
for emissions monitoring, compliance reporting, and permitting are $3.193, $621, and $353, respectively.
The  emissions and monitoring costs include capital and ongoing operations and maintenance costs. The
emissions monitoring unit costs to sources vary depending upon the source type and on the monitoring
requirements that the given source is already meeting  A unit that is already subject to both the Title IV
monitoring requirements and the requirements of the SIP call trading program will not have any additional
administrative burdens imposed by the program, while a coal unit that is not currently subject to any
monitoring and reporting requirements will incur administrative costs
                                              Table 6-32
                    Potential Administrative Costs for Electricity Generating Units, 2007
                                               (1990S)

Average Unit Costs
Annuahzed Costs'
Emissions Monitoring
$3,193
$4.033
Compliance Reporting
$621
$785
Permitting
$353
$445
 'EPA assumes 1.263 units ha\e incremental compliance requirements
 Source ICFAnaKsis
 Numbers ma\ not sum due to rounding


          The number of units determines the need for permits and reporting  All 1.995 units will incur
 compliance reporting and permitting costs  A permit is required for each unit every five years The unit
 permitting costs in Table 6-32 may appear to be low because they have been annuahzed over the five year
 permit cycle  EPA assumes that these costs will not vary among the regulatory alternatives considered
 because the number of units does not vary much by alternative

         Table 6-33 presents the potential transaction costs to owners of electricity generating units for
 trading allowances under the different regulatory alternatives. Transaction costs are estimated to be 1 5
 percent of the total values of the traded allowances
                                               Page 6-40

-------
                                              Table 6-33
              Potential Allowance Trading Transaction Costs for Electricity Generating Units
                              for the NOx SIP Call bv Uniform Alternative
Alternatives
(Ibs/mmbtu)
0 25 Trading
0 20 Trading
0 1 5 Trading
0 1 2 Trading
Transaction Costs'
(million 1990S)
$08
$1 0
$1 7
$33
             1 Total cost of allowance is included in compliance costs
             Source ICFanaKsis
        Tola! Administrative Costs
        Table 6-34 presents the potential total annualized administrative costs to owners of electricity
generating units by alternate e  The costs presented are incremental to the Initial Base Case  The total
admmistratn e costs include monitoring, compliance reporting, permitting, and allowance trading transactions
costs
                                              Table 6-34
                            Potential Total Administrative Costs to Owners of
                                  Electricity Generating Units in 2007
                                            (million 1990S)
Alternatives (Ibs/mmbtu)
0 25 Trading
0 20 Trading
0 1 5 Trading
0 1 2 Trading
Total Annualized
Administrative Costs
$53
$55
$62
$78
             Source ICFanaKsis
             Numbers ma\ not sum due to rounding
6.7     References

Bureau of Economic Analyses. Shipments of Manufacturing Industries, w\v\v bea.doc gov

Energy Information Administration. 1997.  1996 Electric Sales and Revenue  December 1997
                                               Page 6-41

-------
Federal Register. 1986  Emissions Trading Policy Statement. General Principles for Creation. Banking
and Use of Emission Reduction Credits Vol 51. No 233  43814-43858.

Federal Register. 1997 Finding of Significant Contribution and Rulemalang for Certain States in the
Ozone Transport Assessment Group SIP call region for Purposes of Reducing Regional Transport of
Ozone Vol 62. No. 216 60318-60418

Federal Register. 1998  Supplemental Notice for the Finding of Significant Contribution and Rulemalang
for Certain States in the Ozone Transport Assessment Group SIP Call region for Purposes of Reducing
SIP Call regional Transport of Ozone.Vol 63: 17349-

Statistical Abstract of theUmted States, 1997

US Energy Information Administration. 1997   Annual Energy Outlook 1998  December 1997

U S Environmental Protection Agency. 1997a  Proposed Ozone Transport Rulemalang Regulatory
Analysis.  Office of Air  and Radiation. Washington. D.C. September 1997

U.S Environmental Protection Agency. 1997b   "Model for Estimating Emplo>Tnent Changes from Air
Pollution Control Strategies for the Electric Power Industry '" Office of Atmospheric Programs. Washington.
D.C.  June 1997

U S Environmental Protection Agenc\. 1998a  Feasibility of Installing NOx Control Technologies by May
2003 Office of Atmospheric Programs. Acid Rain Dnision.  Washington. D C  September 1998

U S Environmental Protection Agency. 1998b. Information  Collection Request for the Proposed Federal
Implementation Plan Office of Atmospheric Programs. Acid Ram Di\ ision. Washington. D C   September
 1998
                                            Page 6-42

-------
     Chapter 7. RESULTS OF COST, EMISSIONS REDUCTIONS, AND ECONOMIC IMPACT
                   ANALYSES FOR NON-ELECTRICITY GENERATING UNITS
        This chapter presents the results of the cost and economic impact analyses for industrial boilers.
 combustion turbines and other stationary sources   The cost and economic impact analyses (including
 potential impacts to small entities and public sector entities) evaluate the potential impacts associated with
 this SIP call, based on assumptions about how the States will implement the requirements associated with
 meeting their NOx budgets  In an effort to narrow the scope of sources potentially affected by this rule, onh
 large sources are potential!) affected (see Chapter 3)  In addition, the Agenc> considered other factors to
 reduce the number of sources including av ailabihty of control measure data, source category emissions
 relative to total baseline NOx emissions in the SIP call region, and the level of baseline emissions control
 Additional source  categories are eliminated from further consideration based on an analysis of the average
 cost-effectiveness  of emissions control for the source category considering control, monitoring, and
 administrative costs Prior to the cost-effectiveness analysis, the remaining groups of source categories are
 six  (1) industrial  boilers and combustion turbines. (2) stationarv internal combustion (1C) engines. (3)
 cement kilns, (4) process heaters.  (5) glass manufacturing operations, and (6) commercial and industrial
 incinerators  The results of the average cost-effectiveness analysis for each of these remaining source
 category groups are found in Section 7.1. Section  7 2 describes the potential economic impacts of the
 preferred option and selected other options, and  Section 7.3 analvzes the potential impacts  on small entities in
 particular  The economic impact sections focus  on detailed results for the preferred options and provides a
 brief comparison with the range of results for other options considered  More detailed results for the other
 options considered are provided in the economic impact analysis report (Abt. 1998 )  FmalK. references for
 the chapter are pro\ ided in Section 7 4
 7.1     Compliance Costs and Cost-Effectiveness

        Under the final NOx SIP call rulemakmg. four regulatory alternatives are analv/.ed for the trading
 sources and fiv e other alternate es are analyzed for the sources that are not in the trading program   Impacts
 are estimated as \\ell as emissions reductions for the large (as defined in Chapter 3) trading sources
 (industrial boilers and combustion turbines) at regulatory alternatives based on 40%. 50%. 60%. and 70%
 reduction of NOx. respectiv eh. from projected 2007 uncontrolled emissions  These alternatives applv the
 specified control level across the entire SIP call region Impacts are estimated for sources not in the trading
 program under fi\e cost per ton regulator* alternatives $1.500. $2.000. $3,000. $4.000. and $5.000 ' The
 potential costs of complying with the SIP call have two elements implementation (the cost of emissions
 control), and administration (the cost of monitoring emissions, and the associated administrative costs of
 recordkeepmg. and reporting ):  The calculation of administration costs is presented later in Section 7 1 8.
        i
          Analysis of sources in the trading program consider trading occunng across the entire SIP call region
Analysis of sources not in the trading program considers at the source-level applying controls up to a specified cost per
ton cutoff  For more details, refer to Chapter 5

          One categon of costs - the transaction costs associated with trading allowances -- is not included in the cost
estimates discussed in these chapters These costs will depend on the number of sources that elect to engage in trading
Anah sis of HGUs with IPM indicates these costs are 1 5% of compliance costs, so thev are expected to be small

                                              Page 7-1

-------
        For each affected source categon. EPA's estimates of emissions reductions and compliance costs
reflect the Agency's framework for highly cost-effective NOx emissions reductions  These proposed control
measures are selected through a 2 step process  First. EPA examined their technical feasibilitA.
administrative feasibility, and average cost-effectiveness for NOx control applied in the ozone season across
the SIP call region  EPA then determined those measures feasibly achieve the greatest NOx reductions and
are among the most reasonable in light of other actions undertaken or proposed by EPA and States to control
NOx  Based on this process, the Agency considers controls with an average cost-effectiveness, evaluated
across all sources in a categon' group, of less than $2.000 per ozone season ton of NOx removed to be highh
cost-effective and has calculated the amounts of emissions that States must prohibit based on application of
these controls
7.1.1   Results for Industrial Boilers and Combustion Turbines

        In EPA's anah sis. large industrial boilers and combustions turbines are included in the NOx Budget
Trading Program  These sources will be allowed to participate in this interstate emissions trading program if
States elect to include these sources in this program. Currently, the IPM model, discussed in Chapter 4. does
not cover these sources, so EPA has conducted a least-cost analysis for this group of sources  The least-cost
anah sis is EPA's attempt to simulate the outcome of an efficient emissions trading program by assigning
control responsibility based on sources with the lowest control costs  The least cost anah sis only reflects the
efficient allocation of control responsibility among the group of industrial boilers and turbines, and does not.
therefore, take ad\antage of potentially more efficient outcomes that could occur if these sources were
modeled in conjunction \\ith the rest of the utihu sources included in the NOx SIP Budget Trading Program

        Table 7-1 sho\\s the emissions reductions achie\ed in the least-cost anah sis for each regulators
alternatne  The table indicates that the alternatives achieve incremental reductions from the 2007 Clean Air
Act  (CAA) baseline ranging from 31% to 66%
                                               Table 7-1
                  2007 Ozone Season NOx Baseline Emissions and Emission Reductions for
                            Large Industrial Boilers and Combustion Turbines3
Regulator*
Alternate eb
40% Control
50% Control
60% Control
70% Control
Number of
Affected Sources
292
592
803
805
2007 Baseline
Emissions
194,445
194.445
194,445
194,445
2007 Post-Control
Emissions
133.630
108,880
90,193
65.611
2007 Emission
Reductions
60,815
85,565
104,252
128,834
2 The 2007 baseline emissions estimate reflects emissions from all 807 large sources (755 industnal boilers. 52 combustion turbines) in this source
category, both controlled and uncontrolled Emissions estimates for 90 non-fossil fuel fired industrial boilers are not included
b Reductions from controlled 2007 baseline are less than the nominal percentage reduction from an uncontrolled 2007 baseline indicated in the
regulatory- alternative name
                                                Page 7-2

-------
        Table 7-2 shows the annual costs and resulting average cost-effectiveness for each regulator}
alternate e  The annual control costs range from $49.5 to $249 5 million. Annual monitoring and
administrate costs depend on the number of covered sources, and since the number of sources is constant
across the alternatives, this cost is also constant at $26 1 million  The accompanying average cost-
effectiveness results range from $1.243 to $2,140 per ozone season ton   The 60% control level is the most
stringent control level that meets EPA's framework for highly cost-effective ozone season NOx emissions
reductions,  and is selected as the basis for establishing State level emissions budgets
                                              Table 7-2
                           2007 Cost and Cost-Effectiveness Results for Large
                              Industrial Boilers and Combustion Turbines
Regulator}
Alternate e
40% Control
50% Control
60% Control
"0% Control
Annual Control
Cost
(million 1990S)
$495
876
1268
2495
Annual
Monitoring and
Administrative
Costs
(million 1990S)
$26 1
26 1
26 1
26 1
Total Annual Costs
(million 1990S)
$756
1137
1529
2757
Ozone Season Cost
Effecti\ eness
(S/ozone season
ton)
$1.243
1.329
1.467
2.140
7.1.2    Results for Internal Combustion (1C) Engines

        The anak sis of large internal combustion engines is conducted b\ selecting the most cost-effecti\ e
control measure a\ailable for each identified source that does not exceed the cost-effectn eness cut-off
specified in the regulator-} alternative  Table 7-3 shows the emissions reductions achieved in the analysis for
each regulator} alternate e  The table indicates that the alternates achie\e incremental reductions from the
2007 controlled baseline of rough!} 89%
                                              Page 7-3

-------
                                               Table 7-3
                          2007 Ozone Season NOx Emission Reductions for Large
                                         Stationary 1C Engines'
Regulators
Alternative
$l,500/ton
$2.000/ton
$3.000/ton
$4,000/ton
$5,000/ton
Number of
Affected Sources
290
304
304
304
304
2007 Baseline
Emissions
92.424
92.424
92.424
92,424
92.424
2007 Post-Control
Emissions
9.857
9.840
9.840
9.840
9,801
2007 Emission
Reductions
82,567
82.584
82,584
82.584
82.623
a The 2007 baseline emissions estimate reflects emissions from all 305 large sources in this source category both controlled and uncontrolled
        Table 7-4 sho\\s the annual costs and resulting average cost-effectiveness for each regulator.
alternatn e?  All of the regulator.- alternatives achieve similar results and all reflect control measures that
meet EPA's framework for highh cost-effectn e ozone season NOx emission reductions  EPA has selected
the S5000/ton regulator, alternative as the basis for establishing State-level emissions budgets since this
alternatn e provides the greatest emission reduction while being consistent with EPA's framework for highK
cost-effectn e ozone season emissions reduction  This alternatn e results in an  average reduction of 90% from
an uncontrolled 2007 baseline
         3  It should be noted that the monitoring and admmstratn e costs estimated for these sources are
 overstated, since they reflect application of the provisions of Part 75 (primarily, installation of CEMs)
 Stationary 1C engines will be allowed to comply with the less stringent provisions of Part 60

                                                Page 7-4

-------
                                               Table 7-4
                            2007 Cost and Cost-Effectiveness Results for Large
                                         Stationary 1C Engines
Regulatory
Alternative
$1.500/ton
$2.000/ton
$3,000/1011
$4.000Aon
$5.000/ton
Annual Control
Cost
(million 1990S)
$869
869
869
869
S"7 1
Annual
Monitoring and
Administrative
Costs
(million 1990S)
$124
13 3
133
133
133
Total Annual Costs
(million 1990S)
$993
1002
1002
1002
1004
Ozone Season Cost
Effectiveness
(S/ozone season
ton)
$1.203
1.213
1,213
1.213
1,215
7.1.3    Results for Cement Manufacturing (Cement Kilns)

        The anahsis of cement manufacturing operations is conducted b>  selecting the most cost-effective
control measure available for each identified source that does not exceed the cost-effectiveness cut-off
specified m the regulator* alternatn e  Table  7-5 shows the emissions reductions achieved in the anah sis for
each regulators alternative  The table indicates that all the alternatives achieve the same incremental
reductions from the controlled 2007 baseline  This reduction is approximately 38%
                                               Table 7-5
                         2007 Ozone Season NOx Emission Reductions for Large
                           Cement Manufacturing Operations (Cement Kilns)"
Regulator}
Alternathe
$l,500/ton
$2.000/lon
$3.000/ton
$4,000/ton
$5.000/ton
Number of
Affected Sources
57
57
57
57
57
2007 Baseline
Emissions
42.701
42.701
42,701
42.701
42.701
2007 Post-Control
Emissions
26.312
26,312
26,312
26.312
26,312
2007 Emission
Reductions
16.389
16.389
16389
16.389
16,389
 The 2007 baseline emissions estimate reflects emissions from all 58 large sources in this source categop,'. both controlled and uncontrolled
        Table 7-6 shows the annual costs and resulting average cost-effectiveness for each regulatory
alternative  The annual control costs for all alternatives is $23.9 million, and is based on a combination of
                                               Page 7-5

-------
urea-based SNCR and combustion modifications  Annual monitoring and admimstratu e costs for all
alternatives are S3 7 million 4  The accompanying average cost-effectiveness is $1.458 per ozone season ton.
and the reductions from uncontrolled 2007 baseline emissions are just over 40%

        While all the control levels meet EPA's criteria for highly effective ozone season NOx emissions
reductions, there are additional factors that EPA has considered in establishing the final emissions budgets
for this source category  First, the grouping of all cement manufacturing operations (i e . wet. dry, and m-
process-bitummous coal) in Tables 7-5 and 7-6 masks the fact that a significant portion of the large
operations (21 of 58) in the anahsis are not able to achie\e cost-effective reductions with technologies more
stringent than combustion modifications (e g . SNCR and SCR) EPA recened numerous public comments
confirming this result  Based on evidence cited in the cement ACT document and in some comments on the
SIP call proposals. EPA believes that a 30% reduction from uncontrolled levels would be within the cost-
effectiveness range for reducing emissions at all hpes of cement kilns  After reconsidering its own anahsis.
and considering public comments. EPA has decided to base the NOx budget  level on the  combustion
modification technologies which can achieve up to 30% control The reader  should note that due to the
timing of this final decision, the cost and economic impact results presented in the remainder of this report
reflect the $5.000/ton regulator}  alternative rather than the 30% control assumption
                                             Table 7-6
                           2007 Cost and Cost-Effectiveness Results for Large
                           Cement Manufacturing Operations (Cement Kilns)
Regulatory
Alternate e
$1.500/ton
S2.0UO/ton
$3.000/ton
$4.000/ton
$5.000/ton
Annual Control
Cost
(million 1990S)
S2U2
202
202
202
202
Annual
Monitoring and
Administ rathe
Costs
(million 1990S)
$37
37
37
3 7
37
Total Annual Costs
(million 1990S)
$239
239
23 9
23 9
23 9
O/one Season Cost
Effecth eness
(S/ozone season
ton)
$1,458
1.458
1,458
1 ,458
1.458
        4 It should be noted that the monitoring and administrative costs estimated for these sources are
 overstated, since they reflect application of the provisions of Part 75 (primarily, installation of CEMs;
 Cement kilns will be allowed to comply with the less stringent provisions of Part 60
                                              Page 7-6

-------
7.1.4   Results for Glass Manufacturing

        The analysis of large source glass manufacturing operations is conducted by selecting the most cost-
effective control measure available for each identified source that does not exceed the cost-effectiveness cut-
off specified in the regulatory alternative  Table 7-7 shows the emissions reductions achieved in the analysis
for each regulator, alternative  The table indicates that the alternatives achieve incremental reductions from
the 2007 baseline ranging from 46% to 76%
                                               Table 7-7
                         2007 Ozone Season NOx Emission Reductions for Large
                                    Glass Manufacturing Operations3
Regulator*
Alternate e
$1.500/ton
$2.00(x'ton
$3.000/ton
$4.(X)U.'ton
$5.000/ton
Number of
Affected Sources
12
16
25
25
25
2007 Baseline
Emissions
8.545
8.545
8.545
8.545
1 	 8.545
2007 Post-Control
Emissions
4.622
4.361
3,941
3.415
2.029
2007 Emission
Reductions
3.923
4.184
4.604
5.130
6.516
aThe 2007 baseline emissions estimate reflects emissions from all 25 large sources in this source categon, both controlled and uncontrolled
        Table 7-8 sho\\s the annual costs and resulting average cost-effectiveness for each regulatory
alternate e When emissions decreases are considered at all large glass manufacturing sources (regulator,
altematncs greater than S3.000/ton of control), the resulting a\erage cost-effectiveness exceeds EPA's
S2.000 framework  OnK the 12 flat glass manufacturers are able to achieve cost-effective reductions (after
considering administrate e costs) \vith am technology If all large sources in this categon were included in
the NOx Budget Trading Program, the potential inequities in control capability across sources could be
smoothed out but the a\ erage cost effectn eness \\ould exceed EPA's $2.000 frame\\ork  Therefore, this
source category exceeds EPA's cost effectn eness framework and is not included in the assumed NOx
emissions decreases for the Statewide budgets
                                               Page 7-7

-------
                                                Table 7-8
                            2007 Cost and Cost-Effectiveness Results for Large
                                     Glass Manufacturing Operations
Regulatory
Alternative
$1.500/ton
$2.000/ton
$3.000/ton
$4.000/ton
$5.000/ton
Annual Control
Cost
(million 1990S)
na
na
72
98
289
Annual
Monitoring and
Administrative
Costs
(million 1990S)
na
na
2 1
2 1
2 1
Total Annual Costs
(million 1990S)
na
na
93
120
31 0
Ozone Season Cost
Effectiveness
(S/ozone season
ton)
na
na
2.020
2,339
4.758
7.1.5   Results for Process Heaters

        The anal\ sis of large process heaters is conducted b\ selecting the most cost-effectn e control
measure a\ ailable for each identified source that does not exceed the cost-effectn eness cut-off specified in
the regulator, altematne   Table 7-9 sho\\s the emissions reductions achieved in the anah sis for each
regulators alternatn e  The table indicates that the alternatives achie\ e incremental reductions from the 2007
baseline up to 75%
                                                Table 7-9
                          2007 Ozone Season NOx Emission Reductions for Large
                                             Process Heaters3
Regulator}
Alternate e
$1.500/ton
$2.000/ton
$3.000/ton
$4,000/ton
$5,000/ton
Number of
Affected Sources
1
1
18
30
30
2007 Baseline
Emissions
15,147
15,147
15,147
15,147
15.147
2007 Post-Control
Emissions
15.072
15,072
4,099
3,820
3.820
2007 Emission
Reductions
75
75
11.048
11.32"
11.327
 The 2007 baseline emissions estimate reflects emissions from all large sources in this source category, both controlled and uncontrolled
        Table 7-10 sho\\s the annual costs and resulting average cost-effectiveness for each regulator}
alternative  Annual monitoring and administrative costs are not estimated for this category of sources
                                                Page 7-8

-------
because it is evident from Table 7-10 that even without these additional costs there is no regulator},
alternative that meets EPA's criteria for highly cost-effective ozone season N0\ emissions reductions  That
is. when emissions decreases are considered at all large process heating sources (i e.. regulator} alternatives
applying greater than $4.000/ton of control), the resulting average cost-effectiveness clearh  exceeds EPA's
$2.000 framework  Therefore, this source category exceeds EPA's cost effectiveness framework and is not
included in the assumed NOx emissions decreases for the Statewide budgets
                                               Table 7-10
                            2007 Cost and Cost-Effectiveness Results for Large
                                             Process Heaters
Regulatory
Alternative
$1.500/ion
$2.000 ton
$3.000/ton
$4.000-ion
$5.0()0/ton
Annual Control
Cost
(million 1990S)
na
na
31 6
328
328
Annual
Monitoring and
Administrative
Costs'
(million 1990S)
na
na
na
na
na
Total Annual Costs
(million 1990S)
na
na
31 6
328
328
Ozone Season Cost
Effectiveness
(S/ozone season
ton)
na
na
2.860
2.896
2.896
1 Monitoring and administrative costs arc not estimated for these sources since the domain-uide average control cost-effectiveness for the source
categorv exceeds S2 000 per ozone season ton reduced, and therefore is not subject to additional control for this rulemaking
7.1.6   Results of Commercial and Institutional Incinerators

        The anal} sis of large commercial and institutional incinerators is conducted by selecting the most
cost-effectn e control measure available for each identified source that does not exceed the cost-effectneness
cut-off specified in the regulator}  alternate e  Table 7-11 shows the emissions reductions achie\ ed in the
analysis for each regulator} alternate e  The table indicates that the alternative achie\ e from incremental
reductions from the 2007 baseline ranging from 0% to 45%
                                                 Page 7-9

-------
                                                 Table 7-11
                           2007 Ozone Season NOx Emission Reductions for Large
                                  Commercial and Industrial Incinerators'
Regulatory
Alternate e
$1.500/ton
$2,000/ton
$3.000/ton
$4.000/ton
$5,000/ton
Number of
Affected Sources
0
0
30
30
30
2007 Baseline
Emissions
2.852
2.852
2,852
2,852
2.852
2007 Post-Control
Emissions
2.852
2.852
1,577
1.577
1.577
2007 Emission
Reductions
0
0
1,2"5
1.2-5
,.T5
'The 2007 baseline emissions estimate reflects emissions from all 30 large sources in this source categors. both controlled and uncontrolled
        Table 7-12 shows the annual costs and resulting average cost-effectiveness for each regulator*
alternate e   Annual monitoring and administrate e costs are not estimated for this categon, of sources
because it is evident from Table 7-12 that even without these additional costs there is no regulator,
altematne that meets EPA's framework for highh cost-effectne ozone season NOx emissions reductions
Therefore no additional reductions are assumed from this source categon. group for establishing State \e\ el
emissions budgets
                                                 Table 7-12
                             2007 Cost and Cost-Effectiveness Results for Large
                                   Commercial and Industrial Incinerators
Regulatory
Alternative
$1.500/ton
$2.000/ton
$3,000/ton
$4.000/ton
$5.000/ton
Annual Control
Cost
(million 1990S)
$00
00
27
27
27
Annual
Monitoring and
Administrate
Costs
(million 1990S)
naa
na
na
na
na
Total Annual Costs
(million 1990S)
$00
00
27
27
27
Ozone Season Cost
Effectiveness
(S/ozone season
ton)
K/A
N/A
2.118
2,118
2.118
1 Monitoring and administrative costs are not estlmtated for these sources since the region-wide average control cost-effectneness for the source
categon' exceeds S2.000 per ozone season ton reduced, and therefore is not subject to additional reductions in this rulemakmg
                                                  Page 7-10

-------
7.1.7   Summary of Results for Non-Electricity Generating Sources

        Table 7-13 contains of summary emission reductions and costs associated with the final regulator.
decisions arising from the cost-effectiveness analysis  The final combination of regulatory alternatives
achieves an ozone season NOx emission reduction of approximately 203.000 tons beyond the 2007 baseline
This represents approximately a 62% reduction from baseline for these combined sources
                                               Table 7-13
                2007 Ozone Season Emission Reductions and Total Annual Compliance Costs
                         for Non-Electricity Generating Sources Used to Establish
                          State NOx Emissions Budgets under the NOx SIP Calla
Source Category
Industrial Boilers
Combustion Turbines
Internal Combustion Engines
Cement Manufacturing
TOTAL
Baseline Ozone
Season
Emissions
188.636
5 809
92.423
42.701
329.569
Ozone Season
Emissions After
Control
89,065
1.128
9.801
26.312
126.304
Total Reduction
in Ozone Season
Emissions
99.571
4.681
82.623
16.389
203.265
Total Annual
Compliance
Costs
(million 1990S)
1229
40
1004
239
2772
1 Emissions estimates are for large sources onh  Baseline NOx emissions estimates for industrial boilers do not include emissions for the 90 non-fossil
fuel fired boilers  The emissions reduction estimates reflect the preferred altematne for other stationary sources (60°o control of large industrial
boilers and combuMion turbines and control applied up to S5.000 per ton for stationary internal combustion engines and all cement kilns)
        Table 7-14 indicates the control technologies represented by the final regulatory decisions arising
from the cost-effectiveness anaKsis  The table shows that combustion modifications, such as ox\gen trim
and \\ater injection (OT + WI), are the dominant control technologies for boilers and turbines.  Selectne
catalytic reduction (SCR) is the dominant technology for 1C engines, and selectn e non-catalytic reduction
(SNCR) is the dominant  technology for cement manufacturing  Overall. SCR is estimated to be applied to
23% of large non-EGU sources in  EPA's anahsis  SNCR is applied to 21% of large units, and OT+WI is
applied to 33% of large units
                                               Page 7-11

-------
                                              Table 7-14
                   Control Technologies Selected for Non-Electricity Generating Sources
                              for Regulatory Alternatives Used to Establish
                         State NOx Emissions Budgets under the NOx SIP Call"
Control
Technology
SCR
SNCR
OT + WI
OTHER11
TOTAL
60% Control for
Industrial Boilers
and Combustion
Turbines
80
198
392
137
807
S5,000/ton Alternative for:
1C Engines
187
0
0
118
305
Cement Kilns
0
37
0
21
58
Total
267
235
392
276
1.170
* These results represent the number of emissions units for which the specified control technology is applied in the control cost anah.sis State and
source control decisions in response to the SIP call  ma\ differ
b Includes sources that are estimated NOT to appl>  addition controls
7.1.8   Administrative Costs for Non-Electricity Generating Units

        The burden to other stationary source operators potentially resulting from implementation of the
NOx SIP call rule are primarily associated with costs of installing and operating a continuous emissions
monitoring system (CEMS) to monitor NOx mass emissions and demonstrate compliance with limits
established b\ a State  This burden includes the total time, effort, or financial resources expended by an
operator to generate, maintain, retain, or disclose or pro\ ide information to or for a Federal or State agency

        A large proportion of the other stationary sources are expected to install CEMS and/or upgrade their
data acquisition and handling systems (DAHS) in order to participate in the NOx trading program For
trading sources (industrial boilers and combustion turbines) subject to Title IV monitoring in the Ozone
Transport Region (OTR). only administrate e costs of recordkeeping and reporting were estimated  These
units will not potentially experience additional capital or operating and maintenance costs as a result of this
rulemakmg since the> are already equipped with a CEMS meeting Part 75 Subpart H specifications
Howe\er. Title IV units that are not in the OTR will likely require minor upgrades to their DAHS Therefore.
the burden estimate associated with this rulemakmg for these sources includes additional capital and
operations and maintenance costs

        For units not subject to Title IV monitoring but in the OTR, costs are estimated for upgrading the
DAHS and performing annual quality assurance testing  For trading units not subject to the Title IV
monitoring and not in the OTR. additional costs may result from installing a NOx CEMS. or other approved
monitoring system, and a DAHS  The Part 75 NOx monitoring requirements vary depending on the fuel
burned and the hours of operation For example, monitoring requirements are less stringent for gas/oil  fired
sources than for coal-fired sources, and are even less stnngent for peaking units and low NOx mass emissions
sources  Similarly. the non-trading sources are assumed to experience additional costs from installing a NO\
                                               Page 7-12

-------
CEMS. or other approved monitoring system, as well as a DAHS " Although not explicitly required in the
NOx SIP call, it is assumed as a worst-case scenario that if controlled, a State would require a CEMS. or
equivalent, for the non-trading sources

        Table 7-15 presents estimates of the per-source annual administrative costs that other stationary
source operators may experience, based on assumptions of how States will implement administrative
requirements in response to this rulemaking  These estimates are prepared for both trading and non-trading
sources and included in the cost analvsis results
                                             Table 7-15
                          Average Per Source Annual Administrate Costs for
                                   Other Stationary Sources in 2007
                                            (1990 dollars)
Source Category Group
Industrial Boilers and
Combustion Turbines
1C hngines
Cement Manufacturing
Glass Manufacturing
Annual Monitoring
Costs
$27,201
$43.353
$63.540
$83.664
Annual Reporting &
Permitting Costs
$5.141
$253
$253
$253
Total Annual
Administrative Costs
$32,342
$43.606
$63.793
$83.917
7.2     Potential Economic Impacts

        This section presents the results of a screenmg-le\ el economic impact analysis for industrial boilers
and turbines and other stationary  sources potentialh affected by the rule The anah sis estimates the potential
impact on facilities and firms affected b>  the rule b\ comparing compliance costs to estimated sales or
expenditures Facilities and firms for whom costs exceed three percent or sales or expenditures are identified
as the most hkeh to experience significant impacts as a result of the rule Those for whom costs exceed one
percent of sales or expenditures are also highlighted as potentially experiencing significant impacts Costs
that represent less than one percent of sales or expenditures  are not expected to impose significant impacts

        While the RFA as amended b\ SBREFA does not apply to this rulemaking. the Agency elected to
evaluate the potential impacts of the rule on potentialh affected small entities, based on assumptions about
how the States could implement the requirements associated with meeting their NOx budgets

        The analysis of impacts is conducted at three levels  establishment (or facility), firm and industry.
        5  As indicated earlier, it should be noted that the monitoring and administrative costs estimated for
the sources outside the trading program (stationary 1C engines and all cement kilns) will be allowed to comply
with the pro\ isions of Part 60 The monitoring and administrative costs estimated for these sources reflect
the requirements of Part 75 (primarily, installation of CEMs), which are more stringent  than those in Part 60.
                                             Page 7-13

-------
•       Costs at the source level summarized in Section 7 2 are aggregated for each establishment, where an
        establishment owns more than one affected source Establishment-level costs are then compared with
        estimated sales or expenditures for the average sized establishment in the relevant mdustn (4-digit
        SIC) and employee size category (small vs large), as described in Chapter 5

•       Establishment-level impacts are summarized at the industry level, as defined by
                4-digit SIC codes

•       Finalh. costs are further aggregated to the firm level to account for the fact that some firms own
        more than one establishment affected by the rule Firm-level costs are compared with firm sales.
        obtained for the most part from Dun & Bradstreet data, as described in Chapter 5. For go\ernments.
        costs are compared with revenues, and for colleges and universities costs are compared with
        expenditures

        Individual potentially affected establishments and firms max have both industrial boilers and gas
turbines (sources in the trading program) and other stationary sources (sources not in the trading program}
affected by the rule To assess economic impacts more thorough!). it is therefore necessary to consider the
trading and non-trading alternatives in combination

        Section 7 2 1 provides an oveme\\ of the potentially affected firms,  facilities and sources in the
affected unn ersc  Detailed economic impacts are presented m  Section 7 2 2 for the preferred regulator.
alternatne combination - a 60% reduction from uncontrolled 2007 emissions for industrial boilers and
combustion turbines  and a $5.000/ton cost-effectn eness cap for other non-utility  sources (denoted b>
"60%/S5.000")  Section 723 compares these results with results for two additional regulator), alternative
combinations The potential alternate e combinations are shown in Table 7-16. with the regulator}
alternative combinations examined in the economic impact analysis indicated with an ''X"  The alternative
combinations that are analyzed capture the range of stringency  considered b> EPA. from the most stringent
combination (70%/S5.000) to the least stringent combination (40%/S 1.500 )
                                             Table 7-16
                            Potential Regulators  Alternative Combinations for
                        Non-Electricity Generating Unit Economic Impact Analysis

Regulatory
Alternate es
for 1C
Engines and
Cement
Manufacturing
$1.500/ton
$2.000/ton
$3,000/ton
$4.000/ton
$5,000/ton
Regulatory Alternatives for Industrial Boilers and
Combustion Turbines
40% Control
X




50% Control





60% Control




X
70% Control




X
                                              Page 7-14

-------
        Throughout the discussion, economic impacts are presented separately b\ size of the potential!}
affected entit> that owns the affected establishments  Section 7.3 discusses firm-level impacts for potential!}
affected small entities in more detail Chapter 8 discusses impacts on potential!}  affected go-\ ernment-owned
entities in more detail
7.2.1   Overview of Potentially Affected Sources, Establishments and Firms

        Table 7-17 shows the number of firms potentially affected under the preferred alternative, by source
categon 6  Table 7-18 shows the same information by sector and by size of entity
                                              Table 7-17
                Number of Firms and Other Entities Potentially Affected, by Source Category
Source Type
Industrial Boilers
Gas Turbines
Internal Combustion Engines
Cement Manufacturing
Potentially Affected
Firms/Entities
Total
221
22
18
20
Small
28
1
~>
4~
5
          One category of costs -- the transaction costs associated with trading allowances -- is not included in the cost
estimates discussed in these chapters These costs \\ill depend on the number of sources that elect to engage in trading
Analysis of EGUs with IPM indicates these costs are  1 5% of compliance costs, so the\ are expected to be small
                                               Page 7-15

-------
                                              Table 7-18
                         Number of Firms Potentially Affected, bv Sector and Size
Sector and Size of Entitv
Firms
of which, small entities
large entities
entity size unknown "
Federal government
Other government
Utility (SIC 49 11.4931)"
Colleges/universities
TOTAL
Potentially Affected
Firms/Entities
253
36
176
41
1
7
14
6
281
                   ' I'nknown si/e refers to entitle;, sshose emplo\ ee size could not be determined
                   'These are pnmanK cogenerators that suppK less than 50° o of generated power to the electric
                        power grid
7.2.2   Results for the Preferred Alternative Combination (60%/S5,000)

        Firm/Entity-Level Impacts

        Screemng-le\el impact results at the firm le\el are summarized in Table 7-19  This table shows the
number of potential!) affected firms or entities at particular le\els of firm-le\el costs as a percentage of firm
sales, revenues or expenditures

        Table 7-19 sho\\s that, at the firm or entity-le\ el. only a small percentage of the potentially affected
firms or entities for which sales estimates are available (14 of 232 or six percent) experience costs above one
percent of sales, and of these only eight of the 232 (three percent) experience costs abo\e three percent of
sales  Fne out of 36 identified potentially affected small entities may experience costs of three percent of
sales or greater.

        EPA expects that States implementing the SIP call will take these potential impacts into account in
designing their implementation scenarios  Industries with establishments having the potential for significant
impacts are likely candidates to be excluded from control requirements.
                                               Page 7-16

-------
                                              Table 7-19
         Number of Potentially Affected Firms by Firm Costs as a Percentage of Sales/Expenditures:
                                              60%/S5,000

Firms/Non-Profits
Of \\hich, small entities
large entities
entity -size unknown
Federal Go\ ernment a
Other Go\ eminent "
I'tilin b
Colleges Universities
T01AI
<0.5 %
184
22
159
3
na
3
10
6
203
0.5-1.0%
13
5
8
0
na
1
1
0
15
1 - 3%
4
4
0
0
na
1
1
0
6
>3%
6
5
1
0
na
1
1
0
8
Sales
NA"
46
0
8
38
1
1
1
0
49
Total
253
36
176
41
1
7
14
6
L 281
    ^ not a\ailable or (lor the federal go\emment) not applicable
* Co-generation units that suppK less than 5 0°o of generated power to the electric power grid
        Establishment-Level Impacts

        The 281 potential!) affected firms are comprised of 546 establishments  Establishment level impacts
pro\ide additional insights on indnidual facilities, \\hich in some cases are seen as stand-alone profit centers
Table 7-20 summarizes the results of the establishment-level analysis, by sector and firm size

        Table 7-20 shows that impacts of the preferred regulator} alternative are somewhat variable The
large majonrx  of establishments (347) incur costs that represent one percent or less of estimated
sales/expenditures, and of those 286 incur costs less than 0 5 percent of sales/expenditures  Of the 110
establishments potentially incurring costs that exceed three percent of estimated sales/expenditures, only
se\ en are o\\ned by identified small entities  One non-federal government and one utihtx establishment incur
costs greater than three percent of revenues

        EPA expects that States implementing the SIP call will take these potential impacts into account in
designing their implementation scenarios Industries with establishments having the potential for significant
impacts are likely candidates to be excluded from control requirements
                                              Page 7-17

-------
                                              Table 7-20
                                     Number of Establishments b>
                        Costs as a Percentage of Value of Shipments/Expenditures
                                       and Sector and Firm Size:
                                             60%/S5,000

Firms/Non-Profits
Of which, owned by small entities
owned by large entities
entity -size unknown
Federal Government a
Other Government '
Utihn ;
Colleges/Unn ersitics
TOTAL
<0.5 %
269
27
226
2~>
na
3
7
7
286
0.5-1.0%
51
4
49
4
na
1
3
0
61
1 - 3%
71
5
55
11
na
1
3
0
75
>3%
109
7
98
4
na
1
1
0
111
Total
506
3~
42^
41
12
7
14
/
546
1 Revenues nol available for one "other government" and 12 federal government establishments
6 Co-generation units, that suppK less than 50° o of generated pouer to the electric pouer grid
        Industry-Level Impacts

        Table 7-21 shows estimated impacts at the establishment level by industry (2 digit SIC code le\el)
Table 7-21 shows that, for the most part, only a small number of establishments (usualK less than 0 1 percent
of the total) are potentialh significantly impacted in am single industry compared to the total number of
establishments for each potentialh affected industn, \vithin the SIP call region  The 546 affected
establishments represent onh 0 02 percent of all the establishments in the SIP call region (roughh 3 6
million)  In most cases, potential impacts associated with the NOx SIP call are unlikeh to result in am
significant impacts at the industn level for potentially affected industries because the number of affected
establishments are a \ery small proportion of the total in those industries In addition, because only a very
fe\\ establishments ma\ experience potentially significant costs in each industry, the rule is not hkeh to result
in price increases to customers of the affected firms or other indirect economic impacts Furthermore. EPA
expects that States implementing the SIP call will take these potential impacts into account in designing their
implementation scenarios  Industries with potential!} significant impacts are likely candidates to be excluded
from control requirements  EPA has therefore concluded that a more detailed market-level impacts analysis
is not needed for am of these industries
                                               Page 7-18

-------
                         Table 7-21
    Number of Establishments By Establishment-l^vel Costs
as a Percentage of Value of Shipments/Expenditures and Industry:
                        60%/S5,000
SIC
10
14
20
21
22
24
25
26
27
28
29
30
Industry /Sector
Metal mining
Non-metal, non-
fuel
mining/quarrying
Food and kindred
products ml'gr
Tobacco products
mfgr
Textile mill
products
I ,umber & wood
products, exc.
furniture
Furniture &
fixtures
Paper and Allied
Products
Printing &
publishing
Chemicals &
allied products
Petroleum refining
and related
industries
Rubber & plastics
<0.5 %
1
0
35
2
5
0
1
68
1
58
18
7
0.5-1.0%
0
0
1
0
0
0
0
17
0
11
2
1
1-3 %
0
t
±~
0
0
1
0
1
7
1
9
1
0
>3%
0
3
T,
(1
1
1
2
1
0
5
1
1
Total
1
5
39
2
7
1
4
95
2
83
22
9
Percent of Establishments in SIP
Call Region Potentially Affected
at the 2- dicit SIC Code Level
06
02
04
1 6
0 1
<() 1
0 1
20
<() 1
1 0
1 6
0 1
                         Page 7-19

-------

SIC


32

.13

34



35


36

37


38


39

49


M
72

Industry/Sector

Slone, Clay, Glass
and Concrete
Products
Primary metals
Fabricated metal
products, exc
machinery &
trans, equip
Industrial &
commencal
machinery &
computer equip
I electronic & other
elce equip , cxc
computer equip
Transportation
equipment
Measuring instr ,
photo, mcd &
optical goods.
clocks
Miscellaneous
manufacturing
industries
Klcctnc, gas &
sanitary services
Wholesale trade -
nondurable goods
Person nl services

<0.5 %


4

35

4



3


4

11


1


1

12


1
0

0.5-1.0%


6

5

1



1


0

()


0


1

14


0
0

1-3%


23

2

0



1


0

1


0


1

22


0
1

>3%


7

2

0



3


0

1


0


0

75


1
0

Total


43

44

5



8


4

13


1


3

12.3


2
1
Percent of Establishments in SIP
Call Region Potentially Affected
at the 2- dicit SIC Code Level

05

09

<() 1



<0 1


<0 1

0 2


<0 1


0.1

1 2


<() 1
<0 1
Page 7-20

-------

SIC

79

80
89


Industry/Sector

Amusement and
recreation services
! lealth services
Miscellaneous
services
Colleges/universities
Federal government "
Other government "
TOTAI

<0.5 %

0

3
1

7
na
3
286

0.5-1.0%

0

0
0

0
na
1
61

1-3%

1

0
0

0
na
I
75

>3%

0

1
0

0
na
1
111

Total

1

4
1

7
12
6
546
Percent of Establishments in SIP
Call Region Potentially Affected
at the 2- digit SIC Code Le\cl
<() 1

<() 1
<() 1

<() 1
na
na
<() 1
h Includes natural gas transmission establishments (SIC 4922) and electric utilities establishments (SIC 491  1)  I 'tihtics having non-l-,(H ' sources affected bv these alternatives have co-generation units that
supply less than 50% of generated power to the electric power grid
'Revenues not available for one "other government" and  12 federal government establishments
                                                                                         Page 7-21

-------
7.2.3    Comparison by Regulatory Alternative

        Tables 7-22 and 7-23 provide an overview of economic impacts for the three combinations of
regulator} alternatives considered  Table 7-22 presents results at the firm level, and Table 7-23 shous
impacts at the establishment level
                                              Table 7-22
                   Number of Firms by Firm Costs as a Percentage of Sales/Expenditures
                                     and by Regulatory Alternate c

40%/$ 1.500
Preferred
Alternate e
60%/$5.000
70%/S5.000
<0.5 %
208
203
198
0.5-1.0%
12
15
14
1 - 3%
6
6
10
>3%
6
8
10
Sales
NA'
49
49
49
Total
281
281
281
' Sales not axailable or (for federal go\ eminent) not applicable
                                              Table 7-23
 Number of Establishments b> Establishment-Lev el Costs as a Percentage of Value of Shipments/Expenditures
                                     and by Regulatory Alternative

40% '$1.500
Preferred
Alternative
60%/S5.000
70%/$5.000
<0.5 %
333

286

252
0.5-1.0%
31

61

66
1 - 3%
66

75

88
>3%
103

111

127
Sales NA "
13

13

13
Total
546

546

546
1 Sales not a\aiiable or (for federal go\ eminent) not applicable
        The comparison among these regulatory alternatives shows a modest difference in potential economic
impacts between the preferred alternative and either the least or most stringent combination of regulator,
alternatives considered  Only two additional firms and  17 additional establishments may incur costs above
one percent of sales/expenditures for the preferred alternative compared to the  least stringent regulatory
alternative  Applying the most  stringent regulatory alternative results in an increase of six firms and 29
establishments that may incur costs above one percent of sales/expenditures when compared to the preferred
alternative.
                                              Page 7-22

-------
7.3     Small Entity Impacts

        This section discusses potential impacts on small entities that may be affected by requirements
related to the NOx SIP call  Since States are ultimately charged with achieving reductions to meet their
emissions budgets, they should seek to minimize impacts on small entities to the maximum extent
practicable  In this analysis. EPA has simulated State choices, so these impacts may or may not be
respresentative of actual impacts once States make their own choices The information presented in this
section may assist States in selecting control measures that minimize small entity impacts

        Table 7-24 shows  potential small entity impacts for the preferred alternative combination and the
other t\\o regulatory alternative combinations considered
                                              Table 7-24
                            Number of Potentially Affected Small Entities by
                   Cost as a Percentage of Sales/Expenditures by Regulator) Alternative

Regulatory
Alternate e

40%/S 1.500
Preferred
Alternatne
60%/S5.000
70%/%5.000
Total Small
Entities
Potentially
Affected
36

36

36

<1%


29

27

24

1-3%


4

4

5

>3%


3

5

7
This comparison shows that only a small absolute number of small entities is predicted to incur costs above
one percent in am  of the regulator) alternatives  In addition, the preferred alternative only results in 2
additional potential!) affected small entities with compliance costs greater than one percent of firm/entity
sales or re\ enues. compared to the least stringent regulatory alternate e  States should carefully  select control
measures to avoid  ad\ erse impacts on these and other small entities to the extent practicable

        The maximum number of potentially affected small entities that may be significantly impacted under
the preferred alternate e is 9. as shown in Table 7-24 Table 7-25 presents the industries (classified by 4 digit
SIC code) that ha\ e potentially  affected small entities with compliance costs greater than one percent of
firm/entity  sales or revenues for the preferred regulatory alternative, and the  number of small entities in each
industry potentially affected at this level of impact  This calculation excludes entities for which firm/entity -
level sales or revenues are not available (41 entities) and which could not be classified as small or large based
on employment  Some of these firms are likely to be small but the Agency can not estimate how many.
                                              Page 7-23

-------
                                            Table 7-25
                  Potentially Affected Small Entities that May Incur Compliance Costs of
                              Greater Than 1 Percent of Sales/Revenues
                              for the 60%/S5,000 Regulatory Alternative
SIC Code
2033
2075
2434
2869
3241
Description of Affected Industry
Canned fruit, vegetables, presen es.
jams, jellies
Soybean oil mills
Wood kitchen cabinets
Industrial Organic Chemicals, nee.
Cement, hvdraulic
Number of Potentially
Affected Small Entities in
Each Affected Industry
1
1
1
1
5
7.4     References

Abt Associates. 1998  Non-Electricity Generating Unit Economic Impact Analysis for the NOx SIP Call
RJA  Prepared for the U S Em ironmental Protection Agenc>. Office of Air Qualit} Planning and Standards.
September 1998

Pechan-A\ anti Group  Ozone Transport Rulemalang Non-Electricity Generating Unit Cost Analysis
Prepared for the U S Environmental Protection Agenc\. Office of Air Quality Planning and Standards.
September 1998
                                            Page 7-24

-------
                      Chapter 8. IMPACTS ON GOVERNMENT ENTITIES
        This chapter describes the potential impacts on State and local governments and the Federal
government as a result of the NOx SIP call  State and local governments will be required to submit additional
reports and monitoring data to the U S EPA in order to track compliance with the rule.  Section 8.1 discusses
State planning requirements and data collection issues imposed under the NOx SIP call  Section 8.2 presents
impacts from potential administrative costs to States and EPA related to control of electricity generating
units Section 8.3 presents impacts from potential administrate e costs related to control of other stationary
sources  Section 8 4 presents potential compliance (control and administrative) costs to potentially affected
government-owned entities   References are listed in Section 8.5
8.1     State Requirements

        Detailed estimates of incremental reporting and planning requirements, labor hours required to meet
those requirements, and the costs of that labor are described in the Information Collection Request (1CR)
prepared for this rulemaking '

        States arc required to report data annually for those point, area, nonroad mobile, and onroad mobile
sources for \\hich the> adopt control measures to meet their NOx emissions budgets: These annual reports
must include ozone season NOx emission imentones  States must report a statewide inventory of all NOx
sources e\ en. 3 years starting in the year 2003 for the  2002 in\ enton,. and 2008 for the 2007 compliance
demonstration im entor\  While the States are expected to use their existing emission im enton.' data
collection and electronic reporting mechanisms for submitting the data to EPA, some modifications \\ill be
necessary to account for the reporting of ozone season emissions data

        To minimize the reporting burden on State agencies, the reporting requirements for the final rule are
based on existing annual and periodic emission inventors  reporting requirements as much as possible
Howe\ er. since these ne\\ requirements are being established to support an ozone season reduction program
and since existing provisions do not require the collection of ozone season inventories, some additional
reporting \\ill be required  EPA requires that States report annually data for all point sources that are part of
a control measure that is adopted for purposes of meeting the NOx budget  If States  act in accordance with
OTAG recommendations for setting the emissions budgets, the sources controlled will all be point sources
emitting :• 100 tons per year (tp>) of NOx The 100 tpy threshold is consistent with the NOx reporting
threshold for the existing annual emission inventory   However, the rule does allow States the option of
defining the NOx point source threshold to be less than 100 tpy

        EPA will allow the direct reporting of point source data from sources to EPA if the sources are
subject to the monitoring and reporting requirements of Subpart H of 40 CFR Part 75 to satisfy this
         The ICR also estimates burden hours and cost for administrative requirements for the regulated sources
Regulated sources owned by government entities will incur these costs as well  Estimates of government -owned
regulated source administrative requirements are provided in Section 8 2

         Throughout remainder of this chapter, "State" is used to denote relevant state and local air quality planning
organizations responsible for compliance with Clean Air Act requirements

                                              Page 8-1

-------
requirement  The direct reporting of data from sources to EPA will minimize the reporting burden on States
Also, direct reporting will a^o\d duplication of effort for sources subject to the Part 75 requirements

        Currently, there are no existing annual reporting requirements for area, nonroad mobile, and high\va>
mobile source emissions  For the purposes of the final rule, an area source is any anthropogenic source that is
not included in the point, nonroad mobile, or onroad mobile source inventories. The EPA requires that States
report annually area source NOx emissions for only those stationary area source categories for which States
adopt control measures for the purpose of meeting their NOx budget  For nonroad mobile and onroad mobile
sources. EPA requires that States report annually NOx emissions for only those source categories for which
the State adopts control measures that are more stringent than Federal measures for the purpose of meeting
their NOx budget  Based on the recommendations  of OTAG. it is not expected that States will adopt area.
nonroad mobile, or onroad mobile source control measures to meet their NOx budgets However, if one or
more States do adopt such measures, annual reporting of emissions will be necessary to track the States"
progress to\\ ard meeting their NOx budgets

        The rule contains a requirement for States to report statewide point, area, nonroad mobile, and
onroad mobile source NOx emissions data every 3 years starting with the inventor, year 2002   The data
reported would be ozone season emissions data for each third year and would include  data from all source
categories in the State regardless of \\hether sources are being controlled to meet  a NOx budget  This 3-\ear
c\cle reporting requirement coincides with the schedule for the existing periodic emission inventor.' reporting
requirement for the States

        EPA requires that in 2007. States submit to EPA statewide ozone season NOx emissions data from
all NOx sources (point, area, nonroad mobile, and onroad mobile) within the State  The data reporting
requirements are identical to the reporting requirements for the 3-year cycle inventory, but would occur 1 year
prior to the scheduled 3-year c\clc inventory  This one-time reporting requirement for a 2007  statewide
inventor, will allow e\ aluation of whether the NOx budgets are met for 2007 Howe\ er. EPA will \\ork with
the States to minimize the incremental burden associated with preparing both a 2007 and a 2008 statewide
im enton  The incremental burden is associated \\ith the area and mobile source  inventories since States
must alreadx  report point source data annualh   For area, nonroad mobile, and onroad mobile source data.
States may project incremental changes in emissions from 2007 to 2008 to allow the 2008 inventory
requirement to be more easih met and to reduce the burden on the States
8.1.1   Planning Requirements

        Compliance with the reporting requirements falls into three categories of respondent activities  one-
time actions, annual actions, and triennial actions  Respondent States need to understand what activities are
required, when they are to be completed, and decide on data collection methodology  The remainder of this
section discusses what activities are required to be done once, annualh', and tnennially. Section  8.1 2
describes new data required and collection processes

        One-Time Activities

         First. States will need to read and interpret the reporting requirements of the rule. Additionally.
example ozone season emissions calculations must be prepared and submitted to EPA  Depending on the
complexity of these calculations, the amount of time required to comply would vary  Another one-time
activity involves a State modifying its emissions data bases to incorporate seven additional data items for

                                              Page 8-2

-------
point sources, five additional data items for area and nonroad mobile sources, and four additional data items
for on-road mobile sources

        A one-time effort is expected for the States to establish procedures to estimate statewide ozone
season NOx emissions from stationary area sources  Area source NOx emissions in the  OTAG inventory are
due to stationary fuel combustion, incineration and open burning, and wildfires and prescribed burning  It is
assumed that States would develop a spreadsheet or data base containing count} -level activity indicators
(e g . population. emplo\TnenL forest acreage) to allocate activity data typically available at the State le\el to
the counh le\el  It will also be necessary to account for any controls or seasonal restrictions that would
impact NOx emissions

        For onroad mobile sources, it will be necessary for States to prepare a procedure for estimating
count} -level VMT as input to EPA's MOBILE model  It is assumed that States will distribute statewide
VMT a\ailable from the Federal Highway Administration's (FHWA) Highway Performance Monitoring
S> stem (HPMS) to the count}' le\ el using a surrogate activity indicator such as population  In addition, any
onroad mobile source controls applicable to a count} must be identified and accounted for in the emission
estimates for a count}'

        States must prepare and submit a statewide ozone season NOx emissions inventor} of all controlled
and uncontrolled sources  for the \ ear 2007  A 2007 statewide ozone season NOx emissions inventory for all
point, area, nonroad mobile, and onroad mobile sources is required to allo\\ evaluation of whether the NOx
budgets are met for the > ear 2007   This one-time special inventor}  is necessary because the scheduled 3-\ear
reporting c\ cle does not fall in the year 2007  States \\hich must  submit the 2007 inventory may project
incremental changes in emissions from 2007 to 2008 to allo\\ the 2008 inventory requirement to be more
easih met and to reduce the burden on States which must submit  full NOx inventories in consecutne years
(i e /2007 and 2008)

        Finalh. the State must re\ iew a Title V permit revision submitted b\ controlled  sources  The  ICR
lists these one-time actn ities and estimates for the managerial and technical manhours required to complete
the tasks

        Annual Activities

        Annual State activities associated with reporting are as follows  States must  notify- the appropriate
EPA Regional Office when submitting an annual, triennial, and 2007 NOx inventory  Technical staff are
required to prepare and submit an electronic NOx emissions budget report  Most of the data collection
actn ities associated with  the annual inventor} are already being done to meet existing  inventor}
requirements  However, there is additional work associated with compiling and quality-assuring the ozone
season inventor} Estimates for the time needed to accomplish these reports and the costs associated with
these  reports are presented in the ICR

        Triennial Activities

        Ever\ 3 years.  States are required to submit ozone season emissions data for all point, area, nonroad
mobile,  and onroad mobile sources of NOx within the State States are already submitting a statewide
emissions inventory of all point sources under the existing  annual im entory requirements. However.
additional time requirements are expected for States to develop statewide NOx stationary area source.
nonroad mobile source, and onroad mobile source inventories even,  3 years.  Under the existing periodic SIP

                                              Page 8-3

-------
im entory requirements, emissions from these sectors were only determined for ozone nonattainment area
counties  The incremental time for developing statewide ozone season inventories for all controlled and
uncontrolled sources consists of hours allocated to the following activities

•       For stationary area sources, collecting activity data needed to allocate State-level activity data to the
        count} -level and estimating area source emissions.

•       For nonroad mobile sources, estimating emissions according to EPA's NONROAD emission
        im entory model3, and

•       For onroad mobile sources, estimating emissions using EPA's MOBILE model"1

        States must compile a summary report of statewide NOx emissions for submittal to EPA  Depending
upon current reporting practices, this actiMty max require little or no additional hours of labor

        Table 8-1  summarizes the various reporting requirements for State during the period from 2003 to
2008  Estimates for the time needed to accomplish these reports and the  costs associated with these reports
are presented in the ICR
8.1.2   Data Collection

        Mam of the required emissions data elements are already being provided to the EPA under existing
annual point source reporting requirements, as well as periodic SIP inventory reporting provisions for point.
area and nonroad mobile, and onroad mobile sources The EPA is also requiring States to provide an example
ozone season calculation, along with sufficient information for EPA to verify the calculated value of ozone
season emissions This calculation, as well as two additional seasonal data elements (i e . fuel heat content for
point sources. actiMty/throughput le\el). will facilitate quality assurance review of the seasonal emissions
data  Other data fields for providing the source of fuel heat content data, source of emissions data, source of
emission factor, and source of acti\it\ throughput data will also assist EPA in theirNOx budget verification
procedures

        The EPA is also requiring an "Area Designation" element for States that opt to establish an offset
pool composed of actual emission reductions achie\ ed through compliance with the SIP call NOx budgets
where States will need to track whether they are obtaining creditable offsets  as specified in Section 173(c1 of
the Act (which requires that major sources obtain offsets from areas with equal or higher nonattainment
classification) The ICR lists the new data items required for point, area, nonroad mobile, and onroad mobile
sources which are not currently required to be included in emission inventories reported to EPA.
        3 EPA's NONROAD model is currently under development by EPA's Office of Mobile Sources (OMS)  A
 final version of the model is expected to be in use b\ States by 2003

        4 These estimates are based on a State using MOBILE6. which will be the next version of EPA's MOB1LL
 model

                                               Page 8-4

-------
                                                        Table 8-1
              Schedule of Reporting Acti\ ities for Each Year During the Period 2003 through 2008"
Information Collection Activity
2003
2004
2005
2006
2007
2008
One-time (Annualized)
Read the reporting requirements of the rule
Submit example ozone season emissions calculations to EPA [§51 122(g)]
Modifv point, area, nonroad mobile, and onroad mobile source data bases to
add data fields for additional data items [§51 122(c). (d). (e)]
De\elop procedures b\ \\hich to estimate stationary area source NOx
emissions for triennial statewide reporting requirements [§51 122(b)(2). (3)]
De\elop procedure for generating counts -le\ el vehicle miles traveled (VMT)
|§51 122(b)(2).(3)]
Proiect 2007 area, nonroad mobile, and onroad mobile source m\entones to
2008 to satisfx 3-vear cvcle requirement [§51 122(bX2). (3)]
Re\ie\\ Title V permit reusions from controlled sources [§51 121(h)(D]
•
•
•
•
•

•





















Annual
Determine o/onc season emissions for controlled sources [§51 122(c)(l).
a>]
Notifx the appropriate EPA Regional Office \\hen submitting annual.
triennial, and 2007 NOx imenton. [§51 122(h>]
Submit electronic NOx budget emissions report [§51 122(b)(ljj
•
•
•
•
•
•
•
•
•
•
•
•
Triennial
Prepare statewide o/.one season imenton for stationarv area sources.
including a determination of ozone season emissions for all sources
[§51 122(b)(2)]
Prepare statewide ozone season m\entor\ for nonroad mobile sources.
including a determination of ozone season emissions for all sources
(§51 122(b)(2)|
Prepare statewide ozone season m\enton for onroad mobile sources.
including a determination of ozone season emissions for all sources
[§51 122(b)(2)j
Compile summan, report of statewide ozone season NOx emissions and
account for sources that have been reporting directh to EPA
[§51 122(h)(2).(h)]
•
•
•
•








•
•
•
•








•
•
•










•
•

•
•
•





 Activities associated with developing an emissions imenton for a particular year are assumed to take place dunng the year following the inventon
year (e g . activities for compiling a 2002 triennial inventon take place during 2003. and for 2007 inventon during 2008)
Source  Emission Reporting Requirements for Ozone SIP Revisions Relating to Statewide Budgets for SO, Emissions, September. 1998
                                                        Page 8-5

-------
       Several options are currently available for data reporting

       State chooses to continue reporting to the EPA Aerometnc Information Retrieval System (AIRS)
       s\ stem using the AFS format for point sources."1

       State converts its emissions data into the Emissions Inventory Improvement Project/Electronic Data
       Interface (EIIP/EDI) format.6

       State submits its emissions data in a proprietary  format based on the EIIP data model, or

       Annual reporting (except for third year reports) by sources submitting the data directly to EPA   This
       option will be a\ ailable to any source in a State that is both participating in a trading program
       meeting the requirements of Part 96 and that has agreed to submit data in this format The EPA will
       make both the rav\ data submitted in this format and summary data available to any State that
       chooses this option
8.2     Administrative Costs Associated with Units in the Trading Program

        This section presents the estimates of administrative costs for State and Federal governments
associated with control of electricity generating units

        Administrative Costs - State Governments

        Administrative costs to State governments include on-going auditing of sources, certification of
monitoring plans, and handling of permits  Table 8-2 presents the administrative costs to State go\ ernments
for these actn ities if they choose to participate in the model Part 96 trading program These estimates include
all units (both  EGUs and non-EGUs) that are part of the required applicability for the trading program  If a
State chooses not to participate in the trading program, they would incur different administrate e costs to
implement another regulatory means of achieving the required emission reductions  The magnitude of those
costs would depend on the regulatory option chosen by the State  The costs presented here are incremental to
the Initial Base Case  The unit cost for auditing  is SI. 698 (this assumes 60 hours per audit at $28.30 per
hour)  The total annual auditing costs  to States,  assuming that States audit 10 percent of the 1.995
potentially affected sources, is $339.600
        5 This option will continue for point sources for some penod of time after AIRS is reengmeered (before 2002).
 at which time this choice may be discontinued or modified

        6Through the EIIP, EPA is participating in a joint effort with State and local air pollution control acerbic* :o
 establish a common format for exchanging data from one agency to another.

                                               Page 8-6

-------
                                             Table 8-2
                Administrate e Costs Associated with Implementing the Trading Program
                            in States in the SIP call Region in 2007 (1990S)

Unit Costs
Total EGU Costs
Total non-EGU Costs
Total Annual Costs
Emissions Monitoring
Auditing
$1,698
256.568
136,180
$392,748
Certification of
Monitoring Plans
$0-566
238,852
318,658
$557.510
Permitting, Review and
Approval
$115
173,765
92,223
$265.988
Source ICFAnahsis


        The unit costs to States for initial review of monitoring plans range from zero for units with current!)
approved monitoring plans to $566 (20 hours per certification at $28 30 an hour) for units without currently
appro\ ed monitoring plans  Total annual costs to States for certification and recertification of the monitoring
plans for all units potentially affected by the NOx SIP call is $377.852  The unit costs for permitting
actn ities for States includes the costs for States to revie\\ and appro\ e applications for modifications as \\ell
as for ne\\ permits  The annuahzed unit cost to States for permit actn ities is assumed to be $23  The total
annual permitting activit>  costs to States, assuming 1.995 affected sources, is $45.885

        Administrative Costs - EPA

        The primary admimstratn e costs to EPA are associated with administering the trading program  The
t\\o mam tasks imolved in administering the trading program are administering the emissions tracking
s> stem (ETSj. used to track emissions from affected units and administering the allowance tracking system
(ATS) used to track owners of allowances

        EPA estimates that the capital cost in modify ing its existing ETS and ATS tracking systems which
are used to support the federal SO: trading program under Title IV of the Act and the OTC NOx Budget
Trading Program would be $250.000 and $500.000 respective!)  EPA also estimates that there would be
ongoing operational expenses of approximately $100.000 annualh to support each system

        EPA estimates that it would take 0 5 hours to process each allowance transfer and expects
approximately 8.050 transfers a year  This assumes that EPA will have to make 1  transfer at the beginning
of the year and one transfer at the end of the year for each unit. It also assumes that there will be
approximately three private transfers made per affected unit  In addition, it assumed no  additional costs for
units are part of the OTC NOx Budget Trading Program that EPA is already administering This was the
average in 1997 for the SO: trading program   This would require approximately 4.025  hours.

        For processing emissions data. EPA estimates that it will spend 5 hours more for units already
affected by the Acid Ram  Program who are not part of the OTC NOx Budget Trading Program. For these
1.089 units, this will take  5.445 hours.  EPA estimates it will spend  10 hours per unit for units that are not
currently affected by either the Acid Ram Program or the OTC NOx Budget Trading Program  For these 985
units, this will take 9.850  hours
                                              Page 8-7

-------
                                             Table 8-3
                       Administrative Costs Associated with EGUs and non-EGUs
                             in the SIP Call Region to EPA in 2007 (1990S)2

Emissions
Tracking System
Allowance
Tracking System
Total
Capital Costs
$250.000
$500.000
$750.000
Annual Fixed
Operating Costs
$100,000
$100.000
$200,000
Annual Labor
(in hours)
15,295
4.025
19.320
Total Annual Costs
in 2007
$786,000
$327.000
$1,113.000
 These estimates account for participation b\ non-EGU sources in the NO\ Budget Trading Program
        Total Administrative Costs-Electricity Generating Units

        The total EGU-related annual administrative costs of the rule to the States are roughly $800.000. and
roughh  $400.000 to EPA for the 0 15 trading option   The costs presented are incremental to the Initial
Base Case  The differences in these costs between the regulator,' alternatives examined is minimal since
there is little change in the number of affected units overall
8.3     Administrative Costs Associated with Other Stationary Sources Not in the Trading Program

        Mam of the States have the mechanisms in place to support the reporting of emissions data to EPA
under existing emission im entory requirements for these sources  Therefore, the burden for State personnel
to perform these acti\ ities  (e g . collecting emissions data from sources, maintaining emission im enton
records) are not estimated  The States" burden associated with these sources for the final rulemaking is
primarily to estimate and qualit) -assure ozone season emissions, and modifv State data bases to report
additional data items needed to verify ozone season emissions  In addition, there will be additional effort
imolved in compiling statewide area, nonroad mobile, and highway mobile source ozone season NOx
emission imentones e\er\  3 \ears  The information collection activities related to other stationan. sources
imolve an a\erage of 269  hours per >ear at an estimated cost of $7,140 per State, for total of 6.187 hours
and $164,202 for the entire SIP call region This is the annual cost to States associated with this rulemaking
for each year between 2003 and 2005  This is the first 3 year period during which States will be required to
begin reporting under the rule

        No other stationary source-related burden is expected to be imposed on EPA from implementation of
this rulemaking other than additional operation of the emissions tracking system (ETS) and the national
allowance tracking system (NATS)  An estimate of this burden related to non-EGUs is presented in
Table 8-3 along with the EGU-related burden to EPA from operating the ETS and NATS.  The total annual
cost in 2007 for this burden to EPA is estimated at roughly $850,000. based on the results in Table 8-3
                                              Page 8-8

-------
8.4     Government-Owned Entities

        This section summarizes compliance (control and administrate e) costs incurred by government-
o\vned other stationary sources that are assumed to require ne\\ controls under the NOx SIP call  These costs
include both control costs and administrative costs similar to those incurred by other regulated sources,
including costs associated with trading  These costs are a subset of the compliance costs presented in
Chapter 7  The control costs are based on assumptions of ho\\ affected States will implement control
measures to meet their NOx budgets set forth in this rulemaking  While the Unfunded Mandates Reform Act
does not apply to this action, the information on potential compliance costs to government-owned sources
may assist the States m their efforts to develop SIPs that meet these new NOx budgets

        Table 8-4 proMdes an overview of the government entities which own EGUs that may be affected by
the 0 15 trading option
                                             Table 8-4
        2007 Annual Costs To Potentially Affected Government-Owned EGU NOx Emissions Sources
                                 0.15 Trading Regulatory Alternative
Go\ernment Entit\
}• ederal Go\ ernment'1
Slate and Municipal Gcnernment
TOTAL:
Number of
Sources
1
78
79
Annual Control
Costs (thousands
of 1990S)
na
$45.100
$45.100
Annual
Administrative
Costs (thousands
of 1990S)
na
$5.900
$5.900
Total
Compliance
Costs (thousands
of 1990S)
na
$51.000
$51 000
 Control and administrate e costs uere not estimated for this source
       As sho\\n in Table 8-4. there are 79 potentially affected government-owned EGUs that may be
affected under the 0 15 trading regulator}, alternative  These units may experience compliance costs of about
S51 million in 2007

       Table 8-5 pro\ides an oveme\\ of the go\ eminent entities \\hich own non-EGUs that ma\  be
affected under the 60%/S5.000 options  As shown in Table 8-5. there are 30 potentially affected
government-owned non-EGU sources that ma\ experience compliance costs of roughly $3 6 million dollars in
2007
                                             Page 8-9

-------
                                            Table 8-5
          2007 Annual Costs To Potentially Affected Government-Owned NOx Emissions Sources
                                           60%/S5,000
Government Entitv
Federal Government
State Gcnernment - correctional
facilm
Cm' Government- Refuse s> stems
Educational institution
Metropolitan \\ ater s\ siern
Cm. regional se\\erage s\ stems
TOTAL:
Number of
Sources
23
1
1
1
1
3
30
Annual Control
Costs (thousands
of 1990S)
$1,786
602
05
30
54
176
$2.649
Annual
Administrative
Costs (thousands
of 1990S)
$727
46
8
1
46
136
$964
Total
Compliance
Costs (thousands
of 1990S)
$2.513
648
9
31
100
312
$3,613
       Therefore, under the 0 15 trading alternative for EGUs and the 60%/$5.000 per ton alternative for
non-EGUs. 109 sources or units o\vned by Federal. State, and local governments in the SIP call region ma\
potential!) be affected b\ control and administrate e measures at an annual compliance cost of about $55
million in 2007  This compliance cost. ho\ve\er. is only 3 percent of the total compliance cost for these
alternames (SI 69 billion)
8.5
References
Abt Associates. 1998 Non-Electricity Generating Unit Economic Impact Analysis for the NOx SIP Call
Prepared for the U S Em ironmental Protection Agency. Office of Air Quality Planning and Standards.
September 1998

U S  Environmental Protection Agency,  1998a 1CR # 1857 01, Emission Reporting Requirements for
Ozone SIP Revisions Relating to Statewide Budgets for NOX Emissions, September, 1998

US  Environmental Protection Agency.  1998b  Unfunded Mandates Reform Act Analysis for the Proposed
Federal Implementation Rule under the Clean Air Act Amendments Title I  Office of Air and Radiation.
September 1998.
                                            Page 8-10

-------
  Chapter 9. INTEGRATED COST, EMISSIONS, AND SMALL ENTITY IMPACTS SUMMARY


        This chapter presents EPA's estimates of the NOx emission reductions, potential compliance costs,
a\ erage cost-effectiveness, and potential small entity impacts associated with the final NOx SIP call
rulemaking  It brings together the results presented in Chapters 6. 7. and 8  All of these results are based on
the Agenc) "s assumptions of how States in the NOx SIP call region could implement control strategies to
meet the NOx budget le\ els set for them in this rulemaking  The results are then compared to average cost-
effecte eness estimates of other recent regulator, actions that require NOx reductions

        Section 9 1  presents estimates of NOx emission reductions for potentially affected electncit)
generating and non-electncit>  generating sources. Section 9 2 presents estimates of compliance costs (control
and administrate e costs) and average cost-effecte eness for all these sources, and provides a table of average
cost-effecte eness estimates for other recent regulator.- actions that require NOx reductions for purposes of
comparison  Section 9 3 presents an integrated summary of potential small entity impacts
9.1     Emission Reductions

        With this rulemaking. the EPA will establish ozone season NOx budgets for 22 States and the
District of Columbia based on reducing emissions from electricity generating units and other stationan.
sources ;  The analysis of impacts is from a baseline that includes the existing Title IV NOx rules.
ReasonabK A\ ailable Control Technology (RACT) requirements, and Ne\\ Source Performance Standards
(NSPS) and controls for new and recenth -built major NOx sources  The baseline also includes
implementation of Phase I (RACT requirements) of the 0/one Transport Commission (OTC) Memorandum
of Understanding (MOU):

        Table 9-1 sho\\s  the NOx emissions le\els and emissions reductions for selected combinations of
alternate es that EPA has analyzed for potentially affected electricity generating units and other stationan.
sources  These results bring together the six regulatory alternatives analyzed for potential!) affected EGUs;
and the three combination regulator) alternate es for non-electncit) generating sources  The non-EGU
regulator, alternate es combine industrial boilers and combustion turbines with other stationan. sources
(stationary internal combustion engines, cement kilns)  These alternatives are discussed in Chapter 2 and the
results of the analysis  are presented in Chapters 6 and 7   The alternatives selected for the final NOx SIP call
are highlighted
          This category includes industrial (industrial, commercial, and institutional) boilers and combustion turbines,
stationan- internal combustion engines, and cement manufacturing operations (cement kilns [wet. dry, and coal-fired])
For additional details on these source types see Chapter 3

          This baseline is discussed in greater detail m Chapter 4

                                               Page 9-1

-------
                                                                               Table 9-1
                 2007 O/,onc Season NOx Emissions and Emission Reductions for Selected Combinations of Electricity Generating Units and
                                           Other Stationary Source Regulatory Alternatives from the Initial Base Case "
                                                                       (thousands of NOx Tons)
Regulatory Alternatives
Other
Stationary Sources
(330 thousand
baseline tons)
40% Control/
SI 500 per Ton
60% Control/
SSOOO per Ton
70% Control/
$5000j>er Ton
Electricitv Generating Units (1,502 thousand baseline tons)
0.25
Trading
1,111
(722)
1 ,067
(766)
1 ,04 1
(79 1)
0.20
Trading
922
(910)
878
(954)
853
(979)
Regionally
1
847
(985)
803
(1,028)
778
(1,054)
Regionally
2
757
(1,075)
?n
(1,119)
688
(1,144)
0.15
Trading
735
(1,097)
691
(1,141)
666
(1,166)
0.12
Trading
624
(1,208)
580
(1,252)
555
(1,277)
" Emissions reductions arc shown in parentheses  Controls on the electricity generating units occur through a Cap-anil-1 rade program described in the NOx Model 'I radmg Rule and supporting information
Controls on Other Stationary Costs arc determined using Iwo appro.iches I) a least-cost approach th.il approximates a trading program, applied lo large industrial boilers and combustion turbines, and 2) an
approach that applied controls up to a cost cutofl expressed in annual costs per o/one season ton reduced, applied to large stationary 1C engines, and cement manufacturing
                                                                                Page 9-2

-------
9.2     Compliance Costs and Cost-Effectiveness

        Table 9-2 shows annual compliance control costs for selected combinations of regulator, alternatives
that EPA has analyzed for potentially affected electncit> generating units and other stationary sources  Costs
include direct control costs and admmstrative costs (monitoring, recordkeeping. and reporting)  The
alternatives selected for the final NOx SIP call rule are highlighted The costs for EGUs reflect emissions
trading across States  For non-EGUs. costs are determined using two approaches  1) a least-cost approach
that approximates a trading program, applied to large industrial boilers and combustion turbines, and 2) an
approach that applied controls up to a specified cost cutoff expressed in costs per ton reduced, applied to
large stationary 1C engines, and cement kilns

        Table 9-3 provides the resulting 2007 ozone season average control cost-effectiveness values for the
same selected combination of alternatives examined in the previous two tables  The average control cost-
effectiveness of ozone season NOx emission reductions is calculated as the change in total annual compliance
costs relatn e to the Initial Base case divided b\ the change in ozone season NOx emissions relative to the
Initial Base case  This table sho\\ s the increase in average cost-effectiveness values as the combination of
standards considered becomes more stringent  It should be noted that these estimates are only presented to
illustrate the a\erage cost-effectiveness of different combinations of EGU and Other Stationary Source
alternates  The decisions on control levels and the inclusion of individual source categories in this
rulemaking \\ere made b\ e\aluatmg each category separate!}, not b\ using the summary of cost-
effectiveness \ alues in Table 9-3

        OTAG recognized the value of market-based approaches to lowering emissions from power plants
and large industrial sources  The Agencx agrees that using a market-based approach in the emission
reduction program is desirable According!). the Agency has proposed the NOx Model Trading Rule   This
rule pro\ides for an  emissions cap and allows for trading bet\\een sources in the all the jurisdictions co\ered.
which arc essential for this rule to be effectn e and admmistratn eh practicable  The Agenc>  wants to work
with all  affected jurisdictions covered by this rulemaking to establish  such a program  This is a major reason
behind the Agenc\ "s effort at  estimating  NOx control costs across the jurisdictions  in the SIP call region for
electric po\\er generation units and cost minimization across the same domain for Other Stationary Sources
Analytical limitations kept EPA from estimating the costs of a single  cap-and-trade program for electricity
generating sources and large industrial sources m the Other Stationary Sources category (e g . industrial
boilers and combustion turbines)  Gnen that the Agenc\ could not estimate the costs of a single emissions
trading program for these sources, the annual cost estimates for this rulemaking are hkeh to be overstated to
the extent that costs could be reduced b>  trading bet\\een facilities  in  both groups.  However, it should be
noted that mdn idual States ma\ decide to achie\ e their NOx budget with other control techniques, thereb\
affectinc their costs
                                              Page 9-3

-------
                                                                                     Table 9-2
                                               2007 Annual NOx SIP call Compliance Costs for Selected Combinations of
                                           Electricity Generating Unit anil Other Stationary Source Regulatory Alternatives
                                                                            (millions of 1990 dollars)'
Regulatory Alternatives
Other
Stationary Sources
40% Control/
$1,500 per Ton
60% Control/
$5,000 per Ton
70% Control/
$5,000 per Ton
Electricity Generating Units
0.25
Trading
$848
$925
$1,048
0.20
Trading
$1,153
$1,230
$1,353
Regionally
1
$1,323
$1,400
$1,523
Regionally
2
$1,554
$1,631
$ 1 ,754
0.15
Trading
$1,583
SJ,660
$1,783
0.12
Trading
$2,051
$2,128
$2,251
 The decisions on control stringency and the inclusion of"individual source categories in this nilcniakmg were nol made using the summary of cost-effectiveness values in this (able  Control on the cleclricity
generating units occur through a Cap-and-Trade program described in the NOx Model  Trading Rule and supporting information Controls on Other Stationary' Sources arc applied using two approaches  1) a
least-cost approach that approximates a trading program, applied to l.irgc industrial boilers and combustion turbines, and 2) an approach that applied controls up to a cost cutoff expressed as costs per ton
reduced, applied to large stationary 1C engines, and cement manufacturing  Analytical limitations prevented 1,1'A from estimating the costs of a single cap-and-tradc program for electricity generating imils and
large industrial boilers and combustion turbines combined  Costs for these sources are likely lo be lower than has been estimated in this Rl A if States integrate electricity generating units and industrial boiler
and combustion turbine programs in to a single trading program  It should be noted that individual Slates may decide to achieve their NOx budget with other control techniques, thereby afl'ccting their costs

-------
                                                                                   Table 9-3
                                      2007 O/onc Season Average Compliance Cost-Effectiveness for Selected Combinations of
                                          Electricity Generating Unit and Other Stationary Source Regulatory Alternatives
                                                       (1990 dollars per ton of 1NO\ reduced in f he o/onc season)'
Regulatory Alternatives
Other
Stationary Sources
40% Control/
$1,500 per Ton
tfOVo Control/
$5,000 per Ton
70% Control/
S5,000 per Ton
Electricity Generating Units
0.25
Trading
SI, 175
$1,208
$1,325
0.20
Trading
$1.267
$1,2X9
$1,382
Rcgionalitv
1
$1.341
$ 1 .362
$1,445
Rcgionalitv
2
$1,446
$1,458
$1,533
0.15
Trading
$1,443
SI, 455
$1,529
0.12
Trading
$1,698
$ 1 ,700
$1,763
"Controls on the electricity generating units occur through a C'ap-and-1 r.idc progtani described in the \O\ Model 'I radmg Rule and supporting information  Controls on Other Stationary Sources were applied
using two approaches  1) a least-cost approach that approximates a trading program, applied to large industrial boilers and combustion turbines, and 2) an approach that applied controls up to a cost cutoll
expressed in costs per ton reduced, applied to large stationary 1C engine1., and cement manufacturing Analytical limitations prevented Kl'A from estimating the costs of a single cap-and-tradc program for
electricity generating units and large industrial boilers and combustion turbines combined  Costs lor these sources aie hkelv to he lower than has been estimated in this  RIA if States integrate electricity
generating units and industrial holler and combustion turbine programs in to a single trading program  It should be noted that  individual States may decide to achieve their NO\ budget with other control
techniques, thereby affecting their costs
                                                                                    'age 9-5

-------
9.2.1   Cost-Effectiveness Comparisons

        Table 9-4 proudes a reference list of measures that EPA and the States ha\e undertaken to reduce
NOx and their a\ erage cost per ton of NOx reduced  The average annual cost per ton of NOx reduced from
this rulemakmg is included in the table  Most of these measures fall in the $ 1,000 to $2.000 per ton range
With few exceptions, the average cost-effectiveness of these measures is representative of the average cost-
effectiveness of the t\pes of controls EPA and the States have needed to adopt most recently since their
previous planning efforts have already taken advantage of opportunities for even cheaper controls  The
Agenc\ believes that the cost-effectiveness of measures that the cost-effectiveness of measures that it or
States have adopted, or proposed to adopt, forms a good reference point for determining which of the
available additional NOx control measures can most reasonably be interpreted by upwind States or
jurisdictions that significantly contribute to ozone nonattainment
                                                Table 9-4
                            A\ erage Cost-Effectiveness of NOx Control Measures
                              Recently Undertaken or Proposed (1990 dollars)
Control Measure
NOx RACT
Phase II Reformulated Gasoline
State Implementation ol the Oxone Transport
Commission Memorandumof Understanding (OTC
MOU)
Proposed New Source Performance Standards (NSPS)
for Fossil Steam Electric Generating I 'nits
Proposed NSPS for Industrial Boilers
Final NOx SIP Call Rulemakmg - ElectnciU Generating
Units
Final NOx SIP Call Rulemakmg - Other Stationan
Sources
Average Cost per Ton of NOx Reduced
$150- L300
$4.100a
$950-51,600
$1.290
$1,790
$1.468b
$1,365C
'Average cost representing the midpoint of S2J80 to S6.000 per ton. as described in EPA's response to the American Petroleum Institute s petition to
wane the Federal Phase II RFG NOx standard This cost represents the projected additional cost of complying with the Phase II RFC Nox standards
beyond the cost of compKing with the other standards for Phase II RFG
b Estimated a\erage cost-effecli\eness (including compliance costs) associated with the uniform 015 trading alternative
' Estimated average cost-effectiveness (including compliance costs) associated with the preferred alternative (60°'o control - industrial boilers and
combustion turbines, control up to a cost cutoff of S5.000 ton - stationary 1C engines, cement manufacturing)
         There are also a number of less expensive measures recently undertaken by the Agency to reduce
NOx emissions that do not appear in Table 9-4. These actions include (1) the Title IV NOx reduction
program. (2) the federal locomotive standards, (3) the 1997 proposed federal nonroad diesel engine
standards. (4) the federal hea\y duty highway engine 2g/bhp-hr standards, and (5) the federal marine engine
standards  These actions do not provide a meaningful comparison to this rulemakmg because they are
believed to be among the lowest cost options for NOx control  Since these options have been exhausted, the
Agenc>  must now focus on what other measure exist, at a potentially higher average cost-effectiveness \ aluc.
                                                 Page 9-6

-------
that can further reduce NOx emissions  Table 9-4 is thereby useful as a reference for the next higher level of
NOx reduction cost-effectiveness that the Agency considers reasonable to undertake
9.3     Integrated Small Entity Impacts

        The Agency examined the potential economic impacts to small entities associated with this
rulemakmg based on assumptions of how the affected States will implement control measures to meet their
NOx budgets  While the RFA as amended by SBREFA does not apply to this action, these impacts have
been calculated in order to provide additional understanding of the nature of potential impacts, and additional
information to the States as they prepare SIPs designed to meet the NOx budgets set by this rulemakmg  It is
the Agenc\ "s position, however, that the RFA as amended by SBREFA does apply to the proposed NOx FIP
and the proposed response to the section 126 petitions  The Agency has prepared Initial Regulator,
Flexibility Analyses (IRFAs) for both of these actions

        Table 9-5 presents a summary of the potentially affected small entities in EPA's analysis  Of the
191 small entities potentially affected. 41 may experience compliance costs in excess of one percent of
revenues, based on assumptions of how the affected States implement control measures to meet their  NOx
budgets as set forth in this rulemakmg  Potentially affected small entities experiencing compliance costs in
excess of  1 percent of revenues have some potential for significant impact resulting from implementation of
the SIP call   These 41 small entities constitute about 3 percent of the small entities in the SIP call region that
own sources potentialh affected b\ this rulemakmg

        EPA's estimates of other potential economic impacts for entities owning electricity generating units
including changes in capacity, and changes in demand for electricity that could result from implementation of
this SIP call are found in Chapter 6.  Details on EPA's other estimates of potential economic impacts
associated with this rulemakmg to entities owning sources in the other stationary source categon are  found in
Chapter 7
                                             Page 9-7

-------
                                                   Table 9-5
                                 Number of Potentially Affected Small Entities
                                      for the NOx SIP Call Rulemaking
Source Category/
Regulatory
Alternate e
Electncit\
Generating Units
0 15 Trading
Other Stationan
Sources
60% Control/
$5000/ton
TOTAL
Small Entities in
the SIP call Region

500

700^

1,200
Small Entities
Potentially
Affected'

114

77

191
Small Entities in
the SIP Call
Region with
Compliance Costs
> 1% of
Sales/Revenues

32b

9

41
Percentage of
Small Entities in
the SIP Call
Region with
Compliance Costs
> 1% of
Sales/Revenues

6%

1%

3%
'These are small entities that own large sources in the source categories covered under this rulemakmg
'The estimated costs of compliance are calculated assuming all small non-utilit> generators comph through purchasing allowances This approach
tends to overstate compliance costs because cases m which emission reductions can be achieved below the marginal cost of reductions in the SIP call
domain are not considered
cThis represents the number of small entities in the SIP call region owning sources (small and large) in the source categories covered under this
rulemakmg
9.4
References
U S Emironmcntal Protection Agency. 1998a  Initial Regulatory Flexibility Analysis for the Proposed
Federal Implementation Plan  Under the Clean Air Act   September. 1998

U S Environmental Protection Agenc>. 1998a  Initial Regulatory Flexibility Analysis for the Proposed
Section 126 Pennon Rulemaking Under the Clean Air Act  September. 1998
                                                    Page 9-8

-------
           APPENDIX

             to the

REGULATORY IMPACT ANALYSIS
 FOR THE NOx SIP CALL, FIP, AND
    SECTION 126 PETITIONS

Volume 1: Costs and Economic Impacts

-------
                STATE-BY-STATE OZONE SEASON NOx EMISSIONS FOR
         ELECTRICITY GENERATING UNITS BY REGULATORY ALTERNATIVE
      This appendix contains the state-b> -state emissions data used to generate the map figures in Chapter
6 The source for this data is ICF analysis using the latest version of the Integrated Planning Model (IPM)
                                       Page A-1

-------
                                              Table A-l
                2007 Ozone Season Emissions Estimates for the Electric Power Industry for
                                      States in the SIP Call Region
State Name
Alabama
Connecticut
Delaware
District of Columbia
Georgia
Illinois
Indiana
Kentucky
Man land
Massachusetts
Michigan
Missouri
Neu York
Ne\\ Jerse\
North Carolina
Ohio
Pennsylvania
Rhode Island
South Carolina
Tennessee
Virginia
West Virginia
Wisconsin
TOTAL
Initial Base
Case
76,926
5,636
5.838
3
86.455
119,311
136,773
107.829
32,603
16.479
86.600
82,097
39.199
18.352
84,815
163,132
123,102
1,082
36.299
70,908
40,884
115,490
51,962
1,501,775
0.25
Trading
52,084
3.867
6,119
8
55.808
65,988
77.453
57,721
23.317
16.104
58.825
44.388
31.832
13.316
59.255
93.803
81.188
1.071
27.453
46.459
30,632
60.070
32,838
939,599
0.20
Trading
39,937
3,866
4,624
7
46,483
51.650
63.307
47,896
16,194
13.735
42,692
34.662
26.818
12.302
50.088
71.219
71,118
1,071
22.093
24.211
24.280
57.998
24,998
775,529
0.15
Trading
37,440
3,267
3.585
10
37.474
37,928
47.415
38.427
13.864
10,319
34,950
24.037
24,093
8.838
34,556
46,843
46.186
1.071
17,965
23.706
19.276
33.545
18,967
563,762
0.12
Trading
26,689
2,772
3,555
16
26.856
28.917
34.408
30.411
11.193
10216
26.9"0
14.734
21.838
8,198
29.945
39,900
41.800
9~4
13. 6 -'5
20.162
16,215
25.693
18.286
453,443
Numbers ma\ not sum due to rounding
                                               Page A-2

-------
                                              Table A-2
               Comparison of Electric Power Industry 2007 Ozone Season Emissions for the
              Initial Base Case, the 0.15 Budget Component, and the 0.15 Trading Alternative
                                   for States in the SIP Call Region
State Name
Alabama
Connecticut
Delaware
District of Columbia
Georgia
Illinois
Indiana
Kentucky
Man land
Massachusetts
Michigan
Missouri
New York
New Jerse\
North Carolina
Ohio
Penns\l\ama
Rhode Island
South Carolina
I ennessee
Virginia
West Virginia
Wisconsin
TOTAL
Initial BaseCase | 0.15 State Budget
76,926
5,636
5,838
3
86,455
119,311
136,773
107.829
32.603
16.479
86.600
82,097
39.199
18.352
84.815
163.132
123.102
1.082
36.299
70,908
40.884
115,490
51.962
1,501,775
30,644
5,245
4,994
152
32.433
36,570
51,818
38.775
12.971
14,651
29.458
26.450
31,222
8,191
32.691
51.493
45.971
1 .609
19.842
26.225
20.990
24.045
17.345
563,785
0.1 5 Trading
37,440
3.267
3,585
10
37,474
37,928
47,415
38,427
13,864
10.319
34,950
24.037
24.093
8,838
34.556
46,843
46, 1 86
1,071
17,965
23.706
19,276
33,545
18,967
563,762
Numbers ma\ not sum due lo rounding
                                              Page A-3

-------
                                                Table A-3
             Comparison of State-by-State 2007 Ozone Season Emissions for the Initial Base Case,
                                Regional Budgets, and Regional Alternatives
State Name
Alabama
Connecticut
Delaware
District of Columbia
Georgia
Illinois
Indiana
Kentucky
Man land
Massachusetts
Michigan
Missouri
Ne\\ York
New Jerse\
North Carolina
Ohio
Pennsylvania
Rhode Island
South Carolina
Tennessee
Virginia
West Virginia
Wisconsin
Region 1 Subtotal "
Region 2 Subtotal '
Region 3 Subtotal"
TOTAL
Initial Base
Case
76,926
5,636
5,838
3
86,455
119,311
136.773
107.829
32.603
16,479
86.600
82.097
39,199
18.352
84,815
163.132
123.102
1.082
36.299
70.908
40,884
115,490
51.962
—
—
—
1,501,775
Two-Region
Budget
40,835
5,245
4,994
152
43.190
48,418
68,556
51.543
12.971
14.651
38,747
34,998
31,222
8,191
43.093
51.493
45.971
1.609
26.455
34,967
20,990
24,045
23,009
217,034
458,311
—
675^45
Three-Region
Budget
40,835
4,199
4,005
122
43,190
36,570
51.818
38,775
10,380
11.759
29.458
26.450
25,066
6.598
43,093
51.493
36.932
1,290
26.455
34.967
20.990
24,045
17,345
100,351
296,944
188,540
585,835
Two-Region
Alternatrv e
40,048
3,267
3,581
10
46.491
49.977
67.492
49,211
13,236
10.319
42,753
35.939
24,652
8,765
50.365
45.856
59,242
1,071
22.874
23.736
18.684
32,847
24,995
221,530
453,881
—
675,411
Three-Region
Alternative
40,048
2.783
3,320
11
47,438
38.259
42.142
37,201
10,947
10,298
35,994
25.045
22,657
7,902
51,293
46.824
41,363
1 ,07 1
25.239
24.520
19,953
32,659
18,828
100,352
296,905
188,538
585,795
Numbers ma\ not sum due to rounding
1 Regions for the two-region and three-region options are defined in Section 6 2.1 and 622
                                                 Page A-4

-------
          ADDENDUM
              to the
REGULATORY IMPACT ANALYSIS
 FOR THE NOx SIP CALL, FIP, AND
     SECTION 126 PETITIONS

Volume 1: Costs and Economic Impacts
           September 1998

-------
This page intentionally left blank

-------
1.0     Introduction

        This addendum to the Regulatory Impact Analysis for the NOx SIP Call, FIP, and Section 126
Petitions contains final cost and impact estimates for the implementation scenario used by EPA to establish
the State emissions budgets  For the electricity generating units (EGUs), the analysis reflects a revised
emissions cap under the 0.15 trading alternative For the non-EGUs. the analysis reflects a revised emissions
inventorv
2.0     Addendum Analysis of the Revised NOx Cap on the 0.15 Trading Alternative for Electricity
        Generating Units

        Since EPA conducted the main analyses documented in the R1A, the Agency has revised the NOx
budget and recomputed the NOx cap for the 0 15 trading alternative The new emissions cap for the ozone
season is approximately 544 thousand tons of NOx, which is about 20 thousand tons lower than the previous
cap for this option  The Agency made this change in response to comments that it received on the
Supplemental Notice of Proposed Rulemakmg on the 0.15 trading alternative EPA has also decided to allow
banking of NOx emissions in the trading program with flow control  Finally, the Agency has decided to
allow State programs to promote early NOx reductions by both electricity generating units and non-EGUs
(which allows them to make early NOx reductions in 2001 and 2002 that they "bank" for use in 2003) and to
create a process whereby NOx sources  that have problems installing pollution control equipment and
purchasing NOx allowances to co\er their emissions in 2003 can qualify to receive some NOx allowances
from the State

        This analysis examines the emissions, costs, and cost-effectiveness implications of the basic change
to the NOx emissions cap to 544 thousand tons for EGUs for the   15 trading alternative  Generally, for a
trading program like the one that EPA is now working with the States to establish, the Agency expects little
banking to occur based on the consideration of what the direct costs are of NOx reductions are today versus
what they will be in the future How the early NOx reduction credit program and State-managed relief valve
to address reliability concerns will actually work in the future is difficult to predict  Given that they occur
o\ er a short time frame, their consideration is not necessary to gaming an understanding of the NOx SIP
call's annual costs and cost-effectiveness to the electric power industry over time

        Table 1 shows the results of the analysis of emissions and costs of the 0  15 uniform alternative
relative to the Initial Base Case, using the revised cap  Depending on the year, the emission reductions
provided b\ the NOx SIP call under this option range between 919 and 968 thousand summer tons of NOx
per year. These reductions come at an incremental  cost of between $ 1.371 and $ 1.440 million, for  an average
cost-effectiveness of about $1.500 per ozone season ton
                                              ADD-1

-------
                                              Table 1
 Year-by-Year Comparison of the 0.15 Trading Alternative under the Revised Budget to the Initial Base Case:
                 Estimated Emissions, Emission Reductions, Costs, and Cost-Effectiveness

Emissions Under Initial Base Case
(ozone season NOx emissions, thousands of tons)
Emissions, New Budget (ozone season NOx
emissions, thousands of tons)
Emissions Reductions. Relative to Initial Base Case
(Ozone season NOx emissions, thousands of tons)
Incremental Annual Cost. Relative to Initial Base
Case (millions of 1990$)
Cost per Ozone Season Ton of NOx Removed,
Relative to Initial Base Case (1990$)
2003
1,462
544
919
$1,371
$1,493
2005
1,497
544
953
$1,414
$1,484
2007
1,502
544
958
$1.440
$1,503
2010
1,511
544
968
$1.4)1
$1.458
 Source ICF analysis
        Table 2 shows how the analysis of the 0 15 trading alternative changes, in absolute and percentage
terms,  in response to the new cap  Cutting 20 thousand ozone season tons of NOx from the cap lowers
emissions by 3.5 percent and increases emission reductions by 2 1 percent. These additional reductions are
accomplished through an increase in the use of SCR. which rises by about 10.000 MWT from 63.000 MW to
73.000 MW.  As a result, estimated costs increase by about $62 million in 2007. which is an increase of
about 4.5 percent of the incremental cost under the previous cap  The cost per ton of NOx removed is higher
under the new cap by about 2.4 percent. The comparisons are shown only for the year 2007.  they are similar
to the changes and percentage changes for the other years

        Based on this reanalysis of the effects of the NOx SIP call for the 0.15 trading alternative using the
revised budget. EPA has concluded that the comparisons across options presented in the RJA would not be
materially affected b\ the change in the budget
                                              ADD-2

-------
                                               Table 2
        Comparison of the 0.15 Trading Alternative in 2007 under the Original and Revised Budgets:
                      Estimated Emission Reductions, Costs, and Cost-Effectiveness

Emissions, Relative to Initial Base Case
(summer NOx emissions, thousands of tons)
Emission Reductions. Relative to Initial Base Case
(summer NOx emissions, thousands of tons)
Incremental Annual Cost, Relative to Initial Base
Case (millions of 1990$)
Cost per Summer Ton of NOx Removed, Relative
to Initial Base Case (1990S)
Original
Budget
564
938
$1,378
$1,468
Revised
Budget
544
958
$1.440
$1,503
Incremental
Change
(20)
20
$62
$35
Percentage I
Change |
(3.5%)
2.1%
4.5%
24%
 Source ICF anahsis
2.0     Addendum Analysis on the Revised Non-EGU Emissions Inventory for Potentially Affected
        Units

        Since the EPA conducted the mam analysis documented m the RIA. the EPA has revised the non-
EGU emissions inventor,- on the basis of comments received from the States and emissions sources  This
analysis examines the emissions,  costs, and cost-effectiveness implications of the final inventory for the final
regulator) alternatives  The final regulator} alternative for non-EGU trading program sources (i e . industrial
boilers and combustion turbines)  is a 60% reduction from projected 2007 uncontrolled emission rates   The
final regulatory  alternate e for affected sources outside the trading program (i e.. stationary internal
combustion engines, and cement manufacturing operations) is based on a source category-specific evaluation
of the highest emission reduction achievable for less than $5.000 per ozone season ton reduced

        Table 3 contains the original and revised source counts, baseline emissions, and emission reductions
for the final regulatory alternative affecting industrial boilers and turbines  Table 4 shows the original and
revised compliance costs (control costs plus administrative costs), and average cost-effectiveness for these
sources  Table  5 contains the original and revised source counts, baseline emissions, and emission reductions
for the final regulatory alternative affecting 1C engines and cement manufacturing  Table 6 shows the
original and revised compliance costs (control costs plus administrative costs), and average cost-effectiveness
for these sources  As shown, the  differences in emissions, costs, and average cost-effectiveness are minor
The most notable difference is the revised cost-effectiveness for cement manufacturing, which is nearly $200
per ton less than the original analysis

        Based on this reanalysis of the effects of the NOx SIP call for these sources using the emissions
inventory, EPA has concluded that the conclusions reached in the RIA would not be materially affected by the
change in the budget
                                               ADD-3

-------
                                                    Table 3
             Original and Revised Regulatory Alternative Emission Impacts for Industrial Boilers
                                   and Combustion Turbines: 60% Control

Original Anahsis
Revised Analysis
Difference
Number of
Potentially
Affected Sources'
803
799
(4)
2007 Baseline NOx
Emissions (ozone
season tons)
194,445
196,147
1,702
2007 Post-Control
NOx Emissions
(ozone season tons)
90,193
89,653
(540)
2007 NOx
Emission
Reductions (ozone
season tons)b
104,252
106,494
2,242
1 There are a total of 803 large sources now in these categories, both controlled and uncontrolled
6 Reductions from controlled 2007 baseline are less than the nominal percentage reduction from an uncontrolled 2007 baseline indicated in the
regulator^ alternative name
                                                    Table 4
               Original and Revised Regulatory Alternative Cost Estimates for Industrial Boilers
                                   and Combustion Turbines:  60% Control

Original Anahsis
Revised Anahsis
Difference
Annual Control
Costs
(million 1990S)
$1268
$1322
$54
Annual
Monitoring and
Administrative
Costs
(million 1990S)
$26 1
$26 1
$0
Total Annual Costs
(million 1990S)
$1529
$1583
$54
Average Cost-
Effectiveness
(S/ozone season
ton)
$1,467
$1,519
$52
                                                     ADD-4

-------
                                                      Table 5
               Original and Revised Regulatory Alternathe Emission Impacts Source NOT in the
                                    Trading Program: S5,000/ton Alternative

Number of
Potentially
Affected Sources"
2007 Baseline NOx
Emissions (ozone
season tons)
2007 Post-Control
NOx Emissions
(ozone season tons)
2007 NOx
Emission
Reductions(ozone
season tons)6
1C Engines
Original Analysis
Revised Anah sis
Difference
305
302
(3)
92,494
91,599
(895)
9,801
9,717
(84)
82,567
81,882
(685)
Cement Manufacturing
Original Anah sis
Revised Analvsis
Difference
58
57
(1)
42.701
41.580
(1,121)
26,312
25,136
(1,176)
16.389
16.444
55
' There are a total of 302 large stationary 1C engines and 57 large cement manfuctunng sources in the revised analysis, both controlled and
uncontrolled
b Reductions from controlled 2007 baseline are less than the nominal percentage reduction from an uncontrolled 2007 baseline indicated in the
regulator, altematne name
                                                       ADD-5

-------
                                 Table 6
Original and Revised Regulator}- Alternative Cost Estimates for Source NOT in the
                  Trading Program:  S5,000/ton Alternative

Annual Control
Costs
(million 1990S)
Annual
Monitoring and
Administrative
Costs
(million 1990S)
Total Annual Costs
(million 1990S)
Average Cost-
Effectiveness
(S/ozone season
ton)
1C Engines
Original Analysis
Revised Analysis
Difference
$87 1
$863
($08)
$133
$11 1
($22)
$1004
$974
($30)
$1,215
$1.190
($25)
Cement Manufacturing
Original Anahsis
Revised Anahsis
Difference
$202
$200
($02)
$37
$07
($30)
$239
$207
($3 2)
$1.458
$1,259
($199)
                                  ADD-6

-------
                                     TECHNICAL REPORT DATA
                                (Please read Instnicnons on reverse before completing)
   REPORT \o
   EPA-452/R-98-003A
                                                                    3  RECIPIENT'S ACCESSION NO
 4 TITLE AND SUBTITLE
 Regulatory Impact Analysis for the NOx SIP Call, FIP, and
 Section 126 Petitions
     Volume 1:  Costs and Economic Impacts
     Volume 2:  Health  and Welfare Benefits
                  5 REPORT DATE
                  Volume 1:  September 1998
                  Volume 2:  December 1998
                  6 PERFORMING ORGANIZATION CODE
                  OAQPS/AQSSD
 7 AUTHOR(S)
 Office of Air Quality Planning and Standards
 Office of Atmospheric Programs	
                                                                    8  PERFORMING ORGANIZATION REPORT NO
 9 PERFORMING ORGANIZATION NAME AND ADDRESS

    U.S. Environmental Protection Agency
    Office of Air Quality Planning and Standards &
    Office of Atmospheric Programs
    Research Triangle Park, NC  27711
                                                                    10 PROGRAM ELEMENT NO
                  11  CONTRACT/GRANT NO
  12 SPONSORING AGENCY NAME AND ADDRESS

    Director
    Office of Air Quality Planning  and Standards
    Office of Air and Radiation
    U.S. Environmental Protection Agency
    Research Triangle Park, NC  27711	
                                                                    13 TYPE OF REPORT AND PERIOD COVERED
                  14 SPONSORING AGENCY CODE
                  EPA/200/04
  15 SUPPLEMENTARY NOTES
  16 ABSTRACT
  This report contains EPA's estimates of the annual costs and benefits of the final NOx SIP call and the
  proposed NOx FIP and CAA section 126 petition actions.  The report also contains a brief profile of
  potentially affected sources and potential economic impacts.
                                       KEY WORDS AND DOCUMENT ANALYSIS
                    DESCRIPTORS
                                                  b IDENTIFIERS'OPEN ENDED TERMS
                                                                                       c COSATI Field/Group
 Regulatory  impact analysis; benefits-cost
 comparison	
Air Pollution control
 18 DISTRIBUTION STATEMENT
   Release Unlimited
                                                  19 SECURITY CLASS (Report)
                                                     Unclassified
                                                                                       21 NO OF PAGES
                                                  Id SECURITY CLASS (Page)
                                                     Unclassified
                                                                                       22 PRICE
EP\ Form 2220-1 (Rev 4-77)   PREVIOUS EDITION IS OBSOLETE

-------