[Federal Register Volume 63, Number 207 (Tuesday, October 27, 1998)]
[Rules and Regulations]
[Pages 57356-57538]
From the Federal Register Online via the Government Publishing Office [www.gpo.gov]
[FR Doc No: 98-26773]



[[Page 57355]]

_______________________________________________________________________

Part II





Environmental Protection Agency





_______________________________________________________________________



40 CFR Parts 51, 72, 75, and 96



Finding of Significant Contribution and Rulemaking for Certain States 
in the Ozone Transport Assessment Group Region for Purposes of Reducing 
Regional Transport of Ozone; Rule

Federal Register / Vol. 63, No. 207 / Tuesday, October 27, 1998 / 
Rules and Regulations

[[Page 57356]]



ENVIRONMENTAL PROTECTION AGENCY

40 CFR Parts 51, 72, 75, and 96

[FRL-6171-2]
RIN 2060-AH10


Finding of Significant Contribution and Rulemaking for Certain 
States in the Ozone Transport Assessment Group Region for Purposes of 
Reducing Regional Transport of Ozone

AGENCY: Environmental Protection Agency (EPA).

ACTION: Final rule.

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

SUMMARY: In accordance with the Clean Air Act (CAA), today's action is 
a final rule to require 22 States and the District of Columbia to 
submit State implementation plan (SIP) revisions to prohibit specified 
amounts of emissions of oxides of nitrogen (NOX)--one of the 
precursors to ozone (smog) pollution--for the purpose of reducing 
NOX and ozone transport across State boundaries in the 
eastern half of the United States.
    Ground-level ozone has long been recognized, in both clinical and 
epidemiological research, to affect public health. There is a wide 
range of ozone-induced health effects, including decreased lung 
function (primarily in children active outdoors), increased respiratory 
symptoms (particularly in highly sensitive individuals), increased 
hospital admissions and emergency room visits for respiratory causes 
(among children and adults with pre-existing respiratory disease such 
as asthma), increased inflammation of the lung, and possible long-term 
damage to the lungs.
    In today's action, EPA finds that sources and emitting activities 
in each of the 22 States and the District of Columbia (23 
jurisdictions) emit NOX in amounts that significantly 
contribute to nonattainment of the 1-hour and 8-hour ozone national 
ambient air quality standards (NAAQS), or will interfere with 
maintenance of the 8-hour NAAQS, in one or more downwind States. 
Further, by today's action, EPA is requiring each of the affected 
upwind jurisdictions (sometimes referred to as upwind States) to submit 
SIP revisions prohibiting those amounts of NOX emissions 
which significantly contribute to downwind air quality problems. The 
reduction of those NOX emissions will bring NOX 
emissions in each of those States to within the resulting statewide 
NOX emissions budget levels established in today's rule. The 
23 jurisdictions are: Alabama, Connecticut, Delaware, District of 
Columbia, Georgia, Illinois, Indiana, Kentucky, Massachusetts, 
Maryland, Michigan, Missouri, North Carolina, New Jersey, New York, 
Ohio, Pennsylvania, Rhode Island, South Carolina, Tennessee, Virginia, 
West Virginia, and Wisconsin. These States will be able to choose any 
mix of pollution-reduction measures that will achieve the required 
reductions.

EFFECTIVE DATES: This rule is effective December 28, 1998. The 
incorporation by reference of certain publications listed in the 
regulations is approved by the Director of the Federal Register as of 
December 28, 1998.

ADDRESSES: Dockets containing information relating to this rulemaking 
(Docket No. A-96-56 and Docket No. A-9-35) are available for public 
inspection at the Air and Radiation Docket and Information Center 
(6102), US Environmental Protection Agency, 401 M Street SW, room M-
1500, Washington, DC 20460, telephone (202) 260-7548, between 8:00 a.m. 
and 4:00 p.m., Monday through Friday, excluding legal holidays. A 
reasonable fee may be charged for copying.

FOR FURTHER INFORMATION CONTACT: General questions concerning today's 
action should be addressed to Kimber S. Scavo, Office of Air Quality 
Planning and Standards, Air Quality Strategies and Standards Division, 
MD-15, Research Triangle Park, NC 27711, telephone (919) 541-3354; e-
mail: [email protected]. Please refer to SUPPLEMENTARY INFORMATION 
below for a list of contacts for specific subjects described in today's 
action.

SUPPLEMENTARY INFORMATION:

Availability of Related Information

    Documents related to the Ozone Transport Assessment Group (OTAG) 
are available on the Agency's Office of Air Quality Planning and 
Standards' (OAQPS) Technology Transfer Network (TTN) via the web at 
http://www.epa.gov/ttn/. If assistance is needed in accessing the 
system, call the help desk at (919) 541-5384 in Research Triangle Park, 
NC. Documents related to OTAG can be downloaded directly from OTAG's 
webpage at http://www.epa.gov/ttn/otag/. The OTAG's technical data are 
located at http://www.iceis.mcnc.org/OTAGDC. The notice of proposed 
rulemaking for this final action, the supplemental notice of proposed 
rulemaking, and associated documents are located at http://epa.gov/ttn/
oarpg/otagsip.html. Information related to Sections II, Weight of 
Evidence Determination of Covered States, and IV, Air Quality 
Assessment, can be obtained in electronic form from the following EPA 
website: http://www.epa.gov/scram001/regmodcenter/t28.htm. Information 
related to Section III, Determination of Budgets, may be found on the 
following EPA website: http://www.epa.gov/capi. All information in 
electronic form may also be found on diskettes that have been placed in 
the docket to this rulemaking.

For Additional Information

    For technical questions related to the air quality analyses, please 
contact Norm Possiel; Office of Air Quality Planning and Standards; 
Emissions, Monitoring, and Analysis Division; MD-14, Research Triangle 
Park, NC 27711, telephone (919) 541-5692. For legal questions, please 
contact Howard J. Hoffman, Office of General Counsel, 401 M Street SW, 
MC-2344, Washington, DC 20460, telephone (202) 260-5892. For questions 
concerning the statewide emissions budget revisions, please contact 
Laurel Schultz; Office of Air Quality Planning and Standards; 
Emissions, Monitoring, and Analysis Division; MD-14, Research Triangle 
Park, NC 27711, telephone (919) 541-5511. For questions concerning SIP 
reporting requirements, please contact Bill Johnson, Office of Air 
Quality Planning and Standards, Air Quality Strategies and Standards 
Division, MD-15, Research Triangle Park, NC 27711, telephone (919) 541-
5245. For questions concerning the model cap-and-trade rule, please 
contact Rob Lacount, Office of Atmospheric Programs, Acid Rain 
Division, MC-6204J, 401 M Street SW, Washington, DC 20460, telephone 
(202) 564-9122. For questions concerning the regulatory cost analysis 
of electricity generating sources, please contact Ravi Srivastava, 
Office of Atmospheric Programs, Acid Rain Division, MC-6204J, 401 M 
Street SW, Washington DC 20460, telephone (202) 564-9093. For questions 
concerning the regulatory cost analysis of other stationary sources and 
questions concerning the Regulatory Impact Analysis (RIA), please 
contact Scott Mathias, Office of Air Quality Planning and Standards, 
Air Quality Strategies and Standards Division, MD-15, Research Triangle 
Park, NC 27711, telephone (919) 541-5310.

Outline

I. Background
    A. Summary of Rulemaking and Affected States
    B. General Factual Background
    C. Statutory and Regulatory Background
    1. CAA Provisions
    a. 1970 and 1977 CAA Amendments
    b. 1990 CAA Amendments
    2. Regulatory Structure
    a. March 2, 1995 Policy
    b. OTAG

[[Page 57357]]

    c. EPA's Transport SIP Call Regulatory Efforts
    d. Revision of the Ozone NAAQS
    D. Section 126 Petitions
    E. OTAG
    F. Discussion of Comment Period and Availability of Key 
Information
    1. Request for Extension of the Comment Period
    2. Request for Time to Conduct Additional Modeling
    3. Availability of Key Information
    4. Public Hearings
    G. Implementation of Revised Air Quality Standards
    H. Summary of Major Changes between Proposals and Final Rule
    1. EPA's Analytical Approach (Section II.A)
    2. Cost Effectiveness of Emissions Reductions (Section II.D)
    3. Determination of Budgets (Section III)
    4. NOX Control Implementation and Budget Achievement 
Dates (Section V)
    5. SIP Criteria (Section VI.A)
    6. Emissions Reporting Requirements for States (Section VI.B)
    7. NOX Budget Trading Program (Section VII)
    8. Interaction with Title IV NOX Rule (Section VIII)
    9. Administrative Requirements (Section X)
II. EPA's Analytical Approach
    A. Interpretation of the CAA's Transport Provisions
    1. Authority and Process for Requiring SIP Submissions under the 
1-Hour Ozone NAAQS
    a. Authority for Requiring SIP Submissions under the 1-Hour 
NAAQS
    b. Process for Requiring SIP Submissions under the 1-Hour NAAQS
    2. Authority and Process for Requiring SIP Submissions under the 
8-Hour Ozone NAAQS
    a. Authority for Requiring SIP Submissions under the 8-Hour 
NAAQS
    b. Process for Requiring SIP Submissions under the 8-hour 
Standard
    3. Requirements of Section 110(a)(2)(D)
    a. Summary
    b. Determination of Meaning of ``Nonattainment'
    c. Definition of Significant Contribution
    d. Multi-factor Test for Determining Significant Contribution
    e. Air Quality Factors
    f. Determination of Highly Cost-effective Reductions and of 
Budgets
    g. Other Considerations in Determination of Significant 
Contribution
    h. Interfere with Maintenance
    i. Dates
    j. Downwind Areas' Control Obligations
    k. Section 110(a)(2)(D) Caselaw
    B. Alternative Interpretation of Section 110(a)(2)(D)
    C. Weight-of-Evidence Determination of Covered States
    1. Major Findings from OTAG-Related Technical Analyses
    2. Summary of Notice of Proposed Rulemaking Weight-of-Evidence 
Approach
    a. Quantification of Contributions
    b. Evaluation of 1-Hour and 8-Hour Contributions
    c. Comments and Responses on Proposed Weight-of-Evidence 
Approach to Significant Contribution
    3. Analysis of State-specific Air Quality Factors
    a. Overall Nature of Ozone Problem (``Collective Contribution'')
    b. Extent of Downwind Nonattainment Problems
    c. Air Quality Impacts of Upwind Emissions on Downwind 
Nonattainment
    4. Confirmation of States Making a Contribution to Downwind 
Nonattainment
    a. Analysis Approach
    b. States Which Contain Sources That Significantly Contribute to 
Downwind Nonattainment
    c. Examples of Contributions From Upwind States to Downwind 
Nonattainment
    d. Conclusions From Air Quality Evaluation of Downwind 
Contributions
    5. States Not Covered by This Rulemaking
    D. Cost Effectiveness of Emissions Reductions
    1. Sources Included in the Cost-Effectiveness Determination
    a. Electricity Generating Boilers and Turbines
    b. Other Stationary Sources
    2. Sources Not Included in the Cost-Effectiveness Determination
    a. Area Sources
    b. Small Point Sources
    c. Mobile Sources
    d. Other Stationary Sources
    e. Conclusion
    E. Other Considerations
    1. Consistency of Regional Reductions with Attainment Needs of 
Downwind Areas
    a. General Discussion
    b. 8-hour Nonattainment Problems
    c. Commenters' Concerns
    2. Equity Considerations
    3. General Cost Considerations
    4. Conclusion
III. Determination of Budgets
    A. General Comments on the Base Emission Inventory
    1. Quality
    2. Availability
    B. Electricity Generating Units (EGUs)
    1. Base Inventory
    2. Growth
    a. Growth Rates
    b. Use of IPM
    c. Use of ``Corrected'' Growth Rates
    3. Budget Calculation
    a. Input vs. Output
    b. Alternative Emission Limits
    c. Consideration of the Climate Change Action Plan
    C. Non-EGU Point Sources
    1. Base Inventory
    2. Growth
    3. Budget Calculation
    a. Proposed Control Assumptions
    b. Small Source Exemption
    c. Exemptions for Other Non-EGU Point Sources
    d. Sources Without Adequate Control Information
    e. Case-By-Case Analysis of Control Measures
    f. Cost Effectiveness
    g. Industrial Boiler Control Costs
    h. Cement Manufacturing
    i. Stationary Internal Combustion Engines
    j. Industrial Boilers and Turbines
    k. Municipal Waste Combustors (MWCs)
    D. Highway Mobile Sources
    1. Base Inventory
    2. Growth
    3. Budget Calculation
    a. I/M Program Coverage
    b. Emissions Cap
    c. Tier 2 Standards
    d. Low Sulfur Fuel
    e. Conformity
    E. Stationary Area and Nonroad Mobile Sources
    1. Base Inventory
    2. Growth
    3. Budget Calculation
    F. Other Budget Issues
    1. Uniform vs. Regional Controls
    2. Seasonal vs. Annual Controls
    3. Full vs. Partial States
    4. NOx Waivers
    5. Recalculation of Budgets
    6. Compliance Supplement Pool
    a. Size of the Compliance Supplement Pool
    b. State Distribution of the Compliance Supplement Pool
    7. Banking
    a. Banking Starting in 2003
    b. Management of Banked Allowances
    c. Early Reduction Credits
    G. Final Statewide Budgets
    1. EGU
    a. Description of Selected Approach
    b. Summary of Budget Component
    2. Non-EGU Point Sources
    a. Description of Selected Approach
    b. Summary of Budget Component
    3. Mobile and Area Sources
    a. Description of Selected Budget Approach
    b. Summary of Budget Component
    4. Potential Alternatives to Meeting the Budget
    5. Statewide Budgets
IV. Air Quality Assessment
    A. Assessment of Proposed Statewide Budgets
    B. Comments and Responses
    C. Assessment of Alternative Control Levels
    1. Scenarios Modeled
    2. Emissions for Model Runs
    3. Modeling Results
    a. Impacts of Alternative Controls
    b. Impacts of Upwind Controls on Downwind Nonattainment
    c. Summary of Findings
V. NOx Control Implementation and Budget Achievement Dates
    A. NOx Control Implementation Date
    1. Practicability
    a. Combustion Controls
    b. Post-Combustion Controls
    2. Relationship to SIP Submittal Date
    3. Rationale
    B. Budget Achievement Date
VI. SIP Criteria and Emissions Reporting Requirements
    A. SIP Criteria
    1. Schedule for SIP Revision
    2. Approvability Criteria
    a. Source Categories Subject to Additional Approvability 
Criteria

[[Page 57358]]

    b. Pollution Abatement Requirements
    c. Monitoring Requirements
    d. Approvability of Trading Program
    3. Sanctions
    4. FIPs
    B. Emissions Reporting Requirements for States
    1. Use of Inventory Data
    2. Response to Comments
    3. Final Rule
    4. Data Elements to be Reported
    5. 2007 Report
    6. Ozone Season Reporting
    7. Data Reporting Procedures
    8. Confidential Data
    C. Timeline
VII. NOX Budget Trading Program
    A. General Background
    B. NOX Budget Trading Program Rulemaking Overview
    C. General Design of NOX Budget Trading Program
    1. Appropriateness of Trading Program
    2. Alternative Market Mechanisms
    3. State Adoption of Model Rule
    a. Process for Adoption
    b. Model Rule Variations
    4. Unrestricted Trading Market
    a. Geographic Issues
    b. Episodic Issues
    D. Applicability
    1. Core Sources
    a. Commenters Who Felt the Core Group Should Not Be Changed
    b. Commenters Who Felt the Core Group Should Be Expanded
    c. Commenters Who Felt the Core Group Is Overly Inclusive
    2. Mobile/Area Sources
    3. Monitoring
    a. Use of Part 75 to Ensure Compliance with the NOX 
Budget Trading Program
    b. Use of CEMS on Large Units
    c. Commenters Who do not Believe that CEMS are Necessary
    d. Issues Related to Monitoring and Reporting Needed to Support 
a Heat Input Allocation Methodology
    e. Amendments to Part 75
    E. Emission Limitations/Allowance Allocations
    1. Timing Requirements
    2. Options for NOX Allowance Allocation Methodology
    3. New Source Set-Aside
    4. Optional NOX Allocation Methodology in Model Rule
    F. Banking Provisions
    1. Banking Starting in 2003
    2. Management of Banked Allowances
    3. Early Reduction Credits
    4. Optional Methodology for Issuing Early Reduction Credits
    5. Integrating the OTC Program with the NOX Budget 
Trading Program's Banking Provisions
    G. New Source Review
VIII. Interaction with Title IV NOX Rule
IX. Non-Ozone Benefits of NOX Emissions Decreases
    A. Summary of Comments
    B. Response to Comments
    1. Drinking Water Nitrate
    2. Eutrophication
    3. Regulatory Impact Analysis
    4. Justification for Rulemaking
X. Administrative Requirements
    A. Executive Order 12866: Regulatory Impact Analysis
    B. Regulatory Flexibility Act: Small Entity Impacts
    C. Unfunded Mandates Reform Act
    D. Paperwork Reduction Act
    E. Executive Order 13045: Protection of Children from 
Environmental Health Risks and Safety Risks
    1. Applicability of E.O. 13045
    2. Children's Health Protection
    F. Executive Order 12898: Environmental Justice
    G. Executive Order 12875: Enhancing the Intergovernmental 
Partnerships
    H. Executive Order 13084: Consultation and Coordination with 
Indian Tribal Governments
    I. Judicial Review
    J. Congressional Review Act
    K. National Technology Transfer and Advancement Act

Appendix A--Detailed Discussion of Changes to Part 75

CFR Revisions and Additions

Part 51
Sec. 51.121
Sec. 51.122
Part 72
Part 75
Part 96

I. Background

A. Summary of Rulemaking and Affected States

    By notice of proposed rulemaking (NPR, proposal, or ``proposed SIP 
call'') (62 FR 60318, November 7, 1997) and by supplemental notice 
(SNPR or supplemental proposal) (63 FR 25902, May 11, 1998), EPA 
proposed to find that NOX emissions from sources and 
emitting activities (sources) in 23 jurisdictions (hereinafter also 
referred to as States) will significantly contribute to nonattainment 
of the 1-hour and 8-hour ozone NAAQS, or will interfere with 
maintenance of the 8-hour NAAQS, in one or more downwind States 
throughout the Eastern United States. The EPA based these proposals on 
data generated by OTAG, public comments, and other relevant 
information. Today's final action confirms that proposed finding. It 
also requires, under CAA section 110(a)(1) and 110(k)(5), that the 23 
jurisdictions adopt and submit SIP revisions that, in order to assure 
that their SIPs meet the requirements of section 110(a)(2)(D)(i)(I), 
contain provisions adequate to prohibit sources in those States from 
emitting NOX in amounts that ``contribute significantly to 
nonattainment in, or interfere with maintenance by,'' a downwind State. 
The 23 jurisdictions are: Alabama, Connecticut, Delaware, District of 
Columbia, Georgia, Illinois, Indiana, Kentucky, Massachusetts, 
Maryland, Michigan, Missouri, North Carolina, New Jersey, New York, 
Ohio, Pennsylvania, Rhode Island, South Carolina, Tennessee, Virginia, 
West Virginia, and Wisconsin.
    Each of these States and the District of Columbia is required to 
adopt and submit by September 30, 1999, a SIP revision. The SIP 
revision must contain measures that will assure that sources in the 
State reduce their NOX emissions sufficiently to eliminate 
the amounts of NOX emissions that contribute significantly 
to nonattainment, or that interfere with maintenance, downwind. By 
eliminating these amounts of NOX emissions, the control 
measures will assure that the remaining NOX emissions will 
meet the level identified in today's rule as the State's NOX 
emissions budget. For simplicity, this final rule may refer to the 
amounts that such SIP provisions must prohibit in order to meet the 
statute as the ``significant amounts'' of NOX emissions. 
After prohibiting these significant amounts of NOX, the 
remaining amounts emitted by sources in the covered States will not 
``significantly contribute to nonattainment, or interfere with 
maintenance by,'' a downwind State, under section 110(a)(2)(D)(i)(I). 
Section II.C, Weight-of-Evidence Determination of Covered States, 
describes how EPA determined which States include sources that emit 
NOX in amounts of concern (the ``covered'' States), and 
Sections II.D, Cost Effectiveness of Emissions Reductions; II.E, 
Comparison of Upwind and Downwind Costs; and III, Determination of 
Budgets, describe how EPA determined the significant amounts of 
emissions and the resulting statewide emissions budgets for the States 
identified above. Section IV, Air Quality Assessment, discusses air 
quality analyses conducted by EPA which help confirm the decisions and 
requirements set forth in this rulemaking. Section V, NOX 
Control Implementation and Budget Achievement Dates, primarily 
discusses the dates by which (1) the States must submit SIP revisions 
in response to today's action, (2) the sources must implement the 
measures the States choose for the purpose of prohibiting the 
significant amounts of NOX, and (3) the States are projected 
to achieve the budget levels. Section VI, SIP Criteria and Emissions 
Reporting Requirements, describes the SIP requirements themselves.
    The SIP requirements permit each State to determine what measures 
to adopt to prohibit the significant amounts and hence meet the 
necessary emissions budget. Consistent with OTAG's recommendations to 
achieve

[[Page 57359]]

NOX emissions decreases primarily from large stationary 
sources in a trading program, EPA encourages States to consider 
electric utility and large boiler controls under a cap-and-trade 
program as a cost-effective strategy. The recommended cap-and-trade 
program is described in more detail in Section VII, NOX 
Budget Trading Program. The EPA also recognizes that promotion of 
energy efficiency can contribute to a cost-effective strategy. In 
Section VIII, Interaction with Title IV NOX rule, EPA 
explains that it is not adopting proposed revisions to the title IV 
NOX rule concerning the relationship between this rulemaking 
and the title IV NOX rule. The remaining parts of today's 
action include Section IX, Non-Ozone Benefits of NOX 
Reductions, and Section X, Administrative Requirements.
    The EPA also conducted a RIA which is available in the docket to 
this rulemaking as a technical support document (TSD), entitled 
``Regulatory Impact Analysis for the Regional NOX SIP Call'' 
(docket no. VI-B-09). A detailed explanation of how EPA calculated the 
budgets is also available as a TSD entitled ``Development of Modeling 
Inventory and Budgets for the Regional NOX SIP Call'' 
(docket no. VI-B-10). These two TSDs have been revised for the final 
rulemaking. A detailed explanation of the air quality modeling analyses 
is also available, entitled ``Air Quality Modeling Technical Support 
Document for the Regional NOX SIP Call'' (docket no. VI-B-
11) for this final rulemaking. This preamble for today's notice 
responds to some of the comments, but another document, entitled 
``Response to Significant Comments on the Finding of Significant 
Contribution and Rulemaking for Certain States in the OTAG Region for 
Purposes of Reducing Regional Transport of Ozone,'' is included in the 
docket (docket no. VI-C-01).

B. General Factual Background

    In today's action, EPA takes a significant step toward reducing 
ozone in the eastern half of the country. Ground-level ozone, the main 
harmful ingredient in smog, is produced in complex chemical reactions 
when its precursors, volatile organic compounds (VOC) and 
NOX, react in the presence of sunlight. The chemical 
reactions that create ozone take place while the pollutants are being 
blown through the air by the wind, which means that ozone can be more 
severe many miles away from the source of emissions than it is at the 
source.
    The science of ozone formation, transport, and accumulation is 
complex. Ozone is produced and destroyed in a cyclical set of chemical 
reactions involving NOX, VOC and sunlight. Emissions of 
NOX and VOC are necessary for the formation of ozone in the 
lower atmosphere. In part of the cycle of reactions, ozone 
concentrations in an area can be lowered by the reaction of nitric 
oxide with ozone, forming nitrogen dioxide; as the air moves downwind 
and the cycle continues, the nitrogen dioxide forms additional ozone. 
The importance of this reaction depends, in part, on the relative 
concentrations of NOX, VOC and ozone, all of which change 
with time and location.
    At ground level, ozone can cause a variety of ill effects to human 
health, crops and trees. Specifically, ground-level ozone has been 
shown in clinical and/or epidemiologial studies to have the following 
health effects:

     Decreased lung function, primarily in children 
active outdoors
     Increased respiratory symptoms, particularly in 
highly sensitive individuals
     Hospital admissions and emergency room visits for 
respiratory causes among children and adults with pre-existing 
respiratory disease such as asthma
     Inflammation of the lung
     Possible long-term damage to the lungs or even 
premature death.

    The new 8-hour primary ambient air quality standard (62 FR 38856, 
July 18, 1997) will provide increased protection to the public from 
these health effects.
    Each year, ground-level ozone above background is also responsible 
for significant agricultural crop yield losses. Ozone also causes 
noticeable foliar damage in many crops, trees, and ornamental plants 
(i.e., grass, flowers, shrubs, and trees) and causes reduced growth in 
plants. Studies indicate that current ambient levels of ozone are 
responsible for damage to forests and ecosystems (including habitat for 
native animal species).
    As part of the efforts to reduce harmful levels of smog, EPA, 
today, is establishing a requirement for certain States to revise their 
SIPs in order to implement the necessary regional-scale reductions in 
NOX emissions, and, thereby, reduce transported 
NOX and ozone. Since air pollution travels across county and 
State lines, it is essential for State governments and air pollution 
control agencies to cooperate to solve the problem.
    Currently, the following areas, impacted by the 23 jurisdictions 
that are the subject of today's rulemaking, are designated 
nonattainment areas for ozone under the 1-hour NAAQS:

Atlanta, GA
Baltimore, MD
Birmingham, AL
Boston-Lawrence-Worcester (eastern MA), MA-NH
Chicago-Gary-Lake County, IL-IN
Cincinnati-Hamilton, OH-KY
Door County, WI
Greater Connecticut
Kent & Queen Anne's Counties, MD
Lancaster, PA
Louisville, KY-IN
Manitowoc County, WI
Milwaukee-Racine, WI
Muskegon, MI
New York-Northern New Jersey-Long Island, NY-NJ-CT
Philadelphia-Wilmington-Trenton, PA-NJ-DE-MD
Pittsburgh-Beaver Valley, PA
Portland, ME
Portsmouth-Dover-Rochester, NH
Providence (All RI), RI
St. Louis, MO-IL
Springfield (western MA), MA
Washington, DC-MD-VA

    These areas include many of the major urban centers in the eastern 
half of the Nation. The combined population for these areas is 
approximately 61.5 million. As described elsewhere, the reductions 
called for in today's action will reduce ozone levels throughout these 
areas.
    Many more areas currently violate the 8-hour NAAQS. The EPA 
estimates that a total population of approximately 73 million in the 23 
jurisdictions live in counties for which air quality is monitored to be 
in violation of that NAAQS. The reductions called for in today's action 
will reduce ozone levels throughout these areas as well.
    Moreover, as discussed below, many of these areas are expected to 
be classified as ``transitional,'' which means, in most cases, that 
they are expected to come into attainment solely as a result of the 
reductions required by today's action. Thus, for those who live in 
these areas, the reductions required under today's action, in-and-of-
themselves, are expected to mean the difference between unhealthful 
ozone levels and acceptable ozone levels.
    Please note that EPA will not designate ozone nonattainment areas 
for the 8-hour NAAQS until 2000, and these designations will be based 
on the data that are most recently available at that time.

C. Statutory and Regulatory Background

1. CAA Provisions
    a. 1970 and 1977 CAA Amendments. For almost 30 years, Congress has 
focused major efforts on curbing ground-level ozone. In 1970, Congress 
amended the CAA to require, in title I, that EPA issue, and 
periodically review

[[Page 57360]]

and if necessary revise, NAAQS for ubiquitous air pollutants (sections 
108 and 109). Congress required the States to submit SIPs to attain and 
maintain those NAAQS, and Congress included, in section 110, a list of 
minimum requirements that SIPs must meet. Congress anticipated that 
areas would attain the NAAQS by 1975.
    In 1977, Congress amended the CAA by providing, among other things, 
additional time for areas that were not attaining the ozone NAAQS to do 
so, as well as by imposing specific SIP requirements for those 
nonattainment areas. These provisions first required the designation of 
areas as attainment, nonattainment, or unclassifiable, under section 
107; and then required that SIPs for ozone nonattainment areas include 
the additional provisions set out in part D of title I, as well as 
demonstrations of attainment of the ozone NAAQS by either 1982 or 1987 
(section 172).
    In addition, the 1977 Amendments included two provisions focused on 
interstate transport of air pollutants: the predecessor to current 
section 110(a)(2)(D), which requires SIPs for all areas to constrain 
emissions with certain adverse downwind effects; and section 126, 
which, in general, authorizes a downwind State to petition EPA to 
impose limits directly on upwind sources found to adversely affect that 
State. Section 110(a)(2)(D), which is key to the present action, is 
described in more detail below.
    b. 1990 CAA Amendments. In 1990, Congress amended the CAA to better 
address, among other things, continued nonattainment of the 1-hour 
ozone NAAQS; the requirements that would apply if EPA revised the 1-
hour standard; and transport of air pollutants across State boundaries 
(Pub. L. 101-549, Nov. 15, 1990, 104 Stat. 2399, 42 U.S.C., 7401-
7671q). Numerous provisions added, or revised, by the 1990 Amendments 
are relevant to today's proposal.
    (1) 1-Hour Ozone NAAQS. In the 1990 Amendments, Congress required 
the States and EPA to review and, if necessary, revise the designation 
of areas as attainment, nonattainment, and unclassifiable under the 
ozone NAAQS in effect at that time, which was the 1-hour standard 
(section 107(d)(4)). Areas designated as nonattainment were divided 
into, primarily, five classifications based on air quality design 
values (section 181(a)(1)). Each classification carries specific 
requirements, including new attainment dates (sections 181-182). In 
increasing severity of the air quality problem, these classifications 
are marginal, moderate, serious, severe and extreme. The OTAG region 
includes nonattainment areas of all classifications except extreme.
    As amended in 1990, the CAA requires States containing ozone 
nonattainment areas classified as moderate or above to submit several 
SIP revisions at various times. One set of SIP revisions included 
specified control measures, such as reasonably available control 
technology (RACT) for existing VOC and NOX sources (section 
182(b)(2), 182(f)). In addition, the CAA requires the reduction of VOC 
in the amount of 15 percent by 1996 from a 1990 baseline (section 
182(b)(1)). Further, for nonattainment areas classified as serious and 
above, the CAA requires the reduction of VOC or NOX 
emissions in the amount of 9 percent over each 3-year period from 1996 
through the attainment date (the rate-of-progress (ROP) SIP 
submittals), under section 182(c)(2)(B). In addition, the CAA requires 
a demonstration of attainment, including air quality modeling, for the 
nonattainment area (the attainment demonstration), as well as SIP 
measures containing any additional reductions that may be necessary to 
attain by the applicable attainment date (section 182(c)-(e)). The CAA 
established November 15, 1994 as the required date for the ROP and 
attainment demonstration SIP submittals for areas classified as serious 
and above.1
---------------------------------------------------------------------------

    \1\ For moderate ozone nonattainment areas, the attainment 
demonstration was due November 15, 1993 (section 182(b)(1)(A)), 
except that if the State elected to conduct an urban airshed model, 
EPA allowed an extension to November 15, 1994.
---------------------------------------------------------------------------

    (2) Revised NAAQS. Section 109(d) of the CAA requires periodic 
review and, if appropriate, revision of the NAAQS. As amended in 1990, 
the CAA further requires EPA to designate areas as attainment, 
nonattainment, and unclassifiable under a revised NAAQS (section 
107(d)(1); section 6103, Pub. L. 105-178). The CAA authorizes EPA to 
classify areas that are designated nonattainment under the new NAAQS 
and to establish for those areas attainment dates that are as 
expeditiously as practicable, but not to exceed 10 years from the date 
of designation (section 172(a)).
    (3) General Requirements. The CAA continues, in revised form, 
certain requirements, dating from the 1970 Amendments, which pertain to 
all areas, regardless of their designation. All areas are required to 
submit SIPs within certain timeframes (section 110(a)(1)), and those 
SIPs must include specified provisions, under section 110(a)(2). In 
addition, SIPs for nonattainment areas are generally required to 
include additional specified control requirements, as well as controls 
providing for attainment of any revised NAAQS and periodic reductions 
providing ``reasonable further progress'' in the interim (section 
172(c)).
    (4) Provisions Concerning Transport of Ozone and Its Precursors. 
The 1990 Amendments reflect general awareness by Congress that ozone is 
a regional, and not merely a local, problem. As described above, ozone 
and its precursors may be transported long distances across State lines 
to combine with ozone and precursors downwind, thereby exacerbating the 
ozone problems downwind. The phenomenon of ozone transport was not 
generally recognized until relatively recently. Yet, ozone transport is 
a major reason for the persistence of the ozone problem, 
notwithstanding the imposition of numerous controls, both Federal and 
State, across the country.
    Section 110(a)(2)(D) provides one of the most important tools for 
addressing the problem of transport. This provision, which applies by 
its terms to all SIPs for each pollutant covered by a NAAQS, and for 
all areas regardless of their attainment designation, provides that a 
SIP must contain adequate provisions prohibiting its sources from 
emitting air pollutants in amounts that will contribute significantly 
to nonattainment, or interfere with maintenance, in one or more 
downwind States.
    Section 110(k)(5) authorizes EPA to find that a SIP is 
substantially inadequate to meet any CAA requirement. If EPA makes such 
a finding, it must require the State to submit, within a specified 
period, a SIP revision to correct the inadequacy.
    The CAA further addresses interstate transport of pollution in 
section 126, which Congress revised slightly in 1990. Subsection (b) of 
that provision authorizes each State (or political subdivision) to 
petition EPA for a finding designed to protect that entity from upwind 
sources of air pollutants.2
---------------------------------------------------------------------------

    \2\ In addition, section 115 authorizes EPA to require a SIP 
revision when one or more sources within a State ``cause or 
contribute to air pollution which may reasonably be anticipated to 
endanger public health or welfare in a foreign country.''
---------------------------------------------------------------------------

    In addition, the 1990 Amendments added section 184, which 
delineates a multistate ozone transport region (OTR) in the Northeast, 
requires specific additional controls for all areas (not only 
nonattainment areas) in that region, and establishes the Ozone 
Transport Commission (OTC) for the purpose of recommending to EPA 
regionwide controls affecting all areas in that region. At the same 
time, Congress added section 176A, which authorizes

[[Page 57361]]

the formation of transport regions for other pollutants and in other 
parts of the country.
2. Regulatory Structure
    a. March 2, 1995 Policy. Notwithstanding significant efforts, the 
States generally were not able to meet the November 15, 1994 statutory 
deadline for the attainment demonstration and ROP SIP submissions 
required under section 182(c). The major reason for this failure was 
that at that time, States with downwind nonattainment areas were not 
able to address transport from upwind areas. As a result, in a 
memorandum from Mary D. Nichols, Assistant Administrator for Air and 
Radiation, dated March 2, 1995, entitled ``Ozone Attainment 
Demonstrations,'' (March 2, 1995 Memorandum or the Memorandum), EPA 
recognized the efforts made by States and the remaining difficulties in 
making the ROP and attainment demonstration submittals. The EPA 
recognized that development of the necessary technical information, as 
well as the control measures necessary to achieve the large level of 
reductions likely to be required, had been particularly difficult for 
the States affected by ozone transport.
    Accordingly, as an administrative remedial matter, the Memorandum 
indicated that EPA would establish new timeframes for SIP submittals. 
The Memorandum indicated that EPA would divide the required SIP 
submittals into two phases. Phase I generally consisted of (i) SIP 
measures providing for ROP reductions due by the end of 1999, (ii) an 
enforceable SIP commitment to submit any remaining required ROP 
reductions on a specified schedule after 1996, and (iii) an enforceable 
SIP commitment to submit the additional SIP measures needed for 
attainment. Phase II consists of the remaining submittals, beginning in 
1997.
    The Phase II submittals primarily consisted of the remaining ROP 
SIP measures, the attainment demonstration and additional rules needed 
to attain, and any regional controls needed for attainment by all areas 
in the region. The March 2, 1995 Memorandum indicated that the 
attainment demonstration, target calculations for the post-1999 ROP 
milestones, and identification of rules needed to attain and for post-
1999 ROP were due in mid-1997. To allow time for States to incorporate 
the results of the OTAG modeling into their local plans, EPA extended 
the mid-1997 submittal date to April 1998.3
---------------------------------------------------------------------------

    \3\ Guidance for Implementing the 1-hour Ozone and Pre-Existing 
PM10 NAAQS, Memorandum from Richard D. Wilson, dated December 29, 
1997.
---------------------------------------------------------------------------

    b. OTAG. In addition, the March 2, 1995 Memorandum called for an 
assessment of the ozone transport phenomenon. The Environmental Council 
of the States (ECOS) had recommended formation of a national work group 
to allow for a thoughtful assessment and development of consensus 
solutions to the problem. The OTAG was a partnership between EPA, the 
37 easternmost States and the District of Columbia, industry 
representatives, and environmental groups. The OTAG's air quality 
modeling and recommendations formed the basis for today's action.
    c. EPA's Transport SIP Call Regulatory Efforts. Shortly after OTAG 
began its work, EPA began to indicate that it intended to issue a SIP 
call to require States to implement the reductions necessary to address 
the ozone transport problem. On January 10, 1997 (62 FR 1420), EPA 
published a notice of intent that articulated this goal and indicated 
that before taking final action, EPA would carefully consider the 
technical work and any recommendations of OTAG. The EPA published the 
NPR for the NOX SIP call by notice dated November 7, 1997 
(62 FR 60319). The NPR proposed to make a finding of significant 
contribution due to transported NOX emissions to 
nonattainment or maintenance problems downwind and to assign 
NOX emissions budgets for 23 jurisdictions. The EPA 
published a supplemental notice of proposed rulemaking (SNPR) by notice 
dated May 11, 1998 (63 FR 25902) which proposed a model NOX 
budget trading program and State reporting requirements and provided 
the air quality analyses of the proposed statewide NOX 
emissions budgets. The EPA received approximately 700 comments on these 
proposals. The comment periods are described in Section I.F, Discussion 
of Comment Period and Availability of Key Information. Throughout the 
course of the rulemaking, EPA has added information to the docket. By 
notice dated August 24, 1998 (63 FR 45032), EPA published a notice of 
availability listing the additional documents placed in the docket.
    d. Revision of the Ozone NAAQS. On July 18, 1997 (62 FR 38856), EPA 
issued its final action to revise the NAAQS for ozone. The EPA's 
decision to revise the standard was based on the Agency's review of the 
available scientific evidence linking exposures to ambient ozone to 
adverse health and welfare effects at levels allowed by the pre-
existing 1-hour ozone standards. The 1-hour primary standard was 
replaced by an 8-hour standard at a level of 0.08 parts per million 
(ppm), with a form based on the 3-year average of the annual fourth-
highest daily maximum 8-hour average ozone concentration measured at 
each monitor within an area. The new primary standard will provide 
increased protection to the public, especially children and other at-
risk populations, against a wide range of ozone-induced health effects. 
Health effects are described in paragraph I.B, General Factual 
Background. The EPA retained the applicability of the 1-hour NAAQS for 
existing nonattainment areas until such time as EPA determines that an 
area has attained the 1-hour NAAQS (40 CFR 50.9(b)).
    The pre-existing 1-hour secondary ozone standard was replaced by an 
8-hour standard identical to the new primary standard. The new 
secondary standard will provide increased protection to the public 
welfare against ozone-induced effects on vegetation.

D. Section 126 Petitions

    In a separate rulemaking, EPA is proposing action on petitions 
submitted by eight northeastern States under section 126 of the CAA. 
Each petition specifically requests that EPA make a finding that 
NOX emissions from certain major stationary sources 
significantly contribute to ozone nonattainment problems in the 
petitioning State. The eight States are Connecticut, Massachusetts, 
Maine, New Hampshire, New York, Pennsylvania, Rhode Island, and 
Vermont.
    Both the NOX SIP call and the section 126 petitions are 
designed to address ozone transport through reductions in upwind 
NOX emissions. However, the EPA's response to the section 
126 petitions differs from EPA's action in the NOX SIP call 
rulemaking in several ways. In today's NOX SIP call, EPA is 
determining that certain States are or will be significantly 
contributing to nonattainment or maintenance problems in downwind 
States. The EPA is requiring the upwind States to submit SIP provisions 
to reduce the amounts of each State's NOX emissions that 
significantly contribute to downwind air quality problems. The States 
will have the discretion to select the mix of control measures to 
achieve the necessary reductions. By contrast, under section 126, if 
findings of significant contribution are made for any sources 
identified in the petitions, EPA would determine the necessary 
emissions

[[Page 57362]]

limits to address the amount of significant contribution and would 
directly regulate the sources. A section 126 remedy would apply only to 
sources in States named in the petitions.
    Based on the view that the SIP call and section 126 petitions are 
both designed to achieve the same goal, several commenters urged EPA to 
coordinate the two actions to the maximum extent possible. The EPA 
agrees that the two actions are closely related and, therefore, should 
be coordinated. This will help provide certainty for State and business 
planning requirements. In addition, this coordination can help to 
facilitate a trading program among sources in SIP call States that 
choose to participate in the NOX trading program, and any 
section 126 sources that would be subject to a Federal NOX 
trading program.
    The section 126 provisions require that any control remedy be 
implemented within 3 years from the date of the finding that major 
sources or a group of stationary sources emit or would emit in 
violation of the relevant prohibition in section 110(a)(2)(D). Under 
EPA's anticipated rulemaking schedule 4 on the petitions, 
the compliance date for sources for which EPA makes such a finding 
could be April 30, 2002; November 30, 2002; or May 1, 2003. Several 
commenters expressed concern that the compliance deadline under section 
126 was driving EPA's decision on the compliance deadline for the 
NOX SIP call. Therefore, they believed that no changes would 
be made in the proposed NOX SIP call deadline in response to 
comments.
---------------------------------------------------------------------------

    \4\ The eight northeastern States that filed section 126 
petitions also filed suit in the District Court for the Southern 
District of New York, to compel EPA to take action on those 
petitions within prescribed periods. State of Connecticut v. 
Browner, No. 98-1376 (S.D.N.Y., filed Feb. 25, 1998). The EPA and 
the eight northeastern States jointly filed a motion to enter a 
consent order prescribing certain dates for EPA action.
---------------------------------------------------------------------------

    While EPA believes it is advantageous to coordinate the section 126 
and NOX SIP call actions, EPA disagrees that this constrains 
EPA from being responsive to public comments and considering 
alternative compliance dates. See discussion below in Section V, 
NOX Control Implementation and Budget Attainment Dates.
    In the NOX SIP call NPR, EPA proposed that States be 
required to submit SIPs within 12 months of the final SIP call. One 
commenter asserted that the timing and terms of the rulemaking schedule 
for the section 126 petitions precludes EPA from considering public 
comments advocating different SIP due dates for the NOX SIP 
call. The section 126 rulemaking schedule provides several options. One 
option would allow findings on the petitions to be deferred pending 
certain actions by the States and EPA on State submittals in response 
to the NOX SIP call. The premise for the specified schedule 
is that the SIP due date would be September 30, 1999 (i.e., roughly 12 
months from signature of the notice on the final NOX SIP 
call). As discussed below in Section VI, SIP Revision Criteria and 
Schedule, EPA continues to believe 12 months is an appropriate 
timeframe. However, had EPA determined that a longer timeframe for SIP 
submittal was warranted, the section 126 rulemaking schedule would not 
have restricted EPA from establishing a later due date.
    One commenter supported the section 126 rulemaking schedule because 
they thought it had the effect of using the SIP process rather than the 
source-based petitions in that it provides an option of deferring 
section 126 findings if EPA approves a State's NOX SIP. 
Another commenter thought that the conditions for deferring section 126 
findings were too stringent, and, therefore, section 126 would 
inevitably be triggered prior to approval of any SIP provisions. This 
issue is discussed in detail in Section II.A.2.c. in the NPR EPA just 
issued on the section 126 petitions, which appears in the docket.

E. OTAG

    As discussed in the proposed SIP call, OTAG completed the most 
comprehensive analyses of ozone transport ever conducted. The EPA 
participated extensively in this process. The EPA believes that the 
OTAG process was successful and generated much useful technical and 
modeling information on regional ozone transport. This information 
provided EPA with the foundation for this rulemaking.
    The EPA received numerous comments regarding the relationship 
between the OTAG recommendations and EPA's proposed SIP call. Some 
commenters asserted that the Agency's proposal was inconsistent with 
the OTAG recommendations, while others believed that EPA used the 
information and recommendations from OTAG appropriately. Primarily, 
commenters stated that OTAG recommended a range of controls for utility 
sources instead of a uniform level of control for all of the included 
States.
    The OTAG did recommend consideration of a range of controls, and 
although it did not specifically recommend uniform controls across a 
broad region, such a control scheme is within the range of its 
recommendation. The EPA's action today is based on its consideration of 
OTAG's recommendations, as well as information resulting from EPA's 
additional work, and extensive public input generated through notice-
and-comment rulemaking. The EPA continues to believe, for reasons 
explained in Section III.F.1, Uniform vs. Regional Controls, that 
requiring NOX emissions reductions across the region in 
amounts achievable by uniform controls is a reasonable, cost-effective 
step to take at this time to mitigate ozone nonattainment in downwind 
States for both the 1-hour and 8-hour standards.
    Commenters also stated that EPA applied an electric utility control 
level that was more stringent than the upper limit of the OTAG range of 
utility controls. The OTAG recommended a range of utility controls that 
falls between specific CAA-required controls and the less stringent of 
85 percent reduction from the 1990 rate (lb/mmBtu), or 0.15 lb/mmBtu. 
In determining the appropriate level of emissions reductions, EPA 
considered what levels of NOX reductions could be obtained 
by applying, to various source sectors, controls that are among the 
most cost effective and feasible with today's proven pollution control 
technologies. The EPA chose emissions reductions that are equivalent to 
an emission limit from utilities of 0.15 lb/mmBtu. The EPA acknowledges 
that this level may be more protective than the most protective level 
contained in the OTAG recommendation in some cases, but, as discussed 
below in Section IV, Air Quality Assessment, EPA believes that it 
provides the most improvement in air quality while staying within the 
bounds of the most highly cost-effective technology available. (Cost 
effectiveness is discussed in Section II.D.) In addition, by relying on 
actual 1995-1996 continuous emission monitoring data, rather than 
relying on estimated 1990 emission data, this approach provides a more 
accurate way of determining the States' budgets since it minimizes any 
chances of over-or under-estimation of emissions.
    Commenters asserted that OTAG recommended 12 months for additional 
modeling--especially subregional modeling--before promulgating the SIP 
call; and these commenters expressed concern that EPA did not provide 
this amount of time following publication of the NPR. As discussed in 
more detail in Section I.F, Discussion of Comment Period and 
Availability of Key

[[Page 57363]]

Information, the Agency ultimately provided approximately 1 year from 
the conclusion of OTAG for States and other members of the public to 
complete and submit subregional and other types of modeling. The EPA 
has considered this additional modeling in finalizing today's rule.
    Some commenters stated that the goal of OTAG was to address 
attainment of the ozone NAAQS. This is incorrect. The OTAG's goal was 
to reduce ozone transport, which is one of the steps necessary to 
enable attainment; the goal was not to recommend an overall strategy 
that would yield attainment through regional measures alone. The OTAG 
articulated its overall goal as follows:

    * * * 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.5

    \5\ Ozone Transport Assessment Group Policy Paper approved by 
the Policy Group on December 4, 1995.

    It is also EPA's goal to ensure that sufficient regional reductions 
are achieved to mitigate ozone transport in the eastern half of the 
United States and thus, in conjunction with local controls, enable 
nonattainment areas to attain and maintain the ozone NAAQS.
    Commenters indicated that OTAG focused only on the 1-hour standard 
nonattainment problem and did not assess compliance implications of the 
8-hour standard. For this reason, according to commenters, EPA should 
not base today's action on the nonattainment of the 8-hour NAAQS. It is 
true that OTAG was established to address transport issues associated 
with meeting the 1-hour standard. The EPA did not promulgate the 8-hour 
standard until shortly after OTAG concluded; thus, OTAG did not 
recommend strategies to address the 8-hour NAAQS. However, because EPA 
had proposed an 8-hour standard, OTAG did examine the impacts of 
different strategies on 8-hour average ozone predictions.
    In light of OTAG's work and additional information, EPA is able to 
assess ozone transport as it relates to the 8-hour NAAQS and to set 
forth requirements as necessary to address the 8-hour standard in this 
rulemaking. Ozone transport causes problems for downwind areas under 
either the 1-hour or 8-hour standard. The regional reductions of 
NOX that will be achieved through this SIP call for the 1-
hour NAAQS are key components for meeting the new 8-hour ozone standard 
in a cost-effective manner. Therefore, EPA believes that the OTAG 
recommendations for how to address ozone transport are valid for both 
NAAQS.
    Several commenters urged EPA to adopt and implement all Federal 
measures identified in the OTAG recommendations.6 The Agency 
is committed to continue implementing national control measures for 
NOX, as recommended by OTAG. In addition, EPA has adopted 
the following national measures for purposes of reducing VOC: 
architectural and industrial maintenance coatings, consumer/commercial 
products, and autobody refinishing. The EPA has made no decisions 
regarding further VOC reductions beyond the reductions specified as 
phase I in the OTAG recommendations.7
---------------------------------------------------------------------------

    \6\ The OTAG recommendations are located in Appendix B of the 
November 7, 1997 NPR (62 FR 60376).
    \7\ Letter to the Honorable Ken Calvert, Chairman, Subcommittee 
on Energy and Environment, U.S. House of Representatives, from 
Robert D. Brenner, Acting Deputy Assistant Administrator for Air and 
Radiation, U.S. EPA, June 26, 1998, transmitting EPA's responses to 
questions following the May 20, 1998 congressional hearing on EPA's 
proposed rule on paints and coatings.
---------------------------------------------------------------------------

    Other more specific comments concerning the OTAG recommendations 
will be addressed throughout this rulemaking as the issues are 
discussed.

F. Discussion of Comment Period and Availability of Key Information

    The EPA received numerous comments concerning the adequacy of the 
comment period for the November 7, 1997 NPR and May 11, 1998 SNPR. Some 
commenters remarked that the comment period for the NPR should be 
extended to allow for development and review of technical information, 
including inventory data, growth factors, and the resulting budget. 
Commenters stated that the additional time was particularly necessary 
for subregional air quality modeling, which is modeling designed to 
isolate the impacts of emissions from a particular State or group of 
States on downwind areas. Many specifically requested an additional 120 
days, and one requested an additional 9 months. Some commenters 
indicated that EPA did not incorporate their comments from the NPR into 
the SNPR. Other commenters insisted that key information supporting the 
rule is not publicly available. The EPA also received comments that 
additional public hearings should be held in other locations of the 
OTAG region.
1. Request for Extension of the Comment Period
    The EPA allowed a 120-day public comment period for the November 7, 
1997 NPR, which closed on March 9, 1998. By notice (63 FR 17349, April 
9, 1998), EPA reopened the comment period for members of the public to 
submit additional modeling analyses, as well as comments concerning the 
implications that any additional modeling may have for the State NOx 
budgets under consideration in the November 7, 1997 proposal. The 
comment period was reopened through the end of the comment period on 
the SNPR. The SNPR, which was published on May 11, 1998, allowed a 
comment period until June 25, 1998. Thus, for most issues addressed in 
the NPR, including air quality modeling issues, commenters received an 
almost 8-month formal comment period. Indeed, many commenters had 
access to the NPR immediately after October 10, 1997, when it was 
signed and posted on an EPA website. The Agency also received a number 
of comments after June 25, 1998, which were also reviewed and 
considered in developing the final rule.
    The EPA believes this additional opportunity for the public to 
submit comments was reasonable. After March 9, 1998--the initial date 
for close of the comment period on the NPR--EPA received numerous 
comments on various issues raised in the NPR, including air quality 
issues. Many of these comments were extensive, which indicates that 
commenters received adequate time.
    With respect to the concern that EPA did not incorporate comments 
received on the NPR into the SNPR, it would not have been practical for 
EPA to incorporate comments received on the NPR into the SNPR because 
the SNPR was completed soon after the close of the comment period for 
the NPR. In general, the SNPR addressed different aspects of the rule 
than the NPR, and one of the purposes of the SNPR was to take comment 
on several new issues, as noted above. The EPA has addressed comments 
on both the NPR and SNPR in today's action.
    The major issues raised in the comments are responded to throughout 
the preamble of this final rule. A comprehensive summary of all 
significant comments, along with EPA's response to the comments which 
have not been responded to in the preamble (Response to Comments), can 
be found in the docket for this rulemaking (Docket No. A-96-56).

[[Page 57364]]

2. Request for Time to Conduct Additional Modeling
    The OTAG Policy Group, at its June 3, 1997 meeting, recommended 
that States have the opportunity to conduct additional local and 
subregional modeling and air quality analyses, as well as to develop 
and propose appropriate levels and timing of controls. The EPA received 
numerous comments related to OTAG's recommendation. The commenters 
requested that the Agency give States more time to conduct this 
additional modeling so that EPA could more accurately assess each 
State's contribution to downwind nonattainment.
    The EPA signed the NPR on October 10, 1997, and posted it on a 
website at that time, although it was not published in the Federal 
Register until November 7, 1997. As noted above, EPA reopened the 
comment period through June 25, 1998 for submittal of additional air 
quality modeling runs. In effect, this has extended the amount of time 
for modeling analyses to over a year from the date OTAG submitted its 
recommendations, and to over 8 months from the signature date for the 
NPR. By the close of the comment period on June 25, 1998, EPA had 
received numerous comments containing new and extensive air quality 
modeling studies. Accordingly, EPA believes that commenters received 
adequate time.
3. Availability of Key Information
    A number of commenters asserted that EPA failed to make publicly 
available key information, such as modeling and emissions inventory 
data. Specifically, commenters stated that they did not have access to 
the emissions data on which EPA based the air quality modeling for the 
NPR. In addition, according to some commenters, several models used by 
EPA and OTAG are proprietary models and have not been generally 
available to the public.
    In Section III.A.2, Availability, the Agency discusses the 
availability of emissions inventory data to the public.
    The OTAG and EPA conducted air quality modeling runs to determine 
the level of contribution from emissions in upwind areas to ozone 
nonattainment in downwind areas. Some of this modeling employed UAM-
V.8 The UAM-V has generally been available to the public for 
the purpose of analyzing information relevant to today's rulemaking. 
State and local agencies, as well as utility companies and other 
stakeholders, have had access to licenses to use UAM-V.
---------------------------------------------------------------------------

    \8\ Variable-Grid Urban Airshed Model.
---------------------------------------------------------------------------

    Commenters objected that they were obliged either to purchase 
licenses for use of the UAM-V model or to employ as a contractor the 
model owner, and that these financial constraints restricted their 
access to the model. Because this model has, in general, been privately 
developed, EPA believes that reasonable fees for its use should be 
expected. The EPA did not receive information indicating that the 
associated expenses were other than reasonable. To the extent that 
commenters experienced delays in obtaining the UAM-V model, EPA 
believes that the extensions of the comment period resulted in adequate 
time for comment. In any event, any commenter who was not able to gain 
access in the timeframe desired was able to use a comparable model, 
such as the Comprehensive Air Quality Model with Extensions (CAMx), 
which is not proprietary. For the purpose of responding to public 
comments, EPA is considering all information based on CAMx and similar 
models.
    The Agency made available additional modeling runs used to 
determine emissions changes, costs and cost effectiveness for 
electricity generating units (EGUs). These runs were placed on the IPM 
Analyses web site at www.epa.gov/capi, with links to EPA's Office of 
Air and Radiation Policy and Guidance web site.
    On August 10, the EPA placed in the docket and made available on 
the web site, modeling analyses and other information supporting 
today's action. As noted above, by notice dated August 24, 1998 (63 FR 
45032), EPA published a notice of availability which stated that 
throughout the course of the rulemaking, EPA had placed information in 
the docket or made it available on various web sites. This information 
included inventory data and additional modeling runs. By placing those 
materials in the docket and informing the public of their availability, 
EPA provided 4-6 weeks for review and comment by the public. The EPA 
did receive comments concerning this information from the Utility Air 
Regulatory Group on September 9, and EPA is responding to those 
comments in the Response To Comments document. The EPA notes that the 
additional modeling analyses were performed in response to comments 
received on the NPR urging EPA to conduct State-by-State modeling. The 
Agency does not believe it is required to provide for additional 
comment on every action it takes in response to comment, particularly 
where, as here, the new information confirms the Agency's proposed 
conclusions. Therefore, the Agency did not further extend the comment 
period.
4. Public Hearings
    The Agency conducted two hearings in Washington, DC, including a 2-
day hearing on February 3-4, 1998 for the NPR, and a 1-day hearing on 
May 29, 1998 for the SNPR. Some commenters believe that additional 
public hearings should have been held in other locations in the OTAG 
region. The EPA believes these hearings provided reasonable opportunity 
for oral comment on the proposed rulemaking given the timeframes 
associated with this rulemaking. Therefore, the Agency did not schedule 
any additional hearings. The public also had an opportunity to submit 
written testimony within approximately 30 days after each hearing date.

G. Implementation of Revised Air Quality Standards

    On July 18, 1997, EPA published its final rule for strengthening 
the NAAQS for ozone by establishing an 8-hour standard (62 FR 38856). 
Current monitoring data indicate that many areas in the East, Midwest 
and South violate the 8-hour NAAQS. Along with areas violating the 1-
hour NAAQS, areas violating the 8-hour NAAQS are also affected by the 
transport of ozone across the East. The regional NOX 
reduction strategy finalized in today's action will provide a mechanism 
to achieve reductions that will assist States in attaining and 
maintaining this revised standard. In fact, the regional reductions 
alone should be enough to enable the vast majority of the new counties 
violating the 8-hour NAAQS that are located in States throughout the 
East to attain the revised 8-hour standard.\9\
---------------------------------------------------------------------------

    \9\ In the NPR (62 FR 60318, 60363), EPA provided estimates of 
the number of counties expected to attain as a result of the 
NOX SIP call. The EPA will update this list in the coming 
months. The updated estimates of which counties will attain will be 
based on more current air quality data and on the State-by-State 
emissions budgets contained in today's final rule.
---------------------------------------------------------------------------

    On July 16, 1997, President Clinton issued a directive on the 
implementation of the revised air quality standards. This 
implementation policy was described in the NPR (62 FR 60318, 60362-64). 
The EPA received numerous comments on this implementation policy and on 
EPA's plan to create a transitional classification\10\ for 8-hour ozone 
nonattainment areas that meet certain

[[Page 57365]]

criteria. Since these comments concern implementation efforts for the 
revised 8-hour ozone standard and do not relate directly to the 
NOX SIP call on which EPA is taking final action in this 
rulemaking, EPA is not responding in detail to the comments. The EPA 
will address implementation of the revised standard separately. In 
August 1998, EPA issued proposed guidance for public comment to explain 
the implementation policy in further detail and to provide details on 
SIP requirements for transitional areas (63 FR 45060, August 24, 1998). 
The EPA expects to finalize the August 1998 draft guidance, as well as 
guidance for areas other than transitional, by December 1998.\11\
---------------------------------------------------------------------------

    \10\ The ``transitional classification'' EPA intends for 8-hour 
ozone nonattainment areas is further discussed in the NPR (62 FR 
60318, 60363).
    \11\ For a complete listing of the guidance and other actions 
EPA plans to issue to implement the revised ozone and PM NAAQS, see 
a table on EPA's implementation website: http://
ttnwww.rtpnc.epa.gov/implement/actions.htm.
---------------------------------------------------------------------------

H. Summary of Major Changes Between Proposals and Final Rule

    This summary describes the major changes that have occurred since 
the NPR and SNPR in each of the following sections of today's final 
rule.
1. EPA's Analytical Approach (Section II.A)
     The NPR proposed two interpretations for the section 
110(a)(2)(D)(i)(I) provisions concerning the ``significant 
contribution'' test. Under the first, EPA would examine certain factors 
relating to level of emissions and their ambient impact to determine 
whether to make a finding that all of the emissions from a particular 
State's sources contribute significantly to nonattainment or 
maintenance problems downwind. If EPA made such a finding, then EPA 
would examine certain cost factors to determine the extent to which the 
SIP for the State must mitigate (reduce) its emissions. Under the 
second interpretation, EPA would examine all of those factors 
together--level of emissions, ambient impact, and costs--to determine 
whether to make the finding with respect to a specified amount of 
emissions. If EPA made the finding, then it would require the SIP to 
eliminate that amount. In today's final rule, EPA is adopting the 
second interpretation. The EPA indicates, however, that it would adopt 
the same rule if it were instead implementing the first interpretation.
2. Cost Effectiveness of Emissions Reductions (Section II.D.)
     The methodology of determining cost effectiveness has not 
changed. For all sources, the inventory and as a result, the source-
specific costs, in some cases, have changed. This results in a 
different overall budget level and a different overall cost-
effectiveness value. For the non-EGUs, while the methodology has not 
changed, the analysis focuses on large non-EGU sources. The methodology 
in the NPR focused on all non-EGU sources.
3. Determination of Budgets (Section III.)
     For EGU, the EPA maintained the approach to use the 
higher, by State, of 1995 or 1996 heat input data to calculate baseline 
heat input rates for the NFR, and added 577 smaller units to the State 
budget inventories which had erroneously been omitted from the NPR. 
These units included electricity generating sources of 25 megawatts 
(MW) or less of electrical output and additional units not affected 
under the Acid Rain Program. Additional controls are not assumed for 
these sources, but they are added to the budget at baseline levels. The 
Agency has decided to use State-specific growth factors derived from 
application of the IPM using the 1998 Base Case and chose to retain the 
0.15 lbs/mmBtu as the assumed uniform control level for EGU budget 
emissions determination.
     The EPA examined alternatives that focus on non-EGU point 
source reductions from the largest source categories, and within each 
of these categories assumed controls that would result in a regionwide 
average cost effectiveness less than $2000/ton. The resulting budget 
assumes the emissions reductions from large non-EGU sources that are 
among the most cost effective to control and does not include 
reductions from smaller sources and sources that, as a group, are not 
quite as cost effective or efficient to control, or are already covered 
by other Federal measures. As a result, this final rule assumes, for 
purposes of calculating the State NOX budgets, the following 
emissions decreases from uncontrolled levels for the large (generally 
greater than 250 mmBtu or 1 ton/day non-EGU sources (no emission 
reductions are assumed for the smaller sources):

--Non-EGU boilers and turbines--60 percent decrease.
--Stationary internal combustion engines--90 percent decrease.
--Cement manufacturing plants--30 percent decrease.

    It should be noted that point sources with capacities less than 250 
mmBtu/hr but with emissions greater than 1 ton/day are not treated 
differently from sources with capacities greater than 250 mmBtu/hr for 
purposes of calculating the budget. This is a change from the NPR which 
included RACT controls on units with capacities less than 250 mmBtu/hr 
and emissions greater than 1 ton/day (see Section III.G.2.a). As under 
the proposal, the rule allows States to choose control measures other 
than the EPA-assumed controls to meet the numerical budgets.
     The EPA has implemented the following changes that the 
Agency proposed in the NPR for calculating baseline NOX 
emissions from highway vehicles. A 1995 baseline is used for the final 
rule in place of the 1990 baseline used in the NPR. The Highway 
Performance and Monitoring System data were used to estimate States' 
1995 vehicle miles traveled (VMT) by vehicle category, except in those 
cases where EPA accepted revisions offered in the comments. Today's 
action includes those mobile source reductions which EPA has determined 
are appropriate to implement on a national basis, and which have been 
promulgated in final form or are expected to be promulgated in final 
form before States are required to comply with their budgets. The 
highway vehicle budget components include the emission reductions 
resulting from implementation of the National Low Emitting Vehicle 
(NLEV) program, including the phase-in schedule agreed to by the 
States, automobile manufacturers, and EPA. The highway budget 
components do not include the effect of Tier 2 light-duty vehicle and 
truck standards and any associated fuel standards since these standards 
have not yet been proposed. The extent of the reformulated gasoline 
(RFG) and inspection and maintenance (I/M) programs was not assumed to 
change beyond that assumed for the NPR, except for those States that 
were able to demonstrate that the NPR's modeling assumptions did not 
conform to the State's SIP and did not reflect CAA requirements.
     The EPA has chosen to retain the 1990 baseline inventories 
for nonroad mobile sources presented in the NPR for today's action, 
with additional changes made in response to public comments. The 
control strategies assumed for calculating the nonroad and stationary 
area source budget components have not changed from the SNPR.
4. NOX Control Implementation and Budget Achievement Dates 
(Section V)
     The EPA proposed that the SIP revisions require full 
implementation of the necessary State measures by September 2002 and 
took comment on a range of dates from September 2002 through September 
2004. Based on

[[Page 57366]]

public comments and feasibility analyses conducted by EPA, the Agency 
is requiring an implementation date of May 1, 2003. The Agency is also 
providing some compliance flexibility to States for the 2003 and 2004 
ozone seasons by establishing State compliance supplement pools. This 
is described in Section III.F.6.
5. SIP Criteria (Section VI.A)
     The Agency has determined that the additional SIP 
approvability criteria, as proposed in the SNPR, should apply not only 
when States choose to regulate EGUs (63 FR 25912), but also when States 
choose to regulate large steam-producing units (i.e., combustion 
turbines and combined cycle systems with a capacity greater than 250 
mmBtu/hr).
     The Agency proposed revisions to part 51 requiring 
continuous emissions monitoring systems (CEMS) on all large electrical 
generating and steam-producing sources which States elect to subject to 
emissions reduction requirements in response to this rulemaking. The 
EPA took comment on requiring that, if a State chooses to regulate 
these sources to meet the SIP call, the SIP must require these sources 
to use the NOX mass monitoring provisions of part 75, 
subpart H, to demonstrate compliance with applicable emissions control 
requirements. After considering comments, the Agency is requiring that, 
in these circumstances, the SIP specify that large sources comply with 
the monitoring provisions of part 75, subpart H, which includes non-
CEMS monitoring options for units that are infrequently operated or 
units that have low mass emissions.
6. Emissions Reporting Requirements for States (Section VI.B)
     The proposed rule required that States report full-year, 
as well as ozone-season, emissions from all sources for the triennial 
inventories commencing with year 2002 emissions and the 2007 inventory, 
and for those sources for which reports had to be submitted annually 
starting with year 2003 emissions. The final rule requires only ozone-
season emissions reporting for all sources.
     In the SNPR, the EPA proposed, for purposes of reporting 
requirements, to define a point source as a non-mobile source which has 
NOX emissions of 100 tons/year or greater. Under today's 
action, States have the option of establishing a smaller emission 
threshold than 100 tons/year of NOX emissions in defining 
point source. This will allow the definition of point source to remain 
consistent with current definitions in local areas.
7. NOX Budget Trading Program (Section VII.)
     For States that choose to participate in the 
NOX Budget Trading Program, the preamble clarifies the 
intent of the model rule and identifies areas of the rule where States 
have flexibility to include variations in their State rules.
     In the SNPR, the Agency solicited comment on a range of 
options for incorporating banking into the trading program. After 
considering these comments, the Agency is including banking provisions 
in the final rule. The provisions allow for unlimited banking starting 
in 2003 and includes a flow control mechanism to limit the emissions 
variability associated with banking.
     One of the banking approaches presented in the SNPR 
included the option for sources to generate and use early reduction 
credits. Consistent with the provisions of the NOX SIP call 
which provide for State compliance supplement pools, the final rule 
allows States to issue early reduction credits for certain 
NOX emissions reductions achieved between September 30, 1999 
and May 1, 2003.
     The final rule clarifies the timing requirements for State 
submission of allowance allocations to EPA and, as proposed, lays out 
an allocation approach. Each State remains free to adopt the final 
rule's allocation approach or adopt an allocation scheme of its own, 
provided it meets the specified timing requirements, requires new 
sources to hold allowances, and does not allocate more allowances than 
are available in the State trading budget.
8. Interaction with Title IV NOX Rule (Section VIII.)
     In the SNPR, EPA proposed revisions to part 76 addressing 
the interaction between title IV and the NOX SIP call. In 
this final rule, EPA explains that the Agency is not adopting any of 
the proposed revisions to part 76.
9. Administrative Requirements (Section X.)
     NPR Section VIII, Regulatory Analyses, has been replaced 
in the final rule by Section X.A, Executive Order 12866: Regulatory 
Impacts Analysis. The new final rule Section X.A indicates that EPA has 
prepared a RIA for the final rule and cites the cost and benefit 
estimates from that analysis.
     The final rule adds several Sections under X, 
Administrative Requirements, that were absent from the NPR. These 
include: Paperwork Reduction Act; Executive Order 13045: Protection of 
Children from Environmental Health Risks and Safety Risks; Executive 
Order 12898: Environmental Justice; Executive Order 12875: Enhancing 
the Intergovernmental Partnerships; Executive Order 13084: Consultation 
and Coordination with Indian Tribal Governments; Judicial Review; and 
Congressional Review Act. These new Sections provide a more 
comprehensive summary of the Acts and Executive Orders that could apply 
to the final rule. Each Section identifies the requirements of the 
relevant Act or Executive Order, indicates EPA's interpretation of 
whether the Act or Executive Order actually applies to this rulemaking, 
and, if so, indicates how the Agency has addressed the Act or Executive 
Order.

II. EPA's Analytical Approach

A. Interpretation of the CAA's Transport Provisions

    As indicated in the NPR, 62 FR 60323, the primary statutory basis 
for today's action is the ``good neighbor'' provision of section 
110(a)(2)(D)(i)(I), under which, in general, each SIP is required to 
include provisions assuring that sources within the State do not emit 
pollutants in amounts that significantly contribute to nonattainment or 
maintenance problems downwind. This statutory requirement applies to 
SIPs under both the 1-hour ozone NAAQS and the 8-hour ozone NAAQS.
1. Authority and Process for Requiring SIP Submissions Under the 1-Hour 
Ozone NAAQS
    a. Authority for Requiring SIP Submissions under the 1-Hour NAAQS. 
Each State is currently required to have in place a SIP that implements 
the 1-hour ozone NAAQS for areas to which that standard still applies. 
In the NAAQS rulemaking, EPA determined that the 1-hour NAAQS would 
cease to apply to areas that EPA determines have air quality in 
attainment of that NAAQS (40 CFR 50.9(b)). In two recent rulemakings, 
EPA identified numerous areas of the country to which the 1-hour NAAQS 
no longer applies. ``Final Rule: Identification of Ozone Areas 
Attaining the 1-Hour Standard and to Which the 1-Hour Standard is No 
Longer Applicable,'' (63 FR 31014, June 5, 1998); ``Final Rule: 
Identification of Additional Ozone Areas Attaining the 1-Hour Standard 
and to Which the 1-Hour Standard is No Longer Applicable,'' (63 FR 
27247, July 22, 1998).
    The 1-hour NAAQS remains applicable to areas whose air quality 
continues to monitor nonattainment. As noted above in Section I.B, 
General

[[Page 57367]]

Factual Background, these include many major urban areas in the eastern 
half of the United States. States that contain these areas remain 
responsible for meeting CAA requirements applicable to those areas for 
the purpose of attaining the 1-hour NAAQS. For example, States are 
responsible for attainment demonstrations for areas designated 
nonattainment and classified as moderate or higher.
    By the same token, States that are upwind of these areas are 
responsible to meet the ``good neighbor'' requirements of section 
110(a)(2)(D). This responsibility is not alleviated simply because, for 
areas other than the current nonattainment areas, the 8-hour NAAQS has 
replaced the 1-hour NAAQS.
    b. Process for Requiring SIP Submissions under the 1-Hour NAAQS. As 
explained in the NPR, the appropriate route for EPA to require SIP 
submissions under section 110(a)(2)(D)(i)(I) with respect to the 1-hour 
standard is issuance of a ``SIP call'' under section 110(k)(5).\12\ 
Section 110(k)(5) authorizes EPA to find that a SIP is substantially 
inadequate to meet a CAA requirement and to require (``call for'') the 
State to submit, within a specified period, a SIP revision to correct 
the inadequacy. Specifically, section 110(k)(5) provides, in relevant 
part:
---------------------------------------------------------------------------

    \12\ As discussed in the NPR and in greater detail further 
below, the basis for requiring a transport-related SIP revision for 
the 8-hour standard is the requirement in section 110(a)(1) that 
States submit SIPs meeting the requirements of section 110(a)(2) 
within 3 years (or an earlier date established by EPA) of 
promulgation of a new or revised NAAQS. This is discussed in further 
detail below.

    Whenever the Administrator finds that the applicable 
implementation plan for any area is substantially inadequate to 
attain or maintain the relevant [NAAQS], to mitigate adequately the 
interstate pollutant transport described in section 176A or section 
184, or to otherwise comply with any requirement of this Act, the 
Administrator shall require the State to revise the plan as 
necessary to correct such inadequacies. The Administrator shall 
notify the State of the inadequacies, and may establish reasonable 
deadlines (not to exceed 18 months after the date of such notice) 
---------------------------------------------------------------------------
for the submission of such plan revisions.

    By today's action, EPA is determining that the SIPs for the 
specified jurisdictions are substantially inadequate to comply with the 
requirements of section 110(a)(2)(D)(i)(I) because the relevant SIPs do 
not contain adequate provisions prohibiting their sources from emitting 
amounts of NOX emissions that contribute significantly to 
nonattainment in downwind areas that remain subject to the 1-hour 
NAAQS. Based on these determinations, EPA is requiring the identified 
States to submit SIP revisions containing adequate provisions to limit 
emissions to the appropriate amount.
    If a State does not submit the required SIP provisions in response 
to this SIP call, EPA will issue a finding that the State failed to 
make a required SIP submittal under section 179(a). This finding has 
implications for sanctions as well as for EPA's promulgation of Federal 
implementation plans (FIPs). Sanctions and FIPs are discussed in 
Section VI, SIP Criteria and Emissions Reporting Requirements.
    (1) Commenters' Arguments Concerning the Transport Provisions. 
Commenters argued that EPA does not have unilateral authority to issue 
a SIP call under section 110(k)(5) to require States to remedy SIPs 
that do not meet the requirements of section 110(a)(2)(D). The 
commenters noted that when Congress amended the CAA in 1990, Congress 
provided that the sole authority for EPA and States to address 
interstate transport of pollution is through transport commissions. In 
support, the commenters state that Congress: (i) Added sections 176A 
and 184, which authorize the establishment of transport regions and the 
formation of transport commissions; (ii) revised section 110(k)(5) to 
refer to those transport provisions; and (iii) revised section 
110(a)(2)(D)(i) to require that SIP provisions designed to eliminate 
interstate pollutant transport be consistent with other CAA 
requirements. According to the commenters, these provisions, read as a 
whole, mandate that if EPA believes that a transport problem exists, 
EPA's sole recourse is to form a transport region under sections 176A 
and/or 184; EPA may issue a SIP call to mandate compliance with section 
110(a)(2)(D)(i) only in response to a recommendation of the transport 
region. The commenters also claim that this scheme is sensible because 
it provides a consensual forum for States to address interstate 
pollution rather than allowing unilateral action on the part of EPA or 
a State.
    The EPA disagrees with the commenters' conclusion that these 
statutory provisions make clear that EPA cannot require a State to 
address interstate transport without first establishing a transport 
commission and in the absence of a recommendation from the transport 
commission. There is no language of limitation in sections 110(a)(2)(D) 
or (k)(5), or 176A, or 184. Nor is there any support in the legislative 
history for such a narrow reading of the statute. Moreover, under the 
commenters' interpretation, the CAA Amendments of 1990 have placed 
greater constraints on States' and EPA's ability to address the 
interstate transport of pollution. Such an interpretation would be 
inconsistent with the overall purpose of the CAA to ensure healthful 
air. Thus, EPA believes that the transport provisions were added as an 
additional tool to address interstate transport but were not intended 
to preclude other methods of addressing interstate pollution than prior 
to passage of the amendments.
    Under the 1990 Amendments, Congress recognized the growing evidence 
that ozone and its precursors can be transported over long distances 
and that the control of transported ozone was a key to achieving 
attainment of the ozone standard across the nation (Cong. Rec. S16903 
(daily ed. Oct. 27, 1990) (statement of Sen. Mitchell); S16970 
(conference report) S16986-87 (statement of Sen. Lieberman)). Thus, in 
1990, Congress added a new mechanism to address interstate transport. 
Specifically, Congress enacted sections 176A and 184, which provide a 
mechanism for States to work together to address the interstate 
transport problem. However, by their terms, these sections simply 
provide authority for EPA to designate transport regions and establish 
transport commissions. There is nothing in the language of these 
provisions that indicates that they supersede the other statutory 
mechanisms for addressing interstate transport, or that they now 
provide the sole mechanism for resolving interstate pollution 
transport.
    Moreover, although Congress expressly added these two provisions 
through the 1990 Amendments, Congress did not in any way limit section 
110(a)(2)(D), which requires States to address interstate transport in 
their SIPs. The addition of the language providing that States' actions 
under section 110(a)(2)(D) be ``consistent with [title I] of the Act'' 
cannot be read to limit the controls States may adopt to meet section 
110(a)(2)(D) to those recommended by a transport 
commission.13 After all, the transport region provisions are 
only two of many provisions in title I. Rather, this

[[Page 57368]]

language concerning consistency should be read as clarifying that any 
section 110(a)(2)(D) requirement must be consistent with other 
provisions of title I. Similarly, this language makes explicit that SIP 
revisions required in accordance with the procedures of the transport 
provisions would meet the requirements of section 110(a)(2)(D)(i).
---------------------------------------------------------------------------

    \13\ Taken to its logical conclusion, the commenters' argument 
would mean that States are precluded from submitting a section 
110(a)(2)(D) SIP unless it reflects measures recommended through the 
transport commission process. The EPA does not believe that Congress 
would first establish a specific mandate (to submit a SIP to address 
interstate transport) and then limit it in such a cryptic fashion. 
If Congress intended section 110(a)(2)(D) SIPs to only reflect 
transport commission recommendations, Congress could have 
specifically referenced sections 176A and 184 in section 
110(a)(2)(D), rather than generally providing that SIPs be 
``consistent'' with title I of the CAA.
---------------------------------------------------------------------------

    Furthermore, it is significant that Congress did not in any sense 
bind EPA's ultimate discretion to determine whether State plans 
appropriately address interstate transport. Under sections 176A and 
184, the States may only make recommendations to EPA. Thus, under the 
transport provisions, as well as the general SIP requirements of 
section 110(a)(2), EPA must ultimately decide whether the SIP meets the 
applicable requirements of the CAA. If, as the commenters contend, EPA 
is limited to calling on States to address interstate transport only by 
strategies recommended by the State, then EPA would be precluded from 
ensuring that States address interstate transport. For example, EPA 
could establish a transport commission but the commission could fail to 
make recommendations or make insufficient recommendations. (Section 
176A provides that transport commissions may make recommendations to 
EPA only by ``majority vote of all members'' other than those 
representing EPA.) Such a reading of the statute would be absurd in 
light of the growing recognition at the time of the 1990 Amendments 
that transport is a real threat to the primary purpose of title I of 
the CAA--attainment of the NAAQS.
    By the same token, in amending section 110(k)(5) in the 1990 
Amendments, Congress did not add anything that explicitly provides 
that, in the case of interstate transport, section 110(k)(5) would 
apply only when EPA approved (or substituted measures for) a transport 
commission's recommendations. The reference in section 110(k)(5) to the 
transport provisions of sections 176A and 184 does not preclude EPA's 
use of the SIP call provision to call on States to ensure their SIPs 
meet the requirements of section 110(a)(2)(D)(i). Section 110(k)(5) 
also provides for EPA to call on States ``to otherwise comply with 
requirements of this Act;'' among the requirements in chapter I of the 
CAA is the requirement in section 110(a)(2)(D). The reference in 
section 110(k)(5) to the transport provisions simply makes explicit 
that EPA may employ section 110(k)(5) for the additional purpose of 
requiring SIPs to include the control measures as recommended by 
transport commissions and approved by EPA under the transport 
provisions.
    Moreover, there is no indication in the legislative history of the 
1990 Amendments that Congress intended the sections 176A and 184 
transport provisions to supersede the section 110(k)(5) SIP call 
mechanism for ensuring compliance with section 110(a)(2)(D)(i). Reading 
the transport provisions to supersede the SIP call mechanism would 
constitute a significant change from the CAA as it read prior to the 
1990 Amendments. Even if the statute is ambiguous as to whether the 
transport provisions supersede the SIP call mechanism--and EPA believes 
the statute is clear that the transport provisions do not supersede--
congressional silence would suggest that Congress did not intend such a 
significant change (See generally Harrison v. PPG Industries, Inc., 446 
U.S. 578, 602, 100 S.Ct. 1889, 1902, 64 L.Ed.2d 525 (1980) (Rehnquist, 
J., dissenting), cited with approval in Chisom v. Roemer, 501 U.S. 380, 
396 n. 23, 111 S.Ct. 2354, 2364 n. 23, 115 L.Ed.2d 348 (1991)).
    Finally, the commenter asserts that EPA's interpretation of the CAA 
to allow a SIP call in the absence of a transport commission 
recommendation reads out of the CAA the consensual transport commission 
procedures under sections 176A and 184. This is simply not true. The 
EPA interprets the transport commission process to be one tool to 
assess and address interstate transport. In fact, the Northeast Ozone 
Transport Commission, under section 184, has been active since 
enactment of the 1990 Amendments. In 1995, EPA approved a 
recommendation of that commission (60 FR 4712 14). Transport 
commissions remain a viable means for dealing with interstate 
transport. Furthermore, contrary to the general implication of the 
commenter's remark, the OTAG process, though not a formal transport 
commission, provided an opportunity not only for Federal and State 
governments to assess jointly the transport issue, but also involved 
industry, environmental groups and others. The EPA based its SIP call 
on information developed through OTAG, as well as additional analyses 
performed by the Agency and information submitted by a variety of 
groups during the comment period on the proposed rule. Thus, the OTAG 
process contained consensual elements.
---------------------------------------------------------------------------

    \14\ In Commonwealth of Virginia v. EPA, 108 F.3d 1397 (D.C. 
Cir. 1997), the court vacated EPA's SIP call in response to the 
Northeast Ozone Transport Commission's recommendation on the basis 
that the EPA could not require States to adopt a specific control 
measure under its section 110(k)(5) authority and that, in any 
event, EPA could not require States to adopt stricter motor vehicle 
emission standards under either section 110(k)(5) or section 184.
---------------------------------------------------------------------------

    (2) Commenters' Arguments Concerning the Virginia case. Under one 
of the approaches described in the proposed rule, EPA proposed to 
determine, for each of various upwind States, the aggregate ``amounts'' 
of air pollutants (NOX) that contribute significantly to 
nonattainment, and that, therefore must be prohibited by the various 
SIPs. The NOX emissions budget for each State is an 
expression of the amount of NOX emissions that would remain 
after the State prohibits the amount that contributes significantly to 
downwind nonattainment. In the final rule issued today, EPA has 
continued this approach, establishing emissions budgets for each of the 
23 jurisdictions based on required reductions. This determination is an 
important step toward assuring that overall air quality standards are 
met downwind.
    Commenters argue that even if EPA has authority to call on States 
to address interstate transport, EPA does not have the authority under 
section 110(a)(2)(D) to mandate that upwind States limit NOX 
emissions to specified amounts. Rather, according to this view, EPA's 
authority is limited to determining that the upwind States' SIPs are 
inadequate, and generally requiring the upwind States to submit SIP 
revisions to correct the inadequacies. The upwind States would then, 
according to this view, submit a SIP revision that implements what the 
upwind States determine to be the appropriate amount of NOX 
reductions. If EPA believes that those amounts are too small to correct 
the inadequacy, EPA could disapprove the SIP revisions.
    Proponents of this view rely on the recent decision in Virginia v. 
EPA, 108 F.3d 1397, 1406-10 (D.C. Cir. 1997) (Virginia) (citing Train 
v. NRDC), in which the court vacated EPA's SIP call on the basis that 
through it, EPA gave States no choice but to adopt the California low 
emission vehicle (LEV) program. The court found that the language in 
section 110(k)(5) that provides EPA with the authority to call on a 
State to revise its SIP ``as necessary'' to correct a substantial 
inadequacy did not change the longstanding precept that States have the 
primary authority for determining the mix of control measures needed to 
attain the NAAQS.
    The EPA disagrees that the CAA prohibits EPA from establishing an 
emissions budget through a SIP call requiring upwind States to prohibit 
emissions that contribute significantly to downwind nonattainment. 
Section

[[Page 57369]]

110(a)(2)(D) is silent regarding whether States or EPA are to determine 
the level of emission reductions necessary to mitigate significant 
contribution. The caselaw cited by the commenters only provides that 
States are primarily responsible for determining the mix of control 
measures--not the aggregate emission reduction levels that are 
necessary. Moreover, Train v. NRDC, which underlies the Virginia 
court's decision, relied on section 107(a) of the CAA, which specifies 
only that each State is primarily responsible for determining a control 
strategy to attain the NAAQS ``within such State.''
    Section 110(a)(2)(D) does not provide who--EPA or the States--is to 
determine the level of emission reductions necessary to address 
interstate transport. As quoted above, section 110(a)(2)(D)(i)(I) 
requires that SIPs contain ``adequate provisions prohibiting * * * 
[sources] from emitting any air pollutant in amounts which will 
contribute significantly to nonattainment'' downwind. Nor does this 
provision indicate the criteria for determining the ``amounts'' of 
pollutants that contribute significantly to nonattainment downwind. Nor 
does this provision indicate the process for determining those 
``amounts,'' including whether EPA or the States should carry out this 
responsibility. 15 Under Chevron U.S.A., Inc. v. Natural 
Resources Defense Council, 468 U.S. 1227, 105 S.Ct. 28, 82 L.Ed.2d 921 
(1984) (Chevron), because the statute does not answer these specific 
issues, EPA has discretion to provide a reasonable interpretation.
---------------------------------------------------------------------------

    \15\ The EPA is not contending that the ``as necessary'' 
language in section 110(k)(5) provides the basis for EPA's authority 
to identify the emissions budget for upwind States.
---------------------------------------------------------------------------

    Neither the decision in Virginia, nor the body of caselaw upon 
which it relies, addresses this issue. Rather, these cases address 
solely the division between the States and EPA regarding the initial 
identification of control measures necessary to attain the ambient air 
quality standards. The issue before the court in Virginia was whether 
EPA had offered States a choice in selecting control measures or 
instead had mandated the adoption of a specific control measure. 
Relying on Train v. NRDC, 421 U.S. 60, 95 S.Ct. 1470, 43 L.Ed.2d 731 
1975), the Virginia court found that under title I of the CAA, EPA is 
required to establish the overall air quality standards, but the States 
are primarily responsible for determining the mix of control measures 
needed to meet those standards and the sources that must implement 
controls, as well as the applicable level of control for those sources. 
The EPA must then review the State's determination only to the extent 
of assuring that the overall air quality standards are met. If EPA 
determines that the SIP's mix of control measures does not result in 
achieving the overall air quality standards, EPA is required to 
disapprove the SIP and promulgate a FIP, under which EPA selects the 
sources for emissions reductions (Virginia, 108 F.3d at 1407-08, citing 
Train v. NRDC, 421 U.S. 60, 95 S.Ct. 1470, 43 L.Ed.2d 731 (1975); Union 
Electric Co. v. EPA, 427 U.S. 246, 96 S.Ct. 2518, 49 L.Ed.2d 474 
(1976)). This line of cases, which focuses on the selection of 
controls, does not address whether EPA or the States--in the first 
instance--should determine the aggregate amount of reductions necessary 
to address interstate transport.
    Moreover, NRDC v. Train addresses State plans for purposes of 
intrastate emissions planning. In determining that States have the 
primary authority for determining the control measures needed to attain 
the standard, the court relied on section 107(a) of the CAA, which 
provided (and still provides) that:

    Each State shall have the primary responsibility for assuring 
air quality within the entire geographic area comprising such State 
by submitting an implementation plan which will specify the manner 
in which national primary and secondary ambient air quality 
standards will be achieved and maintained within each air quality 
region in such State.''

(421 U.S. at 64, 95 S.Ct at 1474-75 (emphasis added)).

    Thus, the underlying support for the court's determination in Train 
v. NRDC applies only where a State is determining the mix of controls 
within its boundaries, not to the broader task of determining the 
aggregate emissions reductions needed in conjunction with emissions 
reductions from a number of other States in order to address the impact 
of transported pollution on downwind States. 16
---------------------------------------------------------------------------

    \16\ The court's decision in Train v. NRDC appears to rely on 
the plain language of the statute in holding that a State is 
primarily responsible for determining the mix of control measures 
necessary to demonstrate attainment within that State's borders. The 
court in Virginia appears to adopt this ``plain meaning'' 
interpretation without addressing that the language in section 
107(a) applies only to intrastate issues. This issue is not relevant 
in the present case, however, since States are free to decide the 
mix of control measures under today's final action.
---------------------------------------------------------------------------

    Although the cases to date have not addressed directly whether it 
is the province of EPA or the States to determine the aggregate amounts 
of emissions to be prohibited (and hence, the amounts that may remain--
i.e., the emissions budgets), EPA believes it reasonable to interpret 
the ambiguity in section 110(a)(2)(D)(i)(I) to include this 
determination among EPA's responsibilities, particularly in the current 
circumstances. Determining the overall level of air pollutants allowed 
to be emitted in a State is comparable to determining overall standards 
of air quality, which the courts have recognized as EPA's 
responsibility, and is distinguishable from determining the particular 
mix of controls among individual sources to attain those standards, 
which the caselaw identifies as a State responsibility. In Train, a 
State was required to assure that its own air quality attained overall 
air quality standards and to implement emissions controls to do so. 
Under these circumstances, the court clarified that while the 
responsibility for determining the overall air quality standards was 
EPA's, the responsibility for determining the specific mix of controls 
designed to achieve that air quality was the State's. By comparison, as 
stated earlier, a transport case, under section 110(a)(2)(D)(i), does 
not concern any requirement of the upwind State to assure that its own 
air quality attains overall air quality standards. Rather, a transport 
case concerns the upwind State's requirement to assure that its 
emissions are reduced to a level that will not contribute significantly 
to nonattainment downwind. Determining this overall level of reductions 
for the upwind State is analogous to determining overall air quality 
standards, and, thus, should be the responsibility of EPA.
    Once EPA determines the overall level of reductions (by assigning 
the aggregate amounts of emissions that must be eliminated to meet the 
requirements of section 110(a)(2)(D)), it falls to the State to 
determine the appropriate mix of controls to achieve those reductions. 
Unlike the regulation at issue in Virginia, today's regulation 
establishing emission budgets for the States does not limit the States 
to one set of emission controls. Rather, the States will have 
significant discretion to choose the appropriate mix of controls to 
meet the emissions budget. The EPA has based the aggregate amounts to 
be prohibited on the availability of a subset of cost-effective 
controls that are among the most cost effective available. As explained 
elsewhere in this final rule and the NPR, the State may choose from a 
broader menu of cost-effective, reasonable alternatives, including some 
(e.g., vehicle inspection and maintenance programs and reformulated

[[Page 57370]]

gasoline) that may even be more advantageous in light of local 
concerns.
    The task of determining the reductions necessary to meet section 
110(a)(2)(D) involves allocating the use of the downwind States' air 
basin. This area is a commons in the sense that the contributing State 
or States have a greater interest in protecting their local interests 
than in protecting an area in a downwind State over which they do not 
have jurisdiction and for which they are not politically accountable. 
Thus, in general, it is reasonable to assume that EPA may be in a 
better position to determine the appropriate goal, or budget, for the 
contributing States, while leaving to the contributing States' 
discretion to determine the mix of controls to make the necessary 
reductions.
    The EPA's decision to assign the budgets in the final rule is 
particularly reasonable. Today's rulemaking involves almost half the 
States in the Nation, and although these States participated in OTAG 
beginning more than 3 years ago, they still have not agreed on whether 
particular upwind States should be treated as having sources whose 
emissions contribute significantly to downwind nonattainment, what the 
aggregate level of emissions reductions should be, or what the State-
by-State reductions should be. The sharply divergent positions taken by 
the States in their comments on the NPR and SNPR raise doubts that 
those disagreements could ever be resolved by consensus. It is most 
efficient--indeed necessary--for the Federal government to establish 
the overall emissions levels for the various States. This is 
particularly true for an interstate pollution problem such as the one 
being dealt with in this action where the downwind areas at issue are 
affected by pollution coming from several States and the actions taken 
by each of the concerned States could have an effect on the appropriate 
action to be taken by another State. For example, if EPA did not 
specify the emissions to be prohibited from each of the various States 
affecting New York City, each of those States might claim it could 
reduce its emissions less provided other States did more. Or, a State 
close to New York might assert that it could just as effectively deal 
with its contribution to New York through additional VOC, rather than 
NOX, reductions and submit a section 110(a)(2)(D) SIP based 
on a VOC-control rather than NOX-control strategy. These 
choices, however, even assuming they were valid, necessarily relate to 
the choices that would need to be made by the other upwind States 
(e.g., Pennsylvania's choice of a VOC-dominated 110(a)(2)(D) control 
strategy to deal with its contribution to New York could affect what 
Ohio or New Jersey would need to do to deal with their own 
contributions by lowering the overall level of NOX 
reductions being obtained throughout the pertinent region). Where many 
States are involved and the choices of each individual State could 
affect the choices and decisions of the other States the need for 
initial federal action is manifest. The EPA's action to determine the 
amount of NOX emissions that each of the States must 
prohibit in this widespread geographic area is needed to enable the 
States to decide expeditiously how to achieve those reductions in an 
efficient manner that will not undermine the actions of another State. 
By notifying each State in advance of its reduction requirements, EPA 
enables each State to develop its plan with full knowledge of the 
amount and kind of reductions that must be achieved both by itself and 
other affected States. The EPA's action provides the minimum framework 
necessary for a multi-state solution to a multi-state problem while 
preserving the maximum amount of state flexibility in terms of the 
specific control measures to be adopted to achieve the needed emission 
reductions. The reasonableness of EPA's approach to the interstate 
ozone transport problem was recently recognized by a US Court of 
Appeals in the context of upholding EPA's redesignation of the 
Cleveland ozone nonattainment area to attainment in light of EPA's 
approach to the regional transport problem. In the course of doing so 
the court rejected the contention that a separate analysis of the 
current adequacy of the Cleveland SIP under section 110(a)(2)(D) was 
required as a prerequisite to redesignation. The court, after 
describing the November 7, 1997 proposed SIP call and the path EPA was 
on to deal with this multi-state regional problem, upheld EPA's 
redesignation and stated that ``[w]e find that the EPA's approach to 
the regional transport problem is reasonable and not arbitrary or 
capricious.'' Southwestern Pennsylvania Growth Alliance v. Browner, 144 
F.3d 984, 990 (6th Cir. 1998).
    As noted above, commenters have argued that if EPA determines to 
issue any SIP call, the SIP call must be more general (i.e., one that 
simply requires revised SIPs from upwind areas) and not specify the 
amounts of NOX emissions that those areas must prohibit. 
However, if EPA issued a general SIP call and an upwind State responded 
by submitting an inadequate SIP revision, EPA would disapprove that 
SIP, and in the disapproval rulemaking, EPA would be obliged to justify 
why the submitted SIP was unacceptable. Without determining an 
acceptable level of NOX reductions, the upwind State would 
not have guidance as to what is an acceptable submission. The EPA's 
determination, as part of the issuance of the SIP call, of the amounts 
of NOX emissions the SIPs must prohibit obviously provides 
for more efficient and smooth-running administrative processes at both 
the State and Federal levels. For the same reasons that EPA believes it 
is appropriate for the Agency to establish the emissions budgets under 
the authority of section 110(a)(2)(D) and (k)(5), EPA believes that it 
is necessary to do so through a rule under the general rulemaking 
authority of section 301(a). Setting such a rule is necessary, as a 
practical matter, for the Administrator's effective implementation of 
section 110(a)(2)(D). See NRDC v. EPA, 22 F.3d 1125, 1146-48. Without 
such a rule the States could be expected to submit SIPs reflecting 
their conflicting interests, which could result in up to 23 separate 
SIP disapproval rulemakings in which EPA would need to define the 
requirements that each of those States would need to meet in their 
later, corrective SIPs. That in turn would trigger a new round of SIP 
rulemakings to judge those corrective SIPs. The delay attendant to that 
process would thwart timely attainment of the ozone standards.
2. Authority and Process for Requiring SIP Submissions under the 8-Hour 
Ozone NAAQS
    a. Authority for Requiring SIP Submissions under the 8-Hour NAAQS. 
(1) SIP Submissions Under CAA Section 110(a)(1). In the NPR and SNPR, 
EPA proposed to require the 23 upwind jurisdictions to submit SIP 
revisions to reduce emissions that exacerbate ozone problems in 
downwind States under the 8-hour ozone NAAQS, as well as the 1-hour 
NAAQS. The EPA recognized that under the 8-hour NAAQS, areas have not 
yet been designated as attainment, nonattainment, or unclassifiable, 
and are not yet required to have SIPs in place. Even so, EPA proposed 
that upwind areas be required to submit SIPs meeting the requirements 
of section 110(a)(2)(D)(i)(I) with respect to the 8-hour NAAQS.
    In today's action, EPA is confirming its view that it has authority 
under the 8-hour NAAQS to require SIP submittals under section 
110(a)(2)(D)(i)(I) to reduce NOX emissions by the prescribed 
amounts. Section 110(a)(1) provides, in relevant part--


[[Page 57371]]


    Each State shall * * * adopt and submit to the Administrator, 
within 3 years (or such shorter period as the Administrator may 
prescribe) after the promulgation of a national primary ambient air 
quality standard (or any revision thereof) * * * a plan which 
provides for implementation, maintenance, and enforcement of such 
primary standard in each (area) within such State.

    Section 110(a)(2) provides, in relevant part--
    Each implementation plan submitted by a State under this Act 
shall be adopted by the State after reasonable notice and public 
hearing. Each such plan shall [meet certain requirements, including 
those found in section 110(a)(2)(D)].

    The provisions of section 110(a)(1) and (a)(2) apply by their terms 
to all areas, regardless of whether they have been designated as 
attainment, nonattainment, or unclassifiable under section 107. The 
plain meaning of these provisions, read together, is that SIP revisions 
are required under the revised NAAQS within 3 years of the date of 
revision, or earlier if EPA so requires, and that those SIP revisions 
must meet the requirements of section 110(a)(2), including subparagraph 
(D).
    That the SIP submission requirements of section 110(a)(1) are 
triggered by the promulgation of a new or revised NAAQS is made even 
clearer by comparing section 172(b), which applies by its terms only to 
areas that have been designated nonattainment under section 107. 
Section 172(b) provides, in relevant part--

    At the time the Administrator promulgates the designation of any 
area as nonattainment with respect to a [NAAQS] under section 107(d) 
* * *, the Administrator shall establish a schedule according to 
which the State containing such area shall submit a plan or plan 
revision * * * meeting the applicable requirements of subsection (c) 
of this section and section 110(a)(2) * * * Such schedule shall at a 
minimum, include a date or dates, extending no later than 3 years 
from the date of the nonattainment designation, for the submission 
of a plan or plan revision * * * meeting the applicable requirements 
of subsection (c) of this section and section 110(a)(2) * * *

    Section 172(b) establishes the schedule for submissions due with 
respect to nonattainment areas under sections 172(c) and 110(a)(2). The 
section 172(c) requirements apply only with respect to areas designated 
nonattainment.17
---------------------------------------------------------------------------

    \17\ As quoted above, section 172(b) refers to ``applicable 
requirements of * * * section 110(a)(2).'' This reference appears to 
mean those requirements of section 110(a)(2) that either (i) relate 
to all SIP submissions, such as the requirement for reasonable 
notice and public hearing in the language at the beginning of 
section 110(a)(2); or (ii) relate particularly to SIP submissions 
required for nonattainment areas, but that have not yet been 
submitted by the State.
---------------------------------------------------------------------------

    In the NPR, EPA proposed that section 110(a)(1) mandates SIP 
submissions meeting the requirements of section 110(a)(2)(D) and 
provides full authority for EPA to establish a submission date within 3 
years of the July 18, 1997 8-hour ozone NAAQS promulgation date (62 FR 
38856 (NAAQS rulemaking): 62 FR 60325 (NOx SIP call NPR)). The EPA 
further asserted in the NPR that EPA has the authority to establish 
different submittal schedules for different parts of the section 
110(a)(1) SIP revision, and that EPA may require the section 
110(a)(2)(D) submittal first so that upwind reductions may be secured 
at an earlier stage in the regional SIP planning process (62 FR 60325). 
Subsections (ii) and (iii) of this section further elaborates on the 
reasoning underlying EPA's decision to retain its proposal to require 
SIP submissions under section 110(a)(2)(D) for the 8-hour standard.
    (2) Commenters and the Definition of ``Nonattainment.'' Commenters 
challenged several aspects of EPA's proposal to evaluate the 
contribution of upwind areas under the 8-hour NAAQS. Commenters 
asserted that section 110(a)(2)(D)(i) applies to constrain emissions 
from upwind sources only with respect to downwind areas that are 
designated nonattainment. According to these commenters, until EPA 
designates areas nonattainment under the 8-hour NAAQS, EPA has no 
authority to require SIP submissions, under section 110(a)(1), from 
upwind areas with respect to the 8-hour NAAQS. One commenter pointed 
out that the new source review requirements and ozone nonattainment 
requirements enacted in the 1990 Amendments apply only to areas 
designated nonattainment.
    The EPA disagrees with this comment. Section 110(a)(2)(D)(i)(I) 
provides that a SIP must prohibit emissions that ``contribute 
significantly to nonattainment in * * * any other State.'' 
18 The provision does not, by its terms, indicate that this 
downwind ``nonattainment'' must already have been designated under 
section 107 as a nonattainment ``area.'' If the provision were to 
employ the term ``area'' in conjunction with the term 
``nonattainment,'' then it would have to be interpreted to apply only 
to areas designated nonattainment. Other provisions of the CAA do 
employ the term ``area'' in conjunction with ``nonattainment,'' and 
these provisions clearly refer to areas designated nonattainment (e.g., 
sections 107(d)(1)(A)(i), 181(b)(2)(A), 211(k)(10)(D)). Similarly, the 
provisions to which the commenter appeared to refer--section 172(b)/
172(c)(5) (new source review) and section 181(a)(1)/182 (classified 
ozone nonattainment area requirements)--by their terms apply to a 
nonattainment ``area.'' In contrast, section 110(a)(2)(D) refers to 
only ``nonattainment,'' not to a nonattainment ``area.''
---------------------------------------------------------------------------

    \18\ Section 110(a)(2)(D)(i)(I) further provides that a SIP must 
prohibit emissions that ``interfere with maintenance by * * * any 
other State.''
---------------------------------------------------------------------------

    By the same token, section 176A(a) authorizes EPA to establish a 
transport region whenever ``the Administrator has reason to believe 
that the interstate transport of air pollutants from one or more States 
contributes significantly to a violation of a [NAAQS] in one or more 
other States.'' This reference to ``a violation of a [NAAQS]'' makes 
clear that EPA is authorized to form a transport region when an upwind 
State contributes significantly to a downwind area with nonattainment 
air quality, regardless of whether the downwind area is designated 
nonattainment. The EPA believes that section 110(a)(2)(D) should be 
read the same way in light of the parallels between section 
110(a)(2)(D) and section 176A(a). Both provisions address transport and 
both are triggered when emissions from an upwind area ``contribute 
significantly'' downwind. It seems reasonable to apply a consistent 
approach to the type of affected downwind area, which would mean 
interpreting the term ``nonattainment'' in section 110(a)(2)(D) as 
synonymous with the phrase ``a violation of a [NAAQS]'' in section 
176A(a). The CAA contains other provisions, as well, that refer to the 
factual, air quality status of a particular area as opposed to its 
designation status. These provisions include, among others, (i) 
sections 172(c)(2) and 171(1), the reasonable further progress 
requirement, which requires nonattainment SIPs to provide for ``such 
annual incremental reductions in emissions * * * as * * * may * * * be 
required * * * for the purpose of ensuring attainment of the [NAAQS]'' 
(emphasis added); and (ii) section 182(c)(2), the attainment 
demonstration requirement, which mandates a ``demonstration that the 
[SIP] * * * will provide for attainment of the [NAAQS]'' (emphasis 
added). The emphasized terms clearly refer to air quality status. In a 
series of notices in the Federal Register, EPA relied on these 
references to air quality status in determining that areas seeking to 
redesignate from nonattainment to attainment did not need to complete 
ROP SIPs or attainment demonstrations--even though those requirements 
generally applied to areas

[[Page 57372]]

designated nonattainment--as long as the air quality for those 
redesignating areas was, in fact, in attainment. See ``State 
Implementation Plans; General Preamble for the Implementation of Title 
I of the Clean Air Act Amendments of 1990; Proposed Rule,'' 57 FR 
13498, 13564 (April 16, 1992); ``Determination of Attainment of Ozone 
Standard for Salt Lake and Davis Counties, Utah, and Determination 
Regarding Applicability of Certain Reasonable Further Progress and 
Attainment Demonstration Requirements: Direct Final Rule,'' 60 FR 
30189, 30190 (June 8, 1995); and ``Determination of Attainment of Ozone 
Standard for Salt Lake and Davis Counties, Utah, and Determination 
Regarding Applicability of Certain Reasonable Further Progress and 
Attainment Demonstration Requirements: Final Rule,'' 60 FR 36723, 36724 
(July 18, 1995). The EPA's interpretation was upheld by the Court of 
Appeals for the 10th Circuit, in Sierra Club v. EPA, 99 F.3d 1551, 1557 
(10th Cir. 1996).
    Accordingly, EPA believes it clear that the reference in section 
110(a)(2)(D)(i)(I) to ``nonattainment'' refers to air quality, not 
designation status. The EPA believes this matter is clearly resolved by 
reference to the terms of the provision itself, so that under the first 
step of the Chevron analysis, no further inquiry is needed. If, 
however, it were concluded that the provision is ambiguous on this 
point, then EPA believes that, under the second step in the Chevron 
analysis, EPA should be given deference for any reasonable 
interpretation. Interpreting ``nonattainment'' to refer to air quality 
is reasonable for the reasons described above.19
---------------------------------------------------------------------------

    \19\ Similarly, EPA believes that the term ``maintenance'' in 
another clause of section 110(a)(2)(D)(i)(I) refers to air quality 
status as well. This clause includes only the term ``maintenance,'' 
and does not include the term ``area.''
---------------------------------------------------------------------------

    The structure of the schedules for requiring SIP submissions and 
designating areas nonattainment provides support for EPA's 
interpretation. As noted above, section 110(a)(1) requires States to 
submit SIPs covering all their areas--regardless of whether designated, 
or how designated-- within 3 years of a NAAQS revision and requires 
that those SIPs include provisions meeting the requirements of section 
110(a)(2)(D).20 When a new or revised NAAQS is promulgated, 
section 107(d)(1) authorizes a process of up to 3 years for 
designations. States must recommend designations within one year of 
promulgation of a new or revised NAAQS and EPA must designate areas 
within 2 years of promulgation; EPA may take up to 3 years to designate 
areas if insufficient information prevents designations within 2 years. 
In the case of the 8-hour ozone NAAQS, Congress provided specific 
legislation for designations (Pub. L. 105-178 Sec. 6103). Under this 
new legislation, States are provided 2 years to make recommendations 
and EPA must designate areas within 1 year of the time State 
recommendations are due. Because of this legislation, designations must 
occur 3 years following promulgation of the NAAQS (July 2000). The EPA 
believes that it is not sensible to interpret the term 
``nonattainment'' in section 110(a)(2)(D)(i)(I) to refer to 
nonattainment designations because those designations may not be made 
until 3 years after the promulgation of a new or revised NAAQS, and the 
section 110(a)(2)(D) submittals are due within 3 years.
---------------------------------------------------------------------------

    \20\ See ``Re-issue of the Early Planning Guidance for the 
Revised Ozone and Particulate Matter (PM) National Ambient Air 
Quality Standards (NAAQS),'' memorandum from Sally L. Shaver, dated 
June 16, 1998.
---------------------------------------------------------------------------

    Further, interpreting the reference to ``nonattainment'' as a 
reference to air quality, and not designation, is consistent with the 
air quality goals of section 110(a)(2)(D) and the CAA as a whole. In 
the present case, it is clear from air quality monitoring and modeling 
that large areas of the eastern part of the United States are in 
violation of the 8-hour NAAQS, and it is also clear from air quality 
modeling studies that NOX emissions from sources in upwind 
States contribute to those air quality violations. The EPA currently 
has available all the information that it needs to determine whether 
upwind States should be required to revise their SIPs to implement 
appropriate reductions in NOX emissions. The designation 
process will clarify the precise boundaries of the downwind areas, but 
because ozone is a regional phenomenon, information as to the precise 
boundaries of the downwind areas is not necessary to implement the 
requirements of section 110(a)(2)(D)(i). As a result, no air quality 
purpose will be served by waiting until the downwind areas are 
designated nonattainment.
    On the contrary, taking action now is necessary to protect public 
health. As described in Section I.G., the regional NOX 
reductions required under today's action will allow numerous areas 
currently in violation of the 8-hour NAAQS to attain that standard. For 
the millions of people living in those areas, today's action will 
advance the date by which these areas will meet the revised ozone 
standard. Taking action now is particularly important because one of 
the sub-population groups at higher risk to ozone health effects is 
children who are active and spend more time outdoors during the summer 
months when ozone levels are elevated.
    (3) EPA's Authority to Require Section 110(a)(2)(D) Submissions in 
Accordance with section 110(a)(1). Commenters argue that sections 
110(a)(1), (a)(2), and 172(b) should be read so that only requirements 
under section 110(a)(2) that are unrelated to nonattainment are due 
under the section 110(a)(1) timetable. These commenters contend that 
requirements under section 110(a)(2) that are related to 
nonattainment--including section 110(a)(2)(D)--are due under the 
section 172(b) timetable, that is, within 3 years of the designation of 
areas as nonattainment. In support, these commenters rely on language 
in section 110(a)(1) indicating that the submissions are for plans for 
air quality regions ``within such State.'' Finally, certain commenters 
cite as further support for their position the definition of the term 
``nonattainment'' as found in section 107(d)(1)(A), claiming that the 
definition includes interstate transport areas.
    As noted above, section 110(a)(1) provides that States must submit 
SIP revisions providing ``for the implementation, maintenance and 
enforcement'' of the NAAQS in each area of the State within 3 years (or 
a shorter time prescribed by the Administrator) following promulgation 
of a new or revised NAAQS. Section 110(a)(2) then sets forth the 
applicable elements of a SIP. These provisions apply to all areas 
within the State, regardless of designation. Section 172(b) establishes 
a SIP submission schedule for nonattainment areas. It provides that at 
the time EPA designates areas as nonattainment, EPA shall establish a 
SIP submission schedule for the submission of a SIP meeting the 
requirements of section 172(c).
    While EPA agrees that there is overlap between the submission 
requirements under sections 110(a)(1)-(2) and 172(c), EPA believes that 
the plain language of section 110(a)(1)-(2) authorizes EPA to require 
the section 110(a)(2)(D) SIPs on the schedule described today, and that 
there is nothing to the contrary in section 172. Sections 110(a)(2) and 
172 contain cross-references to each other.21

[[Page 57373]]

These cross-references indicate that under certain circumstances, the 
section 110(a)(2)(D) submittal may be required under section 110(a)(1); 
and under other circumstances, the section 110(a)(2)(D) submittal may 
be required under section 172(b). These cross-references are 
particularly relevant with respect to nonattainment areas, which are 
subject to both sections 110(a) (1) and (2) and 172. In the current 
situation, EPA believes that it is appropriate to require the 
submissions to meet section 110(a)(2)(D) in accordance with the 
schedule in section 110(a)(1) rather than under the schedule for 
nonattainment areas in section 172(b).22
---------------------------------------------------------------------------

    \21\ Section 110(a)(2)(D) provides that areas designated 
nonattainment must submit SIPs in accordance with ``part D'' (which 
includes section 172). Section 172(b) requires EPA to establish a 
schedule for designated nonattainment areas to meet the requirements 
of sections 172(c) and 110(a)(2); section 172(c)(7) requires that 
nonattainment SIPs shall meet the requirements of section 110(a)(2).
    \22\ In other situations, EPA has indicated that certain 
elements of section 110(a)(2) would be better addressed in 
accordance with the timeframe established in section 172. See e.g., 
60 FR 12492, 12505 (March 7, 1995) Proposed Requirements for 
Implementation Plans and Ambient Air Quality Surveillance for Sulfur 
Oxides (Sulfur Dioxide) National Ambient Air Quality Standard.
---------------------------------------------------------------------------

    The EPA has provided that, for the revised ozone and particulate 
matter NAAQS, States must assess their section 110 SIPs by July 18, 
2000 to ensure that they adequately provide for implementing the 
revised standards. See Re-issue of the Early Planning Guidance for the 
Revised Ozone and Particulate Matter (PM) National Ambient Air Quality 
Standards (NAAQS), memorandum from Sally L. Shaver, dated June 16, 
1998. The EPA recognized that the section 110 SIP should generally be 
sufficient to address the revised NAAQS. However, the Agency noted 
three areas that the States particularly needed to assess, including 
whether the SIP adequately addressed section 110(a)(2)(D). The EPA also 
provided that the States should submit revisions to address section 
110(a)(2)(D) on the timeframe established by the final NOX 
SIP call, when issued. The submittal date that EPA has specified in the 
final NOX SIP call rule is consistent with both the Early 
Planning Guidance and with section 110(a)(1) and (2) of the CAA.
    The EPA acknowledges that it has not historically required an 
affirmative submission under section 110(a)(2)(D), applicable to 
specific sources of emissions, in response to the promulgation of a new 
or revised NAAQS. In part, this is because sufficient technical 
information was not available to determine which sources ``contribute 
significantly'' to nonattainment in a downwind area. In the absence of 
such a determination, States were unable to regulate sources under this 
provision in any meaningful way. However, based on the many analyses 
performed over the last several years, EPA believes that there is now 
affirmative information regarding significant contribution to ozone 
violations in the eastern portion of the country; in light of that 
evidence, it would not be appropriate to defer action under section 
110(a)(2)(D) until a later time.
    Moreover, as noted above, the section 172(c) SIP submissions apply 
only to areas designated nonattainment. Specifically, section 172(b) 
provides that ``[a]t the time'' EPA designates an area as 
nonattainment, EPA shall set a schedule ``according to which the State 
containing such area shall submit'' SIPs. Section 171(2) provides 
further clarification by providing that for purposes of part D of title 
I of the CAA (CAA sections 171-193) ``[t]he term `nonattainment area' 
means, for any air pollutant, an area which is designated 
`nonattainment' with respect to that pollutant within the meaning of 
section 107(d).'' By its terms then, section 172 does not apply to 
areas designated attainment or unclassifiable (even if such areas are 
not attaining the standard) or for areas not yet designated. Thus, 
section 110(a)(1) provides the only submission schedule for areas not 
designated nonattainment. For those areas, the commenters' argument 
that section 172(b) should establish the timetable for section 
110(a)(2)(D)(i) SIPs clearly fails. Since certain portions of the 23 
jurisdictions covered by this rule likely will not be designated 
nonattainment for the 8-hour standard, EPA believes that the section 
110(a)(1) schedule is the only schedule (and thus is the reasonable 
schedule) to follow for purposes of the SIP call.
    Furthermore, contrary to the commenters' assertions, the definition 
of nonattainment does not broadly include areas that contribute to 
nonattainment in a downwind State. The definition of nonattainment 
includes areas that have monitored violations of the standard and areas 
that ``contribute to ambient air quality in a nearby area'' that is 
violating the standard (section 107(d)(1)(A)(i) (emphasis added)). 
Thus, only ``nearby'' areas that contribute to violations of a standard 
will be included in the nonattainment designation; areas contributing 
to longer-range transport will not be designated nonattainment based 
solely on that longer-range transport. Therefore, they will not be 
subject to section 172(c) requirements and timing.
    The commenters argue that EPA's position that section 110(a)(1) 
governs the section 110(a)(2)(D) SIP submittal schedule leads to the 
absurd result that upwind areas will be required to submit SIPs dealing 
with their contribution to a nonattainment problem downwind before the 
downwind area will be required to submit SIPs under section 172(b). The 
commenters explain that section 110(a)(2) requires SIP submittals on a 
faster timetable (within 3 years from the date of promulgation or 
revision of a NAAQS) than section 172(b) (within 3 years from the date 
of designation as nonattainment). The commenters also contend that 
section 107 provides that States have the primary responsibility for 
ensuring attainment within their boundaries; only after a State 
implements all statutorily required and necessary measures can it 
pursue reductions in other areas through a SIP call or section 126. The 
commenters contend that the SIP call is contrary to the plain language 
of section 107 and congressional intent because it would require upwind 
areas to implement controls before the downwind area has implemented 
all statutorily required or necessary controls.
    While it is true that plans to meet the emissions budget for the 
SIP call will be due prior to nonattainment designations and attainment 
plans for areas designated nonattainment for the 8-hour standard, EPA 
does not consider this result to be absurd in the present case.
    The CAA, at least since its amendment in 1970, has required States 
to regulate ozone. For more than the past 25 years, States have focused 
on the adoption and implementation of local controls for the purpose of 
bringing nonattainment areas into attainment. Thus, historically, the 
downwind nonattainment areas have borne the brunt of the control 
obligations through the implementation of local controls. In 
comparison, areas in attainment of the NAAQS, but upwind of 
nonattainment areas, have not been required to implement controls 
designed to ameliorate the air quality problems experienced by their 
downwind neighbors.
    Since the CAA Amendment of 1977, designated nonattainment areas 
have been subject to specific local control obligations, such as 
vehicle I/M and, for stationary sources, the requirement to implement 
RACT. The CAA Amendments of 1990 tightened these control obligations 
for many areas. Moderate, serious, severe and extreme areas were 
required to reduce emissions by 15 percent between 1990 and 1996. In 
addition, each serious, severe and extreme area is required to achieve 
9 percent reductions over the succeeding 3 year periods until the area 
attains the

[[Page 57374]]

standard. Additional requirements, such as the use of RFG and the use 
of vapor recovery devices on gasoline pumps, are also required for 
certain areas (see generally, CAA section 182 and, e.g., section 
211(k)). Thus, downwind areas with nonattainment problems under the 1-
hour NAAQS are under current obligations to submit SIP revisions 
containing local control measures for that standard. For these areas, 
local reductions needed to meet the 1-hour standard are already 
occurring and will be achieved prior to or on the same schedule as 
reductions States may require in response to the SIP call.
    Furthermore, in many of the downwind areas, States have been taking 
action to reduce ozone levels for many years in order to meet the 1-
hour ozone NAAQS. Although the fact that the 8-hour ozone NAAQS is a 
new form of the ozone standard, however, should not obscure the fact 
that the downwind States have been making efforts to reduce ozone 
levels for decades. The EPA believes that the history of implementation 
by downwind areas of ozone pollution controls further mitigates the 
commenters' argument that it is absurd to require upwind areas to 
implement controls in advance of downwind attainment demonstrations 
under the 8-hour NAAQS.23
---------------------------------------------------------------------------

    \23\ Although the SIP call will provide a benefit to a wide 
number of areas, the focus of the SIP call is to reduce boundary 
conditions for a number of areas that will have difficulty attaining 
either the 1-hour or 8-hour standard (or both) without the benefit 
of reductions from outside the nonattainment area. Based on current 
monitoring data and modeling, EPA predicts that there will be a 
number of areas that are meeting the 1-hour standard that will be 
designated nonattainment for the 8-hour standard. The EPA further 
predicts that many of these areas will come back into attainment due 
solely to the emission reductions achieved by the NOX SIP 
call. However, this incidental benefit--which likely will occur 
without the need for local emission reductions--does not preclude 
EPA from requiring the SIP call reductions, which are needed to help 
other more seriously polluted areas that have long-standing 
pollution problems.
---------------------------------------------------------------------------

    Moreover, virtually all of the downwind States affected by today's 
rulemaking, due to 8-hour ozone nonattainment or maintenance problems, 
are themselves upwind contributors to problems further downwind, and, 
thus, are subject to the same requirements as the States further 
upwind.24 The reductions these downwind States must 
implement due to their additional role as upwind States will help 
reduce their own 8-hour ozone problems on the same schedule as 
emissions reductions for the upwind States. Accordingly, for the most 
part, this rulemaking does not require upwind areas to take action in 
advance of any action by downwind areas to ameliorate the downwind 
problems.
---------------------------------------------------------------------------

    \24\ Maine, New Hampshire, and Vermont are the only downwind 
States that are not subject to today's action.
---------------------------------------------------------------------------

    Finally, even if EPA were requiring upwind States to take action to 
reduce downwind nonattainment and maintenance in advance of action by 
the downwind States, this would simply require upwind areas to take the 
first step by developing SIPs to eliminate their significant 
contribution to the downwind problem. The downwind areas will be 
required to take the next step by developing SIPs that address their 
share. Generally, an agency may resolve a problem (in this case, 
downwind nonattainment) on a step-by-step basis (see e.g., Group 
Against Smog and Pollution, Inc. v. EPA, 665 F.2d 1284, 1291-92 (D.C. 
Cir. 1981)).
    A commenter has observed that under section 110(a)(1), EPA may 
authorize section 110(a)(2) submittals as late as 3 years after 
revision of a NAAQS, which, in this case, would run until July 2000. 
The Early Planning Guidance, described above, indicates that States are 
allowed until July 2000 to make submissions concerning other elements 
of section 110(a)(2). However, as described elsewhere, EPA has 
determined that the section 110(a)(2)(D) submittals should be submitted 
by the end of September 1999 to assure that the required NOX 
reductions will be implemented as expeditiously as practicable, which 
EPA has determined is no later than the May 1 start of the 2003 ozone 
season (see Section V, below).
    Citing section 107(a) of the CAA, the commenters assert that the 
CAA requires downwind areas to fully adopt and implement all 
statutorily required or necessary measures before EPA can require 
upwind areas to control emissions. Section 107 provides that States 
shall have the primary responsibility for assuring air quality within 
the State by submitting a plan that specifies how the NAAQS will be 
achieved and maintained in the State. The commenters attempt to read 
this statement regarding a State's authority to choose the mix of 
control measures within State boundaries as barring the control of 
emissions from upwind States.
    This provision may be read as focusing on the State-Federal balance 
in controlling criteria pollutants, such as ozone, not any upwind-
State, downwind-State balance. The provision indicates that although 
EPA may promulgate Federal measures that provide reductions to help 
States reach attainment, States bear the ultimate responsibility for 
assuring attainment. Further, this provision may be read to indicate 
that States may choose the mix of controls to reach attainment within 
their own boundaries. Nothing in this provision purports to address the 
need for upwind controls. By comparison, section 110(a)(2)(D) 
affirmatively requires States to submit a SIP prohibiting emissions 
that significantly contribute to downwind nonattainment or interfere 
with maintenance of the NAAQS. Thus, the statute, read as a whole, 
contemplates that interstate transport will be addressed as part of the 
downwind States' attainment responsibilities. Indeed, determining the 
upwind area's share of the problem is necessary in order for downwind 
attainment planning. In the absence of the upwind reductions that will 
be achieved, the downwind area would be required to submit an 
attainment plan to demonstrate attainment regardless of cost and 
without benefit of the reduction of upwind emissions that significantly 
contribute to nonattainment. In light of the statute as a whole, it is 
absurd to argue that Congress intended downwind areas to reduce 
emissions at any cost while upwind sources that significantly 
contribute to that nonattainment remain unregulated. Congress attempted 
to balance responsibilities, providing that States could choose the mix 
of controls within the State's borders (CAA section 107(a)) and are 
ultimately responsible for assuring attainment, but also recognizing 
that emissions reductions from upwind States may be needed for 
attainment (CAA section 110(a)(2)(D)(i)).
    b. Process for Requiring SIP Submissions under the 8-Hour Standard. 
The time by which the section 110(a)(2)(D) SIP revision under the 8-
hour NAAQS must be submitted is governed by section 110(a)(1), which 
requires the SIP revision to be ``adopt[ed] and submit[ed] to the 
Administrator, within 3 years (or such shorter period as the 
Administrator may prescribe) after the promulgation of a [NAAQS] (or 
any revision thereof) . . . .'' In the NPR, EPA indicated that the SIP 
revision would be due by the end of September 1999, which EPA expected 
to be 12 months from the date of completing today's final rule. In 
today's action, EPA is confirming that the SIP revision will be due 
September 30, 1999, for the reasons described below in Section VI.A.1, 
Schedule for SIP Revision.
3. Requirements of Section 110(a)(2)(D)
    a. Summary. Today's action is driven by the requirements of CAA 
section 110(a)(2)(D). This provides that each SIP must--


[[Page 57375]]


    * * * contain adequate provisions--(I) prohibiting, consistent 
with the provisions of this title, any source or other type of 
emissions activity within the State from emitting any air pollutant 
in amounts which will--(I) contribute significantly to nonattainment 
in, or interfere with maintenance by, any other State with respect 
to any such national primary or secondary ambient air quality 
standard * * * 

    According to section 110(a)(2)(D), the SIP for each area, 
regardless of its designation as nonattainment or attainment (including 
unclassifiable), must prohibit sources within the area from emitting 
air pollutants in amounts that will ``contribute significantly'' to 
``nonattainment'' in a downwind State, or that ``interfere with 
maintenance'' in a downwind State.
    b. Determination of Meaning of ``Nonattainment'' (1) Geographic 
Scope. In determining the meaning and scope of section 110(a)(2)(D), it 
is useful first to determine the geographic scope of ``nonattainment'' 
downwind.
    At proposal, EPA stated that it--

    * * * proposes to interpret this term to refer to air quality 
and not to be limited to currently-designated nonattainment areas. 
Section 110(a)(2)(D) does not refer to ``nonattainment areas,'' 
which is a phrase that EPA interprets to refer to areas that are 
designated nonattainment under * * * section 107(d)(1)(A)(I) * * * . 
Rather, the provision includes only the term `nonattainment' and 
does not define that term. Under these circumstances, EPA has 
discretion to give the term a reasonable definition, and EPA 
proposes to define it to include areas whose air quality currently 
violates the NAAQS, and will likely continue [to violate in the 
future], regardless of the designation of those areas * * *

(62 FR 60324).
    To determine whether areas would continue to violate in the future, 
EPA proposed to take into account the reductions that would result from 
current CAA control requirements (apart from controls that may be 
required under section 110(a)(2)(D)). To take these reductions into 
account, EPA determined whether the area would be in nonattainment in 
the future based on air quality modeling that assumed CAA-mandated 
reductions and that accounted for growth. If an area would reach 
attainment based on required controls, EPA would not view that area as 
having a nonattainment problem to which any upwind areas may be 
considered to contribute.
    As explained earlier, in today's action, EPA has determined that 
for purposes of the 8-hour NAAQS, the reference to ``nonattainment'' 
should be defined as EPA proposed. Thus, in determining whether an 
upwind area contributes significantly to ``nonattainment'' downwind, 
EPA would evaluate downwind areas for which monitors indicate current 
nonattainment, and air quality models indicate future nonattainment, 
taking into account CAA control requirements and growth.
    For the 1-hour standard, EPA proposed to define nonattainment to 
include all grid cells within a county when a monitor in that county 
indicated nonattainment. Upon further study, EPA found that in some 
instances, a metropolitan area may consist of numerous counties, only a 
few of which contain monitors indicating nonattainment. The EPA 
recognizes that under the 1-hour NAAQS, nonattainment boundaries are 
generally used to describe the area with the nonattainment problem; 
accordingly, EPA believes that this geographic vicinity offers an 
appropriate indication of an area that may be expected to have 
nonattainment air quality. The EPA predicts that many 1-hour 
nonattainment areas that currently monitor nonattainment somewhere 
within the area will remain in nonattainment in 2007, in some cases 
because of predicted violations in counties that currently monitor 
attainment. The EPA believes that the entire area should be considered 
to be in nonattainment until all monitors in the area indicate 
attainment of the NAAQS. Thus, in today's action, EPA used the 
designated nonattainment area in determining the downwind nonattainment 
problem.25
---------------------------------------------------------------------------

    \25\ It should be reiterated that EPA relied on the designated 
area solely as a proxy to determine which areas have air quality in 
nonattainment. This proxy is readily available under the 1-hour 
NAAQS because areas have long been designated nonattainment. The 
EPA's reliance on designated nonattainment areas for purposes of the 
1-hour NAAQS does not indicate that the reference in section 
110(a)(2)(D)(i)(I) to ``nonattainment'' should be interpreted to 
refer to areas designated nonattainment.
---------------------------------------------------------------------------

    As noted above, commenters disagreed with EPA's view that the term 
``nonattainment'' covers areas with air quality that is currently in 
nonattainment, regardless of designation. The EPA's response to those 
comments is also set forth above.
    (2) 2007 Projection Year. In the NPR, EPA indicated that it would 
adopt the year 2007 as the year for determining whether areas achieved 
their required NOX budget levels. Accordingly, in 
determining whether downwind areas should be considered to be, and 
remain in, ``nonattainment,'' EPA would model their air quality in 
2007, based on the implementation of CAA required controls by that 
date, and growth in emissions--generally due to economic growth and 
greater use of vehicles--by that date. At proposal, EPA adopted this 
same approach with respect to both the 1-hour and the 8-hour NAAQS (62 
FR 60325). The EPA is continuing this approach.
    c. Definition of Significant Contribution. As indicated in the NPR, 
neither the CAA nor its legislative history provides meaningful 
guidance for interpreting the term ``contribute significantly'' under 
section 110(a)(2)(D)(i)(I).
    (1) ``Contribute.'' The initial step in defining the ``contribute 
significantly'' term is to determine the meaning of the term 
``contribute.'' In the NPR, EPA stated that it believes this term 
should be defined broadly, so that emissions ``contribute'' to 
nonattainment downwind if they have an impact on nonattainment downwind 
(62 FR 60325). Air quality modeling indicated that emissions from the 
upwind States clearly impact downwind nonattainment problems; as a 
result, EPA generally folded this step of determining whether sources 
``contribute'' to nonattainment downwind into the step of determining 
whether that contribution is ``significant,'' discussed below.
    In addition, section 110(a)(2)(D)(i)(I) requires the SIP to 
prohibit amounts of emissions ``which will contribute significantly * * 
*'' (emphasis added). The EPA believes that the term ``will'' means 
that SIPs are required to eliminate the appropriate amounts of 
emissions that presently, or that are expected in the future, 
contribute significantly to nonattainment downwind.
    Because ozone is a secondary pollutant formed as a result of 
complex chemical reactions involving numerous sources, it is not 
possible to determine the downwind impact on each individual source. In 
addition, ozone generally results from the contributions of numerous 
sources. As indicated in the NPR:

    [U]nhealthful levels of ozone result from emissions of 
NOX and VOCs from thousands of stationary sources and 
millions of mobile sources [and consumer products and other sources] 
across a broad geographic area. Each source's contribution is a 
small percentage of the overall problem; indeed, it is rare for 
emissions from even the largest single sources to exceed one percent 
of the inventory of ozone precursors even for a single metropolitan 
area. Under these circumstances, even complete elimination of any 
given source's emissions may well have no measurable impact in 
ameliorating the nonattainment problem. Rather, attainment requires 
controls on numerous sources across a broad area. Ozone is a 
regional scale

[[Page 57376]]

problem that requires regional scale reductions

(62 FR 60326).
    Accordingly, EPA has adopted a ``collective contribution'' approach 
to determining whether sources ``contribute'' to nonattainment 
downwind: EPA determines the impact downwind of emissions in the 
aggregate from a particular geographic region. If the aggregated 
emissions are considered to contribute to nonattainment downwind, then 
all of the emissions in that region should be considered as 
contributors to that nonattainment problem. In today's action, EPA is 
continuing the same interpretation of the term ``contribute,'' for the 
reasons just described.
    (2) ``Significantly''. (a) Notice of Proposed Rulemaking. In the 
NPR, EPA proposed a ``weight-of-evidence,'' or multi-factor, approach 
for determining whether a contribution is ``significant.''
    The EPA proposed two separate interpretations for the term 
``contribute significantly,'' which had implications as to which 
factors were to be considered in what parts of the analysis. Under the 
first interpretation, significant contribution is determined with 
reference to--

    * * * factors concerning amounts of emissions and their ambient 
impact, including the nature of how the pollutant is formed, the 
level of emissions and emissions density (defined as amount of 
emissions per square mile) in the particular upwind area, the level 
of emissions in other upwind areas, the amount of contribution to 
ozone in the downwind area from the upwind areas, and the distance 
between the upwind sources and the downwind nonattainment problem. 
Under this approach, when emissions and ambient impact reach a 
certain level, as assessed by reference to the factors identified 
above, those emissions would be considered to ``contribute 
significantly'' to nonattainment.

(62 FR 60325).
    Under this interpretation, after identifying amounts of emissions 
that constitute a significant contribution, EPA then determines the 
amount of emissions reductions necessary to adequately mitigate these 
contributions. This determination entails--

    * * * [e]valuation of the costs of available measures for 
reducing upwind emissions * * * as well as to the extent known (at 
least qualitatively), the relative costs of, amounts of reductions 
from, and ambient impact of measures available in the downwind 
areas.

Id.
    Under the second interpretation, EPA considers all of the factors 
under both the significant contribution prong and the mitigation prong 
of the first interpretation, and, once EPA determines an amount of 
emissions that does significantly contribute to downwind nonattainment, 
then EPA would determine that the SIP must contain provisions adequate 
to prohibit that amount of emissions. Id. at 60325-26.
    (b) Today's Action. The EPA has determined that the second 
interpretation should be used; that is, that the determination of 
significant contribution includes both air quality factors relating to 
amounts of upwind emissions and their ambient impact downwind, as well 
as cost factors relating to the costs of the upwind emissions 
reductions. Once an amount of emissions is identified in an upwind 
State that contributes significantly to a nonattainment problem 
downwind, or interferes with maintenance downwind, the SIP must include 
provisions to eliminate that amount of emissions.
    To reiterate, section 110(a)(2)(D)(i)(I) provides that the SIP must 
``prohibit[]'' sources from ``emitting any air pollutant in amounts 
which will contribute significantly to nonattainment in, or interfere 
with maintenance by, any other State.'' The term ``prohibit'' is 
defined as ``to forbid by authority'' or ``prevent,'' or ``preclude.'' 
``The American Heritage Dictionary of the English Language'' (3d ed. 
1992, 1448). The EPA believes that the term ``prohibit'' means that 
SIPs must eliminate those amounts of emissions determined to contribute 
significantly to nonattainment or interfere with maintenance downwind. 
Moreover, EPA believes that whether emissions ``contribute 
significantly'' depends on a multifactor test, as described below. 
Thus, section 110(a)(2)(D)(i)(I) does not require the elimination of 
all upwind source emissions that impact downwind air quality problems, 
but only those amounts of emissions that, based on a multi-factor test, 
significantly contribute to downwind air quality problems.
    d. Multi-factor Test for Determining Significant Contribution. In 
the NPR, EPA proposed a multi-factor test for determining whether 
emissions from an upwind State contribute significantly to a 
nonattainment or maintenance problem downwind. The EPA received 
numerous comments on the factors. Based on the comments and EPA's 
further analysis, EPA, in today's action, is continuing the multi-
factor approach, with some refinements in response to comments, with 
respect to the factors EPA considered and the manner in which EPA 
considered them.
    In determining whether emissions from upwind States affected by 
today's action contribute significantly to downwind nonattainment or 
maintenance problems, EPA specifically considered the following factors 
with respect to each such upwind State. These factors were the primary 
components in EPA's consideration.
     The overall nature of the ozone problem (i.e., 
``collective contribution'')
     The extent of the downwind nonattainment problems to 
which the upwind State's emissions are linked, including the ambient 
impact of controls required under the CAA or otherwise implemented in 
the downwind areas
     The ambient impact of the emissions from the upwind 
State's sources on the downwind nonattainment problems
     The availability of highly cost effective control 
measures for upwind emissions.
    The first three of these factors are related to air quality; the 
fourth is related to costs.
    In addition, EPA generally reviewed several other considerations 
before concluding that upwind emissions contribute significantly to 
downwind nonattainment. The EPA did not consider it necessary, or did 
not have adequate information, to apply each of these factors with 
specificity with respect to each upwind State's emissions. In addition, 
in some instances, EPA did not have quantitative information to assess 
certain of these factors, and instead relied on qualitative 
information. These considerations were secondary aspects of EPA's 
analysis. They include:
     The consistency of the regional reductions with the 
attainment needs of the downwind areas with nonattainment problems
     The overall fairness of the control regimes required of 
the downwind and upwind areas, including the extent of the controls 
required or implemented by the downwind and upwind areas
     General cost considerations, including the relative 
cost-effectiveness of additional downwind controls compared to upwind 
controls
    All of these factors and considerations are described in the 
following sections.
    e. Air Quality Factors. As noted above, EPA specifically considered 
three air quality factors with respect to each upwind State, which 
factors, in conjunction with the cost factor discussed in the next 
section, were the primary components in EPA's consideration:
     The overall nature of the ozone problem (i.e., 
``collective contribution'')
     The extent of the downwind nonattainment problems to 
which the upwind State's emissions are linked,

[[Page 57377]]

including the ambient impact of controls required under the CAA or 
otherwise implemented in the downwind areas
     The ambient impact of the emissions from the upwind 
State's sources on the downwind nonattainment problems
    (1) Collective Contribution. As indicated elsewhere, ozone 
generally results from the collective contribution of emissions from 
numerous sources over a large geographic area. For example, for urban 
nonattainment areas under the 1-hour NAAQS, the downwind sources, 
comprise numerous stationary sources as well as mobile on-road sources, 
mobile off-road sources, and consumer and commercial products. Further, 
additional contributions are made by numerous upwind States, both 
adjacent to and further away from the nonattainment area itself. The 
fact that virtually every nonattainment problem is caused by numerous 
sources over a wide geographic area is a factor suggesting that the 
solution to the problem is the implementation over a wide area of 
controls on many sources, each of which may have a small or 
unmeasureable ambient impact by itself.
    (2) Extent of Downwind Nonattainment Problems, Including Ambient 
Impact of Required Controls. In determining whether a downwind area has 
a nonattainment problem under the 1-hour standard to which an upwind 
area may be determined to be a significant contributor, EPA determined 
whether the downwind area currently has a nonattainment problem, and 
whether that area area would continue to have a nonattainment problem 
as of the year 2007 assuming that in that area, all controls 
specifically required under the CAA were implemented, and all required 
or otherwise expected Federal measures were implemented. If, following 
implementation of such required CAA controls and Federal measures, the 
downwind area would remain in nonattainment, then EPA considered that 
area as having a nonattainment problem to which upwind areas may be 
determined to be significant contributors.
    Thus, this analytical approach assumes that downwind areas 
implement all required controls and receive the benefit of reductions 
from Federal measures, and yet have a residual nonattainment problem 
(prior to the implementation of the regional reductions required by 
today's action). The fact that a nonattainment problem persists, 
notwithstanding fulfillment of CAA requirements by the downwind 
sources, is a factor suggesting that it is reasonable for the upwind 
sources to be part of the solution to the ongoing nonattainment 
problem.
    The EPA undertook a comparable analysis with respect to the 8-hour 
NAAQS. That is, the major urban areas in the northeast, midwest, and 
south that are violating the 8-hour NAAQS are designated nonattainment 
under the 1-hour NAAQS as well. After these areas are designated 
nonattainment under the 8-hour NAAQS, they will become subject to the 
control requirements of section 172(c). However, for these areas, the 
section 172(c) requirements do not, by their terms, impose any specific 
controls other than what these areas have already implemented to 
fulfill the requirements under section 182 attendant to their 
designation and classification under the 1-hour NAAQS. Accordingly, the 
same air quality modeling analyses that shows residual nonattainment 
for at least one of the urban areas linked to each upwind State under 
the 1-hour standard shows residual nonattainment for those areas under 
the 8-hour NAAQS. Indeed, modeling analyses relied on for today's 
action indicate residual nonattainment for the major urban areas even 
after the implementation of regional reductions comparable to those 
required today.26
---------------------------------------------------------------------------

    \26\ The presence of residual nonattainment in major urban areas 
after their implementation of specifically required CAA controls 
supports the regional reductions required under today's action. 
Those regional reductions allow the major urban areas to progress 
towards attainment under the 8-hour NAAQS, and, at the same time, 
significantly ameliorate the nonattainment problems under the 8-hour 
NAAQS for numerous other areas. In fact, EPA projections indicate 
that numerous areas with nonattainment problems will achieve 
attainment of the 8-hour NAAQS as a result of the regional 
reductions.
---------------------------------------------------------------------------

    (3) Ambient Impact of Emissions from the Upwind Sources. In today's 
action, EPA examined the impact of numerous upwind States on numerous 
downwind areas with nonattainment problems.
    Under the 1-hour NAAQS, EPA conducted various air quality modeling 
analyses that examined the impact of emissions from sources in each 
upwind State on ozone levels in downwind nonattainment areas, in light 
of the impact of emissions from sources in other upwind States on the 
downwind area's nonattainment problem. The EPA assessed the frequency 
and magnitude of each upwind State's contribution to downwind 
nonattainment problems. Some of the modeling analyses also permitted 
determining the magnitude of the average contribution and the peak 
contribution from each upwind State, as well as the percentage of each 
upwind State's contribution to the downwind nonattainment problem.
    The EPA determined that for each upwind State affected by today's 
action, its contribution to a downwind nonattainment problem, in 
conjunction with the contribution from other upwind States, comprised a 
relatively large percentage of the nonattainment problem. The EPA 
further determined that, in this context, the impacts from each 
affected upwind State's NOX emissions are sufficiently large 
and/or frequent so that the amounts of that State's emissions should be 
considered to be significant contributions, depending on the cost 
factor and other relevant considerations. For most upwind States, EPA 
conducted two types of modeling--UAM-V and CAMx--that isolated the 
impact of emissions from the upwind State alone on downwind 
nonattainment.
    The EPA also conducted much the same analysis to determine the 
impact of emissions from each upwind State on ozone levels in downwind 
States under the 8-hour NAAQS. Because nonattainment problems under the 
8-hour NAAQS are widespread, and because EPA has not designated 
individual nonattainment areas, EPA focused this part of its inquiry on 
the upwind State's impact on the entire downwind State.
    The EPA's analysis under both the 1-hour and 8-hour NAAQS led EPA 
to conclude that, in light of both the collective contribution nature 
of the ozone problem, and the fact that downwind areas continue to 
suffer a nonattainment problem even after implementation of all 
required CAA measures and Federal measures, emissions from each of the 
affected upwind States have a sufficiently large and/or frequent 
ambient impact such that those emissions contribute significantly to 
nonattainment downwind, depending on the availability of highly cost-
effective measures and on other considerations discussed below.
    f. Determination of Highly Cost-effective Reductions and of 
Budgets. After determining the degree to which NOX 
emissions, as a whole from the particular upwind States, contribute to 
downwind nonattainment or maintenance problems, EPA then determined 
whether any amounts of the NOX emissions may be eliminated 
through controls that, on a cost-per-ton basis, may be considered to be 
highly cost effective. By examining the cost effectiveness of recently 
promulgated or proposed NOX controls, EPA determined that an 
average of approximately $2,000 per ton removed

[[Page 57378]]

is highly cost effective. The EPA then determined a set of controls on 
NOX sources that would cost no more than an average of 
$2,000 per ton reduced. Specifically, EPA determined that one set of 
these controls would include a cap-and-trade program for (i) 
electricity generating boilers and turbines larger than 25 Mwe (``large 
EGUs''), and (ii) large non-electricity generating industrial boilers 
and turbines (``large non-EGU boilers and turbines''). The application 
of an emission rate of 0.15 lb/mmBtu and 1995-1996 utilization for EGUs 
and 60 percent for large non-EGUs to the emissions projected to occur 
in 2007 including growth and CAA measures, led to the determination of 
the amounts to be reduced. The remaining amount is a State's budget.
    The EPA further determined that additional highly cost-effective 
controls are also available for cement manufacturing sources and 
internal combustion engines. On the basis of reasonable assumptions 
concerning growth to the year 2007, EPA then determined the amounts of 
emissions from these source categories that would be eliminated with 
those controls.
    The EPA further determined that there were no other controls on 
other NOX sources that qualify as highly cost effective 
(although several controls are reasonably cost-effective).
    On the basis of the determinations just described for the various 
source categories, EPA determined an amount of NOX emissions 
that may be eliminated through these highly cost-effective measures. 
Because EPA had also determined that the NOX emissions from 
the affected upwind States have a large and/or frequent impact on 
downwind nonattainment or maintenance problems, EPA concludes that the 
amount of NOX emissions from those States that can be 
eliminated through application of highly cost-effective control 
measures contributes significantly to nonattainment or maintenance 
problems downwind.
    Under section 110(a)(2)(D)(i)(I), the SIP must include ``adequate 
provisions prohibiting'' sources from emitting these ``amounts.'' 
Because no highly cost-effective controls are available to eliminate 
the remaining amounts of NOX emissions, EPA concludes that 
those emissions do not contribute significantly to downwind 
nonattainment or maintenance problems. As indicated below and in 
Section III, there are cost-effective alternatives available to States 
that choose not to adopt all of the highly cost-effective measures on 
which EPA based its selection of the significant amounts of 
NOX emissions.
    To implement EPA's determinations, each affected upwind State is 
required to submit for EPA approval SIP controls projected to be 
sufficient, by the year 2007, to eliminate the amount of NOX 
emissions in the State that EPA determined contributes significantly to 
nonattainment. The EPA determined this amount of reductions, for each 
affected upwind State, as follows: EPA first determined the amount of 
NOX emissions in that State by the year 2007, based on 
assumptions concerning both growth and emissions controls that are 
required under the CAA or that will be implemented due to Federal 
actions (the ``2007 base case''). Second, EPA applied the control 
measures identified as highly cost effective to the 2007 base case 
amount for the appropriate source categories. The amount of 
NOX emissions remaining in the State after application of 
controls to the affected source categories constitutes the 2007 budget. 
The difference between the 2007 base case and the 2007 budget is the 
amount of NOX emissions in that State by the year 2007 that 
EPA has determined to contribute significantly to nonattainment and 
that, therefore, the SIPs must prohibit.
    The upwind State's SIP revision due in response to today's action 
must provide controls that, on the basis of the same assumptions 
(including concerning growth) made by EPA in determining the budget, 
would limit NOX emissions in the year 2007 to no more than 
the 2007 budget. The State has full discretion in selecting the 
controls, so that it may choose any set of controls that would assure 
achievement of the budget.
    As EPA stated in the NPR:

    States are not constrained to adopt measures that mirror the 
measures EPA used in calculating the budgets. In fact, EPA believes 
that many control measures not on the list relied upon to develop 
EPA's proposed budgets are reasonable--especially those, like 
enhanced vehicle inspection and maintenance programs, that yield 
both NOX and VOC emissions reductions.[ 27] 
Thus, one State may choose to primarily achieve emissions reductions 
from stationary sources while another State may focus emission 
reductions from the mobile source sector.

    \27\ As indicated in the NPR, EPA considers that measures may be 
reasonable in light of their reduction of VOC and NOX 
emissions, even though their cost-effectiveness in terms of cost per 
NOX emissions removed is relatively high (62 FR 60346-
48).
---------------------------------------------------------------------------

(62 FR 60328).
    The EPA believes that its overall approach derives further support 
from the mandate in section 110(a)(2)(D) that each SIP include 
provisions prohibiting ``any source or other type of emissions activity 
within the State from emitting any air pollutant in amounts' that 
adversely affect downwind areas. The phrase ``any source or other type 
of emissions activity'' may be interpreted to require that the SIP 
regulate all sources of emissions to assure that the total amount of 
emissions generated within the State does not adversely affect downwind 
areas. By its terms, the phrase covers all emitters of any kind because 
every emitter--stationary, mobile, or area--may be considered a 
``source or other type of emissions activity.'' This interpretation is 
consistent with the legislative history of the phrase. Prior to the CAA 
Amendments of 1990, the predecessor to section 110(a)(2)(D), which was 
section 110(a)(2)(E), referred to ``any stationary source within the 
State.'' In the 1990 Amendments, Congress revised the phrase to read as 
it currently does. A Committee Report explained, ``Where prohibitions 
in existing section 110(a)(2)(E) apply only to emissions from a single 
source, the amendment includes ``any other type of emissions 
activity,'' which makes the provision effective in prohibiting 
emissions from, for example, multiple sources, mobile sources, and area 
sources.'' V Leg. Hist. 8361, S. Rep. No. 228, 101st Cong., 1st Sess. 
21 (1989).
    For reasons explained below, if an upwind State chooses to achieve 
all or a portion of the required reductions from large EGUs or large 
non-EGU boilers and turbines, then the SIP must include a mass 
emissions limitation for those sources computed with reference to 
certain growth assumptions and the emission rate limits chosen by the 
State. The EPA recommends that this mass limitation, or cap, be 
accompanied by a trading program. Any such cap-and-trade program must 
be established by May 1, 2003. If the State chooses to achieve all or a 
portion of the required reductions from other sources, then the State 
must implement controls, by the year 2003, on those other sources that 
are projected to achieve the required level of reductions, based on 
certain assumptions (including growth), in the year 2007. The controls 
on these other sources may be rate-based, and no emissions cap on them 
is required. By the year 2007, any applicable mass emissions limitation 
for large EGUs or large non-EGU boilers and turbines must continue to 
be met, and any applicable controls on other sources must continue to 
be implemented. The amount of the 2007 overall budget is used to 
compute the level of controls that would result in the appropriate 
amount of emissions reductions, given assumptions concerning, for 
example,

[[Page 57379]]

growth. To this extent, the 2007 overall budget is an important 
accounting tool. However, the State is not required to demonstrate that 
it has limited its total NOX emissions to the budget 
amounts. Thus, the overall budget amount is not an independently 
enforceable requirement.
    g. Other Considerations in Determination of Significant 
Contribution. The EPA reviewed several other considerations in support 
of its determination that the specified amounts of emissions from the 
affected upwind States contribute significantly to nonattainment 
downwind.
    (1) Consistency of Regional Reductions with Downwind Attainment 
Needs. The EPA conducted modeling analyses of emission reductions of 
virtually the same magnitude as the regional reductions required under 
today's action. Although the impact on any downwind ozone problem of 
each upwind State's emissions reductions alone may be relatively small, 
the impact of those reductions, when combined with the reductions from 
the other States, is substantial. Based on this modeling, EPA 
determined that the regional reductions allow downwind nonattainment 
areas under the 1-hour NAAQS to make appreciable progress towards 
attainment. The EPA further determined that under the 8-hour NAAQS, 
many areas with nonattainment problems are expected to reach attainment 
based solely on the regional reductions, and that other (primarily 
urban) areas would benefit from the regional reductions but are 
expected to experience residual nonattainment. EPA further determined 
that none of the upwind States affected by today's action are affected 
by ``overkill,'' that is, required reductions that are more than 
necessary to ameliorate downwind nonattainment in every downwind area 
affected by that upwind State.
    (2) Fairness. The EPA also considered the overall fairness of the 
control regimes required of the downwind and upwind areas, including 
the extent of the controls required or implemented by the downwind and 
upwind areas. Most broadly, EPA believes that overall notions of 
fairness suggest that upwind sources which contribute significant 
amounts to the nonattainment problem should implement cost-effective 
reductions. When upwind emitters exacerbate their downwind neighbors' 
ozone nonattainment problems, and thereby visit upon their downwind 
neighbors additional health risks and potential clean-up costs, EPA 
considers it fair to require the upwind neighbors to reduce at least 
the portion of their emissions for which highly cost-effective controls 
are available.
    In addition, EPA recognizes that in many instances, areas 
designated as nonattainment under the 1-hour NAAQS have incurred ozone 
control costs since the early 1970s. Moreover, virtually all components 
of their NOx and VOC inventories are subject to SIP-required or Federal 
controls designed to reduce ozone. Furthermore, these areas have 
complied with almost all of the specific control requirements under the 
CAA, and generally are moving towards compliance with their remaining 
obligations. The CAA's sanctions and FIP provisions provide assurance 
that these remaining controls will be implemented. By comparison, many 
upwind States in the midwest and south have had fewer nonattainment 
problems and have incurred fewer control obligations.
    (3) General Cost Considerations. The EPA also considered the fact 
that in general, areas that currently have, or that in the past have 
had, nonattainment problems under the 1-hour NAAQS, or that are in the 
Northeast Ozone Transport Region (OTR), have already incurred ozone 
control costs. The controls already implemented in these areas tend to 
be among the less expensive of available controls. As described in more 
detail below, EPA has determined that, in general, the next set of 
controls identified as available in the downwind nonattainment areas 
under the 1-hour NAAQS would cost approximately $4,300 per ton removed. 
By comparison, EPA has determined that the cost of the regional 
reductions required today would approximate $1,500 per ton removed. 
Thus, it appears that the upwind reductions required by today's action 
are more cost-effective per ton removed than reductions in the downwind 
nonattainment areas. Moreover, under the 1-hour NAAQS, the reductions 
required from each upwind State, in conjunction with reductions from 
other upwind States, result in ambient improvement in at least several 
downwind areas with nonattainment problems.
    The EPA did not have available, and was not presented with, 
meaningful quantitative information indicating the cost-effectiveness 
of the regional reductions required today in light of their ambient 
impact downwind (e.g., the cost of emissions reductions per ppb 
improvement in ambient ozone levels in a downwind nonattainment area). 
This lack of information limited the extent to which EPA could rely on 
this consideration in making its determinations.
    The various considerations just discussed point in the same 
direction as the other factors described above concerning air quality 
and costs. These factors and considerations lead EPA to conclude that 
the amounts of each upwind State's emissions that may be eliminated 
through highly cost-effective measures contribute significantly to 
nonattainment or maintenance problems downwind.
    h. Interfere with Maintenance. Once a nonattainment area has 
attained the NAAQS, it is required to maintain that standard (e.g., 
sections 107(d)(3)(E)(iv), 110(a)(1)). Section 110(a)(2)(D)(i)(I) also 
requires that SIPs contain adequate provisions prohibiting amounts of 
emissions that ``interfere with maintenance by * * * any [downwind] 
State.'' The EPA explained and applied this requirement in the NPR as 
follows:

    This [interfere-with-maintenance] requirement * * * does not, by 
its terms, incorporate the qualifier of ``significantly.'' Even so, 
EPA believes that for present purposes, the term ``interfere'' 
should be interpreted much the same as the term ``contribute 
significantly,'' that is, through the same weight-of-evidence 
approach.
    With respect to the 1-hour NAAQS, the ``interfere-with-
maintenance'' prong appears to be inapplicable. The EPA has 
determined that the 1-hour NAAQS will no longer apply to an area 
after EPA has determined that the area has attained that NAAQS. 
Under these circumstances, emissions from an upwind area cannot 
interfere with maintenance of the 1-hour NAAQS.
    With respect to the 8-hour NAAQS, the ``interfere-with-
maintenance'' prong remains important. After an area has reached 
attainment of the 8-hour NAAQS, that area is obligated to maintain 
that NAAQS. (See sections 110(a)(1) and 175A.) Emissions from 
sources in an upwind area may interfere with that maintenance.
    The EPA proposes to apply much the same approach in analyzing 
the first component of the ``interfere-with-maintenance'' issue, 
which is identifying the downwind areas whose maintenance of the 
NAAQS may suffer interference due to upwind emissions. The EPA has 
analyzed the ``interfere-with-maintenance'' issue for the 8-hour 
NAAQS by examining areas whose current air quality is monitored as 
attaining the 8-hour NAAQS [or which have no current air quality 
monitoring], but for which air quality modeling shows nonattainment 
in the year 2007. This result is projected to occur, notwithstanding 
the imposition of certain controls required under the CAA, because 
of projected increases in emissions due to growth in emissions 
generating activity. Under these circumstances, emissions from 
upwind areas may interfere with the downwind area's ability to 
attain. Ascertaining the impact on the downwind area's air quality 
of the upwind area's emissions aids in determining whether the 
upwind emissions interfere with maintenance


[[Page 57380]]


(62 FR 60326).
    In today's action, EPA is taking the same positions with respect to 
the interfere-with-maintenance test as described in the NPR. Because 
EPA generally interprets the ``interfere-with-maintenance'' test the 
same as the ``contributes-significantly-to-nonattainment'' test, for 
purposes of convenience, in this final rule, EPA sometimes refers to 
``contributes-significantly-to-nonattainment'' to refer to both tests.
    i. Dates. In today's action, EPA is determining that SIP 
submissions required under this rulemaking must be submitted by 
September 30, 1999 (see Section VI.A.1, Schedule for SIP Revision).
    Further, in today's action, EPA is requiring that SIP controls 
required today must be implemented by no later than May 1, 2003, and 
they must achieve reductions computed with reference to an overall 
budget amount determined as of September 30, 2007 (see Section V, 
NOX Control Implementation and Budget Achievement Dates).
    j. Downwind Areas' Control Obligations. Commenters have argued that 
under the CAA, downwind States must implement additional controls 
before EPA may require controls in upwind States. Commenters base this 
argument in part on the provisions of CAA section 107(a), which 
provides,

    Each State shall have the primary responsibility for assuring 
air quality within the entire geographic area comprising such State 
by submitting an implementation plan for such State which will 
specify the manner in which [NAAQS] will be achieved and maintained 
within each air quality control region in such State.

    Commenters further note that downwind States must implement 
additional reductions (beyond those specifically required by the CAA 
28) as needed to attain, under section 182(b)(1)(A)(i) and 
182(c)(2)(A). The commenters add that section 179(d)(2) is a generally 
applicable provision that limits the stringency of required controls to 
what is feasible. The commenters read these provisions together to 
conclude that downwind States must first implement all feasible control 
measures in an effort to reach attainment, and only after EPA 
determines that such States have done so but have not reached 
attainment may EPA require upwind contributors to implement controls. 
The commenters further observe that some of the downwind States in the 
Northeast have not implemented all feasible SIP measures.
---------------------------------------------------------------------------

    \28\ Reductions specifically required by the CAA include, for 
example, the 3 percent-per-year ROP reductions required of ozone 
nonattainment areas classified as serious or higher, under section 
182(c)(2)(B).
---------------------------------------------------------------------------

    The EPA disagrees with this legal analysis. The provision in 
section 107(a) that accords to States the primary responsibility for 
the air quality of their air basins, in essence provides the underlying 
rationale for the requirement of States to submit SIP revisions that 
meet CAA requirements. This phrase clarifies that the requirement of 
assuring attainment does not fall, in the first instance, on EPA. This 
provision does not have implications for apportioning responsibility 
between the downwind State and upwind States for contributions from 
upwind States. Downwind States would still carry the primary 
responsibility of assuring clean air even after the upwind contributors 
have revised their SIPs to meet the requirements of section 
110(a)(2)(D).
    Furthermore, EPA disagrees that section 179(d)(2) has any 
application to today's rulemaking. That provision in essence provides a 
general rule that if a nonattainment area fails to attain by its 
attainment date, EPA may require the State to implement reasonable 
controls that can be ``feasibly implemented.'' This requirement is not 
relevant to today's rulemaking, which addresses the requirements under 
section 110(a)(2)(D)(i)(I) that SIPs include provisions eliminating 
amounts of emissions from their sources that contribute significantly 
to downwind nonattainment.
    In addition, the requirement of downwind States to implement 
reductions beyond minimum CAA requirements if needed for attainment 
does not place the burden of implementing those reductions, in the 
first instance, on the downwind States. This requirement should be read 
to go hand-in-hand with the section 110(a)(2)(D) requirement that 
upwind States include SIP provisions that prohibit their sources from 
emitting air pollutants in amounts that ``significantly contribute'' to 
downwind nonattainment. In today's action, EPA is promulgating criteria 
for interpreting section 110(a)(2)(D) to take into account downwind 
attainment needs.
    As a practical matter, EPA has reviewed the status of Northeast 
States' efforts to comply with the requirements of the 1990 CAA 
Amendments and has found that these States have complied with the vast 
majority of the SIP submission requirements. Even so, EPA is well aware 
that some of the States have not made certain required submissions. 
29-30 However, EPA sees no basis in section 110(a)(2)(D) to 
mandate that downwind areas complete their SIP planning and 
implementation before upwind areas are required to begin that process. 
Upwind areas have been subject to the requirements of section 
110(a)(2)(D)--in some form--since the predecessor to this provision was 
added in the 1977 CAA Amendments. The EPA has determined, through air 
quality modeling, that even after the downwind States fulfill their 
prescribed CAA requirements, they will have areas expected to remain in 
nonattainment. Under these circumstances, the downwind areas continue 
to constitute areas with air quality in ``nonattainment'' under section 
110(a)(2)(D). As a result, upwind areas with emissions in amounts that 
``significantly contribute'' to the nonattainment air quality downwind 
are subject to control requirements whether or not the downwind areas 
they affect have met all of their planning obligations.
---------------------------------------------------------------------------

    \29-30\ If downwind areas fail to meet their planning 
obligations, they are subject to sanctions (See Section VI, below. 
As EPA noted in the NPR, 62 FR 60322-23, in some instances, States 
in the Northeast failed to submit all of their required SIP 
revisions or other commitments under Phase 1 of the March 2, 1995 
Memorandum and as a result, EPA initiated the sanctions process by 
starting sanctions clocks. In general, those States have since made 
the required Phase 1 submissions, and EPA terminated the sanctions 
process by stopping the clocks.
---------------------------------------------------------------------------

    k. Section 110(a)(2)(D) Caselaw. In the NPR, EPA noted that prior 
to the CAA Amendments of 1990, EPA had issued several rulemakings under 
section 110(a)(2)(E), the predecessor to section 110(a)(2)(D), and 
section 126 that addressed the issue of significant contribution in the 
context of pollutant transport. In those rulemakings, EPA generally 
applied a multi-factor test to determine whether the emissions from the 
sources in question constituted a signficant contribution to downwind 
jurisdictions. In each instance, EPA concluded that the emissions at 
issue from the upwind sources were not demonstrated to impact downwind 
air quality in a manner that would constitute significant contribution. 
Several of these determinations resulted in judicial challenges, but in 
each instance the courts upheld the Agency's determination of no 
significant contribution. The EPA indicated in the NPR that the prior 
rulemakings and the related court holdings, provide limited precedents 
for today's action. The EPA noted that these decisions have limited 
relevance because they involved different facts and circumstances, 
including different pollutants, different

[[Page 57381]]

upwind sources, and different downwind effects.
    Several commenters asserted that these prior rulemakings and cases 
are relevant to today's action, and compel EPA to conclude that the 
emissions from the upwind States affected by today's action do not 
contribute significantly to downwind nonattainment or maintenance 
problems. The EPA disagrees that these earlier determinations are 
controlling and that these earlier determinations are inconsistent with 
today's action. The EPA responds to these comments in detail in the 
Response to Comment document.

B. Alternative Interpretation of Section 110(a)(2)(D)

    As discussed above, in the NPR EPA advanced an alternative 
interpretation of section 110(a)(2)(D) (62 FR 60327). Under this 
alternative interpretation, EPA would determine the level of emissions 
that significantly contribute to nonattainment downwind based on 
factors relating to the entire amount of upwind emissions from a 
particular upwind State and their ambient impact downwind. The EPA 
would then determine what emissions reductions must be required to 
adequately mitigate that significant contribution based on factors 
relating to cost effectiveness of reductions and attainment needs 
downwind.
    The EPA continues to believe that this alternative interpretation 
remains a permissible interpretation of the statute for the reasons 
described in the NPR (62 FR 60327). In any event, it should be noted 
that for purposes of today's action, EPA finds no practical difference 
between the requirements that would result from the interpretation of 
section 110(a)(2)(D) adopted today and those that would result from the 
alternative interpretation described in the NPR. That is, even under 
the alternative interpretation, today's rulemaking would contain the 
same findings and require the same SIP revisions as under the 
interpretation adopted today (62 FR 60327).

C. Weight-of-Evidence Determination of Covered States

    As discussed above, EPA applied a multi-factor approach to identify 
the amounts of NOX emissions that contribute significantly 
to nonattainment. The EPA evaluated three air quality factors for each 
upwind jurisdiction (hereafter referred to as ``States'' or ``upwind 
States'') to determine whether each has emissions whose contributions 
to downwind nonattainment problems are large and/or frequent enough to 
be of concern. Further, for those States whose emissions are large and/
or frequent enough to be of concern, EPA applied highly cost-effective 
controls to determine the amount of NOx in upwind States 
which significantly contributes to nonattainment in, or interferes with 
maintenance by, a downwind State. The EPA also generally reviewed 
several other considerations before drawing final conclusions. Even 
though the actual finding of significant contribution applies only to 
the portion of a State's emissions for which EPA has identified highly 
cost-effective controls, for ease of discussion, the term 
``significant'' (or like term) is used in the discussion in this 
section to characterize the emissions of each upwind State that make a 
large and/or frequent contribution to nonattainment in downwind States 
sufficient to warrant eliminating a portion of its emissions equivalent 
to what can be removed through those controls.
    The purpose of this section is to describe the technical analyses 
performed by EPA to (a) quantify the air quality contributions from 
emissions in each upwind State on both 1-hour and 8-hour nonattainment, 
as well as 8-hour maintenance, in each downwind State, and (b) 
determine whether these contributions are significant.
    In the proposed weight-of-evidence approach, EPA specifically 
applied several factors to each upwind State, as discussed in Section 
II.A.3.c, Definition of Significant Contribution. These factors 
include:
     The overall nature of ozone problem (i.e., ``collective 
contribution'');
     The extent of the downwind nonattainment problems to which 
the upwind State's emissions are linked, including the ambient impact 
of controls required under the CAA or otherwise implemented in the 
downwind areas; and
     The ambient impact of the emissions from the upwind 
State's sources on the downwind nonattainment problems.
    As part of the analysis of these factors, EPA considered the 
findings from OTAG's technical analyses, as well as the findings from a 
number of other studies performed by OTAG participants independent of 
OTAG. The major findings from these analyses are described below. This 
is followed by an overview of the approach used by EPA in the proposal 
for considering the above factors to identify States that make a 
significant contribution to downwind nonattainment. The comments and 
EPA's response to comments on EPA's weight-of-evidence proposal are 
then discussed. Following that discussion, the results of additional 
State-by-State UAM-V modeling and State-by-State CAMX 
31 source apportionment modeling performed by EPA in 
response to comments are summarized.32 The EPA's analysis of 
the modeling results in terms of the significance of the contributions 
of upwind States to downwind nonattainment is presented in Section 
II.C.4, Confirmation of States Making a Significant Contribution to 
Downwind Nonattainment.
---------------------------------------------------------------------------

    \31\ Comprehensive Air Quality Model with Extensions.
    \32\ The UAM-V and CAMX models are described in the 
Air Quality Modeling TSD.
---------------------------------------------------------------------------

1. Major Findings From OTAG-Related Technical Analyses
    The major findings from the air quality and modeling analyses by 
OTAG and individual OTAG participants that are most relevant to today's 
rulemaking are as follows:
     several different scales of transport (i.e., intercity, 
intrastate, interstate, and inter-regional) are important to the 
formation of high ozone in many areas of the East;
     emissions reductions in a given multistate region/
subregion have the most effect on ozone in that same region/subregion;
     emissions reductions in a given multistate region/
subregion also affect ozone in downwind multistate regions/subregions;
     downwind ozone benefits decrease with distance from the 
source region/subregion (i.e., farther away, less effect);
     downwind ozone benefits increase as the size of the upwind 
area being controlled increases, indicating that there is a cumulative 
benefit to extending controls over a larger area;
     downwind ozone benefits increase as upwind emissions 
reductions increase (the larger the upwind reduction, the greater the 
downwind benefits);
     a regional strategy focusing on NOX reductions 
across a broad portion of the region will help mitigate the ozone 
problem in many areas of the East;
     both elevated and low-level NOX reductions 
decrease ozone concentrations regionwide;
     there are ozone benefits across the range of controls 
considered by OTAG; the greatest benefits occur with the most emissions 
reductions; there was no ``bright line'' beyond which the benefits of 
emissions reductions diminish significantly;
     even with the large ozone reductions that would occur if 
the most

[[Page 57382]]

stringent controls considered by OTAG were implemented, there may still 
remain high concentrations in some portions of the OTAG region; and a 
regional NOX emissions reduction strategy coupled with local 
NOX and/or VOC reductions may be needed to enable attainment 
and maintenance of the NAAQS in this region.
    The above findings provide technical evidence that transport within 
portions of the OTAG region results in large contributions from upwind 
States to ozone in downwind areas, and that a regionwide approach to 
reduce NOX emissions is an effective way to address these 
interstate contributions.
2. Summary of Notice of Proposed Rulemaking Weight-of-Evidence Approach
    The EPA relied on OTAG data to develop the information necessary to 
evaluate the weight-of-evidence factors identified above. These data 
include emissions (tons) and emission density (tons per square mile), 
air quality analyses, trajectory, wind vector, and ``ozone cloud'' 
analyses, and subregional zero-out modeling. In brief, EPA's proposed 
approach was as follows:
     the OTAG transport distance scale was applied to identify, 
based on the meteorological potential for transport, which States may 
contribute to ozone in downwind States;
     the results of the OTAG subregional modeling runs 
(described below) were used to quantify the extent to which each 
subregion contributes to downwind nonattainment for the 1-hour and/or 
8-hour NAAQS;
     the OTAG 2007 Base Case NOX emissions and 
emissions density were used to identify States which emit large amounts 
of NOX and/or have a high density of NOX 
emissions compared to other States in the OTAG region and, therefore, 
have NOX emissions which may be great enough to contribute 
to downwind nonattainment; and the OTAG 2007 Base Case NOX 
emissions were also used to translate the findings from the subregional 
modeling to a State-by-State basis.
    a. Quantification of Contributions. As part of OTAG's assessment of 
transport, a series of model runs were performed to examine the impacts 
of emissions from each of 12 multistate subregions on ozone in downwind 
areas. The locations of these subregions are shown in Figure II-1.

[GRAPHIC] [TIFF OMITTED] TR27OC98.000

    In each subregional model run, all manmade emissions were removed 
from one upwind subregion and the model was run for the OTAG July 1988 
and 1995 episodes. The ``parts per billion (ppb)'' differences in ozone 
between each subregional zero-out run compared to the corresponding 
2007 Base Case run were used to quantify the air quality impacts of the 
subregion on nonattainment downwind.
    In the proposed NOX SIP call, EPA considered areas as 
``nonattainment'' if air quality monitoring indicates that the area is 
currently measuring nonattainment and if air quality modeling indicates 
future nonattainment, taking into account CAA control requirements and 
growth. In this regard, areas were considered nonattainment for the 1-
hour NAAQS if

[[Page 57383]]

they had 1994-1996 \33\ monitoring data indicating measured 1-hour 
violations and 2007 Base Case 1-hour predictions >=125 ppb. Areas were 
considered to be nonattainment for the 8-hour NAAQS if they had 1994-
1996 monitoring data indicating measured 8-hour violations and 2007 
Base Case 8-hour predictions >=85 ppb. The inconsistency between the 
form of the 8-hour NAAQS, which considers 3 years of data for 
determining the average of the fourth-highest 8-hour daily maximum 
concentration at a monitor, and the limited predictions available from 
the OTAG episodes introduced a complication to the analysis of 8-hour 
contributions. It was not possible to use the model predictions in a 
way that explicitly matched the form of the 8-hour NAAQS. Instead, an 
analysis of seasonal and episodic ozone measurements was performed in 
an attempt to link 8-hour measured concentrations during the OTAG 
episodes to the form of the 8-hour NAAQS, as closely as possible. The 
results of that analysis indicated that the 3-episode average of the 
second highest 8-hour ozone concentrations measured during the OTAG 
1991, 1993, and 1995 episodes corresponded best, overall, to the 3-year 
average of the fourth highest 8-hour daily ambient data. However, since 
OTAG subregional modeling was only available for the 1988 and 1995 
episodes, EPA used the concentrations during these two episodes in 
calculating average second high 8-hour concentrations.\34\
---------------------------------------------------------------------------

    \33\ Data for 1994-1996 were used because these were the most 
recent quality-assured data available at the time the analysis was 
performed.
    \34\ In response to comments, EPA has reexamined the method for 
relating 8-hour model predictions during the OTAG episodes to the 
form of the 8-hour NAAQS. This is discussed further in Section 
II.C.2.c, Comments and Responses on the Proposed Weight of Evidence 
Approach to Significant Contribution.
---------------------------------------------------------------------------

    b. Evaluation of 1-Hour and 8-Hour Contributions. In the proposal, 
EPA summarized the ``ppb'' contributions to downwind nonattainment from 
each subregion in terms of both the frequency and the magnitude of the 
downwind impacts over specific concentration ranges (e.g., 2 to 5 ppb, 
5 to 10 ppb, 10 to 15 ppb, etc.). The results indicate that, in 
general, large contributions to downwind nonattainment occur on 
numerous occasions. Although the level of downwind contribution varies 
from subregion to subregion, a consistent pattern is apparent for both 
1-hour nonattainment and 8-hour nonattainment. Specifically, the 
results of the subregional modeling indicate that emissions from States 
in subregions 1 through 9 produce large 1-hour and 8-hour contributions 
downwind in terms of the magnitude and frequency, including geographic 
extent, of the downwind impacts. In addition, nonattainment areas 
within many States in the OTAG region receive large and/or frequent 
contributions from emissions in these subregions. The EPA proposed to 
find that most of the States whose emissions are wholly or partially 
contained within one or more of these subregions (i.e., Alabama, 
Connecticut, Delaware, Georgia, Illinois, Indiana, Kentucky, Maryland, 
Michigan, Missouri, New Jersey, New York, North Carolina, Ohio, 
Pennsylvania, South Carolina, Tennessee, Virginia, West Virginia, and 
Wisconsin, as well as the District of Columbia) are making a 
significant contribution to downwind nonattainment. In addition to the 
ambient impact demonstrated by the subregional modeling, this proposed 
finding was based on a determination that:
     OTAG strategy modeling and non-OTAG modeling indicate that 
NOX emissions reductions across these States would produce 
large reductions in 1-hour and 8-hour ozone concentrations across broad 
portions of the region including 1-hour and 8-hour nonattainment areas;
     these States are upwind from nonattainment areas within 
the 1- to 2-day distance scale of transport;
     these States form a contiguous area of manmade emissions 
covering most of the core portion of the OTAG region;
     11 of the States that are wholly within subregions 1 
through 9 have a relatively high level of NOX emissions from 
sources in their States; these States are ranked in the top 50 percent 
of all States in the region in terms of total NOX emissions 
and/or have NOX emissions exceeding 1000 tons per day;
     States wholly within subregions 1 through 9 with lesser 
emissions have a relatively high density of NOX emissions;
     for the seven States that are only partially contained in 
one of subregions 1 through 9, the State total NOX 
emissions, as well as each State's contribution to NOX 
emissions in the subregions in which they are located, indicate that 
six of the States each have: NOX emissions that are more 
than 10 percent of the total NOX emissions in one of these 
subregions, NOX emissions in the top 50 percent among all 
States, and/or a majority of its NOX emissions within one of 
these subregions.
    For the New England States that were not included in any of the 
OTAG zero-out subregions, EPA found that two of these States (i.e., 
Massachusetts and Rhode Island) have a high density of NOX 
emissions. Also, the trajectory and wind vector analyses indicated that 
these States are immediately upwind of nonattainment areas in other 
States.
    For the nine States in the OTAG region which are wholly within 
subregions 10, 11, and 12 (i.e., Florida, Kansas, Louisiana, Minnesota, 
Nebraska, North Dakota, Oklahoma, South Dakota, and Texas), and for 
Arkansas, Iowa, and Mississippi, EPA proposed that emissions from each 
of these States should be considered not to significantly contribute to 
downwind nonattainment. These States are further discussed below in 
Section II.C.5, States Not Covered by this Rulemaking.
    c. Comments and Responses on Proposed Weight-of-Evidence Approach 
to Significant Contribution. The EPA received a number of comments on 
various elements of the proposed weight-of-evidence approach. In 
addition, EPA received new modeling and analyses performed by 
commenters which address the issue of significant contribution. The 
following is a summary of the major comments received by EPA and the 
responses to these comments. Additional comments and EPA's response to 
these comments are provided in the Response to Comment document.
    Comment: Some commenters stated that it was inappropriate to use a 
weight-of-evidence approach to determine the significance of upwind 
emissions on downwind nonattainment. Rather, it was argued that EPA 
should use a specific ``bright line'' criterion. Other commenters 
supported the weight-of-evidence approach.
    Response: The magnitude and frequency of contributions from an 
upwind State to downwind nonattainment depend on the extent of the 
nonattainment problem in the downwind area, the emissions in the 
downwind area, the emissions in the upwind State, the distance between 
the upwind State and the downwind area, and weather conditions (i.e., 
winds and temperatures which favor ozone formation and transport). 
Because these factors vary in a complex way across the OTAG region, it 
is not possible to develop a single bright line test for significance 
that will be applicable and appropriate for all potential upwind-State-
to-downwind-area linkages. Therefore, EPA believes that it is more 
appropriate to use a weight-of-evidence approach to account for all of 
these factors than establishing a bright line criterion.
    Comment: Some commented that EPA should not use the trajectory, 
wind vector, and ``ozone cloud'' analyses as a

[[Page 57384]]

basis for determining significant contribution because these techniques 
indicate air movement and do not account for ozone formation and 
depletion due to photochemical reactions and other processes. Other 
commenters argued in favor of using this information as means of 
linking upwind States with downwind nonattainment.
    Response: The EPA agrees that information from such techniques 
should not be used as the sole basis for finding that certain upwind 
States significantly contribute to nonattainment in specific downwind 
States. However, EPA believes that it is important to consider the 
``movement'' of ozone and/or precursors as part of the air quality 
evaluation of contributions from upwind States. This factor is 
incorporated into the air quality models used by EPA for this 
rulemaking. The inclusion of this information, in conjunction with 
numerous other air quality factors in the models, provides for a more 
technically robust analysis than can be provided by the trajectory, 
ozone cloud, and wind vector analyses alone.
    Comment: A number of commenters stated that CAA section 
110(a)(2)(D) requires a State-by-State demonstration that emissions 
within an upwind State make a significant contribution to nonattainment 
in another State and thus, EPA's proposed approach of using subregional 
(i.e., multistate) modeling, together with each upwind State's 
NOX emissions, to establish these linkages is legally 
flawed. These commenters argued that section 110(a)(2)(D) requires 
``each implementation plan submitted by a State'' to contain provisions 
that prohibit any source or other type of emissions activity ``within 
the State'' from emitting air pollutants in amounts that contribute 
significantly to a downwind nonattainment problem. The commenters 
concluded that these provisions require, as a matter of technical 
procedure, that EPA must base its determination that emissions from a 
particular State significantly contribute to nonattainment downwind on 
a technical analysis of that particular State's emissions. According to 
the commenters, section 110(a)(2)(D) by its terms, prohibits EPA from 
making that technical determination by examining the impact of 
emissions from a group of States on a downwind nonattainment problem, 
and then extrapolating from that information to determine whether 
emissions from each State within that group should be considered to 
make a significant contribution.
    As a technical matter, these commenters argue that if emissions 
from more than one State are lumped together in assessing the 
contribution to a downwind State, there is no way to determine the 
amount of emissions in each contributing State that must be reduced. 
The commenters argue that the only way to establish specific upwind 
State to downwind State linkages is through air quality modeling on a 
State-by-State basis. Further, the commenters contend that once an area 
beyond a particular State's boundaries is modeled, there is no way of 
knowing how much farther upwind to go in terms of defining a source 
area. In order to address these issues, many commenters stated that EPA 
must do State-by-State zero-out UAM-V modeling and/or State-by-State 
source apportionment modeling using the CAMx model to determine 
downwind contributions from upwind States.
    Response: On the legal issue, EPA disagrees that the above-
referenced provisions of section 110(a)(2)(D), by their terms, mandate 
the technical procedure for EPA to make the determination of 
significant contribution. These provisions simply indicate that EPA 
must make that determination on a SIP-by-SIP basis, that is, for EPA to 
issue a SIP call with respect to a particular State, EPA must determine 
that the provisions of that SIP fail to adequately control emissions 
from sources within the State. However, these provisions do not mandate 
any particular technical procedure for making that determination. As a 
result, EPA may employ any technical procedure that is sufficiently 
accurate. As discussed below, EPA believes that its subregional 
approach is sufficiently accurate to justify the SIP call. However, in 
response to this and other comments, EPA did conduct State-by-State 
modeling. The results of this modeling, as discussed below, confirm the 
results of the subregional modeling.
    On the technical issue, EPA used the subregional modeling as part 
of the proposed approach because OTAG had developed and relied on this 
modeling as part of its analysis to quantify the impacts of manmade 
emissions in upwind areas on ozone in downwind areas. In addition, in 
conjunction with other information, EPA believes that it is possible to 
make rational extrapolations from the subregional results in order to 
draw conclusions as to the contribution of individual States. The EPA 
believes that it is credible to use NOx emissions in each 
State, along with the subregional modeling results, in the 
determination of significance in view of the results of OTAG modeling 
which indicate that, in addition to local emissions, the level of ozone 
in a downwind State is directly related to the magnitude of 
NOx emissions in upwind areas and the proximity of the 
upwind area to the downwind State. A more detailed discussion of the 
technical validity of the subregional modeling is contained in the 
Response to Comment Document.
    The EPA recognizes that State-by-State modeling would provide some 
additional precision to the magnitude and frequency of individual 
State-to-State contributions. In response to the recommendations for 
additional modeling, EPA performed both State-by-State UAM-V zero-out 
modeling and State-by-State CAMx source apportionment modeling for many 
of the upwind States in the OTAG region which were proposed as 
significant contributors. The EPA's analysis of the contributions to 
downwind nonattainment using the State-by-State modeling confirms the 
overall finding, based on the proposed subregional modeling, that the 
23 jurisdictions identified in the proposal significantly contribute to 
nonattainment in downwind States. Specifically, the subregional 
modeling indicates that manmade emissions from sources in subregions 1 
through 9 make large and/or frequent contributions to 1-hour and 8-hour 
nonattainment in specific downwind States. The EPA's analysis of the 
State-by-State modeling demonstrates that each of the 23 upwind 
jurisdictions identified through subregional modeling significantly 
contribute to nonattainment in specific downwind States. In addition, 
the results of the State-by-State modeling show that the specific 
upwind-State-to-downwind-nonattainment linkages indicated by the 
subregional modeling are confirmed overall by the State-by-State 
modeling. The State-by-State modeling analyses are summarized below and 
more fully documented in the Air Quality Modeling TSD.
    Comment: The EPA received comments that zero-out modeling 
introduces sharp spatial changes in emissions and pollutants along the 
edges of the zero-out area. The commenters contend that this is not 
credible and provides an incorrect assessment of transport.
    Response: The EPA disagrees with this comment, as discussed in the 
Response to Comments document. Also, as indicated above, in response to 
other comments, EPA has performed CAMx source apportionment modeling 
which does not use a zero-out technique for quantifying ozone 
contributions from upwind States. In general, EPA has found that the 
source apportionment technique and zero-out modeling

[[Page 57385]]

provide consistent information on the relative contribution of upwind 
States to downwind nonattainment. In cases where the two techniques do 
not provide consistent results, the source apportionment technique 
tends to indicate larger contributions than the zero-out modeling. The 
differences between these two modeling techniques are described further 
in the Air Quality Modeling TSD.
    Comment: Some comments referenced a study which analyzed the 
``noise'' (i.e., uncertainty) in the UAM-V modeling system. This study 
purports to show that the contributions from some States EPA proposed 
as significant are within the ``noise'' of the model.
    Response: This study focuses on model uncertainty by varying many, 
but not all, inputs to the model. The study does not contend that the 
inputs selected by OTAG are incorrect, but rather that there may be 
other plausible values for these inputs. The results indicate that 
there is a range of uncertainty in predicted ozone associated with the 
range of possible values for the particular inputs studied by the 
commenter. The study does not indicate that there is any bias in the 
model's predictions (i.e., there is no indication that the predictions 
are too high or too low). The specific values for the inputs being used 
by EPA in its air quality modeling are the same values that were used 
by OTAG. These values were selected by the OTAG Regional and Urban 
Scale Modeling Work Group, which included experts in air quality 
modeling from the public and private sector, in conjunction with the 
model's developers, Systems Application International. The predictions 
from OTAG's model runs using these same input values were evaluated 
against ambient measurements and found by OTAG to provide acceptable 
results. The EPA continues to believe that the specific inputs selected 
by OTAG are technically sound and the modeling results are credible. A 
further discussion of EPA's response to this comment is in the Response 
to Comments document.
    Comment: Several commenters stated that emissions from large point 
sources of NOx in specific States do not contribute 
significantly to downwind nonattainment.
    Response: As discussed in Section II.A.3.c, Definition of 
Significant Contribution, under EPA's collective contribution approach, 
if emissions in the aggregate from a particular geographic region or 
State are found to contribute significantly to nonattainment downwind, 
then the emissions in that region or State are considered to be 
significant contributors to that nonattainment problem. Moreover, EPA 
treats emissions as ``contributing significantly'' only to the extent 
they may be eliminated through highly cost-effective reductions. Thus, 
if all emissions from a State, when considered in the aggregate, are 
found to contribute significantly to nonattainment downwind, and if 
there are highly cost-effective controls for NOx emissions 
from sources in the upwind State, then the amount of NOx 
emissions from these sources that can be eliminated with such controls 
are considered to be making a significant contribution. The amount of 
emissions determined through this approach to make a significant 
contribution may be relatively small, compared to the upwind State's 
entire inventory; and the ambient impact downwind of eliminating that 
amount may be relatively small as well. However, this small impact does 
not mean that the emissions themselves are not significant insofar as 
their contribution to nonattainment downwind. Further, as discussed in 
Section IV, Air Quality Assessment, when the amount of emissions 
required to be eliminated from upwind States are combined and modeled 
collectively, their ambient impact downwind is larger.
    Comment: One commenter provided a recommendation for dealing with 
the concern that the spatial resolution of meteorological inputs to the 
air quality model may be too coarse to require that predicted 
exceedences correspond exactly with a county violating the NAAQS. The 
commenter's recommendations were to base the selection of 1-hour 
nonattainment receptors on model predicted exceedences in either (a) 
all counties within the metropolitan statistical area containing the 
nonattainment area or (b) all counties comprising the designated 1-hour 
nonattainment area.
    Response: The EPA believes that the appropriate way to address this 
issue is to use all counties comprising the designated 1-hour 
nonattainment area. That is, all counties in a designated 1-hour 
nonattainment area should be considered as possible nonattainment 
receptors for the purposes of evaluating contributions to nonattainment 
under the 1-hour NAAQS. The EPA recognizes that not all counties within 
a designated nonattainment area have monitors, and that some counties 
may have monitors that indicate attainment in that county. Even so, EPA 
recognizes that under the 1-hour NAAQS, nonattainment boundaries are 
generally used to describe an area with the nonattainment problem. 
Thus, EPA believes that this geographic vicinity offers the best 
indication of an area that may be expected to have nonattainment air 
quality somewhere within its boundaries. The EPA believes that it is 
appropriate to include all counties in the designated nonattainment 
area because the entire nonattainment area is responsible for meeting 
the 1-hour NAAQS, even if only one monitor measures nonattainment at 
any one time. As noted elsewhere, EPA predicts that many 1-hour 
nonattainment areas that currently monitor nonattainment somewhere 
within the area will remain in nonattainment in 2007, in some cases 
because of predicted violations in counties that currently monitor 
attainment. The EPA believes that the entire area should be considered 
to be in nonattainment until all monitors in the area indicate 
attainment of the NAAQS. Thus, in today's rulemaking, EPA used the 
designated 1-hour nonattainment area in selecting the receptors to be 
used to evaluate impacts on downwind nonattainment problems.
    Comment: Several commenters questioned the validity of EPA's 
approach of using the 3-episode average of the second highest 8-hour 
daily maximum concentration to represent the form of the 8-hour NAAQS 
(i.e., the 3-year average of the fourth highest 8-hour daily maximum 
values at a monitor 35). Commenters expressed the concern 
that the average second high may not be representative for all areas 
across the OTAG domain. However, none of the commenters provided any 
suggested alternatives to EPA's approach.
---------------------------------------------------------------------------

    \35\ For the purposes of discussion in this Section, these 
values are referred to as ``design'' values.
---------------------------------------------------------------------------

    Response: The analysis performed by EPA to establish a relationship 
between the air quality during the OTAG episodes and the form of the 8-
hour NAAQS was based upon an analysis of 3 years of monitoring data 
compared to monitoring data during the OTAG episodes. In response to 
comments, EPA performed an analysis to determine how the predicted 
average second high 8-hour values, as well as several alternative 8-
hour values, compared to ambient 8-hour design values, based on 1994 to 
1996 measured data. Based on this analysis, EPA determined that, 
overall, the model-predicted average second high values underestimate 
the corresponding ambient design values for those counties in the OTAG 
domain with 1994-1996 ambient values >=85 ppb. In addition to the 
average second high, EPA also compared six other measures of 8-hour 
model predictions to ambient design values. The six other measures 
include the highest, second

[[Page 57386]]

highest, third highest, and fourth highest ozone predictions across the 
July 1991, 1993, and 1995 episodes; the 3-episode average of the 
highest concentrations; and the 3-episode average of the highest, 
second highest, and third highest concentrations. The EPA also 
developed the same measures using model predictions from all 4 episodes 
for comparison to the ambient design values. The results indicate that 
none of the alternative measures provides a universal best match to 
ambient 8-hour design values in all States. Each of the indicators 
overestimates values in some areas and underestimates values in other 
areas to a varying extent. Furthermore, the best representation of 8-
hour design values using predictions from the OTAG episodes varies from 
State to State. Given that the predicted average second high 
underestimates ambient 8-hour design values and that none of the other 
8-hour indicators examined by EPA provides a ``best'' match to ambient 
values in all cases, EPA has decided to analyze the contributions to 8-
hour nonattainment problems using all 8-hour predictions >=85 ppb. The 
EPA believes that this approach is appropriate given that EPA is using 
modeling results for the 8-hour NAAQS merely as an indicator of the 
likelihood that areas that currently monitor violations of the 8-hour 
NAAQS will continue to be nonattainment for the 8-hour NAAQS and/or 
have 8-hour maintenance problems in 2007.36 Thus, the air 
quality analysis of 8-hour contributions, described below, focuses on 
all 8-hour values >=85 ppb.
---------------------------------------------------------------------------

    \36\ Similarly, the EPA is also using 1-hour model predictions 
>=125 ppb as an indicator that areas currently designated 
nonattainment for the 1-hour NAAQS will continue to be nonattainment 
for the 1-hour NAAQS in 2007.
---------------------------------------------------------------------------

    Comment: Several commenters submitted new State-by-State zero-out 
modeling using UAM-V and CAMx source apportionment modeling 
purporting to show that contributions from particular upwind States are 
insignificant.
    Response: The EPA reviewed the commenters' modeling to determine 
and assess (a) the technical aspects of the models that were applied; 
(b) the types of episodes modeled; (c) the methods for aggregating, 
analyzing, and presenting the results; (d) the completeness and 
applicability of the information provided; and (e) whether the 
technical evidence supports the arguments made by the commenters. 
Overall, the modeling submitted by commenters is viewed by EPA as 
generally technically credible, although not complete in all cases. The 
EPA's ability to fully evaluate and utilize the modeling submitted by 
commenters was hampered in some cases because only limited information 
on the results was provided. For example, a commenter may have provided 
results for only 1 or 2 days in an episode, or for only one of several 
episodes with no information presented on the results for the remaining 
days or episodes that were modeled. As another example, results were 
presented for only the peak ozone day in an episode while greater 
contributions may have been predicted on other high ozone days of the 
episode. For some of the modeling, the information was only presented 
in graphical form which made the results difficult to evaluate in a 
quantitative way. Also, in some cases the model predictions were only 
presented as episode composite values without information on peak 
contributions. The EPA's full assessment of the modeling submitted by 
commenters is provided in the Response to Comments document.
    In light of the absence of complete information in the modeling 
provided by commenters and other comments calling for State-by-State 
analyses, EPA decided to perform additional air quality modeling of the 
type submitted by commenters in order to consider all of the data 
resulting from such model runs. The EPA modeling includes State-by-
State zero-out modeling using UAM-V and State-by-State CAMx 
source apportionment modeling.
    EPA conducted further analysis of other factors included in the 
multi-factor approach for significant contribution. The results of 
EPA's consideration of these factors and EPA's modeling are described 
next.
3. Analysis of State-specific Air Quality Factors
    a. Overall Nature of Ozone Problem (``Collective Contribution''). 
As described above, EPA believes that each ozone nonattainment problem 
at issue in today's rulemaking is the result of emissions from numerous 
sources over a broad geographic area. The contribution from sources in 
an upwind State must be evaluated in this context. This ``collective 
contribution'' nature of the ozone problem supports the proposition 
that the solution to the problem lies in a range of controls covering 
sources in a broad area, including upwind sources that cause a 
substantial portion of the ozone problem. This upwind share is 
typically caused by NOx emissions from sources in numerous States. 
States adjacent to the State with the nonattainment problem generally 
make the largest contribution, but States further upwind, collectively, 
make a contribution that constitutes a large percentage in the context 
of the overall problem. As an example to illustrate the overall nature 
of the ozone problem, EPA discusses below the ozone problem in the New 
York City nonattainment area.
    b. Extent of Downwind Nonattainment Problems. For each downwind 
area to which an upwind State may be linked, EPA also examined the 
extent of the downwind nonattainment problem, including the air quality 
impacts of controls required in downwind areas under the CAA, as well 
as of controls required or implemented on a national basis. As 
indicated elsewhere, EPA determined that a downwind area should be 
considered ``nonattainment'' for purposes of section 110(a)(2)(D)(i)(I) 
under the 1-hour NAAQS if the area currently (as of the 1994-96 time 
period) has nonattainment air quality 37 and if the area is 
modeled to have nonattainment air quality in the year 2007, after 
implementation of all measures specifically required of the area under 
the CAA as well as implementation of Federal measures required or 
expected to be implemented by that date. The EPA determined that each 
such downwind area had a residual nonattainment problem even after 
implementation of all these control measures. The presence of residual 
nonattainment is a factor that supports the need to reduce emissions 
from upwind sources to allow further progress towards 
attainment.38 As an example, the residual nonattainment for 
the New York City area is discussed in more detail below.
---------------------------------------------------------------------------

    \37\ As explained elsewhere, for the 1-hour standard, EPA based 
its determination as to the boundaries of the area with air quality 
violating the NAAQS on the boundaries of the area designated as 
nonattainment.
    \38\ Indeed, the modeling relied on in today's action indicates 
that many downwind nonattainment areas carry a residual 
nonattainment problem even after implementation of regional 
reductions by all the States affected by today's action. Although 
not essential to EPA's conclusions, the presence of this 
nonattainment problem even after implementation of regional 
controls, based on the modeling used in today's rulemaking, 
indicates that even further reductions, regionally or locally, would 
be needed to assure attainment in those downwind areas.

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

[[Page 57387]]

    c. Air Quality Impacts of Upwind Emissions on Downwind 
Nonattainment. As indicated above, in response to comments, additional 
air quality modeling was performed by EPA to confirm the proposed 
approach which relied on subregional modeling to quantify the impacts 
of emissions from upwind States on nonattainment in downwind areas. The 
additional modeling consisted of State-by-State zero-out modeling using 
UAM-V and State-by-State source apportionment modeling using the CAMx 
Anthropogenic Precursor Culpability Assessment (APCA) 
technique.39 A description of these models is contained in 
the Air Quality Modeling TSD. Both models are currently being used by 
the scientific and regulatory community for air quality assessments. 
The EPA is not aware of any information that would indicate that either 
model provides more credible predictions than the other. Each modeling 
technique (i.e., zero-out and source apportionment) provides a 
different technical approach to quantifying the downwind impact of 
emissions in upwind States. The zero-out modeling analysis provides an 
estimate of downwind impacts by comparing the model predictions from a 
Base Case run to the predictions from a run in which the Base Case 
manmade emissions are removed from a specific State. In contrast, the 
source apportionment modeling quantifies downwind impacts by tracking 
formation, chemical transformation, depletion, and transport of ozone 
formed from emissions in an upwind source area and the impacts that 
ozone has on nonattainment in downwind areas. The EPA ran both models 
for all four OTAG episodes (i.e., July 1-11, 1988; July 13-21, 1991; 
July 20-30, 1993; and July 7-18, 1995) using the 2007 SIP Call Base 
Case emissions. The development of emissions for this Base Case 
scenario are described in Section IV, Air Quality Assessment.
---------------------------------------------------------------------------

    \39\ For ease of discussion, EPA is using the term ``UAM-V'' to 
refer to the UAM-V State-by-State zero-out modeling and the term 
``CAMx'' to refer to the CAMx source apportionment modeling.
---------------------------------------------------------------------------

    The EPA selected several metrics in order to evaluate the downwind 
contributions from emissions in upwind States. The metrics were 
designed to provide information on the three fundamental factors for 
evaluating whether emissions in an upwind State make large and/or 
frequent contributions to downwind nonattainment. These factors are (a) 
the magnitude of the contribution, (b) the frequency of the 
contribution, and (c) the relative amount of the contribution. The 
magnitude of contribution factor refers to the actual amount of 
``ppbs'' of ozone contributed by emissions in the upwind State to 
nonattainment in the downwind area. The frequency of the contribution 
refers to how often the contributions occur and how extensive the 
contributions are in terms of the number of grids in the downwind area 
that are affected by emissions in the upwind State. The relative amount 
of the contribution is used to compare the total ``ppb'' contributed by 
the upwind State to the total ``ppb'' of nonattainment in the downwind 
area.
    As indicated above, two modeling techniques (i.e., UAM-V zero-out 
and CAMx source apportionment) were used for the State-by-State 
evaluation of contributions. The EPA developed metrics for both 
modeling techniques for each of the three factors. However, because of 
the differences between the two techniques, some of the metrics used 
for the UAM-V modeling and the CAMx modeling are different. The 
specific UAM-V and CAMx metrics and how they relate to the three 
factors used for the evaluation of contributions are described below.
    The EPA examined the contributions from upwind States to downwind 
nonattainment for several types of nonattainment receptors. 
Nonattainment receptors for the 1-hour analysis include those grid 
cells that (a) are associated with counties designated as nonattainment 
for the 1-hour NAAQS and (b) have 1-hour Base Case model predictions 
>=125 ppb. These grid cells are referred to as ``designated plus 
modeled'' nonattainment receptors. Using these receptors, the metrics 
were calculated for each 1-hour nonattainment area as well as for each 
State. To calculate the metrics by State, all of the 1-hour 
nonattainment receptors in that State were pooled 
together.40 Table II-1 lists the 1-hour nonattainment areas 
that were considered in this analysis, along with the State(s) in which 
the nonattainment area is located. In addition to the areas listed in 
Table II-1, EPA also evaluated the contributions of upwind States to 
ozone concentrations over Lake Michigan because modeled air quality 
over the lake can be indicative, under certain weather conditions, of 
air quality in portions of the States surrounding the 
lake.41
---------------------------------------------------------------------------

    \40\ For ease of discussion in this Section, the 1-hour 
nonattainment areas and the set of nonattainment receptors pooled 
over an entire State are referred to as downwind areas.
    \41\ High measured ozone concentrations in portions of Illinois, 
Indiana, Michigan, and Wisconsin near the shoreline of Lake Michigan 
are often associated with weather conditions which cause ozone 
precursor pollutants to be blown offshore over the lake during the 
morning, where they can form high ozone concentrations which then 
return onshore during ``lake breeze'' wind flows in the afternoon. 
Because the size of the grid cells used in the OTAG modeling is 
relatively large compared to the spatial scale of the lake breeze, 
the high ozone concentrations predicted over the lake may not be 
blown back onshore in the model. Since high concentrations over the 
lake do, in reality, impact air quality along the shoreline of one 
or more of these States, the EPA believes that it is appropriate to 
use predicted contributions to ozone over Lake Michigan as a 
surrogate for contributions to any one of the surrounding States 
(i.e., Illinois, Indiana, Michigan, and Wisconsin).

            Table II-1.--1-Hour Nonattainment Areas Evaluated
------------------------------------------------------------------------
      Nonattainment area                        State(s)
------------------------------------------------------------------------
Atlanta......................  Georgia.
Baltimore....................  Maryland.
Birmingham...................  Alabama.
Boston/Portsmouth 1..........  Massachusetts, New Hampshire.
Chicago/Milwaukee 2..........  Illinois, Indiana, Wisconsin.
Cincinnati...................  Kentucky, Ohio.
Greater Connecticut..........  Connecticut.
Louisville...................  Indiana, Kentucky.
Memphis......................  Mississippi, Tennessee.
New York City................  Connecticut, New Jersey, New York.
Philadelphia.................  Delaware, Maryland, New Jersey,
                                Pennsylvania.
Pittsburgh...................  Pennsylvania.
Portland.....................  Maine.
Rhode Island.................  Rhode Island.
Southwestern Michigan 3......  Michigan.

[[Page 57388]]

St. Louis....................  Illinois, Missouri.
Washington, DC...............  District of Columbia, Maryland, Virginia.
Western Massachusetts........  Massachusetts.
------------------------------------------------------------------------
\1\ For the purposes of this analysis EPA has combined the Greater
  Boston nonattainment area which includes portions of Massachusetts and
  New Hampshire, with the Portsmouth, New Hampshire nonattainment area
  into a single downwind nonattainment receptor area.
\2\ For the purposes of this analysis EPA has combined the 1-hour
  nonattainment counties that are along the shoreline of Lake Michigan
  in the States of Illinois, Indiana, and Wisconsin into a single
  downwind nonattainment receptor area.
\3\ For the purposes of this analysis EPA has combined the 1-hour
  nonattainment counties that are along the shoreline of Lake Michigan
  in the State of Michigan into a single downwind nonattainment receptor
  area.

    For the 8-hour analysis, nonattainment receptors are those grid 
cells that (a) are associated with counties currently violating the 8-
hour NAAQS (based on 1994-1996 data) and (b) have 8-hour Base Case 
model predictions >=85 ppb. These grid cells are referred to as 
``violating plus modeled'' nonattainment receptors. The metrics for the 
8-hour contribution analyses were calculated on a State-by-State basis 
by pooling together the ``violating plus modeled'' receptors in a 
State.
    (1) UAM-V State-by-State Modeling. In the UAM-V zero-out model runs 
all manmade emissions in a given upwind State were removed from the 
Base Case scenario. Each zero-out scenario was run for all 4 episodes 
and the ozone predictions in downwind States were then compared to 
those from the Base Case run in order to quantify the downwind impacts 
of emissions from the upwind State (i.e., the State in which the 
manmade emissions were removed). The EPA performed zero-out runs for 
the following set of States:
     Alabama, Georgia, Illinois, Indiana, Kentucky, 
Massachusetts, Michigan, Missouri, North Carolina, Ohio, South 
Carolina, Tennessee, Virginia, West Virginia, and Wisconsin.
    Zero-out modeling for Massachusetts was performed because this 
State was the only State in the Northeast with relatively large 
NOX emissions that was not included in any of the OTAG 
subregional modeling. The other States listed above were selected for 
zero-out modeling in order to respond to comments that emissions in all 
or portions of each of these States do not contribute significantly to 
downwind nonattainment.
    The EPA analyzed the model-predicted ozone concentrations from the 
zero-out runs using the four metrics described below. The results for 
these metrics are too voluminous to include in the notice in their 
entirety. The full set of results is contained in the Air Quality 
Modeling TSD. Each metric was calculated using 1-hour daily maximum 
concentrations >=125 ppb as well as 8-hour daily maximum concentrations 
>=85 ppb. Model predictions from all 4 episodes were used for 
calculating the metrics.42
---------------------------------------------------------------------------

    \42\ Model predictions from the first few days of each episode 
are considered ``ramp-up'' days and were excluded from the analysis, 
following the procedures adopted by OTAG. The ramp-up days include 
the first 3 days of the July 1988, 1991, and 1995 episodes and the 
first 2 days of the July 1993 episode.
---------------------------------------------------------------------------

    UAM-V Metric 1: Exceedences. This metric is the total number of 
predicted concentrations exceeding the NAAQS (i.e. 1-hour values >=125 
ppb and 8-hour values >=85 ppb) within the downwind area. In 
calculating this metric, EPA summed the number of occurrences of values 
above the applicable standard (i.e., 1-hour or 8-hour) for all 
nonattainment receptors within the downwind area. For example, in 
Downwind Area #1 there are five 1-hour ``designated plus modeled'' 
nonattainment receptors. For this downwind area, the Base Case value 
for Metric 1 is calculated by first counting the number of days, across 
all four episodes, that had 1-hour daily maximum values >=125 ppb at 
each of the five receptors. The result is the total number of 
exceedences at each receptor over all days in all four episodes. The 
total number of exceedences at each receptor is then summed across all 
five receptors to produce the total number of exceedences in Downwind 
Area #1, which is the value for Metric 1 for this area.
    UAM-V Metric 2: Ozone Reduced--ppb. This metric shows the magnitude 
and frequency of the ``ppb'' impacts from each upwind State on ozone 
concentrations in each downwind area. These impacts are quantified by 
calculating the difference in ozone concentrations between the zero-out 
run and the Base Case. The results are then tabulated in terms of the 
number of ``impacts'' within six concentration ranges: >=2 to 5 ppb, 
>=5 to 10, >=10 to 15, >=15 to 20, >=20 to 25, and >=25 ppb. The 
impacts for 1-hour daily maximum values and 8-hour daily maximum values 
are determined by tallying the total ``number of days and grid cells'' 
>=125 ppb or >=85 ppb that receive contributions within the 
concentration ranges. In the analysis of contributions, as described 
below, the data from Metric 2 are used in conjunction with Metric 1 to 
determine the percent of the exceedences in the downwind area that 
receive contributions of >=2 ppb, >= 5 ppb, >=10, ppb, etc. The maximum 
``ppb'' impact within the downwind area is also calculated.
    UAM-V Metric 3: Total ppb Reduced. This metric quantifies the total 
ppb contributed in the downwind area from an upwind State, not 
including that portion of the contribution that occurs below the level 
of the NAAQS. For 1-hour concentrations, Metric 3 is calculated by 
taking the difference between the Base Case predictions in each 
nonattainment receptor and either (a) the corresponding value in the 
zero-out run, or (b) 125 ppb, whichever is greater (i.e., 125 ppb or 
the prediction in the zero-out run). The Base Case vs. zero-out 
differences are summed over all days and across all nonattainment 
receptors in the downwind area. The calculation of this metric is 
illustrated by the following example. If the Base Case 1-hour daily 
maximum ozone prediction is 150 ppb and the corresponding value from 
the zero-out run is 130 ppb, then the difference used in this metric is 
20 ppb. However, if the value from the zero-out run is 115 ppb, then 
the difference used in this metric is 25 ppb (i.e., 150 ppb-125 ppb, 
because 115 ppb is less than 125 ppb).
    For analyzing the contributions using Metric 3, the values of this 
metric are compared to the total amount of ozone above the NAAQS (i.e., 
125 ppb, 1-hour or 85 ppb, 8-hour) in the Base Case. This baseline 
measure of the ``total amount of nonattainment'' (i.e., the total 
``ppb'' of ozone that is above the NAAQS) is calculated by summing the 
``ppb'' values in the Base Case that are above the level of the NAAQS. 
The total contribution from an upwind State to a particular downwind 
area calculated by Metric 3 is expressed in relation to the

[[Page 57389]]

amount that the downwind area is in nonattainment. For example, if 
Upwind State #1 contributes a total of 50 ppb >=125 ppb to Downwind 
Area #2 and the total Base Case ozone >=125 ppb in Downwind Area #2 is 
500 ppb, then the contribution from Upwind State #1 (i.e., 50 ppb) to 
Downwind Area #2 is equivalent to 10 percent of Downwind Area #2's 
nonattainment problem (i.e., 50 ppb divided by 500 ppb, times 100).
    UAM-V Metric 4: Population-Weighted Total ppb Reduced. This metric 
is similar to the ``Total ppb Reduced'' metric except that the 
calculated contributions are weighted by (i.e., multiplied by) 
population. In calculating this metric, the ``ppb'' contributions are 
determined for each nonattainment receptor, then summed across all 
nonattainment receptors in a particular downwind area. During this 
calculation, the population in the nonattainment receptor is multiplied 
by the total contribution in that receptor (i.e., grid cell) and then 
this value is added to the corresponding values for the other receptors 
in the downwind area. The results for this metric are expressed 
relative to the population-weighted Base Case amount similar to the 
approach followed with Metric 3, as described above.
    (2) CAMx Source Apportionment Modeling. In the CAMx modeling, the 
source apportionment technique was used to calculate the contributions 
from upwind States to ozone concentrations above the NAAQS in downwind 
areas. Due to computational constraints, it was not possible for EPA to 
treat each State in the OTAG region as a separate source area. Several 
of the smaller States in the Northeast were grouped together as were 
seven States in the far western portion of the region. The following 
States were treated as individual source areas:
     Alabama, Florida, Georgia, Illinois, Indiana, Iowa, 
Kentucky, Louisiana, Maine, Massachusetts, Michigan, Mississippi, 
Missouri, New Jersey, New York, North Carolina, Ohio, Pennsylvania, 
South Carolina, Tennessee, Texas, Virginia, West Virginia, and 
Wisconsin.
    The following States were grouped together:
     Connecticut and Rhode Island were combined; Maryland, 
Delaware and the District of Columbia were combined; New Hampshire and 
Vermont were combined; and Arkansas was combined with the portions of 
Oklahoma, Kansas, Minnesota, Nebraska, North Dakota, and South Dakota 
that lie within the OTAG region.
    The contributions from each of these source areas to downwind 
nonattainment were evaluated using four metrics. As indicated above, 
the CAMx metrics are calculated for the same types of nonattainment 
receptors as the UAM-V zero-out metrics. The CAMx metrics are 
calculated in a way that is different from the metrics used for the 
zero-out runs in large part because of the differences between the two 
techniques. The zero-out modeling calculates contributions using the 
difference in predictions between two model runs (i.e., a Base Case and 
a State-specific zero-out run). In contrast, the CAMx source 
apportionment technique calculates contributions by internally tracking 
ozone formed from emissions in each source area. In raw form, the 
source apportionment technique produces a ``ppb'' contribution from 
each source area to hourly ozone in each receptor grid cell. The 
individual hourly ``ppb'' contributions were treated in the way 
described below to calculate 1-hour and 8-hour values for the four 
metrics. The approach was based on recommendations to EPA by Environ, 
the developers of CAMx. For 1-hour concentrations the metrics are 
calculated based on contributions to all hourly predictions >=125 ppb. 
For 8-hour concentrations, the metrics are calculated based on the 
contribution to every 8-hour period in a day with an average 
concentration >=85 ppb. In order to provide a link to the way 1-hour 
and 8-hour concentrations were treated for the zero-out runs, EPA also 
calculated the CAMx metrics for 1-hour daily maximum values >=125 ppb 
and 8-hour daily maximum values >=85 ppb. 43 The full set of 
results for all of the CAMx metrics is contained in the Air Quality 
Modeling TSD.
---------------------------------------------------------------------------

    \43\ As described in the Air Quality Modeling TSD, the metrics 
calculated using the hourly contributions >= 125 ppb are consistent 
with the metrics calculated using 1-hour daily maximum contributions 
>= 125 ppb. Similarly, the metrics calculated using all 8-hour 
periods >= 85 ppb are consistent with the metrics calculated using 
8-hour daily maximum values >= 85 ppb.
---------------------------------------------------------------------------

    The CAMx Metrics 1 and 2 provide information on the magnitude and 
frequency of contributions in a form that is similar to UAM-V Metrics 1 
and 2.
    CAMx Metric 3: Highest Daily Average Contribution. This metric is 
the highest daily average ozone ``ppb'' contribution from each upwind 
source area to each downwind nonattainment receptor area over all days 
modeled in all four episodes. The following example illustrates how 
this metric is calculated for 1-hour ozone concentrations. Similar 
procedures are followed for calculating this metric for 8-hour 
concentrations. First, the hourly ``ppb'' contributions from a 
particular upwind source area to each nonattainment receptor in a 
downwind area are summed across all receptors in the downwind area. 
This total daily contribution is then divided by the number of hours 
and grid cells >=125 ppb in the downwind area to determine the daily 
average ``ppb'' contribution. This calculation is performed on a day by 
day basis for each day in the 4 episodes. After the average 
contributions are calculated for each day, the highest daily average 
value across all episodes is selected for analysis. In addition, the 
highest daily average contribution is expressed as a percent of the 
downwind area's average ozone >=125 ppb. That is, the highest daily 
average ``ppb'' contribution is divided by the average of the ozone 
concentrations >=125 ppb on that day (i.e., the day on which the 
highest average ppb contribution occurred). For example, if the highest 
daily average contribution from an upwind State to nonattainment 
downwind is 15 ppb and the average of the hourly ozone values >=125 ppb 
on this day in the downwind area is 150 ppb, then the 15 ppb 
contribution, expressed as a percent, is 10 percent.
    CAMx Metric 4: Percent of Total Manmade Ozone Contribution. This 
metric represents the total contribution from emissions in an upwind 
State relative to the total ozone for all hours above the NAAQS in the 
downwind area. This metric, which is referred to as the ``average 
contribution,'' is calculated for each episode as well as for all four 
episodes combined. The following example is used to illustrate how this 
metric is calculated for a single episode for a particular downwind 
area. In step 1, all predicted Base Case hourly values >=125 ppb in the 
downwind area are summed over all nonattainment receptors and all days 
in an episode. In step 2, the ``ppb'' contributions from a source area 
to this downwind area are summed over all nonattainment receptors in 
the downwind area and all days in the episode to yield a total ppb 
contribution. The total contribution calculated in Step 2 is then 
divided by the total ozone >=125 ppb in the downwind area to produce 
the fraction of ozone >=125 ppb in the downwind area that is due to 
emissions from the upwind source area. This fraction is multiplied by 
100 to express the result as a percent.
4. Confirmation of States Making a Significant Contribution to Downwind 
Nonattainment
    In the proposal, EPA made findings of significant contribution 
based on a

[[Page 57390]]

weight-of-evidence approach that included consideration of air quality 
contributions based on subregional modeling. As discussed in section 
II.C.2, Summary of Notice of Proposed Rulemaking Weight-of-Evidence 
Approach, EPA believes that the subregional modeling provides an 
adequate independent basis for determining which States contribute 
significantly to downwind nonattainment. The evaluation of the State-
by-State modeling confirms the overall findings that were based on the 
subregional modeling and provides more refined information regarding 
the impacts of specific upwind States on nonattainment in individual 
downwind areas. This State-by-State modeling is discussed in more 
detail below.
    a. Analysis Approach. The EPA has analyzed the results of the 
State-by-State UAM-V zero-out modeling and the State-by-State CAMx 
source apportionment modeling for each of the 23 jurisdictions for 
which this modeling is available.44 Both UAM-V and CAMx 
modeling results are available for fifteen States (i.e., Alabama, 
Georgia, Illinois, Indiana, Kentucky, Massachusetts, Michigan, 
Missouri, North Carolina, Ohio, South Carolina, Tennessee, Virginia, 
West Virginia, and Wisconsin). For an additional eight States (i.e., 
Connecticut, Delaware, the District of Columbia, Maryland, New Jersey, 
New York, Pennsylvania, and Rhode Island), CAMx modeling is available. 
Also, as noted above in Section II.C.3, State-by-State Air Quality 
Modeling, Connecticut and Rhode Island were combined as a single source 
area, and Maryland, the District of Columbia, and Delaware were also 
combined as a single source area. Because the NOX emissions 
and/or NOX emissions density is large in each jurisdiction 
within both of these combined source areas, EPA believes that the 
downwind contributions from these combined source areas can be 
attributed to each jurisdiction within the source area.
---------------------------------------------------------------------------

    \44\ The approach for dealing with the 15 States in the OTAG 
domain which were not proposed to make a significant contribution to 
downwind nonattainment are discussed below in Section II.C.5, States 
Not Covered by this Rulemaking.
---------------------------------------------------------------------------

    For the 1-hour NAAQS, EPA evaluated downwind impacts in two ways 
using the factors described in Section II.C.3, State-by-State Air 
Quality Modeling. First, EPA evaluated the contributions from each 
upwind State to nonattainment in each downwind State. Second, the EPA 
evaluated the contributions from each upwind State to nonattainment in 
each downwind 1-hour nonattainment area. In downwind States which only 
contain a single intrastate nonattainment area (e.g., Atlanta), the 
results of the downwind State and downwind nonattainment area analyses 
are the same because the same nonattainment receptors are used in both 
cases. For the 8-hour NAAQS, EPA evaluated the contributions from 
upwind States to 8-hour nonattainment in each downwind State.
    The EPA used the following process in determining whether a 
particular upwind State contributes significantly to 1-hour 
nonattainment in an individual downwind area. First, EPA reviewed the 
extent of the nonattainment problem in the downwind area using ambient 
design values and model predictions of future ozone concentrations 
after the application of (a) 2007 Base Case controls, (b) additional 
local NOX reductions, and (c) regional reductions 
(additional local plus upwind NOX reductions).45 
As indicated above, EPA determined that each downwind area had a 
residual nonattainment problem even after implementation of the control 
measures in the 2007 Base Case.
---------------------------------------------------------------------------

    \45\ Scenarios (b) and (c) refer to the runs used to assess 
transport as described in Section IV.
---------------------------------------------------------------------------

    Second, using the information from CAMx Metric 4 46, EPA 
reviewed (a) the relative portion of the ozone problem in each downwind 
area that is due to ``local'' emissions (i.e., emissions from the 
entire State or States in which the downwind area is located), (b) the 
total contribution from all upwind emissions (i.e., the sum of the 
contributions from manmade emissions in all upwind States, combined), 
and (c) the contribution from manmade emissions in individual upwind 
States. The local versus upwind contributions for each downwind area 
are provided in the Air Quality Modeling TSD. The EPA analyzed this 
information to determine whether upwind emissions are an important part 
of the downwind areas' nonattainment problem. In general, the data 
indicate that, although a substantial portion of the 1-hour 
nonattainment problem in many of the downwind areas is due to local 
emissions, a substantial portion of the nonattainment problem is also 
due to emissions from upwind States. In addition, for most upwind-
State-to-downwind-area linkages there is no single upwind State that 
makes up all of the upwind contribution. Rather, the total contribution 
for all upwind States combined is comprised of individual contributions 
from a number of upwind States many of which are relatively similar in 
magnitude such that there is no ``bright line'' which distinguishes 
between the contributions from most of the individual upwind States.
---------------------------------------------------------------------------

    \46\ This information represents the average contributions 
across all four episodes. In addition to the four-episode average 
contribution, EPA also examined the highest single-episode average 
contribution from each upwind State to each downwind area.
---------------------------------------------------------------------------

    Third, EPA determined whether each individual upwind State 
significantly contributes to nonattainment in a particular downwind 
area using the UAM-V and CAMx metrics to evaluate three aspects, or 
factors of the contribution.47 These factors include the 
magnitude, frequency, and relative amount of the contribution. The 
specific UAM-V and CAMx metrics which correspond to each of the factors 
are identified in Table II-2. As indicated in the table, there is at 
least one metric from each modeling technique that corresponds to each 
of the three factors.
---------------------------------------------------------------------------

    \47\ The factors used to interpret the metrics should not be 
confused with the multi-factor approach used to identify the amounts 
of NOX emissions that contribute signficantly to 
nonattainment.

      Table II-2.--Metrics Associated With Each Contribution Factor
------------------------------------------------------------------------
           Factor                     UAM-V                 CAMx
------------------------------------------------------------------------
Magnitude of Contribution...  Maximum ``ppb''       Maximum ``ppb''
                               contribution          Contribution
                               (Metric 2)            (Metric 2); and
                                                     Highest Daily
                                                     Average
                                                     Contribution
                                                     (Metric 3).
Frequency of Contribution...  Number and percent    Number and percent
                               of exceedences with   of exceedences with
                               contributions in      contributions in
                               various               various
                               concentration         concentration
                               ranges (Metric 1      ranges (Metric 1
                               and 2)                and 2).
Relative Amount of            Total ``ppb''         Four-episode average
 Contribution.                 contribution          percent
                               relative to the       contribution from
                               total ``ppb'' that    the upwind State to
                               the downwind area     nonattainment in
                               is above the NAAQS    the downwind area
                               (Metric 3); and       (Metric 4); and
                               Total population-     Highest single-
                               weighted ``ppb''      episode average
                               contribution          percent
                               relative to the       contribution from
                               total population-     the upwind State to
                               weighted ``ppb''      nonattainment in
                               that the downwind     the downwind area
                               area is above the     (Metric 4).
                               NAAQS (Metric 4)
------------------------------------------------------------------------


[[Page 57391]]

    It should be noted that the relative contributions of individual 
upwind States to a particular downwind area add up to 100 percent for 
the CAMx 4-episode average percent contribution. However, this is not 
the case for the CAMx highest single-episode average percent 
contribution since the value from one upwind State can occur in a 
different episode than the value from another upwind State for the same 
downwind area. In addition, it should be noted that UAM-V Metrics 3 and 
4 are used in combination to express the total contribution above the 
NAAQS relative to the total amount that the downwind area is above the 
NAAQS. The values for each of these metrics also do not add up to 100 
percent when considering contributions from multiple upwind States to 
an individual downwind area.
    The EPA compiled the UAM-V and CAMx metrics by downwind area in 
order to evaluate the contributions to downwind nonattainment. The data 
on 1-hour and 8-hour contributions were compiled and analyzed 
separately. The data were reviewed to determine how large of a 
contribution a particular upwind State makes to nonattainment in each 
downwind area in terms of the magnitude of the contribution and the 
relative amount of the total contribution. The data were also examined 
to determine how frequently the contributions occur.
    The first step in evaluating this information was to screen out 
linkages for which the contributions were very low, as described in the 
Air Quality Modeling TSD. The finding of significance for linkages that 
passed the initial screening criteria was based on EPA's technical 
assessment of the values for the three contribution factors. Each 
upwind State that had large and/or frequent contributions to the 
downwind area, based on these factors, is considered as contributing 
significantly to nonattainment in the downwind area. The EPA believes 
that each of the factors provides an independent legitimate measure of 
contribution. However, there had to be multiple factors that indicate 
large and/or frequent contributions in order for the linkage to be 
significant. In this regard, the finding of a significant contribution 
for an individual linkage was not based on any single factor.
    For many of the individual linkages the factors yield a consistent 
result (i.e., either large and/or frequent contributions or small and/
or infrequent contributions). In some cases, however, not all of the 
factors are consistent. For upwind-downwind linkages in which some of 
the factors indicate high and/or frequent contributions while other 
factors do not, EPA considered the overall number and magnitude of 
those factors that indicate large and/or frequent contributions 
compared to those factors that do not. Based on an assessment of all 
the factors in such cases, EPA determined that the upwind State 
contributes significantly to nonattainment in the downwind area if on 
balance the factors indicate large and/or frequent contributions from 
the upwind State to the downwind area.
    The EPA's evaluation of the contributions to 1-hour nonattainment 
in New York City is presented as an example to illustrate this process. 
The New York City area, which consists of portions of New York, New 
Jersey, and Connecticut, is designated as a severe nonattainment area 
under the 1-hour NAAQS. The ambient 1-hour design value in New York 
City, based on 1994 through 1996 monitoring data is 144 ppb. During the 
four OTAG episodes, 39 percent of the days are predicted to have 1-hour 
exceedences in 2007 after the implementation of all CAA controls and 
Federal measures.48 Moreover, EPA's air quality modeling of 
the benefits of regional NOX strategies, as described in 
Section IV, Air Quality Assessment, indicates that there would still be 
exceedences of the 1-hour NAAQS remaining in New York City even with 
eliminating the significant amounts of emissions required by this 
NOX SIP Call.
---------------------------------------------------------------------------

    \48\ This is further described in the Air Quality Modeling TSD.
---------------------------------------------------------------------------

    In the assessment of contributions to New York City, EPA examined 
the local versus upwind contributions to 1-hour nonattainment in this 
area, as shown in Table II-3. Local emissions in the New York City 
nonattainment area are spread among numerous stationary sources, area 
sources, highway sources, and nonroad sources, each of which 
contributes only a very small, indeed sometimes immeasurable, amount to 
New York City's ozone nonattainment problem. Combined, these emissions 
result in approximately 55 percent of the New York City area's ozone 
problem. Emissions from States upwind of New York, New Jersey, and 
Connecticut, on average across all four episodes, contribute 45 percent 
of the nonattainment problem in New York City is due to. However, no 
single State stands out as contributing most of the total upwind 
contribution. The biggest single contributor is Pennsylvania (18 
percent) followed by Maryland/Washington, DC/Delaware (5 percent). The 
total contribution from all Northeast States is 23 percent. A similar 
amount (22 percent) of the total contribution is due to emissions in 
those States outside the Northeast. The data in Table II-3 indicate 
that 19 percent of the 22 percent is fairly evenly divided among ten 
States, whose contributions range from 1 percent (6 States) to 4 
percent (Ohio and Virginia). The remaining 3 percent (i.e., 19 percent 
vs 22 percent) is from States that each contribute less than 1 percent, 
on average. The highest single-episode contributions from States upwind 
of the Northeast range from 1 percent (Tennessee) to 8 percent 
(Virginia). In general, the contribution data in Table II-3 indicate 
that a substantial amount of New York City's nonattainment problem is 
due to the collective contribution from emissions in a number of upwind 
States both within and outside the northeast. That these upwind 
contributions are a meaningful part of New York City's nonattainment 
problem is particularly evident in light of the fact that the 
contribution to the problem made by New York City itself is comprised 
of the collective contribution of numerous sources.

        Table II-3.--Percent Contribution From Upwind States to 1-Hour Nonattainment in New York City \1\
----------------------------------------------------------------------------------------------------------------
                                                                                   Percent of
                                                                                  total manmade  Highest single-
                          Downwind area: New York City                           emissions over  episode percent
                                                                                   4 episodes      contribution
-------------------------------------------------------------------------------------------------------\2\------
Amount due to ``Local'' Emissions \3\..........................................              55            \4\NA
Total Amount from all ``Upwind'' States........................................              45               NA
Contributions from Individual Upwind States....................................  ..............  ...............
PA.............................................................................              18               19
MD/DC/DE.......................................................................               5                6

[[Page 57392]]

OH.............................................................................               4                6
VA.............................................................................               4                8
WV.............................................................................               3                7
IL.............................................................................               2                3
IN.............................................................................               1                2
KY.............................................................................               1                3
MI.............................................................................               1                4
MO.............................................................................               1                2
NC.............................................................................               1                2
TN.............................................................................               1                1
Total Amount from All Other States, combined...................................               3             NA.
----------------------------------------------------------------------------------------------------------------
\1\ These values are based on CAMx Metric 3 calculated across all 4 episodes.
\2\ These values are based on CAMx Metric 3 calculated for each episode individually. These values do not add up
  to 100 percent.
\3\ 3. Total contribution from the State(s) in which the Nonattainment area is located.
\4\ 4. Not applicable.

    The extent of New York City's nonattainment problem and the nature 
of the contributions from upwind States were considered in determining 
whether the values of the metrics indicate large and/or frequent 
contributions for individual upwind States. Specifically, additional 
controls beyond the local and upwind NOX reductions which 
are part of the regional NOX strategy may be needed to solve 
New York City's 1-hour nonattainment problem. Also, the total 
contribution from all upwind States is large and there is no single 
State or small number of States which comprise this total upwind 
portion. In this regard, the contributions to New York City from some 
States may not appear to be individually ``high'' amounts. However, (as 
described below) these contributions, when considered together with the 
contributions from other States (i.e., the collective contribution) 
produce a large total contribution to nonattainment in New York City.
    The EPA evaluated the magnitude, frequency, and relative amount of 
contribution from emissions in individual upwind States to determine 
which States contribute significantly to 1-hour nonattainment in New 
York City. The UAM-V and CAMx metrics which quantify each upwind 
State's contribution to New York City for each of the three factors are 
provided in the Air Quality Modeling TSD and described below. 
Examination of the values for these metrics indicates that the upwind 
States can be divided into three general groups, based on the 
magnitude, frequency, and relative amount of contribution. The first 
group contains those upwind States for which the UAM-V and CAMx metrics 
all clearly indicate a significant contribution to 1-hour nonattainment 
in New York City. The second group contains those States for which the 
CAMx and UAM-V metrics are not quite as consistent, but overall the 
metrics indicate a significant contribution to 1-hour nonattainment in 
New York City.49 The third group contains those States for 
which the CAMx and UAM-V metrics clearly indicate that the impacts do 
not make a significant contribution to New York City.
---------------------------------------------------------------------------

    \49\ For New York City, each of the ``Group 2'' States were 
found to make a significant contribution. However, this was not the 
case for all of the Group 2 linkages in other nonattainment areas. 
For example, the contribution from Kentucky to Philadelphia and the 
contribution from Tennessee to Baltimore were Group 2 situations in 
which EPA determined that the contributions were not significant.
---------------------------------------------------------------------------

    Group 1  Upwind States:
    The CAMx and UAM-V metrics all clearly indicate that emissions from 
Maryland/Washington, DC/Delaware, Ohio, Pennsylvania, Virginia, and 
West Virginia make large and/or frequent contributions to 1-hour 
nonattainment in New York City. For Pennsylvania the magnitude of 
contribution, as indicated by the highest daily average contribution 
(CAMx Metric 3), is 25 ppb and the relative amount of contribution is 
18 percent (CAMx Metric 4). For the other upwind areas, the magnitude 
of the contributions range from 9 ppb to 15 ppb (CAMx Metric 3, highest 
daily average contributions) with contributions in the range of 5 ppb 
to 10 ppb--from Ohio, Virginia, and West Virginia (UAM-V Metric 2, 
maximum ``ppb'' contribution). In terms of the frequency of the 
contribution, 7 percent to 11 percent of the total number of grid-hours 
>=125 ppb in New York City receive contributions of 10 ppb from each of 
these States (CAMx Metric 1 and 2). Also, the relative amounts of the 
contribution are in the range of 6 percent to 8 percent (CAMx Metric 4, 
highest single-episode average percent contribution) and the total 
contribution from each of three States (i.e., Ohio, Virginia, and West 
Virginia) is large compared to the total amount of nonattainment, 
ranging from 8 percent to 11 percent (UAM-V Metric 3).
    Group 2  Upwind States:
    The CAMx and UAM-V metrics are somewhat less consistent on the 
extent of contributions from each of 5 States: Kentucky, Illinois, 
Indiana, Michigan, and North Carolina. None of the metrics for either 
model indicate extremely low or extremely high contributions. Rather, 
for these States most of the metrics indicate relatively high 
contributions while a few metrics indicate relatively low 
contributions. The rationale used by EPA for evaluating the 
contributions from these States involved comparing and contrasting each 
piece of data for these States on an individual ``upwind State-by-
upwind State'' basis and as a group (i.e., for all 5 States, together) 
in order to weigh the relative magnitude and frequency of the 
contributions for making a determination of significance.
    UAM-V Metrics--For each of these 5 States the ``weakest'' factor is 
the magnitude contribution (UAM-V Metric 2) in that the highest 
contributions are in the range of 2 to 5 ppb. The other UAM-V Metrics, 
however, indicate that the contributions from each State are of a 
larger frequency and relative amount. Specifically, four of these 
States (Kentucky, Indiana, Illinois, and

[[Page 57393]]

Michigan) each contribute 2 to 5 ppb to as many as 3 percent to 4 
percent of the exceedences in New York City (UAM-V Metrics 1 and 2). 
While North Carolina contributes to somewhat fewer exceedences (2 
percent), this slight weakness is out-weighed by the relative amount of 
contribution (UAM-V Metrics 3 and 4) which indicates that the total 
contribution from North Carolina alone is equivalent to 3 percent of 
the total ``ppb'' >=125 ppb and 4 percent of the population-weighted 
``ppb'' >=125 ppb in New York City. For Indiana, Illinois, and Michigan 
the relative amount of contribution (UAM-V Metrics 3 and 4) is also 
relatively high and ranges from 3 percent to 5 percent. The relative 
amount of contribution from Kentucky is somewhat weaker at 2 percent.
    CAMx Metrics--For Illinois, all of the CAMx metrics indicate 
relatively large and/or frequent contributions, as described below. For 
Kentucky, Indiana, Michigan, and North Carolina the magnitude of 
contribution is large, as indicated by the maximum contribution which 
ranges from 6 ppb (Indiana) to 11 ppb (North Carolina). Also, the 
highest daily average contribution from Kentucky, Michigan, and North 
Carolina are all in the range of 5 ppb to 7 ppb. In terms of the 
frequency of contribution, Indiana and North Carolina contribute in the 
range of 5 ppb to 10 ppb to 3 percent and 6 percent of the exceedences, 
respectively, in New York City. For Kentucky, Indiana, Michigan, and 
North Carolina the relative amounts of contribution is somewhat mixed 
in that the 4-episode average percent contribution is only 1 percent, 
but the highest single-episode average percent contributions are higher 
at 2 percent from both Indiana and North Carolina, 3 percent from 
Kentucky, and 4 percent from Michigan (CAMx Metric 4).
    Overall contributions considering UAM-V and CAMx Metrics--
Considering the CAMx and UAM-V metrics, as described below, the 
majority of the contribution factors indicate that, overall, each of 
the Group 2 States contributes significantly to 1-hour nonattainment in 
New York City.
Kentucky--
    Metrics indicating relatively high and/or frequent contributions:

--Magnitude of Contribution: the maximum contribution from CAMx is 9 
ppb (CAMx Metric 2) and highest daily average contribution is 7 ppb 
(CAMx Metric 3);
--Frequency of Contribution: 4 percent of the exceedences receive 
contributions of more than 2 ppb (UAM-V Metrics 1 and 2); and
--Relative Amount of Contribution: the highest single-episode average 
contribution is 3 percent (CAMx Metric 4).

    Metrics indicating relatively low and/or infrequent contributions:

--Magnitude of Contribution: the maximum contribution from UAM-V is 2 
ppb; and
--Relative Amount of Contribution: the 4-episode average percent 
contribution is 1 percent (CAMx Metric 4).
Indiana--
    Metrics indicating relatively high and/or frequent contributions:

--Magnitude of Contribution: the maximum ``ppb'' contribution is 6 ppb 
(CAMx Metric 2);
--Frequency of Contribution: 4 percent of the exceedences receive 
contributions of more than 2 ppb (UAM-V Metrics 1 and 2) ; and
--Relative Amount of Contribution: the total ``ppb'' contribution is 
equivalent to 3 percent of total amount of nonattainment (UAM-V Metric 
3).

    Metrics indicating relatively low and/or infrequent contributions:

--Magnitude of Contribution: the maximum contribution from is 2 ppb 
(UAM-V Metric 2); and
--Relative Amount of Contribution: the 4-episode average percent 
contribution is 1 percent (CAMx Metric 4).
Illinois--
    Metrics indicating relatively high and/or frequent contributions:

--Magnitude of Contribution: the maximum contribution is 8 ppb (CAMx 
Metric 2); the highest daily average contribution is 6 ppb;
--Frequency of Contribution: 3 percent of the exceedences receive 
contributions of more than 2 ppb; and
--Relative Amount of Contribution: the highest single-episode average 
contribution is 3 percent (CAMx Metric 4); the total ``ppb'' 
contribution is equivalent to 3 percent of total amount of 
nonattainment.

    Metrics indicating relatively low and/or infrequent contributions:

--Magnitude of Contribution: the maximum contribution from UAM-V is 2 
ppb.
Michigan--
    Metrics indicating relatively high and/or frequent contributions:

--Magnitude of Contribution: the maximum contribution is 7 ppb (CAMx 
Metric 2); the highest daily average contribution is 5 ppb (CAMx Metric 
3);
--Frequency of Contribution: 3 percent of the exceedences receive 
contributions of more than 2 ppb (UAM-V Metrics 1 and 2); and
--Relative Amount of Contribution: the highest single-episode average 
contribution is 4 percent (CAMx Metric 4); the total ``ppb'' 
contribution is equivalent to 3 percent of the total amount of 
nonattainment.

    Metrics indicating relatively low and/or infrequent contributions:

--Magnitude of Contribution: the maximum contribution from UAM-V is 2 
ppb
--Frequency of Contribution: 1 percent of the exceedences receive 
contributions of 5 ppb or more (CAMx Metrics 1 and 2); and
--Relative Amount of Contribution: the 4-episode average percent 
contribution is 1 percent (CAMx Metric 4).
North Carolina--
    Metrics indicating relatively high and/or frequent contributions:

--Magnitude of Contribution: the maximum contribution is 11 ppb (CAMx 
Metric 2); the highest daily average contribution is 6 ppb (CAMx Metric 
3);
--Frequency of Contribution: 6 percent of exceedences receive 
contributions of 5 ppb or more (CAMx Metrics 1 and 2); and
--Relative Amount of Contribution: the total ``ppb'' contribution is 
equivalent to 3 percent of total amount of nonattainment.

    Metrics indicating relatively low and/or infrequent contributions:

--Relative Amount of Contribution: the 4-episode average percent 
contribution is 1 percent (CAMx Metric 4).

    Group 3  Upwind States: The CAMx and UAM-V metrics clearly indicate 
that the emissions from the following States do not make large and/or 
frequent contributions to 1-hour nonattainment in New York City: 
Alabama, Georgia, Massachusetts, Missouri, South Carolina, Tennessee, 
and Wisconsin. The rationale for this conclusion is as follows:

--Magnitude of Contribution: all of these upwind States individually 
contribute less than 2 ppb to 1-hour daily maximum exceedences in New 
York City (UAM-V Metric 2); the highest daily average contribution was 
1 ppb or less from Alabama, Georgia, and Massachusetts, and 2

[[Page 57394]]

ppb from South Carolina, Tennessee, and Wisconsin (CAMx Metric 3); and
--Relative Amount of Contribution: the 4-episode average contributions 
from Alabama, Georgia, Massachusetts, South Carolina, and Wisconsin are 
less than 1 percent (CAMx Metric 4); the total contributions from 
Missouri and Tennessee are each equivalent to 1 percent of the total 
amount of nonattainment in New York City (UAM-V Metric 3).

    Based on the preceding evaluation, EPA believes that emissions in 
each of the following twelve jurisdictions contribute significantly to 
1-hour nonattainment in the New York City nonattainment area: the 
District of Columbia, Delaware, Illinois, Indiana, Kentucky, Maryland, 
Michigan, North Carolina, Ohio, Pennsylvania, Virginia, and West 
Virginia.
    b. States Which Contain Sources That Significantly Contribute to 
Downwind Nonattainment. The results of EPA's assessment of the State-
by-State UAM-V and CAMx modeling confirms the findings based on 
subregional modeling that the 23 jurisdictions contribute large and/or 
frequent amounts to downwind nonattainment under both the 1-hour and 8-
hour NAAQS and forms an independent basis for those findings. The 
specific upwind States which significantly contribute to nonattainment 
in specific downwind States are listed in Tables II-4 and II-5 for the 
1-hour NAAQS and Table II-6 and Table II-7 for the 8-hour NAAQS. The 
information on the 1-hour contribution linkages are presented by upwind 
State in Table II-4 and by downwind State in Table II-5. In Table II-4 
the upwind States are each listed in the first column and the downwind 
States to which each upwind State contributes significantly are listed 
in the second column. In Table II-5, the same information is presented 
by downwind State. In this table, each downwind State is listed in the 
first column and the upwind States that contribute to that downwind 
State are listed in the second column. The 8-hour contribution linkages 
are presented by upwind State in Table II-6 and by downwind State in 
Table II-7.

  Table II-4.--Downwind States for Which Upwind States Contain Sources
         That Contribute Significantly to 1-Hr Nonattainment \1\
------------------------------------------------------------------------
         Upwind state                       Downwind states
------------------------------------------------------------------------
Alabama......................  GA, IL*, IN*, MI*, TN, WI*.
Connecticut..................  ME, MA, NH.
Delaware.....................  CT, ME, MA, NH*, NJ, NY, PA, RI, VA.
District of Columbia.........  CT, ME, MA, NH*, NJ, NY, PA, RI, VA.
Georgia......................  AL, TN.
Illinois.....................  CT*, IN, MD, NJ*, NY, MI, MO, WI*.
Indiana......................  CT*, DE*, DC*, IL*, KY, MD, NJ*, NY, MI,
                                OH, VA*, WI*.
Kentucky.....................  AL, CT*, DC*, GA, IL*, IN, MD, MI*, NJ,
                                NY, MO, OH, VA, WI*.
Maryland.....................  CT, ME, MA, NH*, NJ, NY, PA, RI, VA.
Massachusetts................  ME, NH.
Michigan.....................  CT, DC*, MD, NJ, NY, VA*.
Missouri.....................  IL, IN, MI, WI*.
New Jersey...................  CT, ME, MA, NH, NY, PA, RI.
New York.....................  CT, ME, MA, NH, NJ, RI.
North Carolina...............  CT*, DC*, GA, KY, MD, NJ, NY, OH, PA,
                                VA*.
Ohio.........................  CT, DE, DC*, KY, MD, MA, NH*, NJ, NY, PA,
                                RI, VA.
Pennsylvania.................  CT, DE, DC, ME, MD, MA, NH, NJ, NY, RI,
                                VA.
Rhode Island.................  ME, MA, NH.
South Carolina...............  AL, GA, TN.
Tennessee....................  AL, GA, IL*, IN, KY, MI*, OH, WI*.
Virginia.....................  CT, DE, DC, KY*, MD, MA, NH*, NJ, NY, PA,
                                RI.
West Virginia................  CT, DE, DC, MD, MA, NJ, NY, PA, RI, VA.
Wisconsin....................  IL*, IN*, MI* .
------------------------------------------------------------------------
\1\ States marked with an asterisk (*) are included because they are
  part of an interstate nonattainment area that receives a contribution
  from the upwind State. New Hampshire is included because it is part of
  the combined Boston/Portsmouth area; Connecticut and New Jersey are
  included because they are part of the New York City area; Kentucky is
  included because it is part of the Cincinnati area; Delaware is
  included because it is part of the Philadelphia area; Illinois is
  included because it is part of the St. Louis area; Illinois, Indiana,
  Michigan, and Wisconsin are included because they are part of the Lake
  Michigan area; and Maryland, Virginia, and the District of Columbia
  are included because they are part of the Washington, DC area.


     Table II-5.--Upwind States that Contain Sources that Contribute
       Significantly to 1-Hr Nonattainment in Downwind States \1\
------------------------------------------------------------------------
        Downwind state                       Upwind states
------------------------------------------------------------------------
Alabama......................  GA, KY, SC, TN.
Connecticut..................  DE, DC, IL*, IN*, KY*, MD, MI*, NJ, NY,
                                NC*, OH, PA, VA, WV.
Delaware.....................  IN*, OH, PA, VA, WV.
District of Columbia.........  IN*, KY*, MI*, NC*, OH*, PA, VA, WV.
Georgia......................  AL, KY, NC, SC, TN.
Illinois.....................  AL*, IN*, KY*, MO, TN*, WI*.
Indiana......................  AL*, IL, KY, MO, TN, WI*.
Kentucky.....................  IN, NC, OH, TN, VA*.
Maine........................  CT, DE, DC, MD, MA, NJ, NY, PA, RI.
Maryland.....................  IL, IN, KY, MI, NC, OH, PA, VA, WV.
Massachusetts................  CT, DE, DC, MD, NJ, NY, OH, PA, RI, VA,
                                WV.
Michigan.....................  AL*, IL, IN, KY*, MO, TN*, WI*.
Missouri.....................  IL, KY.

[[Page 57395]]

New Hampshire................  CT, DC*, DE*, MD*, MA, NJ, NY, OH*, PA,
                                RI, VA*.
New Jersey...................  DE, DC, IL*, IN*, KY, MD, MI, NY, NC, OH,
                                PA, VA, WV.
New York.....................  DE, DC, IL, IN, KY, MD, MI, NJ, NC, OH,
                                PA, VA, WV.
Ohio.........................  IN, KY, TN, NC.
Pennsylvania.................  DE, DC, MD, NJ, NC, OH, VA, WV.
Rhode Island.................  DE, DC, MD, NJ, NY, OH, PA, VA, WV.
Tennessee....................  AL, GA, SC.
Virginia.....................  DE, DC, IN*, KY, MD, MI*, NC*, OH, PA,
                                WV.
Wisconsin....................  AL*, IL*, IN*, KY*, MO*, TN* .
------------------------------------------------------------------------
\1\ Upwind States marked with an asterisk (*) are considered to
  significantly contribute to the downwind State because they contribute
  to an interstate nonattainment area that includes part of the downwind
  State. New Hampshire is included in the Boston/Portsmouth area;
  Connecticut and New Jersey are included in the New York City area;
  Kentucky is included in the Cincinnati area; Delaware is included in
  the Philadelphia area; Illinois is included in the St. Louis area;
  Illinois, Indiana, Michigan, and Wisconsin are included in the Lake
  Michigan area; and Maryland and Virginia are included in the
  Washington, DC area.


     Table II-6.--Downwind States to Which Sources in Upwind States
            Contribute Significantly for the 8-hour Standard
------------------------------------------------------------------------
         Upwind state                       Downwind states
------------------------------------------------------------------------
Alabama......................  GA, IL, IN, KY, MI, MO, NC, OH, PA, SC,
                                TN, VA.
Connecticut..................  ME, MA, NH, RI.
Delaware.....................  CT, ME, MA, NH, NJ, NY, PA, RI, VA.
District of Columbia.........  CT, ME, MD, MA, NH, NJ, NY, PA, RI, VA.
Georgia......................  AL, IL, IN, KY, MI, MO, NC, SC, TN, VA.
Illinois.....................  AL, CT, DC, DE, IN, KY, MD, MI, MO, NJ,
                                NY, OH, PA, RI, TN, WV, WI.
Indiana......................  DE, IL, KY, MD, MI, MO, NJ, NY, OH, PA,
                                TN, VA, WV, WI.
Kentucky.....................  AL, DC, DE, GA, IL, IN, MD, MI, MO, NJ,
                                NY, NC, OH, PA, SC, TN, VA, WV, WI.
Maryland.....................  CT, DE, DC, ME, MA, NH, NJ, NY, PA, RI,
                                VA.
Massachusetts................  ME, NH
Michigan.....................  CT, DC, DE, MD, MA, NJ, NY, OH, PA, WV.
Missouri.....................  IL, IN, KY, MI, OH, PA, TN, WI.
New Jersey...................  CT, ME, MA, NH, NY, PA, RI.
New York.....................  CT, ME, MA, NH, NJ, PA, RI.
North Carolina...............  AL, CT, DE, GA, IN, KY, ME, MD, MA, NJ,
                                NY, OH, PA, RI, SC, TN, VA, WV.
Ohio.........................  CT, DC, DE, IN, KY, MD, MA, MI, NJ, NY,
                                NC, PA, RI, TN, VA, WV.
Pennsylvania.................  CT, DC, DE, ME, MD, MA, NH, NJ, NY, OH,
                                RI, VA.
Rhode Island.................  ME, MA, NH.
South Carolina...............  AL, GA, IN, KY, NC, TN, VA.
Tennessee....................  AL, DC, DE, GA, IL, IN, KY, MD, MI, MO,
                                NC, OH, PA, SC, VA, WV, WI.
Virginia.....................  CT, DE, DC, ME, MD, MA, NJ, NY, NC, OH,
                                PA, RI, SC, WV.
West Virginia................  CT, DC, DE, IN, KY, MD, MA, NJ, NY, NC,
                                OH, PA, RI, SC, TN, VA.
Wisconsin....................  MI.
------------------------------------------------------------------------


     Table II-7.--Upwind States that Contain Sources that Contribute
        Significantly to 8-hour Nonattainment in Downwind States.
------------------------------------------------------------------------
        Downwind state                       Upwind states
------------------------------------------------------------------------
Alabama......................  GA, IL, KY, NC, SC, TN.
Connecticut..................  DE, DC, IL, MD, MI, NJ, NY, NC, OH, PA,
                                VA, WV.
District of Columbia.........  IL, KY, MD, MI, OH, PA, TN, VA, WV.
Delaware.....................  IL, IN, KY, MI, NC, OH, PA, TN, VA, WV.
Georgia......................  AL, KY, NC, SC, TN.
Illinois.....................  AL, GA, IN, KY, MO, TN.
Indiana......................  AL, GA, IL, KY, MO, NC, OH, SC, TN, WV.
Kentucky.....................  AL, GA, IL, IN, MO, NC, OH, SC, TN, WV.
Maine........................  CT, DE, DC, MD, MA, NJ, NY, NC, PA, RI,
                                VA
Maryland.....................  DC, IL, IN, KY, MI, NC, OH, PA, TN, VA,
                                WV.
Massachusetts................  CT, DE, DC, MD, MI, NJ, NY, NC, OH, PA,
                                RI, VA, WV.
Michigan.....................  AL, GA, IL, IN, KY, MO, OH, TN, WI.
Missouri.....................  AL, GA, IL, IN, KY, TN.
New Hampshire................  CT, DE, DC, MD, MA, NJ, NY, PA, RI.
New Jersey...................  DE, DC, IL, IN, KY, MD, MI, NC, NY, OH,
                                PA, VA, WV.
New York.....................  DE, DC, IL, IN, KY, MD, MI, NC, NJ, OH,
                                PA, VA, WV.
North Carolina...............  AL, GA, KY, OH, SC, TN, VA, WV.
Ohio.........................  AL, IL, IN, KY, MI, MO, NC, PA, TN, VA,
                                WV.
Pennsylvania.................  AL, DE, DC, IL, IN, KY, MD, MI, MO, NJ,
                                NY, NC, OH, TN, VA, WV.
Rhode Island.................  CT, DE, DC, IL, MD, NJ, NY, NC, OH, PA,
                                VA, WV.

[[Page 57396]]

South Carolina...............  AL, GA, KY, NC, TN, VA, WV.
Tennessee....................  AL, GA, IL, IN, KY, MO, NC, OH, SC, WV.
Virginia.....................  AL, DE, DC, GA, IN, KY, MD, NC, OH, PA,
                                SC, TN, WV.
West Virginia................  IL, IN, KY, MI, NC, OH, TN, VA.
Wisconsin....................  IL, IN, KY, MO, TN.
------------------------------------------------------------------------

    c. Examples of Contributions From Upwind States to Downwind 
Nonattainment. A full discussion of EPA's analysis supporting the 
determination that specific upwind States contribute significantly to 
individual downwind States under the 1-hour and 8-hour NAAQS is 
provided in the Air Quality Modeling TSD. Examples of the types of 
contributions which link individual upwind States to downwind areas are 
provided below for the 1-hour NAAQS for the 23 upwind jurisdictions.
--Alabama's Contribution to 1-Hour Nonattainment in Atlanta
    Magnitude of Contribution: The maximum contribution is 39 ppb (CAMx 
Metric 2); the highest daily average contribution is 31 ppb (CAMx 
Metric 3).
    Frequency of Contribution: Alabama contributes at least 10 ppb to 
12 percent of the 1-hr exceedences (UAM-V Metrics 1 and 2).
    Relative Amount: The total contribution from Alabama is equivalent 
to 14 percent of the total amount >=125 ppb in Atlanta (UAM-V Metric 
3); Alabama contributes 8 percent of the total manmade ppb >= 125 ppb 
in Atlanta (CAMx Metric 4; 4-episode average percent contribution).
--Connecticut/Rhode Island's Contribution to 1-Hour Nonattainment in 
Western Massachusetts
    Magnitude of Contribution: The maximum contribution is 61 ppb (CAMx 
Metric 2); the highest daily average contribution is 50 ppb (CAMx 
Metric 3).
    Frequency of Contribution: Connecticut/Rhode Island contribute at 
least 10 ppb to 100 percent of the 1-hr exceedences (CAMx Metrics 1 and 
2).
    Relative Amount: Connecticut/Rhode Island contribute 35 percent of 
the total manmade ppb >= 125 ppb in Western Massachusetts (CAMx Metric 
4; 4-episode average percent contribution).
--Georgia's Contribution to 1-Hour Nonattainment in Birmingham
    Magnitude of Contribution: The maximum contribution is 51 ppb (CAMx 
Metric 2); the highest daily average contribution is 24 ppb (CAMx 
Metric 3).
    Frequency of Contribution: Georgia contributes at least 10 ppb to 
11 percent of the 1-hr exceedences (UAM-V Metrics 1 and 2).
    Relative Amount: The total contribution from Georgia is equivalent 
to 12 percent of the total amount >=125 ppb in Birmingham (UAM-V Metric 
3); Georgia contributes 3 percent of the total manmade ppb >= 125 ppb 
in Birmingham (CAMx Metric 4; 4-episode average percent contribution).
--Illinois's Contribution to 1-Hour Nonattainment in New York City
    Magnitude of Contribution: The maximum contribution is 8 ppb (CAMx 
Metric 2); the highest daily average contribution is 6 ppb (CAMx Metric 
3).
    Frequency of Contribution: Illinois contributes at least 5 ppb to 
20 percent of the 1-hr exceedences (CAMx Metrics 1 and 2).
    Relative Amount: The total contribution from Illinois is equivalent 
to 3 percent of the total amount >=125 ppb in New York City (UAM-V 
Metric 3); Illinois contributes 3 percent of the total manmade ppb >= 
125 ppb in New York City (CAMx Metric 4; single highest episode percent 
contribution).
--Indiana's Contribution to 1-Hour Nonattainment in Baltimore
    Magnitude of Contribution: The maximum contribution is 8 ppb (CAMx 
Metric 2); the highest daily average contribution is 6 ppb (CAMx Metric 
3).
    Frequency of Contribution: Indiana contributes at least 5 ppb to 26 
percent of the 1-hr exceedences (CAMx Metrics 1 and 2).
    Relative Amount: The total contribution from Indiana is equivalent 
to 4 percent of the total amount >=125 ppb in Baltimore (UAM-V Metric 
3); Indiana contributes 3 percent of the total manmade ppb >= 125 ppb 
in New York City (CAMx Metric 4; single highest episode percent 
contribution).
--Kentucky's Contribution to 1-Hour Nonattainment in Baltimore
    Magnitude of Contribution: The maximum contribution is 9 ppb (CAMx 
Metric 2); the highest daily average contribution is 8 ppb (CAMx Metric 
3).
    Frequency of Contribution: Kentucky contributes at least 5 ppb to 
24 percent of the 1-hr exceedences (CAMx Metrics 1 and 2).
    Relative Amount: The total contribution from Kentucky is equivalent 
to 3 percent of the total amount >=125 ppb in Baltimore (UAM-V Metric 
3); Kentucky contributes 5 percent of the total manmade ppb >= 125 ppb 
in Baltimore (CAMx Metric 4; single highest episode percent 
contribution).
--Maryland/District of Columbia/Delaware's Contribution to 1-Hour 
Nonattainment in New York City
Magnitude of Contribution: The maximum contribution is 50 ppb (CAMx 
Metric 2); the highest daily average contribution is 15 ppb (CAMx 
Metric 3).
    Frequency of Contribution: Maryland/District of Columbia/Delaware 
contribute at least 10 ppb to 14 percent of the 1-hr exceedences and at 
least 5 ppb to 38 percent of the 1-hr exceedences (CAMx Metrics 1 and 
2).
    Relative Amount: Maryland/District of Columbia/Delaware contribute 
5 percent of the total manmade ppb >= 125 ppb in New York City (CAMx 
Metric 4; 4-episode average percent contribution).
--Massachusetts' Contribution to 1-Hour Nonattainment in Portland, ME
    Magnitude of Contribution: The maximum contribution is 79 ppb (CAMx 
Metric 2); the highest daily average contribution is 67 ppb (CAMx 
Metric 3).
    Frequency of Contribution: Massachusetts contributes at least 10 
ppb to 100 percent of the 1-hr exceedences (UAM-V Metrics 1 and 2).
    Relative Amount: The total contribution from Massachusetts is 
equivalent to 100 percent of the total amount >=125 ppb in Portland, ME

[[Page 57397]]

(UAM-V Metric 3); Massachusetts contributes 56 percent of the total 
manmade ppb >= 125 ppb in Portland, ME (CAMx Metric 4; 4-episode 
average percent contribution).
--Michigan's Contribution to 1-Hour Nonattainment in Baltimore
    Magnitude of Contribution: The maximum contribution is 9 ppb (CAMx 
Metric 2); the highest daily average contribution is 8 ppb (CAMx Metric 
3).
    Frequency of Contribution: Michigan contributes at least 5 ppb to 7 
percent of the 1-hr exceedences (CAMx Metrics 1 and 2).
    Relative Amount: The total contribution from Michigan is equivalent 
to 5 percent of the total amount >=125 ppb in Baltimore (UAM-V Metric 
3); Michigan contributes 5 percent of the total manmade ppb >= 125 ppb 
in Baltimore (CAMx Metric 4; single highest episode percent 
contribution).
--Missouri's Contribution to 1-Hour Nonattainment over Lake Michigan
    Magnitude of Contribution: The maximum contribution is 19 ppb (CAMx 
Metric 2); the highest daily average contribution is 12 ppb (CAMx 
Metric 3).
    Frequency of Contribution: Missouri contributes at least 10 ppb to 
66 percent of the 1-hr exceedences (CAMx Metrics 1 and 2).
    Relative Amount: The total contribution from Missouri is equivalent 
to 22 percent of the total amount >=125 ppb over Lake Michigan (UAM-V 
Metric 3); Missouri contributes 9 percent of the total manmade ppb >= 
125 ppb over Lake Michigan (CAMx Metric 4; 4-episode average percent 
contribution).
--New Jersey's Contribution to 1-Hour Nonattainment in Western 
Massachusetts
    Magnitude of Contribution: The maximum contribution is 30 ppb (CAMx 
Metric 2); the highest daily average contribution is 23 ppb (CAMx 
Metric 3).
    Frequency of Contribution: New Jersey contributes at least 10 ppb 
to 100 percent of the 1-hr exceedences (CAMx Metrics 1 and 2).
    Relative Amount: New Jersey contributes 16 percent of the total 
manmade ppb >= 125 ppb in Western Massachusetts (CAMx Metric 4; 4-
episode average percent contribution).
--New York's Contribution to 1-Hour Nonattainment in Western 
Massachusetts
    Magnitude of Contribution: The maximum contribution is 25 ppb (CAMx 
Metric 2); the highest daily average contribution is 23 ppb (CAMx 
Metric 3).
    Frequency of Contribution: New York contributes at least 10 ppb to 
100 percent of the 1-hr exceedences (CAMx Metrics 1 and 2).
    Relative Amount: New York contributes 18 percent of the total 
manmade ppb >= 125 ppb in Western Massachusetts (CAMx Metric 4; 4-
episode average percent contribution).
--North Carolina's Contribution to 1-Hour Nonattainment in Philadelphia
    Magnitude of Contribution: The maximum contribution is 10 ppb (CAMx 
Metric 2); the highest daily average contribution is 9 ppb (CAMx Metric 
3).
    Frequency of Contribution: North Carolina contributes at least 2 
ppb to 4 percent of the 1-hr exceedences (UAM-V Metrics 1 and 2).
    Relative Amount: The total contribution from North Carolina is 
equivalent to 4 percent of the total amount >=125 ppb in Philadelphia 
(UAM-V Metric 3); North Carolina contributes 2 percent of the total 
manmade ppb >= 125 ppb in Philadelphia (CAMx Metric 4; single highest 
episode percent contribution).
--Ohio's Contribution to 1-Hour Nonattainment in Baltimore
    Magnitude of Contribution: The maximum contribution is 13 ppb (CAMx 
Metric 2); the highest daily average contribution is 12 ppb (CAMx 
Metric 3).
    Frequency of Contribution: Ohio contributes at least 5 ppb to 51 
percent of the 1-hr exceedences (CAMx Metrics 1 and 2).
    Relative Amount: The total contribution from Ohio is equivalent to 
11 percent of the total amount >=125 ppb in Baltimore (UAM-V Metric 3); 
Ohio contributes 4 percent of the total manmade ppb >= 125 ppb in 
Baltimore (CAMx Metric 4; 4-episode average percent contribution).
--Pennsylvania's Contribution to 1-Hour Nonattainment in Greater 
Connecticut
    Magnitude of Contribution: The maximum contribution is 28 ppb (CAMx 
Metric 2); the highest daily average contribution is 23 ppb (CAMx 
Metric 3).
    Frequency of Contribution: Pennsylvania contributes at least 10 ppb 
to 60 percent of the 1-hr exceedences and at least 5 ppb to 98 percent 
of the 1-hr exceedences (CAMx Metrics 1 and 2).
    Relative Amount: Pennsylvania contributes 10 percent of the total 
manmade ppb >= 125 ppb in Greater Connecticut (CAMx Metric 4; 4-episode 
average percent contribution).
--South Carolina's Contribution to 1-Hour Nonattainment in Atlanta
    Magnitude of Contribution: The maximum contribution is 24 ppb (CAMx 
Metric 2); the highest daily average contribution is 23 ppb (CAMx 
Metric 3).
    Frequency of Contribution: South Carolina contributes at least 5 
ppb to 6 percent of the 1-hr exceedences (UAM-V Metrics 1 and 2).
    Relative Amount: The total contribution from South Carolina is 
equivalent to 4 percent of the total amount >=125 ppb in Atlanta (UAM-V 
Metric 3); South Carolina contributes 2 percent of the total manmade 
ppb >= 125 ppb in Atlanta (CAMx Metric 4; single highest episode 
percent contribution).
--Tennessee's Contribution to 1-Hour Nonattainment Over Lake Michigan
    Magnitude of Contribution: The maximum contribution is 12 ppb (CAMx 
Metric 2); the highest daily average contribution is 11 ppb (CAMx 
Metric 3).
    Frequency of Contribution: Tennessee contributes at least 5 ppb to 
14 percent of the 1-hr exceedences (UAM-V Metrics 1 and 2).
    Relative Amount: The total contribution from Tennessee is 
equivalent to 6 percent of the total amount >=125 ppb over Lake 
Michigan (UAM-V Metric 3); Tennessee contributes 10 percent of the 
total manmade ppb >= 125 ppb over Lake Michigan (CAMx Metric 4; single 
highest episode percent contribution).
--Virginia's Contribution to 1-Hour Nonattainment in New York City
    Magnitude of Contribution: The maximum contribution is 25 ppb (CAMx 
Metric 2); the highest daily average contribution is 11 ppb (CAMx 
Metric 3).
    Frequency of Contribution: Virginia contributes at least 10 ppb to 
11 percent of the 1-hr exceedences and at least 5 ppb to 36 percent of 
the 1-hr exceedences (CAMx Metrics 1 and 2).
    Relative Amount: The total contribution from Virginia is equivalent 
to 11 percent of the total amount >=125 ppb in New York City (UAM-V 
Metric 3); Virginia contributes 4 percent of the

[[Page 57398]]

total manmade ppb >= 125 ppb in New York City (CAMx Metric 4; 4-episode 
average percent contribution).
--West Virginia's Contribution to 1-Hour Nonattainment in New York City
    Magnitude of Contribution: The maximum contribution is 14 ppb (CAMx 
Metric 2); the highest daily average contribution is 10 ppb (CAMx 
Metric 3).
    Frequency of Contribution: West Virginia contributes at least 5 ppb 
to 9 percent of the 1-hr exceedences and at least 2 ppb to 28 percent 
of the 1-hr exceedences (UAM-V Metrics 1 and 2).
    Relative Amount: The total contribution from West Virginia is 
equivalent to 9 percent of the total amount >=125 ppb in New York City 
(UAM-V Metric 3); West Virginia contributes 7 percent of the total 
manmade ppb >= 125 ppb in New York City (CAMx Metric 4; single highest 
episode percent contribution).
--Wisconsin's Contribution to 1-Hour Nonattainment Over Lake Michigan
    Magnitude of Contribution: The maximum contribution is 43 ppb (CAMx 
Metric 2); the highest daily average contribution is 8 ppb (CAMx Metric 
3).
    Frequency of Contribution: Wisconsin contributes at least 10 ppb to 
11 percent of the 1-hr exceedences (CAMx Metrics 1 and 2).
    Relative Amount: Wisconsin contributes 4 percent of the total 
manmade ppb >= 125 ppb over Lake Michigan (CAMx Metric 4; 4-episode 
average percent contribution).
    d. Conclusions From Air Quality Evaluation of Downwind 
Contributions. As indicated above, EPA is following a multi-step 
approach for determining whether emissions from an upwind State 
significantly contribute to nonattainment downwind. The first step 
involves an air quality evaluation to determine whether the air quality 
factors, and particularly the extent of the downwind contributions from 
emissions in the upwind State, indicate that those contributions are 
large and/or frequent enough to be of concern under the 1-hour and/or 
8-hour NAAQS. The second step, as described below, employs a cost-
effectiveness analysis to determine which of the upwind emissions may 
be eliminated through highly cost-effective controls. Any emissions 
that may be so eliminated are considered to be emissions that 
significantly contribute to nonattainment downwind. Finally, to confirm 
that the emissions considered to significantly contribute, taken as a 
whole, have a meaningful impact on nonattainment in downwind areas, EPA 
modeled the air quality effects of eliminating that amount of emissions 
(see Section IV, Air Quality Assessment, below).
    The EPA's conclusions from the first step in this process, the air 
quality evaluation, is that emissions from sources in each of the 23 
jurisdictions listed below make a significant contribution to 
nonattainment downwind for both the 1-hour and 8-hour NAAQS and 
interfere with maintenance of the 8-hour NAAQS. This determination was 
based on two independent sets of analyses, each of which EPA believes 
provides an independent basis for these conclusions. These two 
independent analyses are (1) subregional modeling using UAM-V, and (2) 
State-by-State modeling using CAMx and UAM-V. For the subregional 
modeling, EPA examined the frequency and magnitude of the impacts from 
each subregion along with State emissions data and other air quality 
information to evaluate the contributions from upwind States to 
nonattainment in downwind areas. For the UAM-V and CAMx State-by-State 
techniques, a number of measures of ozone contribution, or metrics, 
were used to assess, from several perspectives, the air quality effect 
of contributions from sources in different upwind States.
    The EPA weighed the results of its analysis of these several air 
quality metrics to determine which upwind States contain sources whose 
emissions contribute significantly to downwind nonattainment or 
maintenance problems. By examining the results of several air quality 
metrics, EPA assured that no one metric determined whether a State 
contains sources whose emissions contribute to downwind air quality 
problems. Rather, the determination of whether an upwind State 
contained sources whose emissions contribute significantly to a 
downwind nonattainment problem was based on the extent of the 
contributions reflected by multiple metrics. The EPA concluded that 
each set of modeling (i.e., subregional and State-by-State) when 
considered independently under EPA's weight-of-evidence approach 
provides a sound technical basis for finding that NOX 
emissions from sources in the following 23 jurisdictions make a 
significant contribution to nonattainment of the 1-hour and 8-hour 
NAAQS in, or interfere with maintenance of the 8-hour NAAQS by, one or 
more downwind States:

Alabama
Connecticut
Delaware
District of Columbia
Georgia
Illinois
Indiana
Kentucky
Maryland
Massachusetts
Michigan
Missouri
New Jersey
New York
North Carolina
Ohio
Pennsylvania
Rhode Island
South Carolina
Tennessee
Virginia
West Virginia
Wisconsin

    The remaining 15 OTAG States not covered by this final rule are 
discussed below.
5. States Not Covered by This Rulemaking
    In Section VI of the NPR, EPA proposed to find that emissions from 
sources in the following 15 States in the OTAG region do not 
significantly contribute to downwind nonattainment under the 1-hour or 
8-hour ozone NAAQS, or interfere with maintenance under the 8-hour 
NAAQS: Arkansas, Florida, Iowa, Kansas, Louisiana, Maine, Minnesota, 
Mississippi, North Dakota, Nebraska, New Hampshire, Oklahoma, South 
Dakota, Texas, Vermont (62 FR 60369). The EPA received comments on this 
section of the NPR and has recently conducted some additional CAMx 
analyses.50 The CAMx modeling suggested that further 
analysis using UAM-V State-by-State modeling would be warranted in 
order to have a set of information comparable to that for other States 
that are subject to this rule. In today's rulemaking, EPA is taking no 
action on whether emissions from sources in these 15 States do or do 
not contribute significantly to downwind nonattainment, or interfere 
with maintenance downwind, under either NAAQS. Thus, by today's 
rulemaking, EPA is not requiring these 15 States to submit SIP 
revisions providing for NOX emissions controls to meet a 
statewide NOX emissions budget; nor is EPA determining that 
these States will not be required to make these SIP submissions in the 
future. The EPA is continuing to review available information on the 
downwind impacts of these States, including comments submitted on the 
NPR. In addition, EPA plans to conduct State-by-State modeling to 
determine whether a SIP revision under section 110(a)(2)(D)(i)(I) 
should be required from any of these States in the future.

[[Page 57399]]

The EPA intends to begin this modeling in the fall of 1998.
---------------------------------------------------------------------------

    \50\ See ``Notice of Availability'' 63 FR 45032 (August 24, 
1998).
---------------------------------------------------------------------------

    As discussed in the NPR (62 FR 60318 at 60370), EPA reiterates that 
these 15 States may need to cooperate and coordinate SIP development 
activities with other States that are subject to today's action. Also, 
States with interstate nonattainment areas for the 1-hour standard and/
or the new 8-hour standard should cooperate in reducing emissions to 
mitigate local-scale interstate transport problems (e.g., transport 
from one State in a multi-state urban nonattainment area to another 
State in that area) to provide for attainment in the nonattainment area 
as a whole. The EPA encourages the 15 States to conduct additional 
analyses on ozone transport recommended by the OTAG Policy Group, which 
indicated that these States, ``* * * will, in cooperation with EPA, 
periodically review their emissions, and the impact of increases, on 
downwind nonattainment areas and, as appropriate, take steps necessary 
to reduce such impacts including appropriate control measures.'' 
51
---------------------------------------------------------------------------

    \51\ OTAG Recommendation: Utility NOX Controls, 
approved by the Policy Group, June 3, 1997.
---------------------------------------------------------------------------

    Comment: A number of commenters supported the proposal to exclude 
the proposed States, either in general or for specific States. Others 
opposed the proposal in general, or for specific States.
    Response: Because EPA is taking no action on the 15 States at this 
time, EPA will not respond to comments concerning these States at this 
time. As discussed above, EPA intends to continue to review ambient air 
quality data, air quality modeling results, and other technical 
information on the downwind contribution from all States not found to 
be significant contributors in today's action.
    Comment: Several commenters stated that if EPA revisits which 
States should be included in the rulemaking, EPA must reopen the public 
comment period.
    Response: The EPA agrees. Because today's action does not propose a 
change from the NPR concerning which States should be covered, no new 
comment period is needed at this time. As EPA noted in the NPR, if 
results from additional modeling and technical analyses indicate that 
States other than the 22 States (and the District of Columbia) that are 
the subject of today's action should be required to submit a SIP 
revision under section 110(a)(2)(D)(i)(I), EPA will publish a new NPR 
as to any such States and provide an additional comment period. As also 
stated in the NPR, in 2007, EPA will reassess transport in the full 
OTAG region to evaluate the effectiveness of the regional 
NOX measures and the need, if any, for additional regional 
controls.

D. Cost Effectiveness of Emissions Reductions

    As discussed above, in today's action, EPA considers control costs 
in determining whether, and the extent to which, upwind emissions 
contribute significantly to nonattainment, or interfere with 
maintenance downwind. The EPA considers cost factors in conjunction 
with other factors generally related to levels of emissions.
1. Sources Included In the Cost-Effectiveness Determination
    This subsection describes the rationale used to determine the cost 
effectiveness of emissions reductions measures. The EPA evaluates the 
relative costs of the available control measures using average cost 
effectiveness, measured as dollars per ton of NOX reduced 
relative to a baseline, to identify those emissions reductions that are 
``highly cost-effective.'' In performing this evaluation, EPA considers 
the cost savings of a regionwide NOX emissions trading 
system for large electricity generating boilers and turbines (i.e., 
boilers and turbines serving a generator larger than 25 MWe). As 
described in this section, EPA has determined that these emissions 
reductions are highly cost effective on a regionwide basis.
    To assure equity among the various source categories and the 
industries they represent, EPA considered the cost effectiveness of 
controls for each source category separately throughout the SIP call 
region. Sources are combined into a common source category if they 
serve the same general industry (e.g., boilers and turbines that are 
used by the electricity generation industry are combined in the same 
category). In general, this means that the sources in the same source 
category share the same six-digit source code classification (SCC). One 
exception is in the case of boilers and turbines which are combined and 
then separated into (1) a category of boilers and turbines serving 
generators that produce electricity for sale to the grid; or (2) a 
category of boilers and turbines that exclusively generate steam and/or 
mechanical work (e.g., provide energy to an industrial pump), or 
produce electricity primarily for internal use and not for sale. The 
EPA believes that this categorization better reflects the industrial 
sectors served.
    For each source category, the required emission levels (in tons per 
ozone season) were determined based on the application of 
NOX controls that achieve the greatest feasible emissions 
reduction while still falling within a cost-per-ton-reduced range that 
EPA considers to be highly cost-effective (hereinafter also referred to 
as ``highly cost-effective'' measures). Marginal or incremental costs 
of control are additional cost-effective measures that may provide 
important information about alternatives. In particular, incremental 
cost-effectiveness helps to identify whether a more stringent control 
option imposes much higher costs relative to the average cost per ton 
for further control. The use of an average cost-effectiveness measure 
may not fully reveal costly incremental requirements where control 
options achieve large reductions in emissions (relative to the 
baseline).
    In this rulemaking, EPA has chosen to focus on an average cost-
effectiveness measure in identifying highly cost-effective control 
options for several reasons. Since EPA's determination for the core 
group of sources is based on the adoption of a broad-based trading 
program, average cost-effectiveness serves as an adequate measure 
across sources because sources with high marginal costs will be able to 
take advantage of this program to lower their costs. In addition, 
average cost-effectiveness estimates are readily available for other 
recently adopted NOX control measures.
    The EPA examined a representative sample of potentially available 
controls. NOX controls for this rulemaking were considered 
highly cost-effective for the purposes of reducing ozone transport to 
the extent they achieve the greatest feasible emissions reduction but 
still cost no more than $2,000 per ton of ozone season NOX 
emissions removed (in 1990 dollars), on average, for each source 
category. The discussion below further describes the basis for this 
cost amount and the techniques used for each category. Many may 
consider certain controls that cost more than $2,000 per ton of 
NOX reduced to be reasonably cost-effective in reducing 
ozone transport or in achieving attainment with the ozone NAAQS in 
specific nonattainment areas; however, EPA has determined to focus 
today's rulemaking on only highly cost-effective reductions. In the 
future, as EPA continues to consider the impact of ozone transport and 
the most effective ways to assure downwind attainment, EPA may 
reconsider whether State NOX budget levels should be lowered 
to reflect application of additional controls

[[Page 57400]]

that, although more expensive, are nevertheless cost-effective. In 
addition, as discussed below, in determining whether to assume 
reductions from source categories with only a few sources or relatively 
small emissions, EPA considered administrative efficiency in developing 
conclusions about whether to assume emissions reductions for these 
sources.
    In determining the cost of NOX reductions by large 
electricity generating units (EGUs), EPA assumed an emissions trading 
system. As discussed in Section IV below, EPA evaluated and compared 
the likely air quality impacts of this rulemaking with and without a 
regionwide NOX emissions trading system for electricity 
generating sources. This analysis shows that a regionwide trading 
program causes no significant adverse air quality impacts. Because such 
a program would result in significant cost savings, EPA's cost-
effectiveness determination for large electricity generating boilers 
and turbines assumes that each State will adopt the lowest cost 
approach, i.e., the States will elect to include these sources in a 
regionwide NOX emissions trading program. However, States 
retain the option of choosing other, perhaps more expensive, approaches 
to achieving the necessary reductions. For non-EGU sources in the core 
group of the trading program, EPA used a least cost method which is 
equivalent to an assumption of an intrastate trading program. Inclusion 
of these sources in a regionwide trading program would provide further 
cost savings. For other source categories for which EPA identified 
highly cost-effective controls (i.e., internal combustion engines and 
cement manufacturing), EPA assumed source-specific controls. However, a 
State may choose to include such categories in the trading program and 
realize further cost savings.
    For the purposes of this rulemaking, EPA considers the following 
sizes of point sources to be large: (1) electricity generating boilers 
and turbines serving a generator greater than 25 MWe; or (2) other 
point sources with a heat input greater than 250 mmBtu/hr or which emit 
more than one ton of NOX per average summer day.
    In the NPR, EPA based the cost-effectiveness determination on 
NOX emissions controls that are available and of comparable 
cost to other recently undertaken or planned NOX measures. 
Table 1 provides a reference list of measures that EPA and States have 
recently undertaken to reduce NOX and their average annual 
costs per ton of NOX reduced. Most of these measures fall 
below $2,000 per ton. With few exceptions, the average cost-
effectiveness of these measures is representative of the average cost-
effectiveness of the types of controls EPA and States have needed to 
adopt most recently because their previous planning efforts have 
already taken advantage of opportunities for even cheaper controls. The 
EPA believes that the cost-effectiveness of measures that EPA 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 easily be implemented by upwind States whose 
emissions impact downwind nonattainment problems.

  Table 1.--Average Cost-effectiveness of NOX Control Measures Recently
                               Undertaken
                             [1990 dollars]
------------------------------------------------------------------------
                                                                   Cost
                                                                 per ton
                        Control measure                           of NOX
                                                                 Removed
------------------------------------------------------------------------
NOX RACT.......................................................  150-1,3
                                                                     00
Phase II Reformulated Gasoline.................................  \52\ 4,
                                                                    100
State Implementation of the Ozone Transport Commission
 Memorandum of Understanding...................................  950-1,6
                                                                     00
New Source Performance Standards for Fossil Steam Electric
 Generation Units..............................................   1,290
New Source Performance Standards for Industrial Boilers........  1,790
------------------------------------------------------------------------
\52\ Average cost representing the midpoint of $2,180 to $6,000 per ton.
  This cost represents the projected additional cost of complying with
  the Phase II RFG NOX standards, beyond the cost of complying with the
  other standards for Phase II RFG.

    The Federal Phase II RFG costs presented in Table 1 are not 
strictly comparable to the other costs cited in the table. Federal 
Phase II RFG will provide large VOC reductions in addition to 
NOX reductions. Federal RFG is required in nine cities with 
the nation's worst ozone nonattainment problems; other nonattainment 
areas have chosen to opt into the program as part of their attainment 
strategy. The mandated areas and those areas in the OTAG region that 
have chosen to opt into the program are areas where significant local 
reductions in ozone precursors are needed; such areas may value RFG's 
NOX and VOC reductions differently for their local ozone 
benefits than they would value NOX reductions from RFG or 
other programs for ozone transport benefits.
    Commenters on the proposal generally agreed with basing the cost-
effectiveness determination on the cost effectiveness of other recently 
undertaken measures. Therefore, EPA has considered controls with an 
average cost-effectiveness less than $2,000 per ton of NOX 
removed to be highly cost effective and has calculated the amounts of 
emissions that States must prohibit based on application of these 
controls. Some commenters believed that a more appropriate measure of 
cost effectiveness was incremental--instead of average--dollars per ton 
of NOX removed. Other commenters believed that a more 
appropriate measure was dollars per ppb of ozone removed from a 
nonattainment area. The EPA continues to depend on regionwide average 
dollars per ton of NOX removed when evaluating what control 
measures are highly cost-effective for the purposes of this rulemaking.
    Table 2 summarizes the control options investigated for each source 
category and the resulting average, regionwide cost effectiveness.

[[Page 57401]]



                          Table 2.--Average cost Effectiveness of Options Analyzed \53\
                                             [1990 dollars in 2007]
----------------------------------------------------------------------------------------------------------------
                Source category
----------------------------------------------------------------------------------------------------------------
                                                     Average Cost-effectiveness ($/ozone season ton) for each
                                                                          control option
                                               -----------------------------------------------------------------
Boilers and Turbines Generating Electricity...  0.20 lb/mmBtu.......  0.15 lb/mmBtu.......  0.12 lb/mmBtu.
                                                $1,263..............  $1,468..............  $1,760.
Boilers and Turbines not Generating             50% reduction.......  60% reduction.......  70% reduction.
 Electricity.
                                                $1,235..............  $1,467..............  $2,140.
Other Stationary Sources \54\.................  $3,000/ton maximum    $4,000/ton maximum    $5,000/ton maximum
                                                 per source.           per source.           per source.
Cement Manufacturing..........................  $1,458..............  $1,458..............  $1,458
Glass Manufacturing...........................  $2,020..............  $2,339..............  $4,758.
Incinerators..................................  $2,118..............  $2,118..............  $2,118.
Internal Combustion Engines...................  $1,213..............  $1,213..............  $1,215.
Process Heaters...............................  $2,860..............  $2,896..............  $2,896.
----------------------------------------------------------------------------------------------------------------
\53\ The cost-effectiveness values in Table 2 are regionwide averages. The cost-effectiveness values represent
  reductions beyond those required by Title IV or Title I RACT, where applicable.
\54\ For cement manufacturing, incinerators, internal combustion engines and process heaters, the table
  indicates that the same control technology (at the same cost) would be selected whether the cost ceiling for
  each source is $3,000, $4,000, or $5,000 per ton; thus the average cost-effectiveness number for these source
  categories is the same in each column. For glass manufacturing, the table indicates that additional emissions
  reductions would be obtained from more effective and more costly control technologies as the cost ceiling
  increase.

    The following discussion explains the controls determined by EPA to 
be highly cost-effective for each source category.
    The EPA has analyzed the implications of each State limiting 
trading within its borders compared to entering into a common trading 
program with all other States, provided that States choose to control 
EGUs at an average level of 0.15 lb/mmBtu. In the case of intrastate 
trading, EPA found that the average cost per ton of the resulting ozone 
season NOX reduction was about $1,499 per ton. This result 
from the IPM model was for all the States together considering changes 
in dispatch and other aspects of the future operation of the nation's 
power system. Individual State results were not provided by the model. 
As explained below, EPA expects that individual State cost per ton 
results are likely to be fairly close to this collective result.
    For a regionwide budget based on 0.15 lb/mmBtu, EPA's analyses 
suggest that whether (1) there were individual State trading programs, 
or (2) a single regionwide trading program, all States experienced a 
substantial reduction in summer NOX emissions from Base Case 
emissions levels. For this to occur, there have to be similar 
opportunities throughout the SIP call region for highly cost-effective 
reductions to occur at EGUs. If this were not true, EPA would have 
found, in the case where there is a single trading program across the 
entire SIP call region, that some States reduce a much greater share of 
their NOX emissions than other States do. The fact that 
there are similar opportunities for NOX reductions in each 
of the States indicates that if there were individual State trading 
programs in place they would each generally have an average cost 
effectiveness for reducing ozone season NOX emissions that 
is fairly close to the cost effectiveness of trading programs in other 
States. Therefore, each State is generally likely to have an average 
cost effectiveness of about $1,550 per ton, the amount we found in the 
results of the IPM model run for a scenario where each State ran its 
own trading program.
    a. Electricity Generating Boilers and Turbines. For EGUs larger 
than 25 MWe, the control level was determined by applying a uniform 
NOX emissions rate regionwide. The cost-effectiveness for 
each control level was determined using the IPM. Details regarding the 
methodologies used can be found in the Regulatory Impact Analysis of 
this rulemaking. Table 2 summarizes the control levels and resulting 
cost-effectiveness of three options analyzed.
    A regionwide level of 0.20 lb/mmBtu was rejected because though it 
resulted in an average cost effectiveness of less than $2,000 per ton, 
the air quality benefits were less than those for the 0.15 lb/mmBtu 
level which was also less than $2,000 per ton. The results suggest that 
a regionwide level of 0.15 lb/mmBtu should be assumed for this source 
category when calculating the amount of emissions that should be 
considered significant and therefore prohibited in each covered State. 
This control level has an average cost-effectiveness of $1,468 per 
ozone season ton removed. This amount is consistent with the range for 
cost-effectiveness that EPA has derived from recently adopted (or 
proposed to be adopted) control measures. As discussed later in this 
preamble, EPA has determined that EGU sources are fully capable of 
implementing this level of control by May 1, 2003.
    The EPA estimates that a control level based on 0.12 lb/mmBtu, has 
a cost effectiveness of $1,760 per ozone season ton removed, which is 
within the upper range of cost effectiveness. This estimate is based on 
the Agency's best estimates of several key assumptions on the 
performance of pollution control technologies and electricity 
generation requirements in the future which the Agency thoroughly 
researched over the last two years. Given that the cost per ton 
estimate for 0.12 lb/mmBtu trading is much closer to $2,000 than the 
0.15 lb/mmBtu trading, EPA is not as confident about the robustness of 
the results. Also, although EPA is very comfortable that a 0.15 lb/
mmBtu trading program beginning in 2003 will not lead to installation 
of SCR technology at a level and in a manner that will be difficult to 
implement or result in reliability problems for electric power 
generation, the Agency's level of comfort is not as high in considering 
0.12 lb/mmBtu-based trading.55 With a strong need to 
implement a program by 2003 that is recognized by the States as 
practical, necessary, and broadly accepted as highly cost effective, 
the Agency has decided to base the

[[Page 57402]]

emissions budgets for EGUs on a 0.15 lb/mmBtu trading level of control.
---------------------------------------------------------------------------

    \55\ For reasons explained in Section V., below, EPA has 
determined that May 1, 2003 is the earliest practicable date for 
achieving the level of emissions reductions EPA selected, and 
therefore is the appropriate date for achieving these reductions in 
light of the CAA's attainment date requirements.
---------------------------------------------------------------------------

    It should be noted that the cost-effectiveness values for EGUs were 
calculated using a slightly older version of the final EGU inventory. 
Changes made to the inventory and growth assumptions resulted in 
decreasing the final regionwide allowable emission level for EGUs, 
under the 0.15 option, to 543,825 tons per year from 563,785 tons per 
year. Reducing the allowable regionwide emissions increased the average 
cost-effectiveness value of the 0.15 option from $1,468/ton, to $1,503/
ton.
    b. Other Stationary Sources. The appropriate cost-effective control 
level for large non-EGU source categories was determined by evaluating 
various regulatory alternatives. For industrial boilers and turbines 
(i.e., boilers and turbines greater than 250 mm/Btu per hour or with 
NOX emissions greater than 1 tpd), the control level was 
determined by applying a uniform percent reduction regionwide in 
increments of 10 percent. For all other stationary sources, the control 
level was determined by applying source-category-specific cost-
effectiveness thresholds, because trading was not assumed to be readily 
available for these source categories. Details regarding the 
methodologies used are in the Regulatory Impact Analysis. Table 2 
summarizes the control levels and resulting cost-effectiveness for each 
option under each category.
    Further, for large non-EGUs, the cost-effectiveness determination 
includes estimates of the additional emissions monitoring costs that 
sources would incur in order to participate in a trading program. Some 
non-EGUs already monitor their emissions. In the NPR, EPA had not 
included monitoring costs in the cost-effectiveness determination 
because such costs had not been estimated at that time. Since then, EPA 
has evaluated monitoring system costs. These costs are defined in terms 
of dollars per ton of NOX removed so that they can be 
combined with the cost-effectiveness figures related to control costs. 
Since monitoring costs do not vary with the level of control, the cost 
per ton for monitoring varies in accordance with the amount of control 
being required. For purposes of this analysis, the level of control was 
assumed to be the level of control used to calculate the budget. 
Monitoring costs varied from about $150 to $400 per ton of 
NOX removed, depending on the type of source category.
    The EPA, therefore, determines that: (1) For large non-electricity-
generating industrial boilers and turbines, a control level 
corresponding to 60 percent reduction from baseline levels is highly 
cost-effective (this percent reduction corresponds to a regionwide 
control level of about 0.17 lb/mmBtu); and (2) for large internal 
combustion engines and cement manufacturing sources, a control level 
corresponding to the application of NOX reduction technology 
costing no more than $5,000/ton for each source is, on average, highly 
cost effective. As indicated in Table 2 and described in detail in the 
RIA, these control levels are associated with a cost effectiveness of 
approximately $1,467/ton for boilers and turbines, $1,458/ton for 
cement manufacturing, and $1,215/ton for internal combustion engines. 
This results in an average emissions reduction from uncontrolled 
emissions of 90 percent for internal combustion engines and 30 percent 
for cement manufacturing sources. The EPA notes that States may include 
these source categories in the model NOX budget trading 
program, further assuring that each source would be able to cost-
effectively meet its reduction requirements. The EPA determined that 
controlling glass manufacturing sources, incinerators, and process 
heaters was not highly cost-effective because all the options analyzed 
for these source categories cost more than $2,000 per ton of 
NOX removed. Thus, no additional controls are assumed for 
these sources when determining the significant amounts that must be 
reduced in each State.
2. Sources Not Included In the Cost-effectiveness Determination
    For the following groups of sources, EPA is determining that no 
additional control measures or levels of control should be assumed in 
this rulemaking, for the reasons described.
    a. Area Sources. In the NPR, EPA noted that control levels for area 
sources (i.e., sources other than mobile or point sources) could not be 
determined based on available information concerning applicable control 
technologies. Comments to the NPR did not identify specific 
NOX control technologies that were both technologically 
feasible and highly cost-effective. Because EPA has no new information 
on applicable control technologies for area sources, no additional 
control level is assumed for these sources in this rulemaking. Further 
discussion concerning area sources can be found in Section III, below, 
of this preamble.
    b. Small Point Sources. For the purposes of this rulemaking, EPA 
considers the following sizes of point sources to be small: (1) 
Electricity generating boilers and turbines serving a generator 25 MWe 
or less, and (2) other point sources with a heat input of 250 mmBtu/hr 
or less and which emit less than one ton of NOX per average 
summer day. In the NPR, EPA stated that the collective emissions from 
small sources were relatively small (in the context of this rulemaking) 
and the administrative burden, to the States and regulated entities, of 
controlling such sources was likely to be considerable. As a result, in 
the NPR, EPA proposed not to assume reductions from these sources in 
establishing the State budgets.
    Comments to the NPR did not identify specific approaches that would 
result in significant emission reductions and be administratively 
efficient in controlling these sources. On the contrary, many comments 
encouraged EPA to exclude small point sources from any budget 
calculations for this rulemaking.
    Therefore, in today's action, EPA is not assuming additional 
control levels for these sources. Further discussion concerning small 
point sources may be found in section III, below, of this preamble.
    c. Mobile Sources. In the NPR, EPA noted that it could not identify 
any additional NOX controls that States could implement for 
mobile or nonroad sources beyond those already reflected in the 
proposed State NOX budgets that were both technologically 
feasible and cost-effective, relative to point sources covered by this 
rule, for the purposes of reducing NOX. Several commenters 
stated that the EPA should require States to implement additional 
reductions for mobile sources. However, these commenters did not 
identify specific, new, technologically feasible mobile source 
NOX controls that were highly cost-effective by the 
standards of today's action. The EPA has re-examined the availability 
of mobile source control measures available to States, as discussed in 
more detail in sections III.D. and III.E. below, and has not identified 
any such controls that are both technologically feasible and highly 
cost-effective for NOX control. Therefore, the States' final 
NOX budgets promulgated in today's action do not assume 
implementation of additional highway or nonroad mobile source controls 
or expansion of existing controls beyond those described in the NPR. 
Further discussion concerning mobile sources, including the national 
measures EPA has assumed for purposes of today's rule, can be found in 
Section III, Determination of Budgets.
    d. Other stationary sources. The EPA does not assume, in this 
rulemaking, any additional control measures or

[[Page 57403]]

lower emissions levels for municipal waste combustors because these 
combustors are already being controlled through MACT regulations. 
Moreover, no additional control measures were assumed for source 
categories with relatively small NOX emissions (e.g., iron 
and steel mills, nitric acid manufacturing sources, space heaters, lime 
kilns, recovery plants, and engine test facilities). Further discussion 
concerning why controls were not assumed for these source categories 
may be found in Section III of this preamble.
    e. Conclusion. The above discussion described the controls for 
various source categories that EPA considers to be highly cost-
effective. The next step in the process is to determine the amounts of 
NOX emissions that would be eliminated by applying these 
highly cost-effective controls to the respective source categories. The 
EPA considers those emissions to be the amounts that contribute 
significantly to nonattainment in, or interfere with maintenance by, 
downwind States. By assuming that reductions of this magnitude should 
occur, EPA determined the resulting State-specific ``budget.'' Section 
III, Determination of Budgets describes the process EPA used to 
determine each State's budget and discusses comments received on the 
NPR.

E. Other Considerations

    As described above, EPA determined the amount of emissions that 
significantly contribute to downwind nonattainment from sources in a 
particular upwind State primarily by (i) evaluating, with respect to 
each upwind State, several air quality related factors, including 
determining that all emissions from the State have a sufficiently great 
impact downwind (in the context of the collective contribution nature 
of the ozone problem); and (ii) determining the amount of that State's 
emissions that can be eliminated through the application of cost-
effective controls. Before reaching a conclusion, EPA evaluated several 
secondary, and more general, considerations. These include:
     The consistency of the regional reductions with the 
attainment needs of the downwind areas with nonattainment problems
     The overall fairness of the control regimes required of 
the downwind and upwind areas, including the extent of the controls 
required or implemented by the downwind and upwind areas
     General cost considerations, including the relative cost-
effectiveness of additional downwind controls compared to upwind 
controls This section discusses these additional considerations.
1. Consistency of Regional Reductions With Attainment Needs of Downwind 
Areas
    a. General Discussion. Currently, air quality levels in the eastern 
part of the United States are above the 1-hour NAAQS in various, 
primarily urban, areas. Air quality levels are also above the 8-hour 
NAAQS in those same areas, as well as many others.
    The OTAG, and subsequently EPA, have conducted region-wide air 
quality modeling, using the UAM-V model, which shows that in 
approximately 20 primarily urban areas, the 1-hour nonattainment 
problem will persist by the year 2007, even after all of the controls 
specifically required under the CAA as well as Federal measures are 
implemented.56 This nonattainment problem that remains after 
implementation of those mandated controls may be termed ``residual 
nonattainment.'' For the 8-hour NAAQS modeling shows that under the 
same circumstances, at least one urban area that is linked to each 
upwind State will continue to experience residual nonattainment, and 
significantly more areas will be in nonattainment as well.
---------------------------------------------------------------------------

    \56\ As described elsewhere, the controls specifically required 
under the CAA include the controls identified in the modeling 
baseline, as well as certain Federal controls such as NLEV. These 
controls do not include any additional reductions that may be 
required in the local nonattainment areas as part of their 
attainment demonstrations.
---------------------------------------------------------------------------

    Further, as discussed above, OTAG's subregional modeling as well as 
EPA's CAMx modeling and State-by-State zero-out UAM-V modeling, 
indicate that upwind States contribute significantly to those downwind 
nonattainment problems under both standards. In general, under the 1-
hour standard, emissions from each upwind State affect at least 
several, primarily urban, nonattainment areas downwind. For example, 
each of the midwest/southern States of Ohio, Kentucky, Tennessee, West 
Virginia, Virginia, and North Carolina affects between five and eight 
downwind nonattainment areas. Under the 8-hour standard, emissions from 
each upwind State affect nonattainment problems that comprise an even 
larger geographic area. For example, Ohio, Kentucky, Tennessee, West 
Virginia, Virginia, and North Carolina each affect between eight to 
thirteen downwind States with nonattainment problems.
    As described in section IV below, EPA has conducted additional 
regionwide modeling which shows that upwind reductions comparable to 
those required under today's rule have an appreciable impact on 
downwind nonattainment problems under both NAAQS. The downwind impact 
from each individual upwind State's reductions may be relatively small, 
but the impact from all upwind reductions, collectively, is 
appreciable. This regionwide modeling-- which employs the UAM-V model 
relied upon by OTAG and also used by EPA for today's action--indicates 
that even after implementation of the regional reductions, which help 
downwind areas make progress toward attainment, certain downwind areas 
under the 1-hour NAAQS, and numerous downwind areas under the 8-hour 
NAAQS, will experience residual nonattainment. In addition, under the 
8-hour NAAQS, many other areas with nonattainment problems are expected 
to reach attainment based solely on the regional reductions.
    Furthermore, as mentioned earlier, the above-described modeling 
indicates no upwind States whose required regional reductions, in 
combination with the other regional reductions and CAA required 
controls, provide more ozone reduction than is necessary for every 
downwind nonattainment problem affected by that upwind State to attain 
under each NAAQS. That is, there is no instance of ``overkill,'' so 
that none of the upwind reductions required under today's action is 
more than necessary to ameliorate downwind nonattainment.
    b. 8-Hour Nonattainment Problems. As indicated above, the upwind 
reductions are useful in ameliorating downwind nonattainment under both 
NAAQS, but they are particularly useful in areas with nonattainment 
problems under the 8-hour NAAQS because more areas have such problems 
under that standard. Emissions reductions from each upwind State affect 
a broader swath of downwind 8-hour nonattainment problems, including 
problems adjacent to, and further away from, the upwind State. For 
example, emissions from Ohio affect nonattainment problems in each 
State adjacent to Ohio, as well as numerous States further away. As 
noted above, in some cases, the upwind reductions eliminate the 
downwind nonattainment problem; in other cases, those reductions 
ameliorate the downwind problem but residual nonattainment remains.
    Moreover, under the 8-hour NAAQS, upwind contributions tend to be a 
particularly large percentage of the downwind nonattainment problem. 
For example, along the Northeast corridor, cumulatively upwind States 
including adjacent States, contribute 83 percent of

[[Page 57404]]

Washington, DC's nonattainment problem; 68 percent of Maryland's 
nonattainment problem; 65 percent of Pennsylvania's nonattainment 
problem; and 85-88 percent of each of New Jersey's, New York's, 
Connecticut's, and Massachusett's nonattainment problems. These high 
levels of upwind contributions to widespread nonattainment problems--
both near to, and far from, the upwind State--indicate that the 
regional reductions from the upwind areas may be expected to be useful 
in ameliorating downwind nonattainment under the 8-hour NAAQS.
    c. Commenters' Concerns. Commenters argued that in the NPR that EPA 
failed to demonstrate that the proposed reductions in upwind emissions 
were necessary for downwind areas to demonstrate attainment. Commenters 
pointed out the lack of local attainment demonstrations under the 1-
hour NAAQS.57
---------------------------------------------------------------------------

    \57\ As noted in Section II.A., EPA proposed two analytical 
approaches, the second of which is the same as EPA is today 
promulgating. The commenters's criticisms seem to apply equally to 
both approaches.
---------------------------------------------------------------------------

    The EPA does not believe a local attainment demonstration is 
required before EPA can call on upwind States to reduce emissions 
pursuant to section 110(a)(2)(D). The EPA believes that available 
modeling analyses demonstrate that upwind reductions are necessary to 
help downwind areas come into attainment. The OTAG and EPA subregional 
modeling, UAM-V State-by-State zero-out modeling, and the CAMx 
modeling, described above, link each upwind State's emissions and 
downwind attainment needs, in a manner that is sufficient to support 
today's action. To reiterate, under the 1-hour NAAQS, the emissions 
reductions from each upwind State, combined with other emissions 
reductions, are needed to reduce downwind nonattainment problems. That 
need is underlined by the fact that the modeling relied on for today's 
action indicates residual nonattainment after implementation of all 
required controls and Federal measures. Even after implementation of 
the regional reductions, there is residual nonattainment for at least 
one downwind area linked to each upwind State. The same is true for the 
8-hour NAAQS, as noted above.
    The EPA recognizes that in the future, additional information may 
become available that would shed further light on the amount of 
emissions reductions needed for downwind areas to attain the NAAQS. 
Local-scale modeling may indicate more precisely the ambient impact of 
regional and local reductions on downwind nonattainment areas and the 
amount of any residual nonattainment. Nevertheless, it should be 
emphasized that the models relied on for today's action are state-of-
the-art, and that their various inputs--particularly the inventories--
have recently undergone close scrutiny and careful refinement through 
public comment and expert analysis. Accordingly, EPA believes that the 
overall model results indicating the general impact of upwind emissions 
and reductions in emissions should be viewed as valid. Accordingly, EPA 
believes that it has an adequate base of information to require the 
regional reductions under the 1-hour and 8-hour NAAQS at this time.
2. Equity Considerations
    The EPA believes further justification for today's action is 
provided by overall considerations of fairness related to the control 
regimes required of the downwind and upwind areas, including the extent 
of the controls required or implemented by those areas.
    The OTAG and EPA modeling analyses clearly indicate that upwind 
emissions contribute more than trivial amounts to downwind 
nonattainment problems. As a result, upwind emitters are exacerbating 
the health and welfare risks faced by those who live and work in 
downwind areas afflicted with unhealthful levels of ozone. The EPA 
believes that the principle of simple fairness applies here: upwind 
States should reduce their emissions that visit those health and 
welfare problems upon their downwind neighbors. Otherwise, their 
downwind neighbors would be obliged to pay additional costs to reduce 
local emissions beyond what would otherwise be necessary to protect 
their health from upwind emissions. In EPA's judgment, it is fair to 
require the upwind sources to reduce at least the portion of their 
emissions for which highly cost-effective controls are available. 
Indeed, fairness considerations would point towards requiring upwind 
reductions even if there were some degree of cost inefficiency.
    Further, it should be recognized that the major urban nonattainment 
areas have been required to incur control costs for ozone precursors 
since shortly after the 1970 CAA Amendments. In general, over the past 
quarter of a century, these areas have implemented SIP controls that, 
in combination with Federal measures, place ozone-related controls on 
virtually all portions of their inventory of ozone precursors, 
including VOCs as well as NOx. The Air Quality Modeling TSD 
includes descriptions of the control measures in place for several 
major urban nonattainment areas. Although not every major urban 
nonattainment area has complied with every CAA requirement for ozone 
precursors, the major urban nonattainment areas have complied with 
almost all of these requirements, and the CAA provides remedies to 
assure complete implementation of the required provisions. These 
measures have already lead to substantial reductions in ozone levels. 
By comparison, upwind States have not implemented reductions intended 
to reduce their impact on downwind nonattainment areas.
3. General Cost Considerations
    The EPA also generally considered the cost-effectiveness of 
additional local reductions in the 1-hour ozone nonattainment areas. 
The EPA conducted this analysis as part of its Regulatory Impact 
Analysis, completed under Executive Order 12866, for the rulemaking in 
which EPA revised the ozone NAAQS, 62 FR 38866 (July 18, 1997). The EPA 
surveyed the additional VOC and NOx controls available in 
areas throughout the country that are expected to be nonattainment 
under either NAAQS. The EPA ascertained that nationally, on average, 
these additional measures would cost approximately $4,300 per ton 
removed during the ozone season. See ``Control Measures Analysis of 
Ozone and PM Alternatives: Methodology and Results,'' July 17, 1997, 
table VII-2, p. 56. Although this figure is a national average, it 
provides a basis to conclude that local reductions may be expected to 
be more expensive than the approximately $1,500 in cost per ozone-
season ton removed for the regional NOx reductions required 
in today's rulemaking.
    Commenters criticized EPA's proposal to measure cost-effectiveness 
in terms of cost per ton of emissions removed because it did not take 
into account the ambient impact downwind of the emissions reductions. 
Commenters cautioned that under certain circumstances, a high level of 
emissions reductions upwind may result in high costs (even though cost-
effective on a per-ton basis), but relatively little ambient benefit 
downwind. Commenters emphasized that emissions reductions tend to have 
the greatest ambient benefit when they are within, or adjacent to, the 
area with the nonattainment problem. Commenters also said that 
emissions reductions further upwind have less ambient benefit. 
Accordingly, commenters stated that EPA's cost-effectiveness

[[Page 57405]]

justification did not support its proposed reduction requirements.
    The EPA acknowledges the concerns expressed by the commenters that 
focusing solely on the cost effectiveness, defined in terms of cost per 
ton removed, of the emissions reductions would exclude consideration of 
the total costs incurred by the upwind sources, and would exclude 
consideration of the downwind ambient benefits that those costs 
achieve, compared to the costs of achieving the same ambient impact 
through either local reductions or more extensive reductions in 
adjacent upwind areas. The EPA further acknowledges air quality 
modeling makes clear that reductions in emissions closer to the air 
quality problem have a greater ambient impact.
    However, EPA has not been presented with, nor been able to develop, 
an accurate comparison of the downwind costs of emissions reductions 
that would achieve the same ambient impact as the regional reductions 
required by today's action. The EPA does not have comprehensive 
information concerning available local measures or their costs or 
ambient impacts.
    However, as a qualitative matter, EPA believes that available 
evidence indicates that the upwind costs are reasonable not only in 
light of cost-effectiveness per ton removed, but also in light of the 
downwind ambient impact of the emissions reductions. Under the 1-hour 
NAAQS, emissions from each upwind State generally affect several 
downwind nonattainment urban areas. Thus, matching the total ambient 
impact of the emissions reductions from the upwind State would require 
emissions reductions in several downwind areas.58
---------------------------------------------------------------------------

    \58\ Although the reductions required of any one individual 
upwind State under today's rule may not, by themselves, result in 
large ambient impacts downwind, those reductions, when combined with 
reductions from other upwind States, do result in appreciable 
reductions downwind.
---------------------------------------------------------------------------

    Although presently available information does not permit a useful 
quantitative comparison of total upwind and downwind costs in terms of 
their ambient impact, EPA believes that upwind reductions replace local 
reductions that, on a cost-per-ton removed basis, may be expected to be 
more expensive. Moreover, it should be recognized that for all of the 
nonattainment areas under the 1-hour NAAQS, the residents have already 
incurred substantial control costs to eliminate part of the local 
contribution to the air quality problem. Under these circumstances, EPA 
considers it equitable to require the upwind emitters to offset their 
contribution to the problem through at least the reductions that are 
the most highly cost-effective--in terms of cost-per-ton removed--
rather than require the residents of the downwind area to offset those 
upwind contributions through even more local control measures.
    Furthermore, under the 8-hour NAAQS, the available information--
again, on a qualitative basis--indicates that the upwind emissions 
reductions replace a significantly greater set of local measures. As 
indicated above, emissions from each upwind State affect a wide swath 
of downwind areas with nonattainment problems. As a result, the 
emissions reductions from the upwind State replace local reductions in 
numerous downwind areas. Moreover, some of these downwind areas are 
adjacent to the upwind State, while others are further away. Thus, 
under the 8-hour NAAQS, EPA believes that the qualitative case is even 
more vivid that the upwind emissions reductions replace substantial and 
costly local measures.
    Finally, with respect to the meteorological phenomenon that upwind 
reductions have less ambient impact the further away they are from the 
downwind nonattainment problem: EPA modeled the ambient impact of 
regional variations in the levels of upwind emissions reductions. This 
modeling, and its results, are discussed in the Air Quality TSD. In 
brief, the modeling results indicate that it is neither more cost-
effective nor more beneficial to air quality to pursue subregional 
variations in upwind emissions controls.
4. Conclusion
    For the reasons discussed above, EPA believes that adequate 
information is available to determine, on a qualitative basis, that the 
upwind reductions required by today's action are reasonable in light of 
the attainment needs downwind, and that the costs of those reductions 
are reasonable in light of the costs the downwind areas would otherwise 
face. For these and other reasons noted elsewhere, EPA believes that 
requiring the regional reductions in today's notice is a reasonable 
step to take at this time.
    Of course, as more comprehensive information becomes available 
(including additional modeling, additional information concerning local 
control options and costs, as well as more refined regional air quality 
information), EPA will continue to examine the issue of regional 
transport. In addition, as described in Section III., EPA expects to 
review the issue of regional transport by the year 2007 and may require 
additional steps by either the upwind States or the downwind States, or 
both, to address the issue further. Even so, as noted above, the 
information that is available provides no evidence that the regional 
reductions required today may prove not to be needed.

III. Determination of Budgets

    The EPA used the highly cost-effective measures identified in 
Section II.D. above to calculate the amounts of emissions in each 
covered State that will contribute significantly to nonattainment or 
interfere with maintenance in one or more downwind States (the 
``significant amounts''). This Section further describes issues related 
to cost-effective controls and the role of these controls in the 
calculation of budgets.
    First, as described earlier in this notice, EPA projected the total 
amount of NOX emissions that sources in each covered State 
would emit, in light of expected growth, in 2007 taking into account 
measures required under the CAA (the ``2007 base year emissions 
inventory''). The EPA then projected the total amount of NOX 
emissions that each of those States would emit in 2007 if each such 
State applied these highly cost-effective measures (2007 controlled 
inventory). The difference between the 2007 base inventory and the 2007 
controlled inventory for each covered State is the ``significant 
amount'' that the State's SIP must prohibit to satisfy section 
110(a)(2)(D)(i)(I). Each covered State's 2007 controlled inventory--
referred to in this Section as the State's ``emissions budget''--
expresses the total amount of NOX emissions remaining after 
the State's SIP prohibits the ``significant amount'' of NOX 
emissions in that State. Each covered State must demonstrate that its 
SIP includes sufficient measures (of the State's choice) to eliminate 
those emissions, and thereby meet its budget, in the time frames 
discussed later in this notice.

A. General Comments on the Base Emission Inventory

    Background: In the NPR, EPA solicited comment on technical 
information used in revising the 1996 base year emissions inventories 
and the growth and control assumptions used to develop the 2007 
projection year base inventories. The EPA received over 200 comment 
letters (from industry, associations, States, environmental 
organizations, and U.S. Congressional representatives) on the condition 
of 1996 base year and projected 2007 emission inventories. The EPA 
accepted

[[Page 57406]]

proposed modifications to the extent EPA was able to validate them.
    As discussed in the NPR (62 FR 60318), EPA established a 120-day 
comment period (ending March 9, 1998) to address issues related to the 
proposed rule. In order to develop revised inventories used to 
recalculate the budgets for final rulemaking in a timely manner, EPA 
felt that comments received after the March 9, 1998 deadline would be 
addressed only if time and resources were available and after directing 
attention to comments received prior to the end of the comment period. 
The EPA is legally obligated under the Administrative Procedure Act to 
respond only to comments timely submitted during the public comment 
period. Response to comments timely submitted before the end of the 
comment period fulfills EPA's obligation to 5 U.S.C. 553(c).
    Although the Agency was not able to address all comments submitted 
after March 9, 1998, as discussed in Section III.F.5. of this notice, 
EPA is allowing commenters an additional opportunity to request 
revisions to the source-specific data used to establish each State's 
budget. During this time, EPA will be addressing those comments 
submitted during the NPR and SNPR comment periods which were not 
addressed for reasons indicated above, as well as evaluate comments 
that are submitted per Section III.F.5. of the NFR.
1. Quality
    Comment: Commenters suggested that the OTAG inventory may not be of 
sufficient quality for use in the modeling and budget determinations 
for the non-EGU point, area, nonroad mobile, and highway vehicle source 
sectors. The commenters stated that OTAG originally intended the 
inventories to be used in analyzing ozone transport mechanisms and the 
effect of possible control measures, not for establishing emission 
budgets as EPA has proposed. Additionally, as one commenter mentioned, 
many States had prepared inventories only for their moderate and above 
nonattainment areas, so that the remainder of the State's counties were 
supplemented with USEPA data. In contrast to these criticisms, other 
commenters supported the quality of the inventories and the procedures 
used in their development.
    Response: Under the initial OTAG inventory collection process, the 
37 States in the domain provided emission estimates for each entire 
State. The majority of the supplied data were 1990 State ozone SIP 
emission inventories, but some States supplied data from later years 
that reflected significant improvement over the 1990 data. 
Additionally, OTAG collected point source data from the States to 
update and revise existing emissions inventories used by OTAG. The 
result of these efforts was an improved emissions inventory which OTAG 
utilized for modeling as well as strategy analyses.
    The EPA used the final OTAG version of the inventory for the 
emission estimates in the NPR, and then improved the inventory with 
data supplied by the States and industry through the public comment 
period. As a result, the revised emissions inventory is the most 
accurate available for modeling, strategy analyses, and budget 
calculation purposes. The inventory has been through numerous versions, 
each version reviewed and extensively commented on by States, industry, 
and the public. These inventory data are more accurate than any other 
data used in the past as the basis for the various State-specific SIP 
revisions (such as rate-of progress SIP revisions or attainment 
demonstrations). The EPA considers it sufficiently accurate for 
purposes of determining the budgets.
    The EPA recognizes that emission inventories change as more 
accurate data or methods are developed for estimating emissions. For 
inventory changes that may be necessary after final promulgation of the 
budgets, EPA has a process for determining what changes need to be made 
as well as how the changes would be made to the inventories. This is 
discussed in further detail in Section III.F.5. of this notice.
    Comment: Several commenters were concerned that the initial State 
NOX emissions inventories submitted by the States were never 
quality-assured or commented upon by the States, the regulated 
community, or the public. Some commenters suggested the reevaluation of 
emissions estimates with State, local, and industry support.
    Response: Under the guidance of OTAG, the initial emission 
inventories submitted by the States were quality-assured by technical 
experts, including State and local emission inventory contacts, 
industry, EPA staff and contractors, and the OTAG Emission Inventory 
Technical Committee. As EPA amended and modified the inventory for use 
in the modeling for the NPR, SNPR, and the budget analyses, additional 
quality assurance was completed. The most accurate inventory 
development tools available at the time were used to validate these 
data and to quality assure emission calculations in these data bases. 
Existing data sets, including the NET data, the OTC NOX 
Baseline emission inventory, EPA'S AIRS/AFS major point source 
reporting system, and EPA's Emission Tracking System (ETS), which 
contains data submitted and certified as correct by the States, were 
used for comparison purposes. Where discrepancies were found, either 
before, during, or after the public comment period, States and industry 
were contacted to clarify and support revised emission estimates.
2. Availability
    Comment: Commenters asserted that the emissions inventory used for 
the SIP modeling and budget calculations were not made available for 
public review along with the proposed rule. One commenter stated that 
the emissions inventory that forms the basis for the NPR (the SIP Call 
inventory) did not become available until the first week in February 
1998.
    Response: On October 10, 1997, EPA posted emissions data on the TTN 
for use and review during the public comment period (See NPR, 60318). 
These data, in conjunction with the OTAG inventories, were the basis of 
the initial proposed budgets and modeling analyses in the NPR. Thus, 
these data were available to the public before the beginning of the 
120-day comment period on the NPR, which allowed ample time to develop 
budget, modeling, and cost analyses for submission during the comment 
period. By notice dated January 28, 1998 (63 FR 4206), EPA issued a 
caution that comments on the inventory must be submitted by the March 
9, 1998 close-of-public-comment date, so that EPA could finalize the 
inventories and use them for further analyses.
    On February 3, 1998, in response to initial public comments and 
internal review of the initially released data, draft amendments to the 
emissions inventory were posted on the EPA's TTN site. These changes 
included the addition of EGU sources less than or equal to 25 MWe which 
were excluded from the initial budget calculation, correction of EGU 
growth factors, and the reclassification to the non-EGU file of some 
sources previously erroneously identified by OTAG as EGU sources. 
Erroneously omitted non-EGU point source records were also added to the 
emissions inventory. Area, highway, and nonroad mobile source 
information was not modified in this iteration. By posting this data on 
February 3, 1998, EPA allowed 5 more weeks for public comment on the 
revised data, until the conclusion of the comment period for inventory 
data on March 9, 1998. Because the revisions were fairly minor, EPA 
believes this amount of time was adequate. The EPA did receive

[[Page 57407]]

comments by March 9, 1998 on the revised data it had posted on February 
3, 1998.

B. Electricity Generating Units (EGUs)

    Background: To determine the budget for each State's electricity 
generating sector, EPA developed an inventory of baseline heat input 
(mmBtu) and NOX emissions (tons/season) data for each unit. 
In the NPR, EPA proposed to use the higher, by State, of 1995 or 1996 
heat input data to calculate baseline heat input rates (62 FR 60352). 
The EPA maintained this approach for the SNPR, but added 577 smaller 
units to the State budget inventories, which had erroneously been 
omitted for the NPR. These units included electricity generating 
sources of 25 megawatts of electrical output (MWe) or smaller and 
additional units not affected under the Acid Rain Program.
1. Base Inventory
    Comment: Commenters suggested that using the higher of 1995 or 1996 
utilization rates for setting the baseline for the EGU portion of the 
budget may not be appropriate in all instances. In general, commenters 
argued for various degrees of flexibility in choosing the baseline 
year(s) to be used for calculation of budgets.
    Response: As discussed below, EPA has made corrections to the 
baseline heat input data for a small number of EGUs based on careful 
review of the data supplied with source-specific comments. Using 1997 
CEMS data is not a practical option because EPA has not had time to 
extract from the Acid Rain Emissions Tracking System (ETS) the 5-month 
ozone season heat input values, quality assure them, or publish them. 
(Although EPA's Acid Rain Program intends to publish its 1997 Emissions 
Scorecard later in 1998, this publication will contain only annual, not 
ozone season, data.) Accordingly, EPA has finalized the EGU portion of 
the budget for each State using the higher of the 1995 or 1996 ozone 
season heat input values.
    Comment: Commenters asserted revisions were needed to the published 
heat input data for some EGUs and proposed related additional source-
specific changes. Commenters on this issue stated that inaccurate 
calculations of heat input data resulted in significant errors in the 
Statewide budgets. Several suggested the need for revision before 
calculation of final budgets. Many of these commenters provided 
specific data that they urged EPA to use in the final budget setting 
process.
    Response: The EPA has analyzed the data submitted by these 
commenters and, where warranted, has made the requested adjustments. 
Approximately 200 corrections were made to the baseline heat input data 
for EGU sector inventories.
    Comment: Commenters also noted the need to further correct, for 
some States, the listing of units in the electricity generating sector 
inventory. Commenters listed specific EGUs that EPA should either 
include or remove from the inventory, or for which EPA should correct 
applicable baseline data (e.g., capacity, operating parameters). 
Several commenters argued that substantial revision of the inventory 
was necessary before setting budgets under the final rulemaking.
    Response: The EPA has analyzed the data submitted by these 
commenters, including following up with commenters when needed to 
assure proper interpretation of the data. Where warranted, EPA has 
corrected the State inventories of units and applicable baseline data.
    While the vast majority of corrections consisted of adding small 
units (e.g., municipal generators and peaking diesel units), combustion 
turbines, and independent power producers not affected under the Acid 
Rain Program, some involved deleting units that are no longer 
operational or have been misclassified and, in actuality, are 
industrial non-electricity generating boilers. The net result is that 
EPA has added approximately 800 units to the State EGU inventories. The 
EPA believes that these inventories are sufficiently accurate to 
develop a budget.
    Comment: Commenters suggested types and sizes of sources to include 
or exclude from the electricity generating sector inventory. As to the 
sizes of sources to include in the inventory, commenters on the NPR 
were roughly split on the inclusion of units less than or equal to 25 
MWe. Several noted that emissions from sources below this level were 
negligible and should not be included. One commenter noted, however, 
that these sources should be included in the final budget because they 
tend to operate on peak demand days which frequently correspond to high 
ozone days. Several suggested that 15 MWe be the cutoff for the utility 
component of the budget.
    On a separate concern, a few commenters disagreed with the 
inclusion of non-utility power generators in the utility list of 
sources and proposed that they be included with industrial non-
electricity generating unit sources.
    Response: Many of these comments appear to confuse discussions of 
other related issues (e.g., core sources for NOX cap and 
trade rule, appropriate sources for cost-effective control) with the 
types and sizes of EGUs to be included in the baseline inventory for 
setting the budget. All emissions should be included in the base 
inventory and, thus, in the budget. As noted previously, using 
information supplied by commenters, EPA has agreed to add many small 
units to the base inventories of several States. Concurrently, EPA has 
also decided not to classify EGUs less than or equal to 25MWe as core 
sources for the trading program, as discussed in Section VII of this 
notice, or to assume an emissions decrease for these small units 
(``cutoff level'') as part of Statewide budgets for EGUs.
    The EPA maintains its decision to include industrial units that 
generate electricity in the definition of EGUs is entirely consistent 
with the changing, more competitive, character of today's electric 
power generation industry in the US. Also, these units are amenable to 
the same NOX control technologies, at generally the same 
cost-effectiveness, as utility units.
2. Growth
    Background: In the NPR and SNPR, EPA used forecasts of future 
electricity generation to apply State-specific growth factors in 
calculating the emissions budgets for the electricity generating 
sector. In the SNPR, EPA revised the growth factors (the ``corrected'' 
projections) to account for projected new combustion turbine and 
combined cycle units inadvertently excluded in the analysis developed 
in support of the NPR. The EPA also discussed in the SNPR that 
``revised'' electricity generation projections could lead to lower 
growth rates, and therefore lower budgets, and placed supporting 
information in the docket. However, EPA proposed to use the 
``corrected'' projections in calculating State budgets to provide 
additional compliance flexibility to sources and States (63 FR 25905).
    a. Growth Rates.
    Comment: The EPA received approximately 36 comments in response to 
the NPR and roughly 28 comments in response to the SNPR regarding the 
estimated growth rates that were used to determine the NOX 
budget for each State. These comments were submitted by State agencies, 
associations, utilities, and a public interest group. Commenters 
expressed concern regarding a number of specific issues, including the 
following:
    (i) the appropriateness of using growth factors to determine the 
NOX budget,

[[Page 57408]]

    (ii) use of the IPM model to establish the growth factors for each 
State, and
    (iii) the use of the ``corrected'' instead of the ``revised'' 
projections.
    Some of these commenters opposed growth factors generally, but many 
of them supported the concept of--but not the method proposed for--
applying a growth factor.
    Response: The OTAG's technical analyses of NOX emissions 
suggested that EPA needed to consider the electric power industry's 
future growth in determining the amount of NOX reduction 
that would be reasonable for the power industry to make in the future. 
The OTAG factored the growth of the power industry's emissions from 
1990 to 2007 into the air quality analysis that it performed. The 
results of this analysis were the basis of its recommendations to EPA 
to lower NOX emissions from the power industry in many 
Eastern States. Because the Agency made its predictions about 
attainment in 2007 based on projections of emissions considering 
growth, rather than on historical emissions, the Agency also believes 
that the State budgets to be used up to 2007 should account for growth 
in electricity demand. Not accounting for growth in demand for 
electricity would require States to reduce emissions below the level 
that EPA predicted was necessary to reach attainment. By accounting for 
growth through 2007 and applying that growth beginning in 2003, EPA 
essentially allows sources to emit at a slightly higher level than 0.15 
lb/mmBtu in the years 2003 through 2006.
    In today's action, the Agency has determined to continue to 
incorporate growth out to 2007 in developing State budgets for summer 
NOX emissions. Not accounting for growth would mean that 
additional control measures--to offset growth--would be required, and 
EPA has not determined that those additional control measures would be 
cost-effective. In considering growth, EPA has determined to continue 
to use either 1995 or 1996 State-wide heat input data, for whichever 
year was higher for units over 25 megawatts that burn fossil fuels for 
baseline data. (More details on this approach can be found above in 
Section III.B.1. Base Inventory).
    To estimate growth, EPA considered several options. Ultimately, the 
Agency has decided to use State-specific growth factors derived from 
application of the Integrated Planning Model (IPM) using the 1998 Base 
Case 59 (also referred to as the ``revised'' growth 
factors). This is the same Base Case used for the Regulatory Analysis 
in support of the SNPR. The reasons for using these data are discussed 
below under ``Use of IPM.''
---------------------------------------------------------------------------

    \59\ The Base Case is the condition of the industry in the 
absence of the SIP call.
---------------------------------------------------------------------------

    b. Use of IPM.
    Comment: Many commenters questioned whether use of the IPM model 
was appropriate to derive accurate State-specific growth factors. 
Commenters expressed concern that there was too much variation between 
each State's individual growth rate as determined by the IPM model, and 
suggested that use of region-wide IPM growth factors may be more 
appropriate. They also questioned the reliability and accuracy of the 
IPM model, especially as applied on an individual State basis. A number 
of commenters stated that EPA's growth projections were lower than 
growth rates projected in the context of State utility planning 
efforts. Several commenters suggested that EPA base its growth rates on 
projections other than OTAG, or EPA's IPM forecasts; they especially 
urged the Agency to consider individual State-prepared forecasts. This 
was to avoid problems that commenters believe exist in EPA's use of the 
IPM model for forecasting electricity generation in various areas of 
the country. Specific concerns focused on:
    (i) the effect of IPM projections and associated NOX 
budgets on future growth within each State, and
    (ii) how the IPM model accounts for:

--planned nuclear unit retirements,
--the impact of a deregulated utility marketplace, and
--improvements in energy efficiency and control technology.

    Many commenters also generally expressed concern that there is 
insufficient information or documentation on how EPA used the IPM model 
to determine growth factors.
    Many commenters asserted that EPA should not incorporate the growth 
factors into the budget calculation process. These commenters argued 
that adding growth to baseline activity and subsequently applying 
controls reduces the stringency of the standards, and introduces an 
unacceptable level of uncertainty. They suggested that the budgets 
should be based on historic utilization rates, and that States could 
then determine how to allocate their budgets to provide for growth. 
These commenters recommended that, if a growth factor must be used, 
then EPA should apply a uniform growth rate region-wide to determine 
the NOX budget for each State.
    Response: The EPA initially considered using the OTAG growth rates, 
but found that they were largely based on past, State-specific 
generation trends and did not factor in the more competitive electric 
power market where electricity will be increasingly moving between 
regions in response to the cost of producing electricity. The Agency 
also found that there were several other major limitations that were 
described in the NPR. (62 FR 60352-60353).
    The Agency considered setting the State NOX budgets 
based on past generation levels in States, but this approach also does 
not consider how competition in the industry in the future will alter 
electricity generation practices. It ignores growth and shifts in 
production altogether. A variant of this approach, suggested by several 
commenters, would be to use a uniform growth factor for all States 
based on some projection of future growth through the 23 jurisdictions 
covered by this rule. This approach appears even-handed, but EPA views 
it as unfair and inaccurate with respect to States in which:
    (i) utilities are particularly economical to operate, and
    (ii) the generation of power by these firms is expected to grow at 
a rate greater than average.
    Another similar alternative suggested in the public comments was 
that EPA use a uniform growth factor for all States in the same region, 
e.g., the North American Electricity Reliability Council (NERC) 
regions, or subregions. The problem with this approach is, again, that 
certain States within the same region are expected to vary in their 
rate of growth, given differences in their electric utilities. The fact 
that some States are in several NERC regions also makes this approach 
less practical.
    The Agency looked at several well-recognized forecasts of regional 
electricity generation growth, such as those provided by NERC, the 
Annual Energy Outlook of the Energy Information Administration (EIA), 
and Data Resources Incorporated's (DRI) World Energy Service U.S. 
Outlook. None of these modeling systems provides results at the State 
level. Therefore, the Agency would have to develop ways to apportion 
these regional predictions to States. The EPA knows of no way to 
apportion these regional values to States that would resolve the 
concerns expressed by commenters. Furthermore, the Agency uses the 
growth rates from IPM to calculate the cost-effectiveness of 
NOX emission reductions, as well as to determine 
NOX budgets for States. Therefore, using growth rates that 
are not from IPM would lead the Agency to using one set of State-
specific

[[Page 57409]]

generation estimates to develop NOX budgets and a different 
set of State-specific generation estimates for determining cost-
effectiveness. As a result, EPA's evaluations of future activities of 
the power industry might not be considered consistent. Finally, 
although each of these sources provides reasonable electricity 
generation forecasts, each of the forecasts could be criticized for the 
assumptions they make in a manner similar to the way commenters have 
criticized growth factors from IPM.
    Some commenters suggested that the Agency use individual State 
forecasts instead of IPM forecasts, including projections used for 
State utility planning efforts. The EPA rejected this type of approach 
for two reasons. First, nothing in the comments suggested to EPA that 
the State forecasts are more accurate or more reliable than the IPM 
forecasts. Instead, the State forecasts varied State by State in the 
way they predicted future electricity generation. Adoption of these 
forecasts could result in inconsistencies in setting the State budgets. 
Electricity generation forecasts require making many technical 
assumptions which, admittedly, lead to some uncertainty in the results. 
Accordingly, the Agency believes that the fairest way to determine 
emissions budgets is to handle these assumptions in a consistent way 
for all of the States, as long as a reasonable approach and reasonable 
modeling assumptions are used.
    Therefore, EPA has decided to use the IPM 1998 Base Case emissions 
forecast for deciding State NOX budgets in today's action. 
The Agency finds it to be the fairest and most reliable overall 
approach to estimating growth factors. It deals consistently with the 
technical assumptions that occur in energy forecasting and employs a 
reasonable set of assumptions in the process of making a forecast. As 
an added advantage, it has undergone considerable review by the 
electric power industry over the last two years, and the industry was 
aware that it might be applied as it is in today's rulemaking. Finally, 
EPA's use of IPM for forecasting State growth rates provides for 
overall consistency in forecasting future emissions and estimating the 
cost-effectiveness of reductions in this rulemaking.
    The EPA believes that IPM provides a reasonable forecast of State 
growth rates because it carefully takes into account the most important 
determinants of electricity generation growth that are facing the power 
industry today. These major factors include: regional demands for 
electricity, the impacts of wholesale competition that lead to changes 
in market share for various utilities, changes in fossil fuel prices, 
expected improvements in electricity generation technology, costs of 
emission control technology, expected changes in generation unit 
operations and regional dispatch practices to lower production costs, 
nuclear unit retirements, alteration in planning reserve margins to 
meet peak demand, and limitations in moving power between regions due 
to transmission constraints.
    An explanation of how EPA uses IPM to address these issues and 
other important factors is included in EPA's Analyzing Electric Power 
Generation under the CAAA, March 1998 (Docket no. V-C-3). Because EPA's 
assumptions have been reviewed by the public over the last two years 
and the Agency has worked with EIA and other groups to improve them in 
response to comments and new information, the Agency believes that it 
has made reasonable assumptions for a Base Case forecast of electric 
power generation.
    c. Use of ``Corrected'' Growth Rates.
    Comment: Some comments on the SNPR expressed concern that the new 
``corrected'' growth factors are artificially inflated and will 
compromise efforts to improve air quality throughout the region. Some 
of the commenters suggested that States should have the flexibility to 
determine how to manage emissions from new sources in the context of 
the original growth factors and NOX budgets proposed in the 
NPR. Some of these commenters also stated that it was unclear why EPA 
chose to use the ``revised'' projections in its cost analysis but 
retained the ``corrected'' growth factors in its budget calculations. 
Other commenters, however, were supportive of the new growth factors 
and the use of the ``corrected'' projections. Finally, several 
commenters requested that EPA further explain how the ``corrected'' 
growth factors were derived and subsequently used to generate the 
NOX budgets.
    Response: In the NPR, EPA proposed a set of growth factors based 
upon the 1996 IPM Base Case forecast. In the SNPR, EPA corrected the 
growth factors used in calculating State budgets to account for new 
generation that had inadvertently been left out of the original 
calculations (the ``corrected'' growth factors). On the basis of 
comments that EPA has received on its assumptions for forecasting 
electricity generation throughout the country during the last year, the 
Agency revised a set of key assumptions at the beginning of 1998. These 
assumptions lead to a better projection of electricity generation 
nationally, by region, and by State. Therefore, the Agency has decided 
to use the 1998 IPM Base Case forecast over the 1996 IPM Base Case 
forecast as the basis for its ``revised'' State growth estimates.
    The recent important changes that were incorporated into EPA's use 
of IPM in 1998 include using the most recent NERC estimate of regional 
electricity demand; the latest available EIA and NERC generation unit 
data; updated fuel forecasts; updated assumptions on nuclear, 
hydroelectric, and import assumptions (with special attention to 
differences in summer use); and an increase in the level of detail in 
the model to more accurately capture the transmission constraints that 
exist for moving power between various regions of the country. The 
Agency also updated its assumptions on the size and operation of all 
electricity generation units of utilities and independent power 
producers (with special attention to cogenerators) and updated its 
assumptions on planning reserve margins and the costs of building new 
generation capacity. For this, the Agency relied heavily on information 
compiled from utilities by NERC and the EIA. Each of these agencies has 
regular contact with the power industry and has its data reviewed by 
the power industry. Again, details on these improvements in IPM can be 
found in EPA's Analyzing Electric Power Generation under the CAAA, 
March 1998 (Docket no. V-C-3).
    In the SNPR, EPA used the ``revised'' growth factors in the IPM 
model in its cost analysis but used the higher, ``corrected'' growth 
factors to calculate State budgets. The EPA proposed the higher growth 
factors because the Agency believed that this results in less cost and 
more flexibility for sources to achieve their budget reductions 
beginning in 2003. However, some commenters pointed out that EPA had 
provided sufficient flexibility by accounting for growth to the year 
2007 and applying that growth estimate beginning in 2003. These 
commenters remarked that it was not necessary to add further 
flexibility by using the higher, but less current and less accurate, 
``corrected'' growth rates. They also stated that EPA should use the 
most up-to-date information available. The EPA agrees and is using the 
``revised'' growth rates based upon the 1998 IPM Base Case forecast to 
calculate the State budgets used in today's final rule.
3. Budget Calculation
    a. Input vs. Output.
    Background: In the SNPR, the component of each State's budget 
assigned to electricity generation was determined using the State's 
total heat

[[Page 57410]]

input, applicable emission rate (0.15 lb/mmBtu), and projected growth 
in total heat input to 2007. The Agency solicited comment on an 
alternative approach to calculating the State's budget using each 
State's share of the 23 jurisdiction electricity generation (electrical 
output). The SNPR describes in detail the output-based approach, and 
its possible benefits as advanced by its proponents (63 FR 25907). The 
Agency asked for comments on the appropriateness, legality, rationale, 
and methodology for incorporating the output-based approach when 
calculating the electricity generation component of each State's 
budget.
    Comments: The Agency received comments both supporting and opposing 
output-based State budgets. Supporters of output-based budgets 
asserted:
     An output-based budget would promote competition among 
different types of electricity providers on an equal basis in a 
deregulated electric utility industry.
     An output-based budget would promote CO2, 
mercury, SO2 and off-season NOX reductions beyond 
what would occur under a system that assigns State budgets based upon 
input.
     An output-based budget may result in more cost-effective 
NOX reductions.
     Issuing output-based budgets is legally permissible.
    The commenters opposed to output-based State budgets objected to 
the allocation of allowances to non-NOX-emitting units, such 
as nuclear, hydroelectric, solar, or geothermal power plants. They 
claimed that this would make compliance more difficult and more costly 
for fossil-fuel burning sources because fewer allowances would be 
allocated to them.
    Commenters opposed to output-based budgets also claimed that:
     Output-based budgets would not necessarily improve energy 
efficiency compared to existing incentives, such as fuel costs.
     The output-based State budgets may not result in the same 
geographic distribution of emissions as would occur under the original 
budget allocation.
     There could be significant administrative problems with 
changing the basis of the State budgets.
    In addition, some commenters, though in general supporting 
allocations by output, specifically objected to allocating allowances 
to nuclear-powered units because they believed that this method would 
encourage nuclear-powered electrical generation, which, they further 
believed, would have adverse ancillary impacts on the environment.
    The Agency received additional comments on the method of allocating 
State budgets to sources. Further discussion of these comments can be 
found in Section VI.C.2 of this preamble.
    Response: The EPA has an extensive history of promoting the 
efficient use of natural resources, particularly energy, through both 
voluntary and regulatory measures. Key emissions standards, such as the 
standards for new vehicles and the recently promulgated new source 
performance standards to new power plants, are written as output-based 
fuel-neutral performance standards that promote the efficient use of 
energy. The EPA has begun to work with States to find mechanisms to 
more directly credit the use of energy efficiency measures in SIP. The 
EPA also has a number of programs that encourage the use of energy 
efficient technologies by providing energy users, particularly in the 
residential, commercial and industrial sectors, with information on the 
economic and environmental benefits of such technologies.
    Although the Agency has concluded, for the reasons stated below, 
that heat-input-based budgets to States are more appropriate at this 
time, the EPA intends to work with stakeholders to overcome existing 
obstacles and to design an output allocation system that could be used 
by States as part of their trading program rules in their SIPs and by 
EPA in future allocations to States.
    The EPA considered how State NOX budgets would be 
changed using the output approaches suggested by the commenters. The 
EPA revised its State budget calculations using available electrical 
generation data from the EIA for utility and non-utility generators for 
the higher electrical generation output of either 1995 or 1996, by 
State. In Table III-1 below, Column 2 presents the proposed budgets 
based upon heat input. Column 3 presents the revised budgets based upon 
heat input and the revised growth factors. Column 4 shows output-based 
budgets, based upon all electrical generation. Some commenters 
suggested including fossil-fuel and renewable energy source 
generation--including hydroelectric, solar, wind, and geothermal 
generation--but not nuclear generation. These are included in Column 5. 
One commenter suggested using electrical generation from fossil-fuel 
only, which is included in Column 6.

                               Table III-1.--State Budgets by Energy Source Basis
                                        (Higher of 1995 or 1996 EIA data]
----------------------------------------------------------------------------------------------------------------
            Column 1                 Column 2        Column 3        Column 4        Column 5        Column 6
----------------------------------------------------------------------------------------------------------------
                                  Proposed input- Revised input-                   Output-based
                                   based budgets   based budgets   Output-based    budgets--all    Output-based
              State                fossil fuel-    fossil fuel-     budgets all     generation    budgets fossil
                                      burning         burning       generation    sources except   fuel-burning
                                    generators      generators        sources         nuclear       generators
----------------------------------------------------------------------------------------------------------------
Alabama.........................           30644           29026           34832           35068           32744
Connecticut.....................            5245            2583            7677            5156            4456
Delaware........................            4994            3523            2392            3214            3417
District of Columbia............             152             207             100             133             142
Georgia.........................           32433           30255           32223           31713           30819
Illinois........................           36570           32045           44253           27888           29602
Indiana.........................           51818           49020           32212           43285           45831
Kentucky........................           38775           34923           24847           33389           34166
Maryland........................           12971           15033           13284           12969           13212
Massachusetts...................           14651           14780           11017           13248           13496
Michigan........................           29458           28165           32275           32037           32457
Missouri........................           26450           23923           19790           22700           23498
New Jersey......................            8191           10863           12764           11227           11470
New York........................           31222           30273           39503           39440           32114

[[Page 57411]]

North Carolina..................           32691           31394           32006           30156           29866
Ohio............................           51493           48468           39790           47143           50019
Pennsylvania....................           45971           52006           53450           47014           48476
Rhode Island....................            1609            1118            2242            3012            3202
South Carolina..................           19842           16290           23252           14085           13831
Tennessee.......................           26225           25386           26410           26084           24770
Virginia........................           20990           18258           19091           15700           15567
West Virginia...................           24045           26439           22853           30708           32527
Wisconsin.......................           17345           18029           15745           16637           16324
                                 -------------------------------------------------------------------------------
    Total.......................          563785          542007          542007          542007          542007
----------------------------------------------------------------------------------------------------------------

    The Agency then calculated the effective NOX emission 
rate for each State in terms of lb/mmBtu, assuming that the entire 
electricity generation component of the budgets, as determined by the 
input or output methods, were allocated to the electric generating 
units (EGUs). The Agency wanted to evaluate whether the effective 
NOX emission rate would be too low to prove feasible absent 
participation by the State in an interstate NOX emission 
trading program. The EPA found that under output-based State budgets 
from all generation sources, three States would need to impose an 
effective emission limitation of 0.10 lb/mmBtu or less on their fossil-
fuel burning electricity generators (see Column 3 in Table III-2 
below). One State would need to impose an emission limitation of 0.07 
lb/mmBtu. Such a low effective emission limitation may not be 
technically achievable if a State chooses not to join an interstate 
allowance trading program, unless the State requires some sources to 
shutdown. In contrast, the Agency found that it was feasible and cost-
effective to make reductions even without an interstate NOX 
trading program under an input-based State budget calculated using a 
uniform NOX emission rate of 0.15 lb/mmBtu.

                     Table III-2.--Effective Emissions Rates for Each State by Output Basis
                                        [Higher of 1995 or 1996 EIA data]
----------------------------------------------------------------------------------------------------------------
                    Column 1                         Column 2        Column 3        Column 4        Column 5
----------------------------------------------------------------------------------------------------------------
                                                     Effective                       Effective
                                                   emission rate     Effective     emission rate     Effective
                                                   under input-    emission rate   under output-   emission rate
                                                   based budgets   under output-   based budgets   under output-
                      State                        (Fossil fuel    based budgets       (all        based budgets
                                                      burning          (All         generation     (Fossil fuel-
                                                    generators)     generation)       except          burning
                                                    (lb/mmBtu)                       nuclear)       generators)
----------------------------------------------------------------------------------------------------------------
Alabama.........................................            0.15            0.18            0.18            0.17
Connecticut.....................................            0.15            0.45            0.30            0.26
Delaware........................................            0.15            0.10            0.14            0.15
District of Columbia............................            0.15            0.07            0.10            0.10
Georgia.........................................            0.15            0.16            0.16            0.15
Illinois........................................            0.15            0.21            0.13            0.14
Indiana.........................................            0.15            0.10            0.13            0.14
Kentucky........................................            0.15            0.11            0.14            0.15
Maryland........................................            0.15            0.13            0.13            0.13
Massachusetts...................................            0.15            0.11            0.13            0.14
Michigan........................................            0.15            0.17            0.17            0.17
Missouri........................................            0.15            0.12            0.14            0.15
New Jersey......................................            0.15            0.18            0.16            0.16
New York........................................            0.15            0.20            0.20            0.16
North Carolina..................................            0.15            0.15            0.14            0.14
Ohio............................................            0.15            0.12            0.15            0.15
Pennsylvania....................................            0.15            0.15            0.14            0.14
Rhode Island....................................            0.15            0.30            0.40            0.43
South Carolina..................................            0.15            0.21            0.13            0.13
Tennessee.......................................            0.15            0.16            0.15            0.15
Virginia........................................            0.15            0.16            0.13            0.13
West Virginia...................................            0.15            0.13            0.17            0.18
Wisconsin.......................................            0.15            0.13            0.14            0.14
----------------------------------------------------------------------------------------------------------------


[[Page 57412]]

    Advocates of an output-based approach contend that individual 
sources would have the greatest incentive to improve their efficiency, 
relative to all other sources in the program, if both State budgets and 
individual source allocations were on an output basis and were updated 
periodically. For example, if a company replaces a turbine with a more 
efficient one, the unit supplying the turbine would reduce the amount 
of fuel (heat input) the unit combusts and would reduce NOX 
emissions proportionately, while the associated generator would produce 
the same amount of electricity. Thus, the company would receive the 
same allowances if an output-based allocation were updated after the 
efficiency improvement. This same company would receive fewer 
allowances under a system that reallocates based on heat input after 
the efficiency improvement. The company would keep the same allowance 
allocation if it had a permanent allocation, based upon either heat 
input or output. With a permanent allocation, the company would have 
more allowances available than before its efficiency improvements 
because of its emission reductions, but fewer allowances than if it had 
greater electrical output recognized through an updated allocation. 
Thus, of the four approaches, an updated allocation based upon output 
gives the greatest incentive for improving efficiency in electricity 
generation.
    To provide an incentive within the State budget determinations for 
improving efficiency over time, EPA would need to issue the State 
budgets based upon output and periodically update those State budgets. 
However, many industry commenters wanted long-term or permanent 
allowance allocations to allow for compliance planning. Updates to the 
State budgets would require States to reallocate allowances to their 
sources. In addition, States (both upwind and downwind) would find it 
easier to manage their resources for improving air quality if they 
receive a fixed budget for a period of years. With a fixed budget, a 
State would have the choice of whether to periodically adjust 
allocations rather than being required to periodically reallocate 
allowances to its sources.
    Finally, the Agency continues to have concerns about data available 
to establish the baseline for an output-based State budget. The EIA 
withholds some of the electricity generation information it collects 
from non-utility generators in order to protect source confidentiality. 
Therefore, part of the generation data required to establish State 
budgets is not available to EPA. Thus, EPA would have difficulty in 
computing and defending State budgets.
    In addition, some units are cogenerators, which are electrical 
generators that divert part of their heated steam to provide heat 
(steam output), rather than to generate electricity. Information on 
steam output from cogenerating units or from industrial boilers is not 
currently available to EPA. A cogeneration unit that was included under 
the State budget as an electricity generating unit based upon heat 
input would only have its electrical output included in an output-based 
State budget, ignoring the portion of heat input used to generate steam 
output. Thus, output-based State budgets based on currently available 
data could inadvertently underallocate budgets to States with many 
cogenerators, which are some of the most efficient units. This could 
actually discourage improvements in efficiency through cogeneration.
    For the reasons stated above, the Agency concludes that it is not 
appropriate to develop output-based State NOX emission 
budgets at this time. However, the Agency does believe that output-
based allocations to sources could provide significant benefits. As 
stated earlier in this Section, the EPA intends to work with 
stakeholders to overcome existing obstacles and to design an output 
allocation system based on electricity and steam generation that could 
be used by States as part of their trading program rules in their SIPs. 
In addition, EPA is proposing FIPs for States that do not submit 
adequate SIPs by the deadline required by this final rulemaking. As 
part of its proposal, the Agency is soliciting comment on source 
allocations for each State based upon both input and output. While EPA 
believes that the output data are not sufficiently complete or accurate 
to use for final budgets or for final source allocations at this time, 
the Agency is taking comment on the proposed allocations in order to 
receive public comment and to develop more accurate and more complete 
output data that could be used in the final FIP rulemaking.
    The EPA does believe that, over the long-term, it should continue 
to look at the issues that surround the use of output-based 
allocations. In addition, as stated in Section III.B.5. of this 
preamble, the Agency will review the progress of States in meeting 
their budgets in 2007. In that review, the Agency will consider not 
only whether the SIPs achieved the reductions that had been projected 
to meet the budgets, but also issues such as future budget levels and 
allocation mechanisms including shifting to an output-based allocation 
method.
    b. Alternative Emission Limits.
    Comments: The EPA received numerous comments on the proposed 
uniform control level of 0.15 lbs/mmBtu for the EGU sector assumptions 
across the 23 jurisdictions. Many States supported this proposed 
control assumption. The EPA also received a number of alternative 
proposals. These contain emission-reduction assumptions ranging from 
0.12 lb/mmBtu to be implemented on the schedule proposed in the NPR to 
a phased approach that starts with 0.35 lb/mmBtu to be implemented by 
sector and provides for further evaluation of the need for more 
stringent levels. The latter commenters based their recommendations on 
their views that emissions from upwind States do not have an ambient 
impact that is as important as EPA believes, or that implementation of 
the EGU control levels proposed by EPA would not be feasible by the 
date EPA proposed. In addition, a number of utilities and other 
commenters voiced concern that the proposed control assumption of 0.15 
lb/mmBtu would be too stringent to provide sufficient surplus 
allowances for trading.
    Response: At the time of the proposal, EPA chose 0.15 lb/mmBtu as 
the assumed uniform control level for EGUs because it provided the 
greatest air quality improvements feasible and was cost-effective 
because its cost ($1,700 per ton NOX removed in the 5-month 
ozone season) was, on average, within the cost range of other controls 
that had been recently promulgated or proposed. The EPA also 
investigated the costs of several alternative uniform control options: 
0.25, 0.20, and 0.12 (though 0.12 resulted in lower emission levels, 
its average cost-effectiveness calculated at the time of the proposal 
was $2,100/ton, exceeding EPA's target cost range of $1,000 to $2,000/
ton).
    Subsequent to the NPR and SNPR, EPA updated its EGU costing model 
(IPM) and revised stationary source emission inventories (based on 
public comment). These revisions and corrections lowered the average 
cost of compliance for all the control levels considered. Additionally, 
EPA conducted extensive air quality modeling of a number of alternative 
control levels. The results of the air quality analyses were examined 
using a number of different metrics for both the one-hour and eight-
hour standards. These air quality analyses are discussed in more detail 
in Section IV of this notice.

[[Page 57413]]

    The revised air quality analyses show that there is no ``bright 
line'' to illustrate at what control levels the air quality benefits 
begin to diminish. The air quality metrics suggest there are 
corresponding incremental air quality improvements at every incremental 
control level. For example, tightening the control level improves ozone 
levels in many non-attainment areas and leads to additional counties 
achieving attainment under the one-and eight-hour standards. All 
metrics analyzed show that as the control level moves from 0.25 to 0.20 
to 0.15 to 0.12 lb/mmBtu, air quality benefits increase. The analyses 
also show that none of the alternative control options results in 
attainment of the ozone standard in all nonattainment areas.
    The EPA did not select levels higher than 0.15 lb/mmBtu (such as 
0.20 lb/mmBtu or higher) because the 0.15 lb/mmBtu level offers more 
air quality benefits at a cost that is still highly cost-effective. 
Moreover, EPA did not have information to indicate that these higher 
levels could be implemented meaningfully sooner than controls at the 
0.15 lbs/MmBtu level. The EPA acknowledges that the 0.12 lbs/MmBtu 
emission level is also within the average cost-effectiveness range 
based on the revised cost analysis. The incremental cost-effectiveness 
of this option is $4,200 per ton, an incremental cost per ton which is 
85 percent higher than that for the 0.15 lb/mmBtu level. However, for 
reasons explained Section II.D., the EPA is not relying on this 
emission level.
    The revised IPM analyses project that under the 0.12 control 
option, 54 percent of affected EGU capacity should install selective 
catalytic reduction (SCR) and 41 percent should install selective non-
catalytic reduction (SNCR). The installation requirements for SNCR are 
significantly less extensive than for SCR. The analysis of the 0.15 lb/
mmBtu control option projects 31 percent of affected EGU capacity 
should install SCR and 54 percent should install SNCR. Further, the 
technical record provides many examples in the United States and 
internationally of the ability of coal-fired units to achieve emission 
levels below 0.15 lb/mmBtu with the installation of SCR. The record 
contains fewer international examples, and only one US example, of a 
coal-fired unit's ability to achieve emission levels below 0.12 lb/
mmBtu.
    In terms of the proposed level of control on which the trading 
program budget is based, EPA believes that trading at 0.15 lb/mmBtu is 
feasible because the proposed limit can readily be achieved by gas and 
oil-fired boilers. In fact, more than 50 percent of gas and oil-fired 
boilers already operate at NOX levels below 0.15 lb/mmBtu 
and should readily be able to generate emission credits if affected 
States join a trading program.
    The EPA recognizes that for coal-fired boilers to operate at or 
below a 0.15 lb/mmBtu emission limit, SCR would generally be necessary. 
Under a trading scenario, however, if one coal-fired boiler is able to 
emit below 0.15 lb/mmBtu by installing SCR, it can provide emission 
credits to another coal-fired boiler and obviate the need for that 
second boiler to install SCR.
    A remaining issue is whether SCR can achieve NOX levels 
below 0.15 lb/mmBtu. The EPA believes that SCR technology is capable 
both of reducing NOX emissions by more than 90 percent and 
reducing NOX rates below the proposed 0.15 lb/mmBtu limit, 
provided the appropriate regulatory incentive (i.e., emission limit or 
economic incentive) exists. As discussed in EPA's recent report, 
``Performance of Selective Catalytic Reduction on Coal-Fired Steam 
Generating Units,'' emission rates below 0.15 lb/mmBtu are currently 
being achieved by a number of coal-fired boilers using SCRs. Examples 
include: (1) Three Swedish boilers achieving rates between 0.04 and 
0.10 lb/mmBtu; (2) six German boilers achieving rates between 0.08 and 
0.14 lb/mmBtu; (3) two Austrian boilers achieving rates between 0.08 
and 0.12 lb/mmBtu; and (4) four U.S. boilers achieving rates between 
0.07 and 0.14 lb/mmBtu. The EPA also recognizes that these boilers, 
with the exception of the Swedish boilers, have SCR systems designed to 
achieve target emission limits. As a result, they fail to provide an 
accurate picture of the emission levels which SCR is capable of 
achieving below the target emission threshold. For this reason, EPA 
cannot confidently conclude that enough units can feasibly achieve 
levels at 0.12 lbs/MmBtu. In summary, EPA believes that an emission 
rate of 0.15 lb/mmBtu reflects the greatest emissions reduction that 
EPA can confidently conclude is feasible and that is highly cost-
effective, and provides ample allowances to sustain a market under the 
NOX Budget Trading Program.
    c. Consideration of the Climate Change Action Plan.
    Background: The President's Climate Change Action Plan (CCAP) calls 
for implementation of over 100 voluntary programs aimed at reducing 
greenhouse gas emissions. A large number of them are aimed at reducing 
future electricity demand throughout the country. Already, some of 
these programs have shown striking results in accomplishing their 
energy efficiency objectives.
    Comment: Two commenters noted that it is inappropriate for EPA to 
incorporate assumed reductions in energy use based on the voluntary 
measures of the CCAP, which are not binding like a regulation.
    Response: The EPA believes that it is appropriate to incorporate 
the impact of the voluntary measures in the CCAP on future electricity 
demand. The EPA has always believed that it is appropriate to 
incorporate any reasonable assumptions that the Agency can support that 
will affect future electricity demand, or electricity generation 
practices, into its Base Case forecast. For example, improvements in 
electricity generation technology, fuel prices changes, and other types 
of assumptions that are important elements of EPA's forecast of 
electricity generation and resulting air emissions are also not 
mandated by regulation. The Agency has considered the impact of the 
CCAP in using the IPM model for analysis since 1996, and documentation 
of the assumptions that the Agency has been making have been available 
for public review since April 1996. Until now, there have been no 
challenges to this consideration in the numerous reviews that there 
have been of EPA's documentation of how it uses the IPM model. Also, no 
one has challenged EPA's specific approach to factoring the CCAP into 
its electricity generation forecast. (This can be confirmed by 
examination of the dockets for the Clean Air Power Initiative and the 
Phase II Title IV NOX Rule, records of EPA's Science 
Advisory Board, and the records of the Ozone Transport Assessment Group 
meetings.)
    The EPA updated its assumptions in IPM for the CCAP at the 
beginning of 1998. The EPA updated its assumptions in the same manner 
as it has done in the past--by lowering the most recent NERC demand 
forecast by the amount of electricity demand between 2000 and 2010 that 
the best available analysis suggests will occur due to the activities 
in CCAP. The EPA used the in-depth evaluation of the future 
implications of the CCAP for reducing electricity demand that was the 
basis for the findings in the Administration's Climate Action Report, 
July 1997. The amount of demand reduction that occurs appears in 
Analyzing Electric Power Generation under the Clean Air Act, March 
1998. The Climate Action Report analysis was reviewed extensively 
within the Federal government by EPA, the Department of Energy and 
other Federal agencies, and the report was reviewed publicly before its 
publication. The EPA has not received criticism that it has overstated

[[Page 57414]]

the electricity demand reductions that are the basis for the carbon 
reductions under the CCAP.
    Notably, the electricity demand reductions were distributed evenly 
throughout the United States, and therefore have no influence on the 
share of the total amount of NOX emissions that each State 
receives. Furthermore, the Agency examined the implications on its 
cost-effectiveness determination of not including the CCAP reductions 
in its electricity demand forecast. The EPA found that even if the 
Agency did not assume the CCAP reductions, it was still highly cost-
effective to develop a regional level NOX budget for the 
electric power industry, based on the level of control that EPA has 
assumed. (These results appear in Chapter 6 of the Regulatory Impact 
Analysis for the Regional NOX SIP Call, September 1998.)

C. Non-EGU Point Sources

    Background: The EPA developed the NOX SIP call emissions 
inventory for non-EGU point sources based on data sets originating with 
the OTAG 1990 base year inventory. The OTAG prepared these base year 
inventories with 1990 State ozone SIP emission inventories, and EPA 
supplemented them with either State inventory data, if available, or 
EPA's National Emission Trends (NET) data if State data were not 
available.
    For the SNPR, non-EGU point source inventory data for 1990 were 
then grown to 1995 using Bureau of Economic Analysis (BEA) historical 
growth estimates of industrial earnings at the State 2-digit Standard 
Industrial Classification (SIC) level. These emissions were grown to 
1995 for the purposes of modeling and to maintain a consistent base 
year inventory with the EGU data. Because BEA data are historical 
documentation of industry earnings, EPA considered these to be among 
the best available indicators of growth between 1990 and 1995 (63 FR 
25915). Once the common base year of 1995 was established for these 
source categories, the BEA growth assumptions utilized by OTAG were 
used to estimate the 2007 base case inventory.
1. Base Inventory
    Comment: The majority of comments related to the non-EGU point 
source inventory alleged that these inventories were incomplete or 
inaccurate. The comments generally addressed missing sources, non-
existent or retired sources, incorrect source sizes, mis-classification 
of processes, or emission allocation inconsistencies. Many of these 
commenters provided specific adjustments to be made to the inventories, 
including emissions modifications, activity factors, source sizes, and 
facility name changes. A number of States supplied completely new 
inventories to replace what was in the proposed data sets. Other 
commenters made broad, general categorical comment on the quality of 
the inventories with no supporting data.
    Response: As was followed under the OTAG inventory update 
procedures, all State supplied comments were generally incorporated 
``as is'' with the understanding that each State quality-assured its 
own data before submission. Industry-supplied comments were forwarded 
to respective State agencies for review and where data were deemed 
appropriate for inclusion, integrated into the inventories. In some 
instances, States responded that the data provided by the State should 
override that supplied by industry, or vice-versa. Comments were, in 
some cases, not incorporated when necessary to prevent double counting 
of emissions in point and area source inventories, where base year 
emission modifications were calculated from permitted emission levels 
and not actual operating activity, where additional supporting data 
could not be provided by the commenter, or where comments were general 
characterizations of inventories or inventory sectors. Note that even 
after State review, if the EPA felt that the data, procedures, 
methodologies, or documentation provided with the comment were not 
sufficient, valid, or justifiable, comments, or portions thereof, were 
excluded from the revision.
    Both 1990 and 1995 base year emission and growth modifications were 
submitted and where 1990 data were provided, the methods described 
earlier in this Section were utilized to account for growth to 1995 and 
2007 levels.
2. Growth
    Comment: Several commenters suggest that the growth factors used to 
determine 2007 non-EGU point source base year inventories are 
inaccurate or inconsistent across regions and categories of the 
inventory. They explained that if growth factors are to be used to 
estimate future base year emissions, consistent national or region-wide 
values should be utilized for all categories across all States within 
the domain. This, they continue, would promote equitable potential 
progress to all areas and not penalize those that have shown past poor 
growth rates. Some commenters go on to state that growth rates based on 
past growth automatically disadvantage States which have suffered from 
unusually low growth rates. In addition to growth rates, some 
commenters provided 2007 base year emission estimates either with or 
without the growth and control information needed to validate their 
calculation.
    Response: As noted above, EPA relied on BEA State-specific 
historical growth estimates of industrial earnings at the 2-digit SIC 
level as among the best available indicators of growth for non-EGU 
point sources. The BEA projection factors assume the continuance of 
past economic relationships. These factors are published every five 
years and adjusted to account for recent production and growth trends. 
For this reason, BEA data provide a useful set of regional growth data 
that EPA recommends for use in preparing emission inventory 
projections. It is true that BEA projection factors differ among 
different areas and different source categories because of historical 
differences in industrial growth among those different areas and source 
categories. However, in general, these projection factors offer the 
most reliable indicators of future growth as are available.
    In cases where commenters questioned the use of EPA's growth rates 
but provided no alternative of their own, EPA had little choice but to 
continue to use the BEA-derived growth rates. Some commenters provided 
alternative or supporting information for modification of source 
category or State growth estimates. In those cases where a State or 
industry may have had more accurate information than the BEA forecast 
(e.g., planned expansion or population rates), data were verified and 
validated by the affected States and by EPA, and revisions were made to 
the factors used for that category.
3. Budget Calculation
    Background: In the NPR and SNPR, EPA proposed that EGUs with a 
capacity less than or equal to 25 MWe or 250 mmBtu/hour would be 
considered small sources (``cutoff level'') and, as such, EPA would not 
assume an emissions decrease as part of the Statewide budget for this 
group of sources. At the same time, EPA proposed 2 cutoff levels for 
industrial (non-EGU) boilers and turbines: units with a capacity 
greater than 250 mmBtu/hour were defined as large units subject to a 70 
percent emission reduction assumption; units with a capacity less than 
or equal to 250 mmBtu/hr but with emissions greater than 1 ton/day were 
defined as medium units subject to reasonably available

[[Page 57415]]

control technology (RACT); and units with a capacity less than or equal 
to 250 MmBtu/hr and with emissions less than or equal to 1 ton per day 
were considered small sources for which no reduction would be assumed 
in the budget. In the SNPR, EPA specifically invited comment on the 
size cutoffs and on treating large industrial combustion sources 
(greater than 250 mmBtu or approximately 1 ton per day) at control 
levels equal to that for EGUs (63 FR 25909). As described below, this 
approach has been modified somewhat in response to comments and further 
analysis.
    a. Proposed Control Assumptions.
    Comments: Some comments supported EPA's proposed approach of 
assuming 70 percent and RACT controls in its calculation of the 
budgets. Numerous comments were received stating that the 70 percent 
reduction is inappropriate, may not be cost-effective and may not be 
achievable, especially for the following industries: cement plants; 
municipal waste combustors; certain pulp and paper operations, 
including lime kilns and recovery furnaces; glass manufacturing; steel 
plants; and some industrial boilers. Some comments suggested a control 
level of 60 percent rather than 70 percent. On the other hand, one 
commenter stated that SCR and SNCR are applicable and have been 
installed on hundreds of industrial sources.
    Response: The EPA generally agrees that 70 percent emissions 
reduction is not appropriate for all large sources or all large source 
categories, even though SCR and SNCR are applicable and cost-effective 
for many sources. Instead of applying a one-size-fits-all percentage 
reduction to all large non-EGU sources, the specific emissions 
decreases assigned to each of these source categories for purposes of 
budget calculation in the final SIP Call rulemaking reflect the 
specific controls available for each source category that achieve the 
most emissions reductions at costs less than an average of $2,000 per 
ton. As described elsewhere in this notice, EPA's analysis results in 
calculating budget reductions ranging from 30 percent to 90 percent for 
several source categories and no controls to several other source 
categories.
    b. Small Source Exemption.
    Comments: In general, commenters were supportive of EPA including a 
cutoff level as part of the budget calculation; however, there were 
many suggestions on what the cutoff should be. The EPA received 
numerous comments supporting the proposed cutoff level of 25 MWe for 
EGUs, which is approximately equivalent to 250 mmBtu/hr or one ton per 
day. In addition, EPA received a few comments supporting a 250 mmBtu/hr 
cutoff for non-EGU point sources. Commenters indicated that the levels 
were appropriate and that it was important to be consistent with cutoff 
levels in the OTC's NOX trading program. The Ozone Transport 
Commission (OTC) comprises the States of Maine, New Hampshire, Vermont, 
Massachusetts, Connecticut, Rhode Island, New York, New Jersey, 
Pennsylvania, Maryland, Delaware, the northern counties of Virginia, 
and the District of Columbia. In September 1994, the OTC adopted a 
memorandum of understanding (MOU) to achieve regional emission 
reductions of NOX. These reductions are in addition to 
previous OTC state efforts to control NOX emissions, which 
included the installation of reasonably available control technology. 
The OTC's NOX trading program requires utility and 
nonutility boilers greater than 25 MWe or 250 mmBtu to reduce emissions 
in order to meet a NOX budget and allows emissions trading 
consistent with that budget. These NOX reductions will take 
place in two phases, the first phase beginning on May 1, 1999 and the 
second phase on May 1, 2003.
    Some comments suggested assuming budget controls on units less than 
or equal to 25 MWe at RACT levels without a cutoff level. Others 
supported EPA's proposal of assuming no additional controls on these 
sources. Some comments suggested exempting medium-sized non-EGU 
sources.
    Many commenters supported the general 1 ton per day exemption 
contained in the NPR and SNPR. However, a few comments suggested a more 
stringent cutoff level of 50-100 tons per year, similar to definitions 
of ``major source'' in the CAA. One commenter recommended a less 
stringent level of 5 tons per day cutoff level.
    A few comments suggest using tons per day as the primary criterion 
to define large- and medium-sized non-EGU sources, rather than boiler 
capacity. This approach would exempt, for example, industrial boilers 
that exceed the 250 mmBtu capacity, but which emit less than one ton 
per day on average. The EPA's proposed approach considers a source 
large if heat input capacity data are available and exceed the 250 
mmBtu capacity criterion, regardless of its average daily emissions. In 
support of this approach, commenters stated that industrial operations 
do not usually operate at or near capacity, while EGUs often do.
    A few commenters indicated that the OTAG recommendations for 
turbines and internal combustion engines (in terms of horsepower cutoff 
levels) be used. OTAG had recommended cutoff levels of 4,000 horsepower 
for stationary internal combustion engines and 10,000 horsepower for 
gas turbines.
    Response: For reasons described below and in the NPR (62 FR 60354), 
EPA believes that the cutoff levels of 250 mmBtu/hr and 1 ton per day 
for large non-EGU point sources are appropriate. The EPA selected 250 
mmBtu/hr and 1 ton per day primarily because this is approximately 
equivalent to the 25 MWe cutoff used for the EGU sector. Emission 
decreases from sources smaller than the heat input capacity cutoff 
level, and that emit less than 1 ton of NOX per ozone season 
day, are not assumed as part of the budget calculation; these sources 
are included in the budget at baseline levels.
    The EPA believes that the 1 ton per day exclusion contained in the 
NPR and SNPR is appropriate and necessary. This level allows today's 
rulemaking to focus, for the purpose of calculating the budget, on the 
group of emission sources that contribute the vast majority of 
emissions, while at the same time avoids assuming emissions reductions 
from a very large number of smaller sources (as described in the 
following paragraph). In taking today's first major step towards 
reducing regional transport of NOX, EPA does not believe 
that emission reductions from these small sources need to be assumed. 
This approach provides more certainty and fewer administrative 
obstacles while still achieving the desired environmental results. 
Although other cutoff levels were suggested by commenters, EPA believes 
that the cutoff levels described above strike the appropriate balance 
so that reasonable controls may be applied by States to a sufficient 
but manageable number of sources to efficiently achieve the needed 
emission reductions.
    Most small sources emit less than 100 tons of NOX per 
year. Although their total emissions are low, small sources account for 
about 90 percent of the total number of point sources. Thus, not 
assuming controls on these sources at the present time would greatly 
limit administrative complexity and reporting costs. This common-sense 
approach results in reducing the non-EGU population potentially 
affected by the ozone transport rule from more than 13,000 sources 
estimated in the NPR and SNPR to under 1,200.
    Although a few comments suggested using tons per day, not capacity 
(MWe or mmBtu/hr), for setting cutoff levels, EPA chose primarily to 
use capacity indicators. This approach is consistent

[[Page 57416]]

with the framework of the emissions trading program. In addition, EPA 
is concerned that units could have low average emissions during the 
ozone season but relatively high emissions on some high ozone days. 
Accordingly, EPA is relying on a capacity approach first and a tons per 
day approach second (where capacity data is not available or 
appropriate) to define units for which reductions are assumed in EPA's 
budget calculations.
    As noted in the proposal notices, horsepower data was generally 
absent from the available emissions inventory data. Thus, the OTAG 
recommendation could not be used. Because quality assured data are 
still lacking, EPA used alternative approaches to determine size 
categories as described above. For the purposes of calculating the 
State budgets, the following approach is used to determine whether 
controls should be assumed on a particular source for the purposes of 
calculating the budget:

    1. Use heat input capacity data for each source if the data are 
in the updated inventory.
    2. If heat input capacity data are not available, use the 
default identification of small and large sources developed by EPA/
Pechan for OTAG and also used to develop the NPR and SNPR budgets 
for source categories with heat input capacity fields (``default 
data'').
    3. Emission reductions would be assumed if specific source heat 
input capacity data or default data indicate that a source is 
greater than 250 mmBtu/hr in the updated inventory.
    4. If specific or default heat input capacity data are not 
available in the updated inventory (or not appropriate for a 
particular source category), emission reductions would be assumed if 
the unit's average summer day emissions are greater than one ton per 
day based on the updated inventory.
    5. All others are ``small'' and no emission reductions are 
assumed.

    c. Exemptions for Other Non-EGU Point Sources.
    Comments: Several comments described source categories that might 
be excluded from being assigned assumed emissions decreases for 
purposes of calculation of the NOX budgets. In the NPR, EPA 
assumed a 70 percent reduction from large sources and RACT on medium-
sized sources. Some commented that it is not possible to control lime 
kilns and recovery furnaces or that potential NOX emissions 
reductions are very small. One comment noted that recovery units 
typically emit at a rate of 0.15 lb/mmBtu or less and lime kilns at 
0.20 lb/mmBtu or less and suggested establishing an emissions rate 
floor so that sources emitting less than 0.15 lb/mmBtu (or some other 
floor) would not need to further control. Other commenters suggested 
exempting cyclone boilers less than 155 MWe and all aircraft engine 
test facilities.
    Response: The EPA agrees that for purposes of today's rulemaking 
the State budgets should not reflect assumed reductions in emissions 
from lime kilns, recovery units and aircraft engine test facilities. 
The amount of emissions from these source categories is very small 
relative to other point source categories considered in this 
rulemaking. Further, there is no experience in applying NOX 
control technologies full scale to aircraft engine test cells in the 
U.S. (EPA-453/R-94-068, October 1994).
    The EPA acknowledges that NOX controls may be available 
at costs less than $2,000 per ton for lime kilns, recovery units and 
aircraft engine test cells. However, these source categories include a 
relatively small number of sources with a small amount of emissions. 
The EPA is concerned that assuming controls on these sources for 
purposes of State budgets would encourage States to attempt to regulate 
these sources. The EPA believes State regulation could be inefficient 
because of the relatively high administrative costs of developing 
regulations for these few source categories (particularly for aircraft 
engine test cells because no regulations have been developed for this 
source category).
    Similarly, EPA determined for each of the following non-EGU point 
source categories that the amount of emissions are small relative to 
the total non-EGU point source emissions and, thus, State regulation 
could be inefficient because of the relatively high administrative 
costs of developing regulations for these few source categories: 
ammonia, ceramic clay, fiberglass, fluid catalytic cracking, iron & 
steel, medical waste incinerators, nitric acid, plastics, sand/gravel, 
secondary aluminum, space heaters, and miscellaneous fuel use 
operations. Further, for many of these categories the number of sources 
is small and/or control technology information is limited (e.g., where 
an Alternative Control Techniques document does not exist for that 
category). The EPA believes that it would be an inefficient approach to 
suggest that States consider adopting emissions reduction regulations 
for each of these categories. Therefore, EPA did not calculate 
emissions reductions from these source categories for purposes of 
calculating the budget.
    At this stage in the process to reduce regional transport, EPA 
considers it most efficient to focus State and administrative resources 
on the source categories with greater amounts of emissions. While 
States may choose to control any mix of sources in response to the SIP 
call, EPA is not, in today's rulemaking, assuming reductions from these 
source categories as part of the budget reduction calculation and does 
not believe it is necessary for States to do so.
    It should be noted that EPA is generally treating the non-EGU 
boilers/turbines in the same manner as the EGUs to enable States that 
opt into a trading program to develop a simple and effective trading 
program. Thus, the size cutoffs discussed earlier in this section are 
identical. Further, the regulatory definition of a unit has been 
revised to make it clear that only fossil-fuel fired boilers and 
turbines are affected; this is discussed in detail in the trading 
program section later in today's notice. In addition, it should be 
noted that EPA is not excluding reductions from cyclone boilers, 
whether EGU or non-EGU, between 25-155 MWe from the calculation of the 
State budgets in this rulemaking. Such sources can be large emitters of 
NOX and EPA expects the control costs will be less than 
$2000/ton on average through participation in the emissions trading 
program.
    d. Sources Without Adequate Control Information.
    Comments: As described in the SNPR, there are many sources in the 
emissions inventory which lack information EPA would need to determine 
potentially applicable control techniques. The SNPR proposed to leave 
these sources in the budget without assigning any emissions reductions. 
The EPA received comments that generally supported the SNPR approach 
not to assign emissions reductions to the diverse group of sources 
where the Agency lacked sufficient information to identify potential 
control techniques (63 FR 25909).
    Response: This group of sources is diverse and does not fit within 
the categories set out by EPA, but total emissions are low for this 
group. The EPA believes that the effort needed to collect adequate 
information concerning controls for those sources (about 6,000 small 
and 260 medium or large) would be time consuming, the quality of the 
information may be uncertain, and it would potentially affect only a 
small amount of NOX emissions. Therefore, for purposes of 
today's action, EPA continues not to assume decreases in emissions for 
these sources for purposes of calculation of the State budgets, but to 
keep them in the budgets at baseline levels. In the future, as more 
information becomes available, and if additional NOX control 
is needed to further reduce ozone transport, further

[[Page 57417]]

consideration of these sources may be necessary. Of course, States with 
adequate information may choose to control these sources to meet their 
budgets.
    e. Case-By-Case Analysis of Control Measures.
    Comments: Some commenters suggested that EPA simply assume 
reasonably available control technology (RACT) for medium and, in some 
comments, large sources in all upwind States on a case-by-case basis 
and assure that marginally stringent source-specific reduction levels 
are rejected. Many commenters stated that RACT default levels used by 
EPA were not sufficiently accurate and that case-by-case analysis was 
needed because every industrial source is different. Other comments 
generally stated that control level decisions should only be made on a 
case-by-case basis because each affected unit may have unique features 
that alter its cost-effectiveness.
    Response: In the final budget calculation procedure EPA does not 
calculate RACT requirements for medium-sized sources. The assumption of 
RACT or other controls on industrial boilers and turbines between 100-
250 mmBtu/hr would have been inconsistent with EPA's approach for 
utility boilers and turbines, which exempts units less than or equal to 
250 mmBtu/hr. To be consistent with the way EPA treats EGUs and because 
data is often lacking for the smaller size sources, EPA redefined 
``affected'' non-EGU units to primarily include those greater than 250 
mmBtu. In cases where heat input data are not available, affected non-
EGU units are those greater than 1 ton per day; this level is also 
consistent with the EGU cutoff because it is approximately equivalent 
to the 250 mmBtu level. Consistency with the EGU approach is important 
because it provides equity, especially among the smaller boilers and 
turbines and simplifies the model trading program. Therefore, the final 
rule does not calculate budget reductions for the medium size non-EGUs.
    For the above reasons and as described below, EPA has examined the 
non-EGU sources on a category-by-category basis and determined 
appropriate control level assumptions for the large units. There are 
several reasons why EPA did not choose to calculate the budget by 
examining sources on a case-by-case basis. First, such an approach 
would be inefficient since all large sources would need to be examined, 
rather than some source categories being eliminated due to category 
specific cost-effectiveness limitations or amount of emissions. Second, 
it would be very difficult for the States to complete a case-by-case 
analysis of their large sources, develop rules, and respond to the SIP 
call within the 12 month time frame (or the statutory maximum 18 
months). States needed much more time to respond to a similar 
requirement, the 1990 CAA NOX RACT program. The CAA allowed 
a 2-year period before the NOX RACT rules were due from the 
States; however, few States met this time frame and several adopted 
generic RACT rules which, in practice, resulted in much longer time 
frames before the case-by-case RACT analyses were completed and State 
rules adopted. Third, the option of participating in a trading program 
should mitigate cost impacts on some sources that may have unique 
configurations or other constraints. Fourth, EPA has often issued 
standards on a category-wide basis (e.g., New Source Performance 
Standards) which have proved workable even though some individual units 
have higher costs than the average. Fifth, the results of such case-by-
case analyses may not be perceived to be as equitable as the 
categorical approach because the control levels resulting from the 
case-by-case approach are likely to vary from source-to-source and 
State-to-State. Finally, the category-by-category approach selected by 
EPA is preferred because it will achieve air quality benefits sooner 
than the case-by-case approach.
    f. Cost-Effectiveness.
    Comments: The EPA received numerous comments on cost-effectiveness. 
Those comments related to uniform control levels or cost per air 
quality improvement are addressed elsewhere in this notice. Some 
comments supported EPA's proposed $2,000 per ton approach. Some 
commented that EPA should use incremental costs, which are the costs 
and reductions associated with obtaining further control from a unit 
that already has some level of controls installed. Several commenters 
suggested using marginal costs, defined as the cost of the last ton of 
NOX removed by a control strategy. Many stated that the 
costs for non-EGUs should be no greater than for utilities on a $/ton 
basis. One commenter noted that non-EGU costs will be considerably 
lower than EPA estimates. One comment suggested that EPA assume no 
further controls if the source has BACT, LAER, MACT or RACT already in 
place. One comment supported a command-and-control approach instead of 
the least cost for the non-EGUs, and asserted that controlling 13,000 
sources through this rulemaking may not be feasible. Several commenters 
suggested that CEMS costs for non-utilities should be included in the 
cost-effectiveness determinations and that alternative monitoring 
methodologies should be considered.
    Response: The EPA believes that the approach of average cost-
effectiveness described in the proposal notices is appropriate for this 
rulemaking. In establishing the upper limit of the cost-per-ton range 
that EPA considers highly cost-effective for this rulemaking, EPA 
relied on average cost-effectiveness values estimated for recently 
proposed or promulgated rulemakings. The marginal cost-effectiveness 
for the level of control decided upon in the other programs and 
rulemakings was not always estimated or readily available. The EPA's 
latest assessment of cost-effectiveness does account for the level of 
existing or planned control in the baseline case. Therefore, when EPA 
refers to average cost-effectiveness it is the average incremental cost 
between the base and the more stringent level of control.
    For the non-EGU point sources, in the NPR and SNPR EPA had 
aggregated the non-EGUs as one group, which meant that a few source 
categories with relatively low costs and high percentage emissions 
decreases dominated overall average cost-effectiveness. For today's 
final action, EPA revised its approach and analyzed individual source 
categories to determine if control techniques are available at average 
costs less than $2,000 per ton. Further, EPA included in this cost-
effectiveness approach the costs related to CEMS, because this is a new 
and potentially high cost to some of the non-EGU source categories. As 
described in the RIA that supports this final rulemaking, EPA's 
analysis determined that the following non-EGU source category 
groupings could achieve substantial emissions decreases at average 
costs less than $2,000 per ton: industrial boilers and turbines, 
stationary internal combustion engines, and cement manufacturing. As 
further described in the RIA, controls for sources grouped in the 
following categories exceed $2,000 per ton: glass manufacturing, 
process heaters, and commercial and industrial incinerators.
    The EPA believes that, over time, costs for non-EGU point sources 
will be lower than current EPA estimates; however, the changes cannot 
be quantified at this time. As discussed below, EPA agrees that one 
source category that has a NOX standard set through the MACT 
process should not be assumed to implement further controls.
    g. Industrial Boiler Control Costs.
    Comments: Several comments were submitted indicating that 
industrial

[[Page 57418]]

boiler costs are generally higher than utility boiler costs. The 
comments cited factors of load variability, smaller size/economies of 
scale, firing of multiple fuels, and the ability to finance new 
controls and pass on costs. Some comments stated that most industrial 
boilers are one-seventh the size of utilities and, thus, EPA should 
recognize that the costs of controls would generally be higher due to 
economies of scale.
    Response: The EPA agrees that industrial boiler sources are 
generally smaller than utility boiler sources; however, some individual 
industrial sources are larger than some utility sources. The EPA agrees 
that costs, on average, to the industrial sector are expected to be 
somewhat greater than that expected by the utilities due, in part, to 
economies of scale and the need for CEMS (which are already in place at 
utilities). Primarily due to the costs related to continuous emissions 
monitoring systems, EPA's reanalysis of cost-effectiveness for 
industrial boilers resulted in a control level of 60 percent, which is 
less stringent on average than that for utilities.
    h. Cement Manufacturing.
    Comments: In the NPR, EPA proposed a 70 percent control assumption 
on large sources and RACT on medium sources, including cement plants. 
Some commenters suggested that cement manufacturing should be excluded 
because in the SIP Call area, there are only a few cement plants and 
they have low emissions. Several commenters noted that many cement 
plants had already implemented NOX RACT controls. Some 
comments disagreed with the costs and controls contained in EPA's 
Alternative Control Techniques document (EPA-453/R-94-004, March 1994) 
and added that EPA should not assume the same controls for different 
types of cement plants. Several commenters stated that 70 percent 
control is not feasible and SCR costs would be greater than $4,500 per 
ton, but that 20-30 percent control is possible. One commenter stated 
that the SIP call would provide a major competitive advantage to plants 
outside the region, and that multi-plant companies may shut down 
facilities inside the SIP call region and increase output at plants 
outside.
    Response: Over 50 cement manufacturing units together emit more 
than twenty percent of emissions from large point sources not in the 
trading program (about 40,000 tons per season). The EPA believes that 
the emissions from this one industry are sufficiently high that it is 
appropriate to examine the availability of cost-effective controls.
    The cost and control estimates in the Alternative Control 
Techniques (ACT) document were peer reviewed and, as such, are 
considered by EPA as the best data available. Consistent with the ACT 
document for this industry, EPA generally agrees with the commenters 
that a 70 percent control level would exceed the $2,000 per ton level 
used as EPA's cost-effectiveness framework. But, with the evidence 
cited in the cement ACT document and in some comments, EPA believes 
that a 30 percent reduction from uncontrolled levels would be within 
the cost-effectiveness range for reducing emissions at all types of 
cement manufacturing facilities. Therefore, the budget calculations 
assume a 30 percent control level for this source category. The EPA 
does not anticipate that, if States were to choose to apply a 30 
percent control level to cement plants, this would be a major 
competitive disadvantage for plants located in the SIP call area 
because many cement plants in the region have already successfully 
implemented such controls in State RACT programs.
    i. Stationary Internal Combustion Engines.
    Comments: One comment suggested EPA set RACT levels at 25 percent 
for this category.
    Response: As noted above, EPA is not using a RACT approach in the 
final rulemaking, but has examined each non-EGU point source category 
separately to determine the maximum available emissions reductions from 
controls that would cost less than $2,000 per ton on average. As 
described in the RIA, this process of looking at source categories 
individually resulted in EPA changing the control level assumption for 
this category from 70 percent in the NPR to 90 percent control in 
today's final rule. As described elsewhere in this notice, EPA also 
changed the control level assumptions for other source categories 
through this more detailed approach.
    For this source category, EPA determined based on the relevant ACT 
document, that post-combustion controls are available that would 
achieve a 90 percent reduction from uncontrolled levels at costs well 
below $2,000 per ton. (EPA-453/R-93-032, 1993.) Therefore, the budget 
calculations include a 90 percent decrease for this source category 
from uncontrolled levels.
    For spark ignited rich-burn engines, non-selective catalytic 
reduction (NSCR) provides the greatest NOx reduction of all 
technologies considered in the ACT document and is capable of providing 
a 90 to 98 percent reduction in NOX emissions. The control 
technique for spark ignited lean burn, diesel, and dual fuel engines is 
selective catalytic reduction (SCR). The SCR provides the greatest 
NOX reduction of all technologies considered in the ACT 
document for these engines and is capable of providing a 90 percent 
reduction in NOX emissions.
    j. Industrial Boilers and Turbines.
    Comments: Several commenters indicated that boilers using SNCR may 
achieve 40-60 percent reduction, but not 70 percent. Other comments 
supported the 70 percent control level proposed.
    Response: The EPA examined the category of industrial boilers and 
turbines to determine the largest emissions reductions that would 
result from controls costing less than $2,000 per ton on average, 
including costs related to CEM systems. As described in the RIA, for 
this source category, EPA determined that controls, including SCR and 
SNCR, are available that would achieve a 60 percent reduction from 
uncontrolled levels at costs less than $2,000 per ton on average. For 
those sources that participate in the trading program, EPA believes 
that the costs would be further reduced. Therefore, the budget 
calculations include a 60 percent reduction for this source category 
from uncontrolled levels.
    k. Municipal Waste Combustors (MWCs).
    Comments: Several comments suggested that State budgets should not 
reflect emissions decreases for MWCs beyond those already required by 
the MACT rules.
    Response: The NPR did not assume reductions for MWCs in the 
calculation of the budgets. However, since MACT reductions are 
required, and will be achieved well before 2007, those reductions 
should be accounted for in the 2007 baseline emissions inventory. The 
EPA agrees that additional emissions decreases beyond MACT levels are 
not warranted for this source category at this time because they would 
exceed the $2,000 per ton framework for highly cost-effective controls. 
Therefore, EPA has incorporated the NOX emissions decreases 
due to the MACT requirements into the 2007 baseline levels and not 
assume any further reductions.

D. Highway Mobile Sources

    Background: For the NPR and SNPR, highway vehicle emissions were 
projected to 2007 from a base year of 1990. The NPR used the 1990 OTAG 
inventory as its baseline. The 1990 OTAG inventory was based on actual 
1990 vehicle-miles-traveled (VMT) levels for each State, based on State

[[Page 57419]]

submittals to OTAG where available, or on historical VMT data obtained 
from the Highway Performance Monitoring System (HPMS) if State data 
were not available. The EPA proposed to switch to historical 1995 VMT 
levels from the HPMS; States were encouraged to submit their own 1995 
VMT estimates where those estimates differed from HPMS.
    In today's notice, EPA has implemented the changes it proposed in 
the NPR in calculating baseline and projected future NOX 
emissions from highway vehicles. A 1995 baseline is used for today's 
notice in place of the 1990 baseline used in the NPR. The HPMS data 
were used to estimate States' 1995 VMT by vehicle category, except in 
those cases where EPA accepted revisions per the comments. These VMT 
estimates reflect the growth in overall VMT from 1990 to 1995, as well 
as the increase in light truck and sport-utility vehicle use relative 
to light-duty vehicle use. The 1995 NOX emissions 
inventories also reflect the type and extent of inspection and 
maintenance programs in effect as of that year and the extent of the 
Federal reformulated gasoline program. The EPA is continuing to use the 
growth factors developed by OTAG for the purpose of projecting VMT 
growth between 1995 and 2007. These growth factors were revised with 
appropriately explained and documented growth estimates submitted 
during the comment period for the NPR.
    The 2007 highway vehicle budget components presented in today's 
notice are based on EPA's MOBILE5a emission inventory model with 
corrected default inputs, which represents the most current EPA 
modeling guidance to States when developing their SIPs.60
---------------------------------------------------------------------------

    \60\ Both MOBILE5a and MOBILE5b are official EPA models. States 
can use either model in their SIPs, provided they use the corrected 
default inputs with MOBILE5a. For the control programs evaluated in 
today's action, MOBILE5a with corrected default inputs gives the 
same emission estimates as MOBILE5b. Because both models are 
considered valid by EPA and give the same emission estimates, the 
EPA has determined that the choice of which model to use in 
calculating highway vehicle emission budget components is a matter 
of convenience. The EPA has chosen to retain the use of MOBILE5a for 
today's action in order to maintain consistency with the OTAG 
process, in which MOBILE5a with corrected default inputs was used to 
construct its highway vehicle emission inventories and to calculate 
the effectiveness of highway vehicle emission control options.
---------------------------------------------------------------------------

1. Base Inventory
    Comment: The EPA received a number of comments on baseline highway 
vehicle emission inventories. Most of these commenters proposed changes 
to baseline VMT estimates or to control factors related to highway 
vehicle emissions.
    Response: In the NPR and SNPR, EPA asked commenters to provide 
sufficiently detailed information to permit revision to county-level 
emission inventories, in order to allow airshed modeling to be 
performed using the revised inventories. A number of proposed VMT 
revisions submitted by commenters were not sufficiently detailed to 
permit county-level inventory revisions and therefore these revisions 
were rejected. Other commenters provided sufficiently detailed data, 
which were incorporated into the base year VMT inventory, with two 
exceptions. Two States submitted 1995 VMT estimates that were 
inconsistent with EPA and U.S. Department of Transportation information 
on the relative contribution of light-duty trucks to total VMT. The EPA 
chose to use the HPMS default data for these two States.
    Comment: One commenter asked the EPA to use VMT from the 1996 
Periodic Emissions Inventory (PEI) or 1996 National Emissions Trends 
(NET), rather than 1995 Highway Performance Modeling System (HPMS) data 
when calculating baseline inventories. Several other commenters 
supported EPA's use of 1995 HPMS data to calculate baseline VMT 
inventories.
    Response: Guidance on how to construct the 1996 PEI was not 
released until July 1998 and State PEI submittals are not expected 
until 1999. The EPA has determined for this reason that the 1996 PEI is 
not suitable for calculating the baseline VMT inventory. The EPA 
considered using 1996 NET VMT data in its base inventories, but those 
data were based on estimated 1995 HPMS inputs. The EPA has chosen to 
use the actual 1995 HPMS data rather than estimates in order to reduce 
the uncertainties associated with estimating baseline and 2007 emission 
inventories.
    Comment: One commenter suggested using a multi-year VMT activity 
average to establish the highway emission baselines to smooth out 
abnormal patterns, instead of relying solely on 1995 activity.
    Response: The EPA proposed using 1995 VMT in order to shorten the 
time period over which VMT growth would have to be projected. The EPA 
is not aware of any evidence that suggests that 1995 was an abnormal 
year in terms of VMT activity. Furthermore, States did not submit 
multi-year VMT averages in response to the EPA's invitation to submit 
their own VMT data. If the EPA were to construct multi-year averages, 
it is not clear what time frame would be appropriate. The EPA believes 
that the uncertainty related to having to project VMT growth estimates 
over a longer time period is at least as great as the uncertainty 
related to the representativeness of 1995 VMT. For these reasons, EPA 
has chosen to use 1995 VMT for base year and projection year 
inventories.
    Comment: A number of commenters raised various issues about the use 
of the MOBILE5 emission factor model for this analysis. Most of these 
comments focused on specific assumptions or estimates incorporated in 
MOBILE5 which may need to be modified or updated to account for new 
information.
    Response: The EPA is currently developing an updated emission 
factor model called MOBILE6. When final, this model will supersede the 
MOBILE5 model used by the EPA to develop baseline and 2007 emission 
inventories and States' highway vehicle budget components. The concerns 
raised by commenters are being evaluated as part of the MOBILE6 
development process. At the present time, however, MOBILE5 remains 
EPA's official emission factor model. The EPA currently is not able to 
determine whether the highway vehicle emission modeling concerns raised 
by commenters are valid or whether the changes they suggest would raise 
or lower emission estimates; EPA is also not able to quantify the 
effects of commenters' concerns using its current emission models. Some 
of the changes EPA expects to make in its next official emission factor 
model, such as the effects of aggressive driving and air conditioner 
use, are likely to raise emission estimates; others, such as less-rapid 
deterioration of emissions performance than previously forecast, are 
likely to lower emission estimates. Because the overall effect of these 
and other changes cannot yet be determined, the EPA has chosen to 
continue using its current official emission model in today's action.
    As discussed in Section III.F.5, the budgets presented in today's 
action serve as a tool for projecting in advance whether States have 
adopted measures that would produce the required amount of emissions 
reductions, as indicated by the initial demonstration submitted in 
September 1999. The budgets are also a means for determining from 2003 
to 2007 whether States are fully implementing those measures. Thus, the 
budgets are an accounting mechanism for ensuring that the upwind States 
have adopted and implemented control measures that prohibit the 
significant amounts of NOX emissions targeted by section 
110(a)(2)(D)(i)(I). Although EPA's

[[Page 57420]]

projections of emissions from highway vehicles will change as the 
Agency improves its emission models, these changes will not in and of 
themselves require changes in the actions States undertake to reduce 
ozone transport under today's action.
2. Growth
    Comments: The EPA received numerous comments concerning its 
projection of States' 2007 highway vehicle budget components. In 
addition to the changes in baseline VMT discussed previously in Section 
III.D.1 of this notice, the EPA received from a number of States 
proposed revisions to VMT growth estimates and the effectiveness of 
emission control programs.
    Response: In today's action, EPA has implemented the following 
changes it proposed in the NPR in calculating States' 2007 highway 
vehicle budget components. The EPA has used State projections of VMT 
growth from 1995 through 2007 for States that submitted appropriately 
explained projections of VMT growth from 1995 to 2007. For other 
States, EPA projected 2007 VMT levels from the 1995 baseline VMT levels 
using the OTAG projected growth rates.
    As proposed in the NPR, neither the highway vehicle budget 
components nor the overall NOX budgets promulgated in 
today's action alter the existing conformity process or existing SIPs' 
motor vehicle emissions budgets under the conformity rule. The EPA has 
determined that Federal agencies or Metropolitan Planning Organizations 
(MPOs) operating in States subject to today's action do not have to 
demonstrate conformity to the SIP Call budgets or the highway vehicle 
budget component levels used to calculate the budgets. However, areas 
will be required to conform to the motor vehicle emissions budgets 
contained in the attainment SIPs for the new eight-hour standard. For 
their attainment SIPs for transitional ozone nonattainment areas, 
States might seek to rely on the modeling performed for the SIPs 
submitted in response to today's action. To the extent that this 
occurs, the VMT projections and motor vehicle emissions inventories 
associated with today's action could have a role in the conformity 
process, beginning when transitional areas are designated and 
classified in 2000.
3. Budget Calculation
    Background: The EPA proposed highway budget components based on 
projected highway vehicle emissions in 2007 from a base year of 1990, 
assuming implementation of CAA measures, such as inspection and 
maintenance programs and reformulated fuels, measures already 
implemented federally, and those additional measures expected to be 
implemented federally by 2007. The additional Federal measures included 
the National Low Emission Vehicle Standards and the 2004 Heavy-Duty 
Engine Standards. The emission effects of revisions to the Federal 
Emissions Test Procedure, which had also been promulgated in final 
form, were not reflected in the projected 2007 emissions presented in 
the proposal because neither the emissions that this measure is 
designed to control nor the reductions in those emissions expected from 
the test procedure revisions had been incorporated in the projected 
2007 emission estimates or in peer- and stakeholder-reviewed EPA 
emission models. The proposal also did not incorporate any benefits 
from Tier 2 light-duty vehicle standards since the EPA had not yet 
proposed or promulgated regulations concerning the level and 
implementation schedule for Tier 2 standards. Seasonal emissions were 
calculated by estimating emissions for a specific weekday, Saturday and 
Sunday during the ozone season and multiplying by the number of days of 
each type in the ozone season. These estimates were based on 
temperatures and temperature ranges recorded for actual ozone episodes. 
In the NPR, EPA proposed to change this approach to substitute monthly 
average temperatures and temperature ranges for ozone episode-specific 
temperatures when constructing the 2007 budgets. The highway vehicle 
budget components presented in today's notice reflects this change.
    Comment: A number of commenters suggested that the EPA change its 
assumptions regarding emission control programs from those used in the 
NPR. One commenter claimed that the NPR did not include a number of 
cost-effective highway and nonroad mobile source NOX 
reduction programs in its budget calculations. Other commenters 
suggested that the EPA focus more on expanding the RFG and I/M 
programs, adopting gasoline sulfur controls, implementing a 
reformulated diesel fuel program, or implementing the Tier 2 program. 
Contrary to these positions, a number of commenters agreed with the 
EPA's decision not to assume any expansion of the RFG or I/M programs, 
while still other commenters argued that the EPA should not include the 
emission effects of gasoline sulfur controls or reformulated diesel 
fuel in its calculation of State NOX budgets. One commenter 
suggested that the EPA change its NLEV phase-in assumptions to match 
the final NLEV agreement. One commenter asked EPA to include the effect 
of the recent Revised Federal Test Procedure rule, which is aimed at 
reducing excess emissions from aggressive driving or air-conditioner 
use, in its budget calculation.
    Response: Both the NPR and today's action include those mobile 
source reductions which EPA has determined or proposed to determine are 
technologically feasible, highly cost-effective, and appropriate to 
implement on a national basis, and which have been promulgated in final 
form or are expected to be promulgated in final form before States are 
required to submit revised SIPs. The highway vehicle budget components 
include the emission reductions resulting from implementation of the 
NLEV program, including the phase-in schedule agreed to by the States, 
automobile manufacturers, and EPA. The highway budget components do not 
include the effect of Tier 2 light-duty vehicle and truck standards and 
any associated fuel standards since these standards have not yet been 
proposed.
    The extent of the RFG and I/M programs was not assumed to change 
beyond that assumed for the NPR, except for those States who were able 
to demonstrate that the NPR's modeling assumptions did not conform to 
the State's SIP and did not reflect CAA requirements. As discussed 
elsewhere in today's notice and in the NPR, the NOX 
reductions alone from these measures do not appear to be highly cost 
effective in all of the areas that would be subject to reduced budgets. 
Because these measures offer additional benefits beyond NOX 
reductions, specific local areas may determine that these measures are 
appropriate and cost effective given their full range of benefits.
    The baseline and budget calculations include neither the increased 
emissions from aggressive driving or air conditioner use, nor the 
reductions in those emissions resulting from the Revised Federal Test 
Procedure rule. These emission effects are not reflected in EPA's 
MOBILE5a model; they are being evaluated for inclusion in MOBILE6. 
While the EPA has developed a modified version of its MOBILE5 model to 
estimate these effects for its Tier 2 study, this modified model has 
not been used in any regulatory actions and is still subject to 
revision as part of EPA's model development process. As discussed above 
and in Section III.F.5. below, any

[[Page 57421]]

changes by EPA in its emission models will not in and of themselves 
alter the emission reductions States must achieve to comply with the 
requirements of today's action.
    Comment: One commenter suggested that the EPA not split VMT using 
weekend and weekday travel fractions when calculating monthly and 
seasonal total VMT. Another State commenter proposed an alternative 
method for calculating monthly and seasonal VMT from average daily VMT 
which did not rely on the EPA weekend/weekday travel fractions, but 
instead used monthly travel fractions specific to that State. Other 
commenters supported the weekend/weekday inventory modeling approach 
proposed by the EPA.
    Response: The EPA and other organizations have amassed considerable 
evidence that weekend and weekday travel patterns differ significantly. 
The OTAG Final Report requested day-specific inventories for developing 
day-of-the-week activity levels used in emission inventory development 
and episode-specific modeling. Given this requirement, EPA has 
determined that the approach outlined in the NPR is appropriate and 
reasonable. The alternative method using State-specific monthly travel 
fractions as proposed by one State is a reasonable alternative. 
However, because EPA does not have the necessary information to apply 
this method to all other States, EPA did not incorporate this method in 
its analysis.
    a. I/M Program Coverage.
    Comment: One commenter urged the EPA to expand I/M programs to 
cover all urbanized areas with populations above 500,000 as recommended 
by OTAG. Other commenters also requested that EPA expand the I/M 
program or require specific States to adopt specific types of I/M 
programs. By contrast, other commenters supported the I/M approach 
taken by the EPA in the NPR.
    Response: The OTAG recommended that States consider expanding I/M 
programs to cover all urbanized areas with populations above 500,000. 
The EPA has considered this recommendation but does not believe it to 
be appropriate to assume broader I/M implementation in calculating 
State budgets for the reasons outlined in the NPR (62 FR 60355). The 
State budgets promulgated in today's action reflect full implementation 
of I/M as required by the CAA and State SIPs.
    b. Emissions Cap.
    Comment: One commenter suggested that the EPA consider capping 
mobile source emissions, arguing that the proposed rule would place an 
undue burden on stationary sources.
    Response: The State NOX budgets promulgated in today's 
action include the projected emission benefits of those NOX 
controls that the EPA has determined are technologically feasible and 
highly cost effective, as well as additional controls whose 
implementation is not dependent on this rule. While the EPA's analysis 
indicates that certain categories of stationary sources offer the 
potential for large, highly cost-effective NOX emission 
reductions, the State NOX budgets also reflect the emission 
effects of a number of mobile source controls (See Table IV-2). The EPA 
believes that it has applied its criteria for determining which 
controls to assume in State NOX budgets equitably to both 
mobile and stationary sources. In contrast to EGUs and large non-EGUs, 
EPA has not concluded that a mass cap (which would effectively require 
offsets for VMT growth) is highly cost effective. For these reasons, 
EPA does not believe that today's action places an undue burden on any 
emission sector and does not believe that a separate cap on mobile 
source emissions is necessary.
    c. Tier 2 Standards.
    Comment: One commenter requested that EPA include the effects of 
Tier 2 light-duty vehicle standards when calculating State budgets if 
the NLEV program fails. Another commenter suggested that States not be 
permitted to adjust their budgets in case the NLEV program fails.
    Response: This issue is not yet ``ripe'' because NLEV is currently 
being implemented and there are no signs that the program will fail. 
The EPA will consider whether to adjust State budgets if automakers 
representing a significant portion of new vehicle sales withdraw from 
the NLEV program, as discussed in Section III.F.5.
    d. Low Sulfur Fuel.
    Comment: One commenter stated that the EPA disregarded OTAG's call 
for reducing sulfur levels in fuel, which would have the effect of 
reducing NOX emissions.
    Response: The EPA's proposed rule and other actions match the OTAG 
recommendations on fuels, contrary to the commenter's suggestion. The 
OTAG gasoline recommendation stated, ``The USEPA should adopt and 
implement by rule an appropriate sulfur standard to further reduce 
emissions and assist the vehicle technology/fuel system [to] achieve 
maximum long term performance.'' It did not request that EPA implement 
a specific sulfur reduction proposal. The EPA is evaluating the costs 
and benefits of reducing gasoline sulfur levels as part of its proposed 
rulemaking to implement Tier 2 light-duty vehicle and truck standards. 
The EPA is also evaluating the relationship between diesel fuel 
standards and the emission standards as part of (i) its 1999 technology 
review for its 2004 highway heavy-duty diesel engine standards and (ii) 
its 2001 technology review for the Tier 3 and Tier 2 nonroad diesel 
engine standards. Until these evaluations are complete, EPA believes it 
is premature to assume any changes in fuel properties when calculating 
States' highway vehicle budget components.
    e. Conformity.
    Comment: One commenter recommended that NOX 
transportation conformity waivers should lapse in the wake of today's 
action.
    Response: Conformity waivers were granted on an area-by-area basis, 
given the facts of the situation in each local area. Any withdrawal 
should be based on similar local analysis, or upon submittal of a valid 
attainment plan. Today's action is not based on this kind of local 
analysis. Thus, there is no basis for any withdrawal of existing 
NOX transportation conformity waivers. Furthermore, any such 
withdrawal would not alter the Statewide NOX budgets set 
forth in today's action. For these reasons, the EPA has concluded that 
today's action does not alter existing conformity requirements, 
including any NOX conformity waivers.
    Comment: One commenter expressed concern that if current conformity 
budgets do not incorporate the same control assumptions as the States' 
budgets submitted in response to today's rulemaking, the growth in 
areas currently subject to conformity budgets could threaten the 
ability of States to meet the SIP call budgets. The commenter continued 
that failure to tie conformity budgets to transport budgets would allow 
these areas to grow to pre-SIP call control budget levels that could 
cause an exceedance of the Statewide budget. The commenter also stated 
that to address local ozone problems, transportation conformity plans 
should reflect the mobile source controls assumed in the SIP call.
    Response: Conformity budgets cannot be tied directly to the SIP 
Call budgets because the latter are statewide and the former are 
nonattainment-area-specific. The Statewide NOX budgets will 
be enforced as described in today's action, regardless of the 
conformity budgets in specific areas within the affected States. These 
budgets should reflect the actual level of motor vehicle emissions 
which States expect to occur.

[[Page 57422]]

    As noted elsewhere in this section, conformity budgets will reflect 
the mobile source controls assumed in the SIP Call budgets to the 
extent that the attainment SIP ultimately relies upon those controls. 
Today's action does not change the rules governing generation and use 
of emission reduction credits to offset further growth in the 
transportation sector as part of a local area's conformity 
demonstration.

E. Stationary Area and Nonroad Mobile Sources

    Background: The EPA developed the NOX SIP call emissions 
inventory for area and nonroad mobile sources based on data sets 
originating with the OTAG 1990 base year inventory. These base year 
inventories were prepared with 1990 State ozone SIP emission 
inventories supplemented with either State inventory data, if 
available, or EPA's National Emission Trends (NET) data if State data 
were not available. The OTAG 1990 nonroad emission inventories were 
based primarily on estimates of actual 1990 nonroad activity levels 
found in the October 1995 edition of EPA's annual report, ``National 
Air Pollutant Emission Trends.'' In the NPR, EPA proposed switching to 
EPA's 1997 ``Trends'' estimate of 1995 nonroad activity levels.
    For the SNPR, area and nonroad mobile source inventory data for 
1990 were then grown to 1995 using Bureau of Economic Analysis (BEA) 
historical growth estimates of industrial earnings at the State 2-digit 
Standard Industrial Classification (SIC) level. Because BEA data are 
historical documentation of industry earnings, EPA considered these to 
be among the best available indicators of growth between 1990 and 1995 
(63 FR 25915). Once the common base year of 1995 was established for 
these source categories, BEA growth assumptions utilized by OTAG were 
used to estimate the 2007 base case inventory.
1. Base Inventory
    Comment: The EPA received several comments on baseline area and 
nonroad mobile source emission inventories. Several commenters 
submitted estimates of their 1990 nonroad activity levels that differed 
from NPR estimates. One commenter provided statewide 2007 base year 
emissions estimates for numerous area source categories, while others 
provided similar information for 1990 or 1995 emission estimates. Many 
commenters expressed concern with existing area source inventory 
estimates and provided revised county-level area source inventories. 
One commenter suggested using a multi-year activity average to 
establish the nonroad emission baseline, arguing that a multi-year 
average would provide a more representative baseline than would a 
single year's data alone.
    Response: In the NPR and SNPR, EPA asked commenters to provide 
sufficiently detailed information to permit revision to county-level 
emission inventories, in order to allow airshed modeling to be 
performed using the revised inventories. Some proposed area and nonroad 
inventory revisions submitted by commenters were State-wide revisions 
and did not contain sufficient detail to permit the EPA to revise 
county-level nonroad emission inventories. Because the EPA could not 
use these submittals to revise the county-level inventories used as 
inputs to its air quality modeling analyses, these submittals were not 
accepted. Other commenters did provide sufficiently detailed data, and 
EPA revised the appropriate emission inventories to reflect the 
commenters' estimates. These revised inventories were then grown to 
1995 using BEA-derived growth factors, as described above.
    Although EPA proposed in the NPR to switch to a 1995 inventory in 
calculating baseline NOX emissions from nonroad mobile 
sources, EPA has chosen not to do so in today's action. Using the 1995 
inventory presented in the ``Trends'' report as the baseline for 
today's action would have required the use of geographic allocation 
methods that have not undergone peer review and have not been made 
available for public comment by affected interests. The EPA has 
concluded that the use of these unreviewed methods in today's action 
would have deprived stakeholders of adequate opportunity to review, 
understand, and comment on their baseline inventories and the methods 
used to construct them. Hence, EPA has chosen to retain the 1990 
baseline inventories for nonroad mobile sources presented in the NPR 
for today's action, with the changes made in response to comments.
    As discussed above, EPA has chosen to use 1990 nonroad activity 
level estimates as the basis for its nonroad inventory projections. The 
EPA is not aware of any evidence that suggests that 1990 was an 
abnormal year in terms of nonroad activity. Furthermore, States did not 
submit multi-year nonroad activity averages in response to EPA's 
invitation to submit their own nonroad activity data. If EPA were to 
construct multi-year averages, it is not clear what time frame would be 
appropriate. To reduce the impact of unusual years, EPA would have to 
take a long-term average. However, doing so would require EPA to use an 
even earlier year as its base year for nonroad activity and inventory 
projections. The EPA believes that the uncertainty related to having to 
project nonroad activity growth estimates over a longer time period is 
at least as great as the uncertainty related to the representativeness 
of 1990 nonroad activity.
2. Growth
    Comment: Several commenters suggest that the growth factors used to 
determine 2007 stationary area and nonroad mobile source base year 
inventories are inaccurate or inconsistent across regions and 
categories of the inventory. They explained that if growth factors are 
to be used to estimate future base year emissions, consistent national 
or region-wide values should be utilized for all categories across all 
States within the domain. This, they continue, would promote equitable 
potential progress to all areas and not penalize those that have shown 
past poor growth rates. Some commenters go on to state that growth 
rates based on past growth automatically disadvantage States which have 
suffered from unusually low growth rates. In addition to growth rates, 
some commenters provided 2007 base year emission estimates either with 
or without the growth and control information needed to validate their 
calculation.
    Response: As noted above, EPA relied on BEA State-specific 
historical growth estimates of industrial earnings at the 2-digit SIC 
level as among the best available indicators of growth for stationary 
and nonroad area sources. BEA projection factors assume the continuance 
of past economic relationships. These factors are published every five 
years and adjusted to account for recent production and growth trends. 
For this reason, BEA data provide a useful set of regional growth data 
that EPA recommends for use in preparing emission inventory 
projections. It is true that BEA projection factors differ among 
different areas and different source categories because of historical 
differences in industrial growth among those different areas and source 
categories. However, in general, these projection factors offer the 
most reliable indicators of future growth as are available.
    In cases where commenters questioned the use of EPA's growth rates 
but provided no alternative of their own, EPA had little choice but to 
continue to use the BEA-derived growth rates. Some commenters provided 
alternative or supporting information for modification of source 
category or State

[[Page 57423]]

growth estimates. In those cases where a State or industry may have had 
more accurate information than the BEA forecast (e.g., planned 
expansion or population rates), data were verified and validated by the 
affected States and by EPA, and revisions were made to the factors used 
for that category.
3. Budget Calculation
    Background: The EPA proposed nonroad mobile source budget 
components based on projected nonroad mobile source emissions in 2007 
from a base year of 1990. These projections were developed by 
estimating the emissions expected in 2007 from all nonroad engines, 
assuming implementation of those measures incorporated in existing 
SIPs, measures already implemented federally, and those additional 
measures expected to be implemented federally. The additional Federal 
measures include: the Federal Small Engine Standards, Phase II; Federal 
Marine Engine Standards (for diesel engines of greater than 50 
horsepower); Federal Locomotive Standards; and the Nonroad Diesel 
Engine Standards. In the NPR, EPA used the estimates developed by the 
OTAG for nonroad mobile source baseline emissions and growth rates.
    Comments: The EPA received comments to use a State-specific set of 
growth rates for nonroad mobile source emissions.
    Response: The EPA has used State estimates of 1990 nonroad activity 
levels and growth rates for 1990 through 2007 received during the 
comment period to revise its estimates of nonroad NOX 
emissions in 2007, where those State estimates were appropriately 
explained and documented. For other States, the EPA has retained the 
baseline activity levels and growth rates used in the NPR, which in 
turn were based on the growth rates developed for OTAG.

F. Other Budget Issues

1. Uniform vs. Regional Controls
    Background: In the NPR, EPA bases the State budgets upon assumed 
application of reasonable, highly cost-effective NOX control 
measures. These measures were uniform across the 23 affected 
jurisdictions. They consisted of 0.15 lbs/MmBtu for the EGU sector; and 
70 percent control for large, and RACT for medium-sized, non-EGU point 
sources.
    Comments: A number of commenters opposed calculating budgets based 
on uniform emissions reductions and cited the fact that OTAG 
recommended a range of control levels. These commenters offered no 
specific alternatives, such as varying the assumed control levels by 
State or by groups of States, or alternative methods for determining 
different control levels. Numerous comments were received supporting 
the proposed uniform level of emissions reductions.
    Response: The EPA has determined that each of the 23 jurisdictions 
has sources that emit NOX in amounts that significantly 
contribute to downwind nonattainment problems. Moreover, EPA has 
determined that specified levels of control on certain sources in all 
of the jurisdictions would be highly cost-effective. This analysis 
applies with equal force to each of the 23 jurisdictions. It may be 
that emissions from some States have greater ambient impact on downwind 
nonattainment areas than emissions from more distant States. Even so, 
each of the States' NOX emissions have a sufficient ambient 
impact downwind to conclude that those amounts are significant 
contributions and that NOX emissions from all the upwind 
jurisdictions collectively contribute significantly to nonattainment 
downwind. Differentiating the contributions of individual upwind States 
on multiple downwind nonattainment areas is a highly complex task. The 
contributions of individual States are likely to vary from downwind 
area to downwind area, from episode to episode, and from NAAQS to 
NAAQS. Accordingly, it would be extremely complex to develop a budget 
for each State that would reflect the different impacts of its sources' 
emissions on different downwind States.
    Among many factors that EPA considered in weighing whether to 
finalize a uniform control level or regional control levels in 
calculating States' emission budgets was the concern that different 
controls in one part of the SIP call area in combination with an 
interstate emissions trading program may lead to increases in pollution 
within areas having more restrictive controls. That is, if unrestricted 
interstate emissions trading were allowed on an one-for-one basis, 
emissions reductions might be expected to shift away from States 
assigned more restrictive controls to States which received less 
restrictive control requirements due to the lower control costs likely 
to exist in States with less restrictive controls. This may result in 
emissions above the budget level in areas with more restrictive 
controls.
    There are two alternatives for addressing the problem of shifting 
emissions. The first is to allow trading only within uniform control 
regions, but not between regions with NOX budgets reflecting 
different levels of control. The advantage to this approach is that it 
provides a straightforward way of preventing trades of excess emissions 
into regions with more stringent standards. However, a trading program 
that covers a smaller market area will provide less flexibility and 
reduce the possible savings for the affected sources as compared with 
larger trading programs. The second alternative is to establish a 
trading ratio for trades between regions, to reflect the differential 
impact of the emissions on nonattainment. The trading ratio should 
reflect the relative contribution of emissions to downwind non-
attainment problems. The advantage to this approach is that it provides 
the flexibility for trades between regions when the benefits of such 
trades are large, while discouraging a shift of excess emissions into 
regions with more stringent standards. However, none of the comments on 
the proposal included a justification or description for trading 
ratios, which would reflect the differential environmental implications 
and discourage inappropriate shifting of excess emissions.
    The ozone problem in the Eastern United States is the result of a 
large number of different types of sources which affect widely 
distributed nonattainment areas at different times under changing 
weather patterns such that a broadly-established control program is 
necessary. The EPA believes a reasonable strategy is to apply the most 
cost-effective control strategies uniformly in contributing States in 
order to eliminate the combined significant contribution from these 
multiple sources in multiple States.
    The EPA analyzed costs and air quality benefits for two regional 
control level options that were based on a varying level of controls in 
different parts of the 23 jurisdictions. The analysis did not show that 
these two regional control alternatives would provide either a 
significant improvement in air quality or a substantial reduction in 
cost. An analysis of the costs and benefits of different control 
options can be found in the docket. On the basis of the analysis, EPA 
believes an alternative approach with differentiated NOX 
budgets and regionally differentiated trading would not yield 
significant additional air quality benefits or cost savings vis a vis a 
regionwide trading program based on uniform NOX budgets.
2. Seasonal vs. Annual Controls
    Comments: One commenter suggested that controls should be required 
for the

[[Page 57424]]

entire year rather than just during the 5-month ozone season as 
proposed.
    Response: The EPA recognizes that control of nitrogen oxide 
emissions would likely produce non-ozone benefits, as well as ozone 
benefits. For example, NOX control would likely reduce 
surface water acidification or eutrophication of surface waters. Annual 
control of NOX may have a greater impact on winter and 
spring NOX emissions, and therefore on acidification and 
eutrophication, than ozone season (summer) NOX control to 
the extent that acidification and eutrophication result from the 
release of nitrogen compounds from snowpack during snowmelt and rain in 
the spring. Control of NOX emissions also reduces fine 
particulates and regional haze, so that annual control of 
NOX emissions would result in greater non-ozone benefits. 
However, the commenter's suggestion that EPA analyze the costs of, and 
assume in calculating the budgets, annual NOX control to 
address non-ozone problems is outside the scope of this rulemaking 
proceeding. Here, EPA has proposed a NOX SIP call to address 
the failure of certain SIPs to prohibit sources from emitting 
NOX in amounts that contribute significantly to 
nonattainment (or interfere with maintenance of attainment) of the 
ozone NAAQS during the ozone season.
    In analyzing the benefits of ozone season NOX control 
under the proposed NOX SIP call for purposes of the RIA 
(though not as a basis for the decisions in today's rule), EPA 
considered both the ozone and non-ozone benefits. Non-ozone benefits 
include the impact of ozone season NOX control on 
acidification and eutrophication. In particular, emission modeling 
performed by EPA indicates that the SIP Call would reduce wintertime 
NOX emissions. This results in part because, once installed 
to comply with the NOX SIP call, some NOX control 
systems (e.g., low NOX burners which alter the combustion 
process and cannot simply be turned off) would reduce emissions 
throughout the year, even though the NOX limits would be 
seasonal. Also see Section IX.
3. Full vs. Partial States
    Background: In the NPR, the Agency indicated it was proposing to 
include entire States rather than exempting portions of States in the 
development of emissions budgets. The Agency's decision to include full 
States was based upon three major points: (1) The division of 
individual States by OTAG was based, in part, on computational 
limitations in OTAG's modeling analyses; (2) the additional upwind 
emissions from full, as opposed to partial, States would provide 
additional benefit to downwind nonattainment areas; and, (3) Statewide 
emissions budgets create fewer administrative difficulties than a 
partial-State budget.
    Comments: During the two comment periods, 43 comments were received 
which specifically addressed some or all of the major points outlined 
above. The underlying theme throughout the comments on this issue was 
that the States and EPA had undertaken a comprehensive, scientifically 
credible modeling/analysis study during the OTAG, and that the Agency 
should follow OTAG's recommendations on this issue (i.e., allow for 
partial-State emission budgets). Another common theme was that the 
administrative difficulties outlined by the Agency in the NPR were 
exaggerated, and that the affected States should be allowed to generate 
partial-State, as opposed to statewide, emissions budgets, if their 
State considered it feasible to do so. Comments were received that 
portions of Alabama, Georgia, Michigan, Missouri, North Carolina, and 
Wisconsin should be excluded from the SIP Call.
    Response: The underlying concepts for responding to these comments 
are (a) that the atmosphere is constantly in motion and has no 
limitations at geo-political boundaries, and (b) that the larger the 
geographic area that is controlled, the greater the downwind benefits. 
For the States requesting partial-State emissions budgets, there are 
NOX emissions throughout these entire States. The EPA did 
State-specific modeling for each of the affected States, and these 
additional modeling analyses support the concept of statewide emissions 
budgets for each of the affected States. Furthermore, it is a 
reasonable assumption, given the nature of ozone chemistry, that if 
emissions from part of a State contribute significantly to downwind 
nonattainment or maintenance problems, emissions from the entire State 
contribute significantly to downwind nonattainment or maintenance 
problems. In each of the affected States, there is no peculiar 
meteorological phenomenon that would indicate that emissions from some 
portion of that State would not impact downwind nonattainment or 
maintenance problems. Thus, based on additional EPA modeling analyses 
and their technical interpretation, EPA is not promulgating partial-
State emissions budgets. Since each State has the flexibility to 
determine which sources to control in order to meet the budget, a State 
can structure its control strategy to require fewer reductions in 
certain portions of the State and greater controls in other areas, as 
long as the significant amounts of emissions are eliminated.
4. NOX Waivers
    Comments: The EPA received several comments supporting the approach 
outlined in the NPR in which EPA would treat areas that had previously 
received NOX waivers under section 182(f) of the CAA in the 
same manner as other areas in the SIP call. The comments stated that 
(1) special treatment (i.e., higher budget) for the waiver areas would 
increase the burden on downwind States; (2) numerous modeling efforts, 
including OTAG's, have shown that such disbenefits are generally minor 
and occur on days with low ozone concentrations; (3) disbenefits are 
small when upwind NOX reductions are modeled; (4) 
disbenefits are better addressed at the local level; and (5) States 
already have the flexibility to deal with NOX disbenefits, 
if any, through the budget and trading by meeting the budget through 
NOX emission decreases in other areas of the State or 
acquiring allowances through trading. In addition, some commenters 
requested EPA to revoke waivers previously granted. Commenters also 
noted that the localized disbenefits are no less of a problem in the 
Northeast than in the Midwest.
    Numerous comments were also submitted which oppose the approach 
outlined in the NPR. The comments generally stated that in States with 
NOX waiver areas, the NOX budget should be 
increased where NOX decreases lead to ozone increases; 
otherwise States might seek reductions disproportionately outside the 
sensitive areas, resulting in cost-effectiveness levels greater than 
the $2000 per ton framework described in the SIP call proposals. 
Comments referred to disbenefits in Cincinnati, Louisville and the 
Chicago/Gary areas. Many commenters suggested that EPA wait for further 
modeling analyses to be completed and that the zero-out runs are 
inappropriate for evaluating the NOX disbenefit issue. Some 
stated that the NOX budget might interfere with local 
attainment and harm local public health. Other comments recommended 
that EPA consider the impact of additional VOC costs that might be 
needed to offset local ozone increases.
    Response: In today's final rulemaking, EPA is setting 
NOX emissions budgets for each of the jurisdictions affected 
by this action. These budgets are set in the same manner for areas 
without NOX waivers as areas with NOX waivers, 
except in the case of NOX waivers granted for I/M programs. 
Although

[[Page 57425]]

adverse comments were submitted, none of them provided any modeling 
analysis or support documentation showing how a State or States with 
NOX waiver areas should be assigned a larger budget or 
proposing a specific alternative approach for assigning those budgets. 
In contrast, modeling described by EPA in the NPR and SNPR as well as 
additional modeling conducted by the Agency and some commenters 
continues to show that the benefits of NOX emissions 
decreases greatly outweigh any disbenefits. These findings are 
discussed in Section IV, and summarized below.
    The EPA considered the strengths and limitations in the commenters' 
modeling analyses in evaluating whether the technical evidence 
presented in the comments supports the arguments made by the 
commenters. The EPA's review of the commenters' modeling indicates that 
in general (a) downwind ozone benefits increase as greater 
NOX controls are applied to sources in upwind States, (b) 
the net benefits of NOX control at the level of the SIP Call 
outweigh any local disbenefits, and (c) upwind NOX 
reductions tend to mitigate local disbenefits in downwind areas.
    One commenter, the Lake Michigan Air Director's Consortium (LADCO), 
submitted air quality modeling directed toward investigating the 
disbenefits in nonattainment areas around Lake Michigan due to the 
NOX controls in the SIP Call proposal. The commenter's 
general finding was that the greatest ozone decreases with these 
NOX controls occur on high ozone days, while the greatest 
disbenefits occur on low ozone days. The EPA concurs with this finding, 
based on a review of the technical information provided by the 
commenter. Specifically, there were no predicted increases in ozone 
(i.e., disbenefits) in peak 1-hour ozone on any of the 4 days modeled 
by LADCO that had daily maximum 1-hour concentrations >=125 ppb in the 
Base Case. Also, on the 3 low ozone days which had predicted 
disbenefits, the increases were not large enough to result in a peak 
value >=125 ppb. Concerning 8-hour concentrations, only 1 of the 9 days 
with a predicted 8-hour daily maximum concentration >=85 ppb had an 
increase in peak ozone due to the SIP Call NOX controls. 
Also, there was a small disbenefit on the one day modeled which had an 
8-hour daily maximum concentration <85 ppb, but the magnitude of the 
disbenefit on this day was relatively small and did not cause the 8-
hour peak value to exceed 85 ppb. Thus, based on this evaluation, EPA 
generally found that the submitted modeling did not refute the overall 
conclusions EPA has drawn concerning the impacts of NOX 
emissions in the relevant geographic areas.
    As described in the NPR, the OTAG process included lengthy 
discussions on the potential increase in local ozone concentrations in 
some urban areas that might be associated with a decrease in local 
NOX emissions. The OTAG modeling results indicate that urban 
NOX emissions decreases produce increases in ozone 
concentrations locally, but the magnitude, time, and location of these 
increases generally do not cause or contribute to high ozone 
concentrations. That is, NOX reductions can produce 
localized, transient increases in ozone (mostly due to low-level, urban 
NOX reductions) in some areas on some days, but most 
increases occur on days and in areas where ozone is low. In the SNPR, 
EPA documented the estimated ozone benefits of the proposed Statewide 
NOX budgets based on an air quality modeling analysis. The 
major findings of that analysis include: Any disbenefits due to the 
NOX reductions associated with the budgets are expected to 
be very limited compared to the extent of the air quality benefits 
expected from these budgets.
    The results of EPA's assessment of the comments and available 
modeling corroborate and extend the findings presented in the SNPR. 
Thus, with respect to regional ozone transport and today's final 
action, EPA believes it is not appropriate to give special treatment to 
areas with NOX waivers.
    Several nonattainment areas in the 23 jurisdictions were granted 
waivers from certain NOX requirements in past rulemaking 
actions. In the Federal Register notices granting the waivers, EPA 
stated that the continued approval of these waivers is contingent on 
the results of the final ozone attainment demonstrations and plans (See 
61 FR 2428 January 26, 1996, LADCO). The attainment plans will 
supersede the initial modeling information which was the basis for 
waivers EPA granted (e.g., the LADCO waiver). The attainment plans were 
due in April 1998 and were to incorporate the results of the OTAG 
process. The EPA's rulemaking action to reconsider the initial 
NOX waiver may occur simultaneously with rulemaking action 
on the attainment plans. Therefore, as these new modeling analyses are 
submitted to EPA, they will be reviewed to determine if the 
NOX waiver should be continued, altered, or removed.
    As discussed above, EPA has accounted for the continued presence of 
NOX waivers for I/M programs in modeling States' 
NOX budgets. Historically, EPA gives States considerable 
latitude in designing their I/M programs. This latitude is granted in 
recognition of the unique economic and air quality circumstances faced 
by each State. States have used this latitude to develop a range of I/M 
program designs. Some States have adopted EPA-recommended enhanced I/M 
programs; other States have adopted different I/M program designs.
    The EPA acknowledges that some of the States granted NOX 
waivers may be able to modify their programs to obtain NOX 
reductions at minimal cost. However, some of the States which have been 
granted an I/M NOX waiver have developed unique I/M program 
designs in terms of the model years covered, the emission testing 
equipment used, and possibly the number, location, and design of the 
testing and repair stations. The cost for these States to modify their 
I/M programs to obtain NOX reductions are likely to exceed 
the level that EPA has determined to be highly cost-effective for the 
purpose of reducing ozone transport. As a result, the EPA has chosen to 
not include additional emissions reductions due to I/M NOX 
programs when calculating NOX budgets.
5. Recalculation of Budgets
    In the NPR, the EPA made proposals concerning what would happen if 
additional information becomes available after EPA's final rulemaking 
action. Examples of such information might include: (a) Source-specific 
information useful in determining RACT, (b) revised growth or other 
assumptions, (c) revised models and inventory estimates, (d) 
unexpectedly low implementation rates for NLEV, and (e) other new 
federal measures, i.e. Tier 2 controls. In the Recalculation of Budgets 
Section of the NPR, EPA proposed that if additional data become 
available after EPA's final rulemaking action, such data could be 
considered prior to State submittal of revised SIPs. The EPA asked for 
comments on this approach.
    Most of the comments received were in favor of allowing States to 
adjust their emission budgets based on the most recent available data 
on emissions and RACT levels. There were several comments that any new 
calculation methodologies should be applied across all States and be 
approved at EPA Headquarters, and that all States should use the same 
methodology.
    A few commenters did not agree, however. One said that EPA should 
not recalculate the budgets upward. Another said there should be no 
downward ratcheting of budgets. One

[[Page 57426]]

commenter said that it would be premature to assume that as new 
information becomes available the budget should be adjusted to reflect 
this. According to this commenter, it would be more appropriate to 
perform a complete air quality modeling analysis to determine if an 
adjustment in States' NOX budgets is in order.
    The divergent views reflected in these comments has convinced EPA 
that it should clarify the role of the budgets in this rule. In light 
of that role, as explained below, EPA has decided to allow only a 
limited opportunity to revise the budgets in the very near term. 
However, under the approach the Agency is following, the rule would not 
penalize States for not ultimately achieving the budgets, if the State 
initially projected compliance using the data set forth in this rule, 
and the State has fully implemented all of the measures reflected in 
those initial projections, and the measures are as effective in 
reducing NOX emissions as they were projected to be in the 
State plan.
    As explained in the NPR, SNPR, and above, EPA based the budgets on 
its choice of measures that are highly cost-effective and therefore are 
the easiest for upwind States to implement to reduce transport. 
However, EPA sought to structure the rule to give the upwind States a 
choice of which mix of measures to adopt to achieve the aggregate 
amount of required NOX emissions reduction.
    To offer the States this choice, EPA employed a multi-step approach 
leading to a numerical budget for each State. In the first step, EPA 
projected the mass emissions for EGUs and industrial boilers out to 
2007, taking into account measures required under the CAA and projected 
growth. The result was a base case 2007 subinventory for each of those 
two categories. Next, EPA projected the 2007 mass emissions for other 
sectors of the emission inventory (e.g., mobile sources), again taking 
into account projected growth and measures required under the CAA and 
existing SIPs, thereby creating a base case 2007 subinventory for each 
of them as well. The aggregation of all of the base case 2007 
subinventories is the complete base case 2007 inventory. The EPA then 
applied cost-effective control measures to the EGU, industrial boiler 
and other non-EGU source categories as explained in section III., to 
determine the amount of the reductions from these categories. The EPA 
applied control measures to the base case inventory to develop the 
final budget. Thus, the final budget is the sum of (1) the emissions 
remaining after application of the cost-effective control measures to 
the subinventories for the categories for which controls are assumed 
for purposes of budget calculation and (2) the emissions in the base 
case 2007 subinventories for the categories for which EPA assumed no 
controls.
    The rule then requires each upwind State to use the same base case 
2007 inventory in its 1999 SIP submittal as EPA used in developing the 
State's budget. In that SIP submittal, the State must show that the 
measures it has adopted will achieve the same aggregate emissions 
reductions as the control strategies assumed by EPA in developing the 
State's budget. More specifically, to demonstrate compliance with the 
SIP call, a State must adopt and implement control measures that are 
projected to achieve the aggregate emissions reductions determined by 
EPA based on the application of highly cost-effective controls to EGUs, 
industrial boilers and other affected non-EGUs. While a State may 
choose to achieve those reductions through application of measures 
other than those used by EPA in calculating required reductions, any 
measures it adopts must achieve the reductions assumed by EPA in the 
development of its budgets.
    The control measures that the State chooses to require will become 
the enforceable mechanism under the NOX SIP call. If a State 
elects to regulate boilers, turbines or combined cycle units that are 
greater than 250 mmBtu/hr-- regardless of whether they are connected to 
an electrical generator of any size--or to regulate boilers, turbines 
and combined cycle units that serve electrical generators greater than 
25 Mwe, regardless of the heat input capacity of the unit, the State 
must provide mass emissions limits or their equivalent (see section 
VI.A.2) for these sources or source categories. The mass emissions 
limits may be set on a source-by-source basis or may be set for an 
entire group of sources allowing trading between the sources. These 
mass emission limits must assume growth no greater than EPA's 
calculations. Any growth that occurs in that category would have to be 
accommodated within the mass emission allocations provided by the State 
for that category, even if the growth in that category should prove to 
exceed EPA's projections. This is appropriate because as discussed in 
the SNPR and Section VI.A.2. of today's preamble, EPA believes that the 
control approaches, growth assumptions, and monitoring for this group 
of sources have advanced to the point that complying with, tracking, 
and enforcing a maximum mass emissions limit is reasonable. 
Furthermore, based on the analyses in the RIA, EPA believes that mass 
emission limits remain highly cost-effective for these categories when 
growth is accommodated within the limits. The EPA modeled the expected 
growth in capacity and capacity utilization of the source categories 
listed above based on growth assumptions in the IPM that have been 
subject to extensive public comment and refinement over a several-year 
period. On the basis of their growth, assumptions and assumed emissions 
rates, EPA determined that mass emission limits would remain highly 
cost-effective when new sources are covered within the limits. EPA 
projects that even if actual growth for this group of sources exceeds 
the projected growth by over one-third, mass emission limits would 
remain highly cost-effective according to the criteria used for this 
rule.
    For other categories, EPA will not require a State to remain within 
a mass emission allocation. Today's rule does require a State to use 
the base case 2007 inventory in its budget demonstration. However, the 
rule does not require States to obtain additional reductions in cases 
where a State's 2007 emissions exceeds its budget due to higher than 
expected emissions from source categories other than the categories 
listed above (certain boilers, turbines, and combined cycle units). 
These exceedances may be the result of growth that exceeds projections 
for those source categories. However, if a State elects to control 
these other source categories to achieve the required reductions in 
whole or part, the adopted measures must be as effective in reducing 
NOX emissions as they were projected to be in the State 
plan. Any failure by a State to adopt measures adequate to achieve 
reductions equal to the required amount would be treated as 
noncompliance with this rule. Any failure by the State to implement 
these measures by the appropriate date would be considered a failure to 
implement those measures.
    In contrast, the overall budget number itself is not enforceable 
against the State. The budget serves as a tool for projecting in 
advance whether a State has adopted measures that would produce the 
required amount of emissions reductions, as indicated by the initial 
demonstration submitted in September 1999. The budgets are also a means 
for determining from 2003 to 2007 whether States are fully implementing 
those measures. Thus, the budgets are an accounting mechanism for 
ensuring that the upwind States have adopted and implemented control 
measures that prohibit the significant

[[Page 57427]]

amounts of NOX emissions targeted by section 
110(a)(2)(D)(i)(I).
    Given that States will not be subject to enforcement actions if 
emissions in 2007 from uncontrolled sectors exceed the base case 2007 
inventory projections, EPA does not intend to revise those projections 
merely because such new information becomes available over time. 
Rather, EPA intends to allow commenters an additional opportunity to 
request revisions to the source-specific data used to establish each 
State's budget in this SIP call. This opportunity will be made 
available during the first sixty days of the 12-month period between 
signature of today's rule and the deadline for submission of the 
required SIP revisions (i.e., November 23, 1998). Commenters would need 
to submit any proposed changes in their inventories to the EPA Air and 
Radiation docket (A-96-56) within that sixty day period. Individuals 
interested in modifications requested by commenters may review the 
materials as they are submitted and available in the docket. At the end 
of this period, EPA will, within sixty days, evaluate the data 
submitted by commenters and, if it is determined to be technically 
justified, revise this rule to incorporate it into the State budget 
determinations. For a comment to be considered, the request for 
modification must be submitted in electronic format containing, at a 
minimum, the data elements listed below for each source category. 
Additionally, no comment will be considered unless information is 
provided to corroborate and justify the need for the requested 
modification. For example, corroborating information in the case of the 
EGUs can be the inclusion of copies of each source's official same year 
EIA 860 or 861 form submissions that support the requested change. For 
non-EGUs, corroborating information can include 1995 operational and 
emissions information officially submitted (during that time period) by 
the source to a federal, State, or local government regulating entity.
    Each request for modification of data for EGU sources must include 
the following information:
     Federal Information Placement System State Code.
     Federal Information Placement System (FIPS) County Code.
     Plant name.
     Plant ID numbers (ORIS code preferred, State agency 
tracking number also or otherwise).
     Unit ID numbers (a unit is a boiler or other combustion 
device).
     Unit type (also known as prime mover; e.g., wall-fired 
boiler, stoker boiler, combined cycle, combustion turbine, etc.).
     Primary fuel on a heat input basis.
     Maximum rated heat input capacity of unit.
     For electrical generating units, nameplate capacity of the 
largest generator the unit serves.
     For 1995 and 1996 ozone season heat inputs.
     1996 (or most recent) average NOX rate for the 
ozone season.
     Latitude and longitude coordinates.
     Stack parameter information (height, diameter, flow, 
etc.).
     Operating parameters (hours per day, seasonal throughput, 
etc.).
     Identification of specific change to the inventory, and
     The reason for the change.
    Each request for modification of data for non-EGU point sources 
must include the following information:
     Federal Information Placement System State Code.
     Federal Information Placement System (FIPS) County Code.
     Plant name.
     Facility primary standard industrial classification code 
(SIC).
     Plant ID numbers (NEDS, AIRS/AFS, and State agency 
tracking number also or otherwise).
     Unit ID numbers (a unit is a boiler or other combustion 
device).
     Primary source classification code (SCC).
     Maximum rated heat input capacity of unit.
     1995 ozone season or typical ozone season daily 
NOX emissions.
     1995 existing NOX control efficiency.
     Latitude and longitude coordinates.
     Stack parameter information (height, diameter, flow, 
etc.).
     Operating parameters (hours per day, seasonal throughput, 
etc.).
     Identification of specific change to the inventory, and
     The reason for the change.
    Each request for modification of data for stationary area and 
nonroad mobile sources must include the following information:
     Federal Information Placement System State Code.
     Federal Information Placement System (FIPS) County Code.
     Primary source classification code (SCC).
     1995 ozone season or typical ozone season daily 
NOX emissions.
     1995 existing NOX control efficiency.
     Identification of specific change to the inventory, and
     The reason for the change.
    Each request for modification of data for highway mobile sources 
must include the following information:
     Federal Information Placement System State Code.
     Federal Information Placement System (FIPS) County Code.
     Primary source classification code (SCC) or vehicle type.
     1995 ozone season or typical ozone season daily vehicle 
miles traveled (VMT).
     1995 existing NOX control programs.
     Identification of specific change to the inventory, and
     The reason for the change.
    After this initial ``shake out'' period before submission of the 
SIP revisions, EPA will not adjust inventories or the resulting State 
budgets merely because some new information on a segment of EPA's 
projections comes to its attention. However, when EPA reviews each 
State's reports, it will pay special attention to the causes for any 
exceedance of the portions of the inventory that the State is 
controlling as a means to meet today's rule. If a State exceeds its 
budget because of greater-than-expected growth in areas not having 
additional controls, EPA would not penalize the State by requiring the 
State to offset those increased emissions. Rather, EPA would use the 
base case projections for all sectors (as revised after the initial 
period described above) and focus on whether the State had implemented 
the measures that its 1999 demonstration had shown would, based on 
those base case inventories, achieve the budget levels. Similarly, the 
rule would not penalize the State if components in the budget prove 
inaccurate because of changes in models (e.g., the release of an 
updated MOBILE model) or because of technical errors (e.g., the size of 
a unit was incorrectly identified in the inventory, a unit was double-
counted, or the RACT level assumed in the base is different from what 
the State ultimately selected as RACT with EPA approval).
    In the NPR, EPA also raised the question of what would happen if 
EPA adopts national measures beyond what EPA already assumed in the 
base case 2007 inventory. The EPA indicated that it could use either of 
two approaches in response: (1) States could receive credits for the 
real emission reductions that result from the new Federal measures and, 
therefore, implement a smaller portion of its planned emission 
reductions, or (2) States would be required to continue to implement 
the measures in their revised SIPs because affected States are required 
to continue to achieve emissions reductions equivalent to those which 
can be achieved through application of highly cost-effective control 
measures.

[[Page 57428]]

    One commenter supported the emission reduction credit for State 
SIPs resulting from new Federal national measures adopted after the 
State emission budgets are defined but before 2007. According to this 
commenter, in such a case the State could implement a smaller portion 
of its planned emission reductions because of the reduction brought 
about by the Federal national rule. Another commenter said the EPA 
should allow full credit for all Federal measures and encouraged the 
EPA to timely implement and adopt all Federal measures. A State said 
States should be allowed to take full SIP credit for Federal measures 
which are implemented in these States. According to one commenter, not 
allowing States to take credit for new Federal measures would have the 
effect of downward ratcheting of NOX budgets. Other States 
said new Federal measures not accounted for in the SIP call should not 
be used to offset State measures required to achieve the mandated 
NOX emissions reductions.
    The EPA has decided to adopt the second approach described above. 
Thus, EPA's adoption of a national measure not reflected in the base 
case 2007 inventory would not allow the State to avoid a measure that 
would otherwise be needed to demonstrate that the State will achieve 
the required reductions. As stated above, the SIP must prohibit all 
emissions that contribute significantly to downwind nonattainment and 
maintenance problems. The State therefore is required to eliminate an 
amount of emissions corresponding to what is achievable with the highly 
cost-effective measures identified in this notice. The comments 
received have not provided an adequate basis for concluding that EPA's 
adoption of an additional national measure justifies scaling back on 
that requirement. For that reason, EPA will not allow States to adjust 
the base case 2007 inventory inventories to reflect any such additional 
national measures. Rather, for these reports the States should continue 
to use the base case 2007 inventory set forth in this rule.
    In the SNPR, EPA also discussed establishing a process for 
reassessing the State budgets for the post-2007 timeframe. Today's 
final rule is based on analyses using the most complete, 
scientifically-credible tools and data available for the assessment of 
transport. The EPA expects that there will be a number of updates and 
refinements in air quality methodologies and emissions estimation 
techniques over the next 10 years. Therefore, EPA intends to reassess 
ozone transport using the latest emissions and air quality monitoring 
data and the next generation of air quality modeling tools. The 
reassessment will include an evaluation of the effectiveness of the 
regional NOX measures States have implemented in response to 
today's final rule. Modeling analyses will be used to evaluate whether 
additional local or regional controls are needed to address residual 
nonattainment in the post-2007 timeframe. The assessment will also 
examine differences in actual growth versus projected growth in the 
years up to 2007 as well as expected future growth throughout the 
entire OTAG region. The reassessment will also review advances in 
control technologies to determine what reasonable and cost-effective 
measures are available for purposes of controlling local and regional 
ozone problems. In addition, EPA will continue to look at the issues 
that surround the use of output-based State budget allocations. Based 
on this reassessment, EPA may establish new budget levels and 
allocation mechanisms for the post-2007 timeframe. The current budget 
levels and the measures used to comply with today's final rule will 
remain in effect until EPA takes action on establishing new State 
budgets.
6. Compliance Supplement Pool
    The EPA has received comments expressing concern that some sources 
may encounter unexpected problems installing controls by the compliance 
deadline that, in turn, could cause unacceptable risks for a source and 
its associated industry. More specifically, commenters have expressed 
concerns related to the electricity industry. If unexpected problems 
arise for specific sources that are used to generate electricity, some 
commenters believe that compliance with the May 1, 2003 deadline could 
adversely impact the reliability of the electricity supply. Commenters 
that raised concerns regarding the compliance deadline generally 
supported additional compliance flexibility for the SIP call.
    In both the NPR and SNPR, EPA solicited comment on a number of 
provisions that would provide additional flexibility to both States and 
sources for the requirements of the NOX SIP call. In the 
NPR, EPA proposed that the NOX SIP call would require full 
implementation of controls by no later than September 2002, but 
solicited comment on the range of implementation dates from between 
September 2002 and September 2004. In addition to the compliance 
deadline, EPA also solicited comment on the role of banking as a 
separate compliance flexibility for the NOX SIP call. 
Banking may generally be defined as allowing sources that make 
emissions reductions beyond current requirements to save and use these 
excess reductions to exceed requirements in a later time period. 
Depending upon the design of a trading program, banking provisions can 
provide companies greater latitude for when controls are installed at 
particular sources. In the SNPR, EPA presented a range of options for 
incorporating banking in the NOX Budget Trading Program 
including early reduction provisions and phasing in controls. The EPA 
received many comments supporting banking in the NOX Budget 
Trading Program and also as a general flexibility mechanism that should 
be permissible for any State program used to comply with the 
NOX SIP call.
    In response to comments supporting an extended compliance deadline, 
EPA has moved the deadline from the proposed date of September 2002 in 
the NPR to May 1, 2003. As discussed further in Section V, this change 
provides sources 7-8 additional months for implementing control 
requirements while ensuring that controls are fully implemented by the 
2003 ozone season. The EPA believes that the compliance date of May 1, 
2003 for NOX controls to be installed to comply with the 
NOX SIP call is a feasible and reasonable deadline. See 
Section V.A.1. and the technical support document ``Feasibility of 
Installing NOX Control Technologies By May 2003'' for 
further discussion.
    To provide additional flexibility to States and sources for 
complying with the NOX SIP call beyond the extension of the 
compliance deadline, EPA is establishing banking provisions and a 
compliance supplement pool in today's final rule. The banking 
provisions are outlined in Section III.F.7. The compliance supplement 
pool is a voluntary provision that provides flexibility to States in 
addressing concerns associated with full compliance by May 1, 2003. 
Each State will be able to use the pool to cover excess emissions of 
sources that are unable to meet the compliance deadline during the 2003 
and 2004 ozone seasons. The pool may be used to credit sources that 
make early reductions and to directly delay the compliance deadline for 
specific sources. Credits issued from the compliance supplement pool 
will not be valid for compliance past the 2004 ozone season. The EPA 
established the compliance supplement pool by calculating one pool for 
the entire NOX SIP call region. The pool was then allocated 
to the States in proportion to the size of the emissions reduction they 
are required to achieve under the NOX SIP call so that each

[[Page 57429]]

State has its own compliance supplement pool. The size of each State's 
compliance supplement pool and the procedures that will apply to the 
use of the pool are described below.
    a. Size of the Compliance Supplement Pool. The EPA believes it is 
important for the size of the pool to be capped. Capping the pool makes 
it possible to estimate the potential impact that the compliance 
supplement pool may have on NOX emissions during the 2003 
and 2004 ozone seasons. Furthermore, EPA does not anticipate problems 
for sources in meeting the May 1, 2003 deadline. If there are such 
cases, they should be relatively few in number. Therefore, the size of 
the pool only needs to be large enough to cover the limited potential 
for unexpected compliance delays.
    Today's final rule sets the size of the regional compliance 
supplement pool at 200,000 tons. The EPA believes this is a reasonable 
size for the pool given the analyses that were used in establishing the 
State NOX budgets for today's final rule. As discussed in 
Section V.A.1., EPA believes the most cost-effective control strategies 
available to comply with the proposed budgets include post-combustion 
controls (Selective Catalytic Reduction [SCR] and Selective Non-
catalytic Reduction [SNCR]) and combustion controls (e.g., low 
NOX burners, overfire air, etc.) on large electric 
generating units and large non-electric generating units. For the 
reasons cited in Section V.A.1., EPA estimates that the implementation 
of SCR controls is potentially more complicated and requires more time 
than SNCR or combustion controls and, therefore, would determine what 
the longest schedule would be for full implementation of the assumed 
NOX controls. Since EPA estimates that a single SCR 
installation will take about 23 months, EPA expects the first SCR 
installations to be completed in 2001. Since compliance is required by 
2003, one can assume 33 percent of SCR capacity will be installed each 
year from 2001 to 2003. The 200,000 ton number is sufficient to cover 
the excess emissions that must be offset if one year's worth of SCR 
installations were delayed by a year. Table III-3 shows each State's 
compliance supplement pool. The 200,000 tons were allocated to States 
in proportion to the size of the emissions reduction they are required 
to achieve under the NOX SIP call. The EPA used this 
allocation methodology based on the assumption that the need for the 
pool would be directly related to the magnitude of the emissions 
reductions required in each State to comply with the NOX SIP 
call.

                                 Table III-3.--State Compliance Supplement Pools
                                                     [Tons]
----------------------------------------------------------------------------------------------------------------
                                                                                                    Compliance
                      State                            Base           Budget          Tonnage       supplement
                                                                                     reduction         pool
----------------------------------------------------------------------------------------------------------------
Alabama.........................................         218,610         158,677         59,933           10,361
Connecticut.....................................          43,807          40,573          3,234              559
Delaware........................................          20,936          18,523          2,413              417
District of Columbia............................           6,603           6,792           (189)               0
Georgia.........................................         240,540         177,381         63,159           10,919
Illinois........................................         311,174         210,210        100,964           17,455
Indiana.........................................         316,753         202,584        114,169           19,738
Kentucky........................................         230,997         155,698         75,298           13,018
Maryland........................................          92,570          71,388         21,182            3,662
Massachusetts...................................          79,815          78,168          1,648              285
Michigan........................................         301,042         212,199         88,842           15,359
Missouri........................................         175,089         114,532         60,557           10,469
New Jersey......................................         106,995          97,034          9,960            1,722
New York........................................         190,358         179,769         10,590            1,831
North Carolina..................................         213,296         151,847         61,450           10,624
Ohio............................................         372,626         239,898        132,728           22,947
Pennsylvania....................................         331,785         252,447         79,338           13,716
Rhode Island....................................           8,295           8,313            (18)               0
South Carolina..................................         138,706         109,425         29,281            5,062
Tennessee.......................................         252,426         182,476         69,950           12,093
Virginia........................................         191,050         155,718         35,332            6,108
West Virginia...................................         190,887          92,920         97,967           16,937
Wisconsin.......................................         145,391         106,540         38,851            6,717
                                                 ---------------------------------------------------------------
    Total.......................................       4,179,751       3,023,113  ..............         200,000
----------------------------------------------------------------------------------------------------------------

    b. State Distribution of the Compliance Supplement Pool. States 
have two options for making the pool available to sources. One option 
is to distribute some or all of the pool to sources that generate early 
reductions during ozone seasons prior to May 1, 2003. The second option 
is to run a public process to provide tons to sources that demonstrate 
a need for a compliance extension. A State wishing to use the 
compliance supplement pool may divide the State pool and make some of 
it available to sources through both options, or may use only one of 
the options for distributing the pool to sources prior to May 1, 2003 
according to the procedures discussed below. Tons that are not 
distributed by a State prior to May 1, 2003 will be retired by EPA.
    (1) Early Reduction Credits. The EPA encourages States to consider 
making the compliance supplement pool available to sources through an 
early reduction credit program. States may use early reduction credits 
as an incentive for sources to make NOX emissions reductions 
prior to the 2003 ozone season that would otherwise not occur. By 
generating early credits or acquiring them from other sources, 
companies will be able to use the early reduction credits to extend the 
timeframe for achieving actual emissions reductions at specific sources 
that may require additional time. To establish an early credit program, 
States that participate in the NOX Budget Trading Program 
may use the provisions

[[Page 57430]]

set forth in that trading program (See Section VII.F). States not 
participating in the NOX Budget Trading Program are also 
free to develop their own rules for granting early reduction credits 
and recognizing the credits for compliance during the 2003 and 2004 
ozone seasons. The procedures for establishing an early credit program 
are presented below in Section III.F.7.c.
    (2) Direct Distribution to Sources. States may also distribute the 
compliance supplement pool directly to sources that demonstrate a need 
for the compliance supplement. Under this approach, sources would be 
responsible for demonstrating to the State and public that achieving 
compliance by May 1, 2003 would create undue risk either to its own 
operation or its associated industry. Before granting a direct 
distribution to a source, the State must provide the public an 
opportunity to comment on the validity of the need for direct 
distribution of the compliance supplement. The direct distribution 
process must be initiated and completed between September 30, 2002 and 
May 1, 2003. States which choose to grant early reduction credits 
cannot conduct the direct distribution until all early reduction 
credits have been issued by the State. By postponing the direct 
distribution until after September 2002, sources will have the maximum 
opportunity to achieve compliance, either through installation of 
controls or with early reduction credits, before using this option. 
States and the public will also be better positioned to determine 
legitimate requests after September 2002.
    To ensure that direct distribution of the compliance supplement is 
only provided to sources that truly need a compliance extension, States 
are only permitted to give credits to an owner or operator of a source 
that demonstrates the following:
     The process of achieving compliance by May 1, 2003 would 
create undue risk for the source or its associated industry. For 
electric generating units, the demonstration should show that 
installing controls would create unacceptable risks for the reliability 
of the electricity supply during the time of installation. This 
demonstration would include a showing that it was not feasible to 
import electricity from other systems during the time of installation. 
Non-electric generating sources may also be eligible for the compliance 
supplement based on a demonstration of risk comparable to that 
described for the electricity industry.
     For a source subject to an early reduction credit program, 
it was not possible to compensate for delayed compliance by generating 
early reduction credits at the source or by acquiring credits generated 
by other sources.
     For a source subject to an emissions trading program, it 
was not possible to acquire allowances or credits for the 2003 ozone 
season from sources that will make reductions beyond required levels 
during the 2003 ozone season.
7. Banking
    As noted in the NPR and SNPR, States have the flexibility to choose 
their own set of control measures to meet their Statewide 
NOX budget established under the NOX SIP call. 
States and sources have supported the use of emissions trading programs 
as a control measure for complying with the NOX SIP call 
requirements. EPA has provided a model cap-and-trade program 
(NOX Budget Trading Program) for large stationary sources 
that States can adopt as one option for establishing an emissions 
trading program. A number of commenters (both States and sources) have 
also expressed interest in pursuing alternative trading programs in 
addition to or as a substitute for the NOX Budget Trading 
Program. One possible flexibility mechanism available to sources 
subject to an emissions trading program is the ability to bank 
emissions reductions. Banking may generally be defined as allowing 
sources that make emissions reductions beyond required levels to save 
and use these excess reductions to compensate for emitting emissions 
above required levels in a later time period. In the SNPR, EPA 
requested comment on whether and how banking should be incorporated 
into the design of the NOX Budget Trading Program. In the 
proposal, four banking options were presented: (1) Banking would not be 
a feature; (2) banking would begin when the trading program begins (May 
2003); (3) sources would be allowed to generate early reductions 
credits for use after the start of the program and banking would 
continue after the program begins; (4) banking would begin with the 
first phase of a two-phase trading program and continue thereafter 
(i.e., phased-in control requirements). The EPA also requested comment 
on options for managing the use of banked allowances in order to limit 
the potential for emissions to be significantly higher than budgeted 
levels because of banking. The EPA specifically proposed using a ``flow 
control'' mechanism in the latter two banking options where the 
potential exists for a large amount of banked allowances to be 
available for use at the start of the program.
    a. Banking Starting in 2003. Comments for the NOX Budget 
Trading Program were generally supportive of including banking in the 
trading program. Commenters noted that allowing sources to make excess 
reductions in one year and use these reductions to emit above required 
levels in a later year encourages early and cost-saving emission 
reductions, helps avoid end-of-season emissions spikes (because unused 
allowances retain their value for compliance in future years), and 
encourages more expedient development and implementation of 
NOX control technology. Commenters pointed out that banking 
also provides sources flexibility in achieving emission reduction 
goals, allowing them to save allowances in years when the cost of 
achieving a given emission level is relatively low for use in years 
when the cost is relatively higher (for example, a year characterized 
by low availability of nuclear and hydro generation capacity would be a 
higher cost year). Thus, banking was seen by many commenters as a 
critical tool for sources to respond to uncertainty. Some commenters, 
however, expressed caveats along with their support for banking. They 
cited the need for some form of bank management to ensure that the use 
of banked allowances does not detract from the environmental goal of 
the NOX SIP call. At least one commenter recommended that 
EPA identify banking as an area to be reviewed for problems during 
audits of the program to ensure it did not have a detrimental impact.
    The EPA also received comments supporting banking that were not 
specific to the NOX Budget Trading Program. Many commenters 
addressed the concept of banking when proposing alternative strategies 
for establishing and implementing the State budgets that were proposed 
in the NOX SIP call. These comments regarded banking as a 
fundamental factor in establishing the timing and control level for the 
State budgets. With all other factors being equal, a NOX SIP 
call that allows banking provides additional flexibility and cost 
savings to affected sources than a NOX SIP call without 
banking. For this reason, many commenters included banking in their 
alternative proposals.
    In order to provide additional flexibility to States and sources 
under the NOX SIP call as discussed in section III.F.6., and 
recognizing that States may pursue alternative trading programs other 
than the NOX Budget Trading Program, the Agency believes it 
is important to establish criteria for banking that would apply to all 
programs that States may use to comply with requirements of the 
NOX SIP call.

[[Page 57431]]

Therefore, EPA is setting forth provisions in today's final rule that 
will allow banking in the NOX Budget Trading Program and 
other State trading programs. Trading programs used to comply with the 
NOX SIP call may allow banking to start in the first control 
period of the program, May 1 through September 30, 2003. Beginning in 
that control period, States may allow sources included in these 
programs to bank NOX emissions reductions not otherwise 
required by the State's SIP, for compliance in future control periods. 
As outlined below, the banking provisions also require the use of a 
flow control mechanism beginning in 2004 and allow States to credit 
early reductions generated by sources prior to 2003 that may be used 
for compliance only in the 2003 and 2004 ozone seasons. The final rule 
for the NOX Budget Trading Program conforms with these 
banking provisions. Additionally, alternative emissions trading 
programs used to comply with the SIP call will be subject to these 
banking criteria as well other applicable criteria in Sec. 51.121 and 
any other applicable EPA guidance such as the Economic Incentive 
Program rules and guidance.
    b. Management of Banked Allowances. Many utility and industry 
commenters generally opposed the use of discounts or constraints on 
banked allowances, arguing that such measures would reduce the 
incentives to control emissions beyond required levels. In addition, 
commenters felt the measures were overly complex and restrictive, as 
well as unnecessary, since the stringent control level proposed would 
serve as a barrier to overcontrol, precluding the establishment of a 
sizeable bank. Several commenters remarked that any decision regarding 
whether and to what extent a trading program should impose restrictions 
on the use of banked allowances should proceed from an analysis of the 
air quality effects of that use; in the absence of such an analysis, 
there would be little basis for imposing restrictions or for deciding 
what restrictions would properly address air quality effects. However, 
these commenters did not provide analyses demonstrating that the use of 
banked allowances in any given season would not be a problem in the 
context of the NOX SIP call. One commenter pointed out 
specifically that the sheer magnitude of the SIP call region should 
preclude EPA from implementing a flow control management scheme similar 
to that used under the Ozone Transport Commission's (OTC) trading 
program, since protection of problem areas would not be feasible on 
such a large scale.
    Several commenters who were opposed to the management of banked 
allowances, however, stated that if restrictions were to be imposed, 
they would favor flow control as the most cost-effective, least rigid 
means of management. A few commenters added that, if implemented, flow 
control should be applied on a source-by-source basis so as to avoid 
penalizing all of the participants in the trading program for the 
excess banking of individual participants. One commenter stated that if 
EPA concludes that there is an adequate basis for imposing some type of 
restriction, it should avoid placing any absolute limit on the amount 
of banked allowances that can be used in a given season. Another 
commenter suggested that if EPA chooses to propose managed banking, it 
should consider establishing an initial period without managed banking 
upon which a managed banking program can later be based if it turns out 
that ``trading contributes to nonattainment.'' Several additional 
commenters, most notably northeastern States and a few environmental 
groups, supported the use of a flow control management system to 
discourage excess use of banked allowances in any one ozone season. One 
such commenter suggested that EPA conduct an analysis similar to that 
used by the OTC in determining the appropriate level of flow control 
for the SIP call region.
    Based on the stated goal of the NOX SIP call, to achieve 
specified limits on NOX emissions for the purpose of 
reducing NOX and ozone transport across State boundaries in 
the eastern half of the United States, EPA believes it is appropriate 
to place some limitation on the amount of emissions variability that 
may occur with banking, and therefore, occur with the transport of 
NOX. At the same time, any limitations on banking should 
still fit within the market-based structure of trading programs, rather 
than imposing overly stringent limits that would potentially eliminate 
the advantages of having banking in the first place. For these reasons, 
EPA is including a provision in today's final rule requiring any State 
program used to comply with the requirements of the NOX SIP 
call that allows banking to limit the potential effects of banking 
through a flow control mechanism as described below. The flow control 
mechanism will be applicable starting in the 2004 ozone season. In this 
year, unused credits from the compliance supplement pool as well as 
unused credits or allowances from the 2003 ozone season would be 
considered banked.
    The EPA believes that the flow control mechanism serves as an 
important insurance policy against emissions variability in emissions 
trading programs used to comply with the NOX SIP call. The 
mechanism as described below would only restrict the use of banked 
allowances or credits when a significant amount are used for compliance 
in a specific ozone season. Based on the analyses in the RIA, EPA 
believes that the flow control mechanism is set at a level that will 
allow sources to use banking without restriction. However, the flow 
control mechanism provides the extra security to downwind areas that 
banking will not result in significant increases of emissions above 
budgeted levels. The EPA also recognizes that a wide variety of 
emissions trading programs may be used by States. Therefore, the 
requirements for the flow control mechanism described below are 
intended to be general, thus allowing States the flexibility to adjust 
the flow control mechanism to fit the specific needs of each program. 
Section VII.F. also provides further discussion of the flow control 
mechanism and describes how it is incorporated into the NOX 
Budget Trading Program.
    The flow control mechanism allows the unlimited banking of 
emissions reductions by sources during and after 2003, but discourages 
the ``excessive use'' of banked allowances or credits by establishing 
either an absolute limit on the number of banked allowances or credits 
that can be used each season or a rate discounting the use of banked 
allowances or credits over a given level. The key issue with flow 
control is to establish the level at which flow control is triggered. 
In the SNPR, EPA solicited comment on establishing the level at 10 
percent of the ozone season budget for the sources included in the 
trading program. This level was proposed because 10 percent seems to be 
a reasonable number that would allow a significant amount of banked 
allowances or credits to be used, but not so many as to jeopardize the 
intended effects of the NOX SIP call in a given season. The 
EPA also proposed the 10 percent number because it is the level used 
for flow control in the OTC's trading program. Although some commenters 
questioned whether this number is appropriate for the NOX 
SIP call region, commenters did not provide explicit analyses or 
recommendations for a different number. Thus, EPA continues to believe 
that 10 percent is a reasonable number and is including this in today's 
final rule. Based on the analyses in the RIA, EPA does not

[[Page 57432]]

anticipate sources to bank above the 10 percent level. Therefore, this 
level should prevent significant emissions increases resulting from 
banking without restricting sources normal operations. The effect of 
flow control set at 10 percent of the trading program budget is that 
for a given season, sources may use banked allowances or credits for 
compliance without restrictions in an amount up to 10 percent of the 
NOX budget for those sources in the trading program. Banked 
allowances or credits that are used in an amount greater than 10 
percent of the NOX budget for those sources will have 
restrictions that are described below.
    The EPA believes it is necessary to provide flexibility to States 
for determining how to apply the 10 percent flow control in individual 
trading programs and for determining the appropriate restrictions for 
banked allowances or credits that are used in an amount greater than 
the 10 percent number. States have the flexibility to apply the flow 
control mechanism to specifically control the use of banked allowances 
or credits at each source or to apply the mechanism more broadly across 
the entire trading program. For example, by applying flow control at 
the source level, a State would allow each source participating in the 
trading program to use banked allowances without restrictions in an 
amount not greater than 10 percent of its allowable NOX 
emissions for the ozone season. Conversely, flow control could be 
applied so that individual sources may use banked allowances or credits 
in an amount more than 10 percent without restrictions, but the total 
number used throughout the entire trading program (i.e., total number 
of banked credits or allowances used for compliance throughout all 
States participating in the trading program) could not exceed 10 
percent of the allowable NOX emissions for all sources in 
the trading program without restrictions. The net effect is the same 
under either approach--banked allowances or credits may be used each 
year without restrictions in an amount that does not exceed 10 percent 
of the allowable NOX emissions for all sources covered by 
the trading program. The NOX Budget Trading Program uses the 
latter approach. See Section VII.F. for more details.
    The second issue for the flow control mechanism is to determine 
what restrictions should be placed on banked allowances or credits that 
are used in an amount greater than 10 percent of the allowable 
NOX emissions for all sources covered by the trading 
program. Again, EPA is providing flexibility for the restrictions that 
States may use. States may use a discount that is no less than two-for-
one, requiring sources to retire one additional banked allowance or 
credit for each banked allowance or credit used for compliance in an 
amount greater than the 10 percent level. Or States may set the 10 
percent level as a hard cap and not allow any banked allowances or 
credits to be used in an amount greater than the 10 percent level. 
Although the discount option provides more flexibility to sources and 
more uncertainty regarding NOX emissions in a given year, 
EPA believes both options serve as an acceptable restriction for 
limiting the variability of emissions associated with banking. As 
described in Section VII.F, the NOX Budget Trading Program 
uses the 2-for-1 discount as the applicable restriction.
    c. Early Reduction Credits. The majority of commenters for the 
NOX Budget Trading Program generally supported the option of 
awarding early reduction credits. Commenters noted that the issuance of 
credits will provide cost savings and environmental benefits by 
encouraging early reductions, facilitate compliance with the budget by 
allowing sources to earn allowances that may be used to delay more 
stringent emission reductions, and stimulate the market by ensuring 
allowances are available for trading at the program start. Several 
commenters advocated making early reduction credits available for any 
reductions that exceed baseline controls, whereas other commenters 
supported early reduction credits only if they exceed the controls 
required under the SIP call, as was proposed by EPA. A few other 
commenters suggested levels between these two options. A few OTC States 
suggested that OTC allowances banked in Phase II (between 1999-2003 for 
reductions beyond an approximate 0.20 lb/mmBtu rate) could be used as 
early reduction credits in the NOX Budget Trading Program, 
either one-for-one or at a discount ratio, depending on the level 
beyond which credits were awarded in the latter program. A few 
remaining commenters, concerned about the potential for creating or 
exacerbating ozone violations, supported early reduction credits and 
banking only if coupled with flow control.
    Regarding the appropriate length of the period in which early 
reductions could be earned, some commenters supported EPA's proposed 
option in the SNPR of a two-year early reduction period, while others 
favored a three or four-year period. At least one commenter 
specifically recommended that the early reduction period start in 
January 1995, while another suggested September 1998. Several 
commenters rejected EPA's suggestion that early reduction credits be 
calculated as a set-aside from the first five years of allowances, 
arguing that treating the credits as set-asides would be inconsistent 
with the nature of early reduction credits. Conversely, a few other 
commenters felt the credits should be awarded from within State budgets 
to avoid budget inflation. Additional commenters criticized EPA's 
suggestion that if early reduction credits were awarded, they be 
awarded at the company level, arguing instead for individual source 
awards. One commenter stated that awards on a company basis would not 
address the load shifting concerns EPA cited, while another thought EPA 
could address the load shifting concern by basing credits on activity 
levels in a historic period rather than by shifting to a company-level 
award. Finally, at least one commenter felt that States should be able 
to independently establish parameters for awarding voluntary early 
reductions.
    For the reasons set forth in Section III.F.7, Compliance Supplement 
Pool, EPA is allowing, but not requiring, States to grant early 
reduction credit to sources that reduce their ozone season 
NOX emissions below levels specified by the State prior to 
the 2003 control period. The early reduction credits may be used by 
sources for compliance during the 2003 and 2004 ozone seasons. EPA 
believes that an early credit program can be helpful to encourage 
emissions reductions prior to the 2003 ozone season that would not be 
made without an economic incentive for the sources to act. Furthermore, 
the early credit program will provide additional allowances or credits 
for use during the 2003 and 2004 ozone seasons. By generating early 
credits or acquiring early credits from other sources that generated 
credits, companies would have greater latitude in determining when 
actual emissions reductions are achieved at specific sources. As 
discussed in Section III.F.7, this may be beneficial to some companies 
that are concerned about the time and effort required to install all 
necessary emissions controls prior to May 2003. States will be limited 
in the amount of early reduction credits that they may grant by the 
amounts set forth in Section III.F.7 Compliance Supplement Pool. The 
potential pool of credits that is available to each State is intended 
to be large enough to provide a real incentive for early reductions and 
enough flexibility to allow the installation of some control equipment, 
if necessary, past May 2003.

[[Page 57433]]

    Section VII.F. of today's preamble outlines how the early credit 
program is being incorporated into the NOX Budget Trading 
Program and how banked allowances from the OTC program may be 
integrated with this provision. States that develop alternative trading 
programs may craft their early reduction program to meet the needs of 
their specific trading program. The following outlines the general 
requirements that any early reduction program used to comply with the 
NOX SIP call should meet. For an emission reduction to be 
eligible as an early reduction credit, it must meet the following 
criteria:
     Surplus--The reduction is not contained in the State's SIP 
or otherwise required by the CAA.
     Verifiable--The reduction can be verified as actually 
having occured.
     Quantifiable--The reduction is quantified according to 
procedures set forth by the State and approved by EPA. Early reduction 
credits generated by sources serving electric generators with a 
nameplate capacity greater than 25 MWe or greater or boilers, 
combustion turbines and combined cycle units with a maximum design heat 
input greater than 250 mmBtu/hr, should be quantified according to the 
monitoring provisions of part 75, subpart H as required in 
Sec. 51.121(h)(1)(iv).
    Beyond the above requirements, States are free to develop an early 
credit program that meets the needs of their specific trading program 
provided the State does not issue credits in an amount greater the size 
of the credit pool presented in Section III.F.7. A State's early credit 
program may be established for any ozone season occurring after a 
State's early credit rule is approved by EPA into the State's SIP 
revision and before May 1, 2003.
    To ensure that a State does not issue an amount of early credits 
beyond the amount specified in each State's compliance supplement pool, 
EPA recommends that a State develop procedures to be used in case there 
is an over-subscription of the early credit pool. Possible options 
include granting early credits on a first-come, first-served basis or 
waiting until all applications are submitted and then discounting the 
early credits on a pro-rata basis so that the amount of early credits 
issued equals the size of the State's pool. States may also influence 
the amount of early credits that sources generate by considering what 
level of emissions reductions the State will recognize as early 
reductions. For example, a State may choose to issue early reduction 
credits for any reductions below applicable requirements. However, the 
State may choose to make the demonstration more stringent by requiring 
early reduction credits to be generated by reductions that are below a 
limit that is tighter than applicable requirements (e.g., grant early 
reductions that are 30 percent below applicable requirements or below a 
fixed level such as 0.20 lb/mmBtu).
    In the SNPR, EPA also solicited comment on a phased-in 
NOX Budget Trading Program that would begin in 2001, two 
years prior to the compliance date for the NOX SIP call. In 
response to the proposal, most commenters that discussed the phase-in 
program option were generally opposed to it. Their primary argument was 
that such a program would effectively accelerate the compliance date 
for NOX controls under the SIP call. A few commenters, 
however, still supported the phase-in approach as a means of mitigating 
the uncertainties inherent in the allowance market that would develop 
for the 2003 control period, allowing sources to gain experience prior 
to 2003. Some commenters specifically favored a phase-in approach only 
if it does not interfere with the 2003 ozone season compliance 
schedule, whereas others supported a phase-in approach as a means of 
reducing the burdens of the 2003 ozone season compliance schedule.
    Today's final rule requires States to achieve the necessary 
emissions reductions by May 2003 and does not require States to phase-
in controls prior to 2003. States that wish to phase-in controls prior 
to 2003 as a part of a State trading program may do this, but they are 
not required to do so to comply with the NOX SIP call. 
States that establish a phased-in trading program in order to allow 
sources to generate early reduction credits will be subject to the 
requirements for early reductions as described above, including the 
requirement that a State may not grant an amount of early reductions in 
excess of the State's compliance supplement pool. For a discussion of 
how the Ozone Transport Commission's trading program may be integrated 
with the compliance supplement pool and the early reduction provisions, 
see Section VII.F, which describes the banking provisions of the 
NOX Budget Trading Program.

G. Final Statewide Budgets

1. EGU
    a. Description of Selected Approach. As described in Section 
III.B.3. of this notice, the EGU budget component is calculated based 
on applying a 0.15 lb/mmBtu emission limit to sources greater than 25 
MWe. This limit is applied uniformly across all States that are covered 
by this SIP call. The higher of 1995 or 1996 heat input, grown to 2007 
is used to calculate the budget component.
    b. Summary of Budget Component. Both the 2007 electricity 
generating Base Case and the electricity generating Budget component 
were revised from the levels in the SNPR based on the changes described 
in Section III.B.3. of this notice. These revisions are shown in Tables 
III-4 and III-5. The difference between the revised 2007 Base Case and 
Budget emissions from the SNPR and the final Base Case and Budget 
emissions is shown in Table III-4. Negative changes indicate decreases. 
The final percent reduction from the 2007 Base Case to the Budget is 
shown in Table III-5.

     Table III-4.--Changes to Revised SNPR Base Case and Budget Components for Electricity Generating Units
                                                [Tons NOX/season]
----------------------------------------------------------------------------------------------------------------
                                                                Percent      Revised       Final       Percent
            State              Revised base     Final base       change       budget       budget       change
----------------------------------------------------------------------------------------------------------------
Alabama.....................          85,201          76,900          -10       30,644       29,051           -5
Connecticut.................           7,048           5,600          -21        5,245        2,583          -51
Delaware....................          10,727           5,800          -46        4,994        3,523          -29
District of Columbia........             236              *0         -100          152          207           36
Georgia.....................          84,890          86,500            2       32,433       30,255           -7
Illinois....................         119,756         119,300            0       36,570       32,045          -12
Indiana.....................         159,917         136,800          -14       51,818       49,020           -5
Kentucky....................         130,919         107,800          -18       38,775       36,753           -5

[[Page 57434]]

Maryland....................          37,575          32,600          -13       12,971       14,807           14
Massachusetts...............          24,998          16,500          -34       14,651       15,033            3
Michigan....................          73,585          86,600           18       29,458       28,165           -4
Missouri....................          81,799          82,100            0       26,450       23,923          -10
New Jersey..................          17,484          18,400            5        8,191       10,863           33
New York....................          43,705          39,200          -10       31,222       30,273           -3
North Carolina..............          86,872          84,800           -2       32,691       31,394           -4
Ohio........................         167,601         163,100           -3       51,493       48,468           -6
Pennsylvania................         120,979         123,100            2       45,971       52,000           13
Rhode Island................           1,351           1,100          -19        1,609        1,118          -31
South Carolina..............          57,146          36,300          -36       19,842       16,290          -18
Tennessee...................          83,844          70,900          -15       26,225       25,386           -3
Virginia....................          51,113          40,900          -20       20,990       18,258          -13
West Virginia...............          76,374         115,500           51       24,045       26,439           10
Wisconsin...................          45,538          52,000           14       17,345       17,972            4
                             -----------------------------------------------------------------------------------
    Total...................       1,568,655       1,501,800           -4      563,784      543,825           -4
----------------------------------------------------------------------------------------------------------------
*The base case for DC is actually projected to be 3 tons per season. The base case values in this table are
  rounded to the nearest 100 tons.


        Table III-5.--Final NOX Budget Components and Percent Reduction for Electricity Generating Units
                                                  [tons/season]
----------------------------------------------------------------------------------------------------------------
                                                                                                      Percent
                              State                                 Final base     Final budget      reduction
----------------------------------------------------------------------------------------------------------------
Alabama.........................................................          76,900          29,051              62
Connecticut.....................................................           5,600           2,583              54
Delaware........................................................           5,800           3,523              39
District of Columbia............................................              *0             207              NA
Georgia.........................................................          86,500          30,255              65
Illinois........................................................         119,300          32,045              73
Indiana.........................................................         136,800          49,020              64
Kentucky........................................................         107,800          36,753              66
Maryland........................................................          32,600          14,807              55
Massachusetts...................................................          16,500          15,033               9
Michigan........................................................          86,600          28,165              67
Missouri........................................................          82,100          23,923              71
New Jersey......................................................          18,400          10,863              41
New York........................................................          39,200          30,273              23
North Carolina..................................................          84,800          31,394              63
Ohio............................................................         163,100          48,468              70
Pennsylvania....................................................         123,100          52,000              58
Rhode Island....................................................           1,100           1,118              -2
South Carolina..................................................          36,300          16,290              55
Tennessee.......................................................          70,900          25,386              64
Virginia........................................................          40,900          18,258              55
West Virginia...................................................         115,500          26,439              77
Wisconsin.......................................................          52,000          17,972              65
                                                                 -----------------------------------------------
    Total.......................................................       1,501,800         543,825              64
----------------------------------------------------------------------------------------------------------------
*The base case for DC is actually projected to be 3 tons per season. The base case values in this table are
  rounded to the nearest 100 tons.

2. Non-EGU Point Sources
    As indicated in the proposal and discussed earlier in this notice, 
EPA continues to believe that technically feasible control measures 
costing between an average of $1,000 to $2,000 per ozone season ton 
(1990 dollars) are highly cost-effective and therefore should be the 
basis for determining the significant amounts that must be eliminated 
by each covered jurisdiction. In the SNPR, EPA committed to examining 
alternatives that would limit the number of affected non-EGU sources 
for the purpose of establishing emissions budgets, yet still achieve 
the environmental objective of mitigating broad-scale ozone transport. 
The EPA examined alternatives that target reductions from the largest 
non-EGU source category groupings, and within each of the largest 
groupings applied the cost-effectiveness criteria. The resulting 
emissions budget covers the majority of emissions from large non-
utility sources, and does not include reductions from small sources and 
sources that, as a group, are not efficient to control, or are already 
covered by other Federal measures (e.g., CAA Sec. 112 MACT). The 
description below summarizes the budget approach for non-EGU point 
sources.
    a. Description of Selected Approach.
    (1) NOX Budget Sources. The following approach is used 
to determine if a unit's emissions would be decreased as part of the 
budget calculation.

[[Page 57435]]

Industrial boilers, turbines, stationary internal combustion engines 
and cement manufacturing are the only non-EGU sources for which 
reductions are assumed in the budget calculation.
    1. Use heat input capacity data for each source if the data are in 
the updated inventory.
    2. If heat input capacity data are not available, use the default 
identification of small and large sources developed by EPA/Pechan for 
OTAG and also used to develop the NPR and SNPR budgets for source 
categories with heat input capacity fields (``default data'').
    3. Emission reductions would be assumed if specific source heat 
input capacity data or default data indicate that a source is greater 
than 250 mmBtu/hr in the updated inventory.
    4. If specific or default heat input capacity data are not 
available in the updated inventory (or not appropriate for a particular 
source category), emission reductions would be assumed if the unit's 
average summer day emissions are greater than one ton per day based on 
the updated inventory.
    5. All others are ``small'' and no emission reductions are assumed.
    It should be noted (as described earlier in this section) that no 
emissions reductions are assumed for point sources with capacities less 
than or equal to 250 mmBtu/hr but with emissions greater than 1 ton/day 
for purposes of calculating the budget. This is a change from the NPR 
which assumed RACT controls on units with capacities less than or equal 
to 250 mmBtu/hr and emissions greater than 1 ton/day.
    (2) Control Levels. For purposes of calculating the State 
NOX budgets for the relevant sources (described above), the 
following emissions decreases from uncontrolled levels were assumed:
    1. Non-EGU boilers and turbines--60% decrease.
    2. Stationary internal combustion engines--90% decrease.
    3. Cement manufacturing plants--30% decrease.
    These controls result in an overall reduction in emissions from all 
affected large non-EGU point sources of almost 40 percent (187,800 tons 
per season decrease).
    Each State's budget is based on application of these controls 
beginning on May 1, 2003. The EPA recognizes that if States include 
these source categories in a regionwide trading program, as EPA 
encourages States to do, each State will comply with its budget through 
compliance of its sources with the requirements of the regionwide 
trading program. Of course, under the trading program, sources in a 
State may acquire or sell allowances that will, in turn, allow for 
higher or lower emissions levels for that State than assumed in this 
action. Because EPA has determined that the ambient effect of such a 
trading program across the region is consistent with the basis for 
including States in the SIP call (see discussion below at Section IV), 
EPA has structured its rule to allow a State to meet its budget by 
including the amount of emissions for which sources in the State hold 
allowances from out-of-State sources. Overall, total NOX 
emissions in the region will be within the budget.
    b. Summary of Budget Component. Both the 2007 Base Case and Budget 
component for non-electricity generating point sources were revised 
based on the changes described above. Changes to the 2007 base reflect 
changes in the base year (1995) emissions and changes in growth 
factors. Changes to the budget components reflect these changes as well 
as the change in level of control. These resulting budget components 
are shown in Tables III-5 and III-6. The difference between the 2007 
Base Case and Budget emissions as revised in the SNPR and the final 
Base Case and Budget emissions for non-electricity generating point 
sources is shown in Table III-6. Negative changes indicate decreases. 
The final percent reduction from the 2007 Base Case to the Budget is 
shown in Table III-7.

                      Table III-6.--Changes to Revised Base Case and Budget Components for Non-Electricity Generating Point Sources
                                                                    [Tons NOX/season]
--------------------------------------------------------------------------------------------------------------------------------------------------------
                                                           Revised base     Final base    Percent change  Revised budget   Final budget   Percent change
--------------------------------------------------------------------------------------------------------------------------------------------------------
Alabama.................................................          48,187          49,781               3          24,416          37,696              54
Connecticut.............................................           5,254           5,273               0           3,103           5,056               3
Delaware................................................           5,276           1,781             -66           2,271           1,645             -28
District of Columbia....................................             311             310               0             259             292              13
Georgia.................................................          33,939          33,939               0          14,305          27,026              89
Illinois................................................          65,351          55,721             -15          40,719          42,011               3
Indiana.................................................          51,839          71,270              37          29,187          44,881              54
Kentucky................................................          19,019          18,956               0          11,996          14,705              23
Maryland................................................          10,710          10,982               3           5,852           7,593              30
Massachusetts...........................................           9,978           9,943               0           6,207           9,763              57
Michigan................................................          61,656          79,034              28          35,957          48,627              35
Missouri................................................          12,320          13,433               9           9,012          11,054              23
New Jersey..............................................          22,228          22,228               0          12,786          19,804              55
New York................................................          20,853          25,791              24          14,644          24,128              65
North Carolina..........................................          34,412          34,027              -1          19,267          25,984              35
Ohio....................................................          53,329          53,241               0          30,923          35,145              14
Pennsylvania............................................          74,839          73,748              -1          41,824          65,510              57
Rhode Island............................................             327             327               0             327             327               0
South Carolina..........................................          34,994          34,740              -1          18,671          25,469              36
Tennessee...............................................          67,774          60,004             -11          34,308          35,568               4
Virginia................................................          25,509          39,765              56          10,919          27,076             148
West Virginia...........................................          42,733          40,192              -6          21,066          31,286              49
Wisconsin...............................................          21,263          22,796               7          11,401          17,973              58
                                                         -----------------------------------------------------------------------------------------------
    Total...............................................         722,101         757,281               5         399,416         558,618              40
--------------------------------------------------------------------------------------------------------------------------------------------------------


[[Page 57436]]


  Table III-7.--Final NOX Budget Components and Percent Reduction for Non-Electricity Generating Point Sources
                                                  [Tons/season]
----------------------------------------------------------------------------------------------------------------
                                                                                                      Percent
                                                                    Final base     Final budget      reduction
----------------------------------------------------------------------------------------------------------------
Alabama.........................................................          49,781          37,696              24
Connecticut.....................................................           5,273           5,056               4
Delaware........................................................           1,781           1,645               8
District of Columbia............................................             310             292               6
Georgia.........................................................          33,939          27,026              20
Illinois........................................................          55,721          42,011              25
Indiana.........................................................          71,270          44,881              37
Kentucky........................................................          18,956          14,705              22
Maryland........................................................          10,982           7,593              31
Massachusetts...................................................           9,943           9,763               2
Michigan........................................................          79,034          48,627              38
Missouri........................................................          13,433          11,054              18
New Jersey......................................................          22,228          19,804              11
New York........................................................          25,791          24,128               6
North Carolina..................................................          34,027          25,984              24
Ohio............................................................          53,241          35,145              34
Pennsylvania....................................................          73,748          65,510              11
Rhode Island....................................................             327             327               0
South Carolina..................................................          34,740          25,469              27
Tennessee.......................................................          60,004          35,568              41
Virginia........................................................          39,765          27,076              32
West Virginia...................................................          40,192          31,286              22
Wisconsin.......................................................          22,796          17,973              21
                                                                 -----------------------------------------------
    Total.......................................................         757,281         558,618              26
----------------------------------------------------------------------------------------------------------------

3. Mobile and Area Sources
    a. Description of Selected Budget Approach. As discussed in Section 
III.D.3 of the notice, EPA proposed highway budget components based on 
projected highway vehicle emissions in 2007 from a base year of 1990, 
assuming implementation of those measures incorporated in existing 
SIPs, such as inspection and maintenance programs and reformulated 
fuels, measures already implemented federally, and those additional 
measures expected to be implemented federally by 2007. As discussed in 
Section III.E of this notice, EPA proposed nonroad mobile source budget 
components based on projected nonroad mobile source emissions in 2007 
from a base year of 1990. These projections were developed by 
estimating the emissions expected in 2007 from all nonroad engines, 
assuming implementation of those measures incorporated in existing 
SIPs, measures already implemented federally, and those additional 
measures expected to be implemented federally. For area sources, no 
cost-effective control measures were identified in the NPR. Because no 
comments were received that demonstrate that additional controls for 
highway, nonroad, or area sources are both feasible and highly cost-
effective, the final budgets are based on the same levels of controls 
that were proposed.
    b. Summary of Budget Component. Changes were made to the baseline 
stationary area, nonroad and highway mobile source budget data as 
discussed in Sections III.D. and III.E. of this notice. Budget 
components were calculated using the updated baseline and the controls 
discussed above. The resulting final budget components for these 
sectors are contained in Tables III-7, III-8, and III-9 below, along 
with the difference between the proposed Budget emissions and the final 
Budget emissions. The budget components are not compared to the 2007 
base because no reductions were calculated beyond the base case. In the 
NPR and SNPR, EPA used a 2007 CAA baseline for these source sectors. 
Because the measures that are assumed in the budgets for these sectors 
are measures that would occur in the absence of the SIP call, EPA 
believes that it is more appropriate to use the budget level for these 
source sectors as the baseline and compare the total budgets to this 
revised baseline.

                      Table III-8.--Final NOX Budget Components for Stationary Area Sources
                                                  [Tons/season]
----------------------------------------------------------------------------------------------------------------
                                                                     Proposed
                                                                      budget       Final budget   Percent change
----------------------------------------------------------------------------------------------------------------
Alabama.........................................................          25,229          25,225               0
Connecticut.....................................................           4,587           4,588               0
Delaware........................................................           1,035             963              -7
District of Columbia............................................             741             741               0
Georgia.........................................................          11,901          11,902               0
Illinois........................................................           7,270           7,822               8
Indiana.........................................................          25,545          25,544               0
Kentucky........................................................          38,801          38,773               0
Maryland........................................................           8,123           4,105             -49
Massachusetts...................................................          10,297          10,090              -2

[[Page 57437]]

Michigan........................................................          28,126          28,128               0
Missouri........................................................           6,626           6,603               0
New Jersey......................................................          11,388          11,098              -3
New York........................................................          15,585          15,587               0
North Carolina..................................................           9,193          10,651              16
Ohio............................................................          19,446          19,425               0
Pennsylvania....................................................          17,103          17,103               0
Rhode Island....................................................             420             420               0
South Carolina..................................................           8,420           8,359              -1
Tennessee.......................................................          11,991          11,990               0
Virginia........................................................          25,261          18,622             -26
West Virginia...................................................           4,901           4,790              -2
Wisconsin.......................................................          10,361           8,160             -21
                                                                 -----------------------------------------------
    Total.......................................................         302,350         290,689              -4
----------------------------------------------------------------------------------------------------------------


               Table III-9.--Final NOX Budget Components and Percent Reduction for Nonroad Sources
                                                  [Tons/season]
----------------------------------------------------------------------------------------------------------------
                                                                     Proposed
                                                                      budget       Final budget   Percent change
----------------------------------------------------------------------------------------------------------------
Alabama.........................................................          18,727          16,594             -11
Connecticut.....................................................           9,581           9,584               0
Delaware........................................................           4,262           4,261               0
District of Columbia............................................           3,582           3,470              -3
Georgia.........................................................          22,714          21,588              -5
Illinois........................................................          56,429          47,035             -17
Indiana.........................................................          27,112          22,445             -17
Kentucky........................................................          22,530          19,627             -13
Maryland........................................................          18,062          17,249              -4
Massachusetts...................................................          19,305          18,911              -2
Michigan........................................................          24,245          23,495              -3
Missouri........................................................          19,102          17,723              -7
New Jersey......................................................          21,723          21,163              -3
New York........................................................          30,018          29,260              -3
North Carolina..................................................          18,898          17,799              -6
Ohio............................................................          42,032          37,781             -10
Pennsylvania....................................................          29,176          25,554             -12
Rhode Island....................................................           2,074           2,073               0
South Carolina..................................................          12,831          11,903              -7
Tennessee.......................................................          47,065          44,567              -5
Virginia........................................................          25,357          21,551             -15
West Virginia...................................................          10,048          10,220               2
Wisconsin.......................................................          15,145          12,965             -14
                                                                 -----------------------------------------------
    Total.......................................................         500,018         456,818              -9
----------------------------------------------------------------------------------------------------------------


              Table III-10. Final NOX Budget Components and Percent Reduction for Highway Vehicles
                                                  [Tons/season]
----------------------------------------------------------------------------------------------------------------
                                                                     Proposed
                                                                      budget       Final budget   Percent change
----------------------------------------------------------------------------------------------------------------
Alabama.........................................................          56,601          50,111             -11
Connecticut.....................................................          17,392          18,762               8
Delaware........................................................           8,449           8,131              -4
District of Columbia............................................           2,267           2,082              -8
Georgia.........................................................          77,660          86,611              12
Illinois........................................................          77,690          81,297               5
Indiana.........................................................          66,684          60,694              -9
Kentucky........................................................          46,258          45,841              -1
Maryland........................................................          28,620          27,634              -3
Massachusetts...................................................          23,116          24,371               5
Michigan........................................................          81,453          83,784               3
Missouri........................................................          55,056          55,230               0
New Jersey......................................................          39,376          34,106             -13
New York........................................................          94,068          80,521             -14

[[Page 57438]]

North Carolina..................................................          73,056          66,019             -10
Ohio............................................................          92,549          99,079               7
Pennsylvania....................................................          73,176          92,280              26
Rhode Island....................................................           5,701           4,375             -23
South Carolina..................................................          49,503          47,404              -4
Tennessee.......................................................          67,662          64,965              -4
Virginia........................................................          79,848          70,212             -12
West Virginia...................................................          21,641          20,185              -7
Wisconsin.......................................................          41,651          49,470              19
                                                                 -----------------------------------------------
    Total.......................................................       1,179,477       1,173,163              -1
----------------------------------------------------------------------------------------------------------------

4. Potential Alternatives to Meeting the Budget
    The EPA believes that there are additional control measures and 
alternative mixes of controls that a State could choose to implement by 
May 1, 2003. Examples of such measures are described below and 
illustrate that options are potentially available in several source 
categories.
    The EPA believes that, with respect to EGUs, there is a large 
potential for energy efficiency and renewables in the NOX 
SIP call region that reduce demand and provide for more 
environmentally-friendly energy resources. For example, if a company 
replaces a turbine with a more efficient one, the unit supplying the 
turbine would reduce the amount of fuel (heat input) the unit combusts 
and would reduce NOX emissions proportionately, while the 
associated generator would produce the same amount of electricity. 
Renewable energy source generation includes hydroelectric, solar, wind, 
and geothermal generation. EPA recognizes that promotion of energy 
efficiency and renewables can contribute to a cost-effective 
NOX reduction strategy. As such, EPA encourages States in 
the NOX SIP call region to consider including energy 
efficiency and renewables as a strategy in meeting their NOX 
budgets. One way to achieve this goal is by including a provision 
within a State's NOX Budget Trading Rule that allocates a 
portion of a State's trading program budget to implementers of energy 
efficiency and renewables projects that reduce energy-related 
NOX emissions during the ozone season. Another is to include 
energy efficiency and renewables projects as part of a State's 
implementation plan.
    The EPA is working to develop guidance on how States can integrate 
energy efficiency into their SIPs by both of these mechanisms. The 
guidance will present EPA's current thinking on the important elements 
to include in a functional system that allocates a portion of a State's 
trading program budget to implementers of energy efficiency and 
renewables projects within the context of the NOX Budget 
Trading Program. In addition, EPA will issue guidance outlining 
procedures for including energy efficiency and renewables projects in a 
State's SIP as control strategies for achieving the State's 
NOX budget, separate from the NOX Budget Trading 
Program. EPA plans to issue these guidance documents in the Fall of 
1998 so that they will be available to States early in their SIP 
planning process.
    With respect to non-EGUs, individual States could choose to require 
emissions decreases from sources or source categories that EPA exempted 
from the budget calculations. For example, there are many large sources 
for which EPA lacked enough information to determine potential controls 
and emissions reductions; States may have access to such information 
and could choose to apply cost-effective controls. In addition, States 
could choose to regulate one or more of the non-EGU stationary sources 
or source categories which EPA had exempted because emissions were 
relatively low considering other source categories in the 23 
jurisdictions. In individual States, emissions from such sources could 
be a high percentage of uncontrolled emissions and, thus, be subject to 
efficient, cost-effective control for that particular State. Further, 
States may take other approaches to developing their budgets, such as 
cutoffs based on horsepower rather than tons per day, since they might 
have access to data that EPA did not have for all 23 jurisdictions.
    With respect to mobile sources, States could implement other 
NOX control measures in lieu of the controls described 
earlier in this section. For example, vehicle inspection and 
maintenance programs can provide significant NOX reductions 
from highway vehicles. Additional NOX reductions can be 
obtained by opting into the reformulated gasoline program, by 
implementing measures to reduce the growth in VMT, and by implementing 
programs to accelerate retirement of older, higher-emitting highway 
vehicles and nonroad equipment.
5. Statewide Budgets
    The revised Statewide budgets that reflect the changes to the base 
year inventory and growth factors for all sectors and the revised 
control levels for the non-EGU point source sector described above are 
shown in Table III-11. For the 23 jurisdictions combined, the budgets 
result in a 28 percent reduction from the base case. In the NPR and 
SNPR the percent reduction was 35 percent. The difference in the 
percent reduction is due to several factors. First, in the NPR and SNPR 
reductions from certain highway and nonroad controls were assumed to 
occur as a result of measures implemented between promulgation of this 
rule and 2007. These measures include National Low Emission Vehicle 
Standards, the 2004 Heavy-Duty Engine Standards, the Federal Small 
Engine Standards, Phase II, Federal Marine Engine Standards (for diesel 
engines of greater than 50 horsepower), Federal Locomotive Standards, 
and the Nonroad Diesel Engine Standards. These controls were reflected 
in the budget but were not included in the base case. For the final 
rule, EPA determined that these measures should be included in the base 
case, rather than the budgets, because the measures would be 
implemented even in the absence of this rulemaking. Based on the 
emission levels that were used in the SNPR, the effect of using this 
approach to setting the base case is to decrease the percent reduction 
from 35 percent to approximately 31 percent.

[[Page 57439]]

The additional change in the percent reduction (from 31 percent to 28 
percent) is primarily due to EPA's decision not to assume controls for 
several non-EGU source categories and to change the level of control 
for those non-EGU categories for which controls are assumed. Although 
the overall percent reduction went from 35 percent to 28 percent, the 
difference between the budget proposed in the SNPR and the final 
budgets in today's notice is less than 3 percent.

                                  Table III-11.--Revised Statewide NOX Budgets
                                                  [Tons/season]
----------------------------------------------------------------------------------------------------------------
                                                                                                      Percent
                              State                                    Base           Budget         reduction
----------------------------------------------------------------------------------------------------------------
Alabama.........................................................         218,610         158,677              27
Connecticut.....................................................          43,807           40,57              37
Delaware........................................................          20,936          18,523              12
District of Columbia............................................           6,603           6,792              -3
Georgia.........................................................         240,540         177,381              26
Illinois........................................................         311,174         210,210              32
Indiana.........................................................         316,753         202,584              36
Kentucky........................................................         230,997         155,698              33
Maryland........................................................          92,570          71,388              23
Massachusetts...................................................          79,815          78,168               2
Michigan........................................................         301,042         212,199              30
Missouri........................................................          75,089         114,532              35
New Jersey......................................................         106,995          97,034               9
New York........................................................         190,358         179,769               6
North Carolina..................................................         213,296         151,847              29
Ohio............................................................         372,626         239,898              36
Pennsylvania....................................................         331,785         252,447              24
Rhode Island....................................................           8,295            8,31              30
South Carolina..................................................         138,706         109,425              21
Tennessee.......................................................         252,426         182,476              28
Virginia........................................................         191,050         155,718              18
West Virginia...................................................         190,887          92,920              51
Wisconsin.......................................................         145,391         106,540              27
                                                                 -----------------------------------------------
Total...........................................................       4,179,751       3,023,113              28
----------------------------------------------------------------------------------------------------------------

IV. Air Quality Assessment

A. Assessment of Proposed Statewide Budgets

    In the SNPR, EPA documented the estimated ozone benefits of the 
proposed Statewide NOX budgets based on an air quality 
modeling analysis. The major findings of that analysis are as follows:
    (1) The emissions reductions associated with the proposed Statewide 
budgets are predicted to produce large reductions in both 1-hour and 8-
hour concentrations in areas which currently violate the NAAQS and 
which would likely continue to have violations in the future without 
the SIP call budget reductions.
    (2) Looking at individual ozone ``problem areas'' considered by 
OTAG shows similar results, based on the available metrics.
    (3) Any ``disbenefits'' due to the NOX reductions 
associated with the budgets are expected to be very limited compared to 
the extent of the benefits expected from these budgets.
    (4) Even though the budgets are expected to reduce 1-hour and 8-
hour ozone concentrations across all 23 jurisdictions, nonattainment 
problems requiring additional local control measures will likely 
continue in some areas currently violating the NAAQS.

(63 FR 25903)

B. Comments and Responses

    The EPA received numerous comments on the air quality modeling of 
the proposed NOX budgets. The following is a summary of the 
main comments and EPA's responses.
    Comment: Commenters stated that the emissions inventories used for 
modeling were flawed because EPA's projection of the base year 
emissions to 2007 improperly treated growth for certain electric 
generation units by growing these units beyond their design capacity.
    Response: The EPA agrees with this comment and has revised the 2007 
emissions projections for modeling to take this factor into account. 
For the modeling described in the SNPR, EPA applied State-level growth 
factors uniformly to existing sources in each State. This did not 
account for maximum capacity and could have resulted in sources being 
modeled with emissions that were higher than their actual capacity 
would allow. For the modeling described in this notice, EPA has revised 
the projection procedures to use IPM to allocate growth to existing 
units considering their design capacity. As described below, EPA has 
remodeled the 2007 Base Case and the Statewide budgets using this 
revised inventory and found that the conclusions from the revised runs 
do not differ from those based on the SNPR model runs of these budgets.
    Comment: Commenters stated that EPA's modeling in the SNPR examined 
the impacts of the budgets applied regionwide (i.e., for each State for 
which a budget is required), rather than the impacts on downwind 
nonattainment of the budgets applied only in upwind States. Therefore, 
according to the commenters, this modeling is not useful for indicating 
the impact of the State budgets on downwind nonattainment or 
maintenance problems.
    Response: The EPA is well aware that many States in the SIP Call 
region are both upwind and downwind States, that is, they are upwind of 
certain nonattainment areas and downwind from other States. For 
example, Pennsylvania is upwind of New York City, and emissions from 
Pennsylvania sources significantly contribute to this nonattainment 
problem; and

[[Page 57440]]

Pennsylvania is downwind of several States, emissions from which 
significantly contribute to Philadelphia's nonattainment problem.
    The EPA is further aware that modeling analyses that evaluate 
emissions reductions in each State affected by today's rulemaking do 
not isolate the precise impact of emissions reductions from each upwind 
State on nonattainment in a State that is itself both an upwind and 
downwind State. That is, the emissions reductions in that upwind/
downwind area impact its own nonattainment problems. To return to the 
example noted above, because emissions reductions in Pennsylvania 
affect Philadelphia's air quality, modeling Pennsylvania's emissions 
reductions along with emissions reductions in all other affected States 
does not isolate the impact of emissions reductions from States upwind 
of Pennsylvania on Philadelphia's air quality. As a result, EPA is 
aware that the regionwide modeling of different budget levels does not 
indicate the differential impact on downwind areas of higher budget 
levels as compared to lower budget levels in upwind areas.
    Nevertheless, EPA believes that regionwide modeling of the State 
budgets is a useful indication of the overall impacts of various budget 
levels. Today's rulemaking requires regionwide emissions reductions, 
which will carry certain costs and will have certain impacts viewed on 
a State-by-State basis and on a regionwide basis. The multi-State 
budgets promulgated today mean that in a State that is both upwind and 
downwind of other States, such as Pennsylvania, the air quality will, 
in fact, be improved by the emissions reductions in upwind States and 
by the reductions within the States that are required to improve air 
quality further downwind. Thus, it is necessary to consider the upwind 
emissions reductions together with the downwind emissions reductions in 
order to fully evaluate the air quality impacts of the Statewide 
budgets. Regionwide modeling is the only available approach to indicate 
these ``real world'' impacts in individual States, as well as allow an 
assessment of those impacts in light of their costs. Accordingly, this 
modeling is useful in evaluating the overall impacts of the alternative 
budget levels considered in the course of the rulemaking. The EPA 
believes that a comparison of the overall impacts of alternative budget 
levels, in turn, serves as a means to confirm whether the budget levels 
promulgated in today's rulemaking yield meaningful air quality 
benefits. Moreover, EPA has conducted other modeling which indicates 
the impact of budget-level emissions on air quality downwind, as 
discussed below.
    Comment: Commenters stated that EPA should have modeled the 
proposed budgets on a State-by-State basis in order to assess the 
downwind benefits of applying the budgets in each State.
    Response: The EPA performed a multi-factor analysis to determine 
the amount of a State's emissions that significantly contribute to 
downwind nonattainment and what the resulting State budget should be. 
This is discussed in detail in Section II.C., Weight of Evidence 
Determination of Covered States. Specifically, EPA determined that 
emissions from all sources in certain States contribute to downwind 
problems, but that only a portion of those emissions--in some cases, a 
relatively small portion--may be reduced through highly cost-effective 
controls. The EPA established a budget for each State based on the 
elimination of these emissions. After EPA established the budgets, EPA 
performed air quality modeling to quantify the overall ozone benefits 
of the budgets applied in all upwind States on selected downwind areas. 
This modeling is described below. The EPA considered the results of 
this modeling as an additional piece of evidence in the analysis to 
confirm that the amount of emissions reductions from upwind States 
collectively provide meaningful reductions in nonattainment downwind.
    For the purposes of this modeling it is sufficient to model the 
budgets collectively, and not State-by-State, to demonstrate that the 
intended benefits of the budgets are achieved. Commenters who 
recommended State-by-State modeling generally argued that it would 
indicate that the reductions from a particular State would have a 
relatively small impact downwind, particularly compared to the impact 
of local reductions or reductions from other upwind States. In general, 
such a modeling result could stem from the relatively small amount of 
emissions reductions required of a particular upwind State under the 
SIP Call, due to EPA's decision to base the budgets on cost-effective 
controls rather than, more expensive controls. However, EPA's air 
quality modeling of the ambient impact of the required budgets in the 
upwind States on downwind nonattainment (discussed below) shows that 
even if the downwind ambient impact of the required reductions from a 
particular upwind State were small, that impact, when combined with the 
impact from the reductions required from other upwind States, provides 
meaningful downwind benefits. Ozone air quality problems are caused by 
the collective contribution from numerous sources over a large 
geographic area, so that it is appropriate to assess the impact of 
reductions from a particular upwind State in combination with 
reductions from other upwind States. The downwind air quality benefits 
from these upwind reductions confirm the appropriateness of the 
promulgated budgets.
    Comment: Commenters stated that EPA should have modeled alternative 
control options to determine if less stringent controls, either applied 
uniformly or on a subregional basis (i.e., multi-State subregional 
variations in control levels), would provide air quality benefits 
essentially equivalent to EPA's proposal. In addition, commenters 
submitted a considerable number of new modeling analyses intended to 
show that (a) sufficient downwind ozone benefits can be achieved with 
control levels less stringent than those associated with EPA's 
proposal; (b) controls applied in certain upwind States, when examined 
on a State-by-State basis, do not provide ``significant'' benefits in 
any downwind nonattainment area; and/or (c) NOX controls 
increase ozone locally in some areas and these increases are greater 
than the predicted decreases. In addition to new control strategy 
modeling, commenters submitted modeling that pertains to the finding of 
significant contribution. The EPA's responses to this modeling are 
discussed in Section II.C., Weight of Evidence Determination of Covered 
States and in the Response to Comment document.
    Response: In response to the comments on the need to model 
alternative controls, EPA has modeled alternative budgets based on 
several EGU and non-EGU control options. For the most part, these 
alternative budgets were modeled regionwide in order to assess, as 
discussed above, the benefits considering both downwind and upwind 
emissions reductions, collectively. Further, as discussed below, EPA 
modeled several other types of scenarios including runs to assess the 
impacts of the proposal applied in upwind States on several downwind 
areas. The EPA's modeling analyses are summarized below and described 
in detail in the Air Quality Modeling TSD.
    Regarding the new control strategy modeling submitted by 
commenters, EPA has reviewed this information in the same way it 
reviewed the new modeling on ``significant contribution'', as described 
in Section II.C., Weight of Evidence Determination of Covered States. 
Specifically, EPA reviewed the commenters' modeling to determine and

[[Page 57441]]

assess (a) the technical aspects of the models that were applied; (b) 
the treatment of emissions inventories; (c) the types of episodes 
modeled; (d) the methods for aggregating, analyzing, and presenting the 
results; (e) the completeness and applicability of the information 
provided; and (f) whether the technical evidence supports the arguments 
made by the commenters. A summary of this review is discussed next. For 
the most part, the commenters used either the UAM-V model and/or the 
CAMX model to assess the relative impacts of various 
NOX control strategies. As discussed in Section II.C. Weight 
of Evidence Determination of Covered States, modeling results from both 
models are viewed by EPA as technically acceptable. Concerning the 
emissions used for modeling, most commenters stated that they used the 
EPA SNPR or IPM-derived 2007 Base Case emissions as a starting point 
for developing emissions for the control scenarios. However, the 
commenters did not provide emissions data summaries in order for EPA to 
confirm which inventories were used in the modeling. Also, the 
commenters did not document in detail how they applied the controls to 
the emissions inventory.
    Most of the control strategy modeling submitted by commenters was 
performed for the July 1995 episode although a few commenters performed 
modeling for all four OTAG episodes and one commenter provided modeling 
for a non-OTAG episode in June of 1991. As discussed in Section II.C., 
and in the Response to Comment document, EPA's ability to fully 
evaluate and utilize the modeling submitted by commenters was hampered 
in some cases because only limited information on the results was 
provided.
    The EPA considered the strengths and limitations in the commenters' 
modeling analyses in evaluating whether the technical evidence 
presented in the comments supports the arguments made by the 
commenters. A detailed review of the commenters' modeling is contained 
in the Response to Comment document. In general, this review indicates 
that (a) downwind ozone benefits increase as greater NOX 
controls are applied to sources in upwind States, (b) emissions 
reductions at the level of the SIP Call, even when evaluated on an 
individual State-by-State basis, reduce ozone in downwind nonattainment 
areas, (c) the net benefits of NOX control at the level of 
the SIP Call outweigh any local disbenefits, and (d) upwind 
NOX reductions tend to mitigate local disbenefits in 
downwind areas. Thus, based on this evaluation, EPA generally found 
that the submitted modeling did not refute the overall conclusions EPA 
has drawn concerning the impacts of NOX emissions in the 
relevant geographic areas. However, because the extent and level of 
detail in the information presented by the commenters was, in many 
cases, limited and/or qualitative, the EPA decided to model a number of 
alternative control scenarios for all four OTAG episodes. The results 
of EPA's modeling of the impacts of alternative NOX controls 
are described next.

C. Assessment of Alternative Control Levels

    As indicated above, EPA has remodeled the Base Case and Statewide 
budgets using updated EGU emissions which do not exceed the capacity of 
individual units. In addition, EPA has performed modeling of various 
alternative EGU and non-EGU control options. Further, EPA has modeled 
the benefits in selected downwind areas of the budgets applied in 
upwind States. The results of EPA's modeling analyses are summarized 
below and described in more detail in the Air Quality Modeling TSD.
1. Scenarios Modeled
    As part of EPA's assessment, a 2007 SIP Call Base Case (hereafter 
referred to as the ``Base Case'') and eight emissions scenarios were 
modeled, as listed in Table IV-1. The first four scenarios (i.e. 
``0.25'', ``0.20'', ``0.15t'', and ``0.12'') were designed to evaluate 
alternative EGU and non-EGU controls applied uniformly in all 23 
jurisdictions. For each of these four scenarios, EGU emissions were 
determined assuming a cap-and-trade program across all 23 
jurisdictions. The 0.15t scenario reflects the SIP Call proposal for 
both non-EGU and EGU sources. Note that non-EGU controls were modeled 
at the level of the proposal for all scenarios except for the 0.25 
scenario for which less stringent controls were assumed.

                Table IV-1.--Emissions Scenarios Modeled
Base Case:
    2007 SIP Call Base Case \1\
        Point Sources: CAA Controls.
        Area Sources: OTAG ``Level 1'' Controls.
        Highway Vehicles: OTAG ``Level 0'' Controls.


0.25.........................................  0.25 lb/mmBtu, interstate         60% reduction for large
                                                trading.                          sources.
0.20.........................................  0.20 lb/mmBtu, interstate         70% reduction for large
                                                trading.                          sources, RACT for medium
                                                                                  sources\2\.
0.15t........................................  0.15 lb/mmBtu, interstate         70% reduction for large
                                                trading.                          sources, RACT for medium
                                                                                  sources.
0.12.........................................  0.12 lb/mmBtu, interstate         70% reduction for large
                                                trading.                          sources, RACT for medium
                                                                                  sources.
0.15nt.......................................  0.15 lb/mmBtu, intrastate         70% reduction for large
                                                trading.                          sources, RACT for medium
                                                                                  sources.


Downwind Scenarios for Analysis of ``Transport'':
    (1) 0.15nt EGU and non-EGU controls in the Northeast \3\; 2007 Base
     Case emissions elsewhere.
    (2) 0.15nt EGU and non-EGU controls in Georgia; 2007 Base Case
     emissions elsewhere.
    (3) 0.15nt EGU and non-EGU controls in Illinois, Indiana, and
     Wisconsin; 2007 Base Case emissions elsewhere.
\1\ See Table IV-2 for a listing of Base Case control measures.
\2\ Reductions are from 2007 ``uncontrolled'' emissions. Non-EGU sources
  >250mmBtu/hr are considered as ``large''; sources <250mmBtu/hr, but
  >1tpd are considered as ``medium''. The non-EGU point source controls
  assumed for purposes of this modeling do not match the levels assumed
  for the purpose of calculating the final budgets.
\3\ Northeast includes Connecticut, Delaware, District of Columbia,
  Maryland, Massachusetts, New Jersey, New York, Pennsylvania, and Rhode
  Island.


[[Page 57442]]

    The EPA also modeled a 0.15 intrastate trading scenario, 
``0.15nt'', which was constructed with EGU emissions that meet each 
State's budget without interstate trading. In developing the EGU 
emissions for this scenario, intrastate trading among sources in a 
State was allowed to occur. The benefits of the 0.15nt scenario 
compared to those from the 0.15t scenario were examined to determine 
whether an interstate trading program would affect the overall benefits 
of the proposal.
    The last three scenarios in Table IV-1 were designed to evaluate 
the downwind benefits resulting from reductions in transport due to the 
budgets in upwind States. Each of these scenarios constitutes a 
separate modeling run that applies the 0.15nt scenario in a different 
downwind area. For example, in the ``nt15NE'' scenario, the 0.15nt 
emissions budgets were applied only in those Northeast States subject 
to the SIP Call. The predictions from each of these three modeling runs 
for specific downwind areas were compared to the Base Case to estimate 
the impacts of the budgets applied only within the downwind area. The 
predictions from these three runs were then compared to the 0.15nt 
scenario across all 23 jurisdictions to estimate the additional 
benefits in each downwind area due to reductions in transport resulting 
from the budgets applied in both upwind and downwind States.
2. Emissions for Model Runs
    As indicated in Table IV-1, Base Case emissions for area sources 
(including nonroad), highway vehicles, and non-EGU sources represent a 
combination of OTAG emissions data for various control levels. This 
includes CAA controls on non-EGU point sources, OTAG ``level 1'' 
controls on area sources, and ``level 0'' controls on highway vehicles. 
The control measures included in the Base Case for each source category 
are listed in Table IV-2. These modeling runs were performed before 
changes were made to the inventory in response to comments. For the 23 
jurisdictions as a whole, the Base Case NOX emissions that 
were modeled are 2 percent higher than the final Base Case emissions 
that reflect changes made in response to comments.

              Table IV-2.--2007 SIP Call Base Case Controls
------------------------------------------------------------------------
EGUs:
    Title IV Controls [ phase 1 and 2 ].
    --250 Ton PSD and NSPS.
    --RACT & NSR in non-waived NAAs.
Non-EGU Point:
    --NOX RACT on major sources in non-waived NAAs.
    --250 Ton PSD and NSPS.
    --NSR in non-waived NAAs.
    --CTG and Non-CTG VOC RACT at major sources in NAAs and OTR.
    --New Source LAER.
Stationary Area:
    --Two Phases of VOC Consumer and Commercial Products and One Phase
     of Architectural Coatings controls.
    --VOC Stage 1 and 2 Petroleum Distribution Controls in NAAs.
    --VOC Autobody, Degreasing and Dry Cleaning controls in NAAs.
Nonroad Mobile:
    Fed Phase II Small Eng. Stds.
    --Fed Marine Eng. Stds.
    --Fed Nonroad Heavy-Duty (=50 hp) Engine Stds--Phase 1.
    --Fed RFG II (statutory and opt-in areas).
    --9.0 RVP maximum elsewhere in OTAG domain.
    --Fed Locomotive Stds (not including rebuilds).
    --Fed Nonroad Diesel Engine Stds--Phases 2 and 3.
Highway Vehicles:
    --National LEV.
    --Fed RFG II (statutory and opt-in areas).
    --9.0 RVP maximum elsewhere in OTAG domain.
    --High Enhanced I/M (serious and above NAAs).
    --Low Enhanced I/M for rest of OTR.
    --Basic I/M (mandated NAAs).
    --Clean Fuel Fleets (mandated NAAs).
    --On-board vapor recovery.
    --HDV 2 gm std.
Rate of Progress Requirements:
    --Effectively, ROP through 1999.
------------------------------------------------------------------------

    Note that area and mobile source emissions were held constant at 
Base Case levels in all scenarios. The Base Case emissions for EGUs 
were obtained from simulations of IPM which projected 1996 electric 
generation to 2007 based on economic assumptions, unit specific 
capacity, and the requirements in Title I and Title IV of the CAA. The 
Base Case emissions that were modeled for the EGU sector are 4 percent 
higher than the final Base Case emissions for this sector. The EGU 
emissions estimates for each of the control scenarios in Table IV-1 
were also derived using the IPM. Table IV-3 summarizes the emissions 
reductions provided by the control scenarios compared to the Base Case. 
The development of emissions data for air quality modeling is further 
described in the Air Quality Modeling TSD.

[[Page 57443]]



                                Table IV-3.--Summary of NOX Emissions Reductions
----------------------------------------------------------------------------------------------------------------
           Region \1\                  0.25            0.20            0.15t           0.12           0.15nt
----------------------------------------------------------------------------------------------------------------
                  Percent Reduction in Point Source NOX Emissions From 2007 SIP Call Base Case
----------------------------------------------------------------------------------------------------------------
Northeast.......................              29              39              49              52              46
Midwest.........................              40              51              59              65              58
Southeast.......................              35              49              54              61              56
SIP Call \2\....................              37              48              57              62              57
----------------------------------------------------------------------------------------------------------------
                      Percent Reduction in Total NOX Emissions From 2007 SIP Call Base Case
----------------------------------------------------------------------------------------------------------------
Northeast.......................              13              18              22              24              21
Midwest.........................              22              28              33              36              32
Southeast.......................              19              26              29              32              30
SIP Call \2\....................              20              26              30              33             30
----------------------------------------------------------------------------------------------------------------
\1\ The Northeast includes Connecticut, Delaware, District of Columbia, Maryland, Massachusetts, New Jersey, New
  York, Pennsylvania, and Rhode Island; the Midwest includes Illinois, Indiana, Kentucky, Michigan, Missouri
  Ohio, West Virginia, and Wisconsin; the Southeast includes Alabama, Georgia, North Carolina South Carolina,
  Tennessee and Virginia.
\2\ ``SIP Call'' includes the total percent reduction over all 23 jurisdictions subject to budgets as part of
  this notice.

3. Modeling Results
    The EPA applied UAM-V for each of the four OTAG episodes to 
simulate ozone concentrations for the Base Case and each scenario. The 
results for the uniform regionwide scenarios are presented first. This 
is followed by the results comparing interstate and intrastate trading. 
The results for the assessment of overall downwind benefits of the 
budgets applied in upwind States is presented last.
    The analysis of model predictions focused 1-hour daily maximum 
values and 8-hour daily maximum values predicted for all 4 episodes. 
The rationale for analyzing the model predictions in this way is 
discussed in Section II.C. Each of the control scenarios was evaluated 
using the four ``metrics'' listed in Table IV-4. Note that the model 
predictions used in calculating the metrics were restricted to those 1-
hour values >=125 ppb and 8-hour values >=85. Model predictions less 
than these concentrations were not included in the analysis.

                    Table IV-4.--Air Quality Metrics
------------------------------------------------------------------------
Metric 1: Exceedances........  The number of values above the
                                concentration level of NAAQS.\1\
Metric 2: Ozone Reduced-ppb..  The magnitude and frequency of the
                                ``ppb'' reductions in ozone.
Metric 3: Total ppb Reduced..  The total ``ppb'' reduced by a given
                                scenario, not including that portion of
                                the reduction that occurs below the
                                level of the NAAQS.
Metric 4: Population-Weighted  The same as Metric 3, except that the
 Total ppb Reduced.             ozone reductions are weighted by the
                                population in the grid cell in which the
                                reductions occur.
------------------------------------------------------------------------
\1\ 1-hour values >=125 ppb; 8-hour values >=85 ppb.

    A full description of these metrics and the procedures for 
selecting ``nonattainment'' receptors for calculating the metrics can 
be found in the Air Quality Modeling TSD. In brief, ``nonattainment'' 
receptors for the 1-hour analysis include those grid cells that (a) are 
associated with counties designated as nonattainment for the 1-hour 
NAAQS and (b) have 1-hour Base Case model predictions >=125 ppb. These 
grid cells are referred to as ``designated plus modeled'' nonattainment 
receptors. Using these receptors, the metrics were calculated for each 
1-hour nonattainment area as well as for each State. To calculate the 
metrics by State, the ``nonattainment'' receptors in that State were 
pooled together.
    For the 8-hour analysis, ``nonattainment'' receptors include those 
grid cells that (a) are associated with counties currently violating 
the 8-hour NAAQS and (b) have 8-hour Base Case model predictions >=85 
ppb. These grid cells are referred to as ``violating plus modeled'' 
nonattainment receptors. The metrics were calculated on a State-by-
State basis for the 8-hour analyses.
    In general, the four metrics lead to similar overall conclusions. 
The results for the full set of receptor areas (i.e., ``designated plus 
modeled'' for the 1-hour NAAQS and ``violating plus modeled'' for the 
8-hour NAAQS) are provided in the Air Quality Modeling TSD for all four 
metrics. In this preamble, Metrics 1 and 3 are presented to illustrate 
the results.
    a. Impacts of Alternative Controls. The impacts on ozone 
concentrations of the 0.15t scenario and each of the alternative 
scenarios are provided by region (i.e., Midwest, Southeast, and 
Northeast) in Tables IV-5 and IV-6 for Metrics 1 and 3, respectively. 
The complete set of data for individual States and 1-hour nonattainment 
areas is provided in the Air Quality Modeling TSD. Table IV-5 shows the 
percent reduction in the number of exceedances across all four episodes 
between each control scenario and the Base Case. Table IV-6 shows the 
percent reduction in total ozone above the NAAQS provided by each 
scenario, compared to the total ozone above the NAAQS in the Base Case.

[[Page 57444]]



                            Table IV-5.--Results for Metric 1: Number of Exceedances
----------------------------------------------------------------------------------------------------------------
                                       0.25            0.20            0.15t           0.12           0.15nt
----------------------------------------------------------------------------------------------------------------
                  Percent Reduction in the Number of Exceedances 1-Hour Daily Maximum >=125 ppb
----------------------------------------------------------------------------------------------------------------
Midwest.........................              25              32              38              43              38
Southeast.......................              23              33              34              40              36
Northeast.......................              24              31              36              39              36
SIP Call Total..................              24              31              36              40              37
----------------------------------------------------------------------------------------------------------------
                  Percent Reduction in the Number of Exceedances 8-Hour Daily Maximum >=85 ppb
----------------------------------------------------------------------------------------------------------------
Midwest.........................              35              44              50              54              49
Southeast.......................              30              40              46              51              48
Northeast.......................              26              34              41              44              41
SIP Call Total..................              30              39              45              49              45
----------------------------------------------------------------------------------------------------------------


                            Table IV-6.--Results for Metric 3: Total ``ppb'' Reduced
----------------------------------------------------------------------------------------------------------------
                                       0.25            0.20            0.15t           0.12           0.15nt
----------------------------------------------------------------------------------------------------------------
 Total ``ppb'' Reduced Compared to the Total ``ppb'' Above NAAQS in Base Case \1\ 1-Hour Daily Maximum >=125 ppb
----------------------------------------------------------------------------------------------------------------
Midwest.........................              31              39              45              49              44
Southeast.......................              27              37              39              44              41
Northeast.......................              25              32              37              40              37
SIP Call Total..................              27              35              40              43              40
----------------------------------------------------------------------------------------------------------------
   Total ``ppb'' Reduced Compared to the Total ``ppb'' Above NAAQS in Base Case 8-Hour Daily Maximum >=85 ppb
----------------------------------------------------------------------------------------------------------------
Midwest.........................              35              42              48              52              47
Southeast.......................              33              44              49              53              50
Northeast.......................              28              37              43              46              43
SIP Call Total..................              31              40              46              50              46
----------------------------------------------------------------------------------------------------------------
\1\ The values in this table were calculated by dividing the Total ``ppb'' Reduced in the control scenario by
  the Total ``ppb'' above the NAAQS in the Base Case. These values represent the percent of total ozone above
  the NAAQS in te Case that is reduced by the control scenario.

    The results indicate that the 0.15t scenario provides substantial 
reductions in both 1-hour and 8-hour ozone concentrations in all three 
regions.
    In the Midwest the 0.15t scenario provides a 38 percent reduction 
in 1-hour exceedances and a 45 percent reduction in ``total ozone'' 
>=125 ppb. The regionwide Midwest reductions in 8-hour exceedances and 
``total ozone'' >=85 ppb are 45 percent and 50 percent, respectively. 
Considering individual 1-hour nonattainment areas in this region, the 
reduction in exceedances due to the 0.15t controls are 36 percent over 
Lake Michigan,61 73 percent in Southwest Michigan, and 54 
percent in Louisville. The corresponding reductions in ``total ozone'' 
>=125 ppb are 44 percent over Lake Michigan, 81 percent in southwest 
Michigan, and 64 percent in Louisville. The results for other areas are 
contained in the Air Quality Modeling TSD.
---------------------------------------------------------------------------

    \61\ The rationale for analyzing the impacts over Lake Michigan 
is discussed in Section II.C, Weight of Evidence Determination of 
Covered States.
---------------------------------------------------------------------------

    In the Southeast, 1-hour exceedances are reduced by 39 percent and 
the ``total ozone'' >=125 ppb by 34 percent. Considering individual 
nonattainment areas in the Southeast, the 0.15t scenario provides a 36 
percent reduction in 1-hour exceedances in Atlanta and a 39 percent 
reduction in exceedances in Birmingham. The reduction in ``total 
ozone'' >=125 ppb is 41 percent in Atlanta and 54 percent in 
Birmingham. The overall regionwide ozone benefits across the Southeast 
are also large for the 8-hour NAAQS. For example, the number of 8-hour 
exceedances in this region is reduced by 46 percent with the 0.15t 
scenario.
    In the Northeast, 0.15t provides a 37 percent reduction in 1-hour 
exceedances and a 34 percent reduction in ``total ozone'' >=125 pp. For 
individual nonattainment areas in the Northeast, the reductions in both 
Metrics 1 and 3 range from approximately 25 percent in Washington, DC 
up to 100 percent in Pittsburgh. For the serious and severe 1-hour 
nonattainment areas along the Northeast Corridor from Washington, DC to 
Boston, the 1-hour reductions vary from city to city, but are generally 
in the range of 25 percent to 55 percent. The regionwide reductions in 
8-hour exceedances and ``total ozone'' >=85 ppb in the Northeast are 
above 40 percent.
    In general, results from the scenarios evaluated demonstrate that 
the larger the reduction in NOX emissions, the greater the 
overall ozone benefit. As indicated in Table IV-5 and IV-6, the 0.25 
and 0.20 scenarios generally do not provide the same level of reduction 
as the 0.15t scenario in any of the three regions, whereas the 0.12 
scenario provides additional ozone benefits beyond 0.15t in all three 
regions. Also, the results indicate that even with the most stringent 
control option considered, nonattainment problems requiring additional 
local controls may continue in some areas currently violating the 
NAAQS.
    The impact on ozone reductions of a trading program versus meeting 
the budgets in each State can be seen by comparing the results for the 
0.15t and 0.15nt scenarios. The data in Tables IV-5 and IV-6 indicate 
that there is no overall loss of ozone benefits for either 1-hour or 8-
hour concentrations across the 23 jurisdictions due to trading. On a 
regional basis, the benefits of interstate and intrastate trading at 
the 0.15 control level are essentially the same in the Northeast and 
Midwest and slightly less with interstate trading in the Southeast.

[[Page 57445]]

    As indicated in the summary of comments, several commenters stated 
that there would be local disbenefits due to the EPA proposal that 
would outweigh any benefits. The modeling runs discussed here shed 
light on the issue. Of the four metrics examined by EPA, Metrics 3 and 
4 (i.e., ``Total ppb Reduced'' and ``Population-Weighted Total ppb 
Reduced'') are most appropriate for identifying any net disbenefits 
because the ozone decreases and any increases (disbenefits) are 
considered in calculating each of these metrics. The metrics will have 
negative values for situations in which the total disbenefits are 
greater than the total benefits. The EPA examined the 1-hour estimates 
for these metrics for each 1-hour nonattainment area and the 8-hour 
estimates by State to identify any areas in which the modeling 
indicated a net disbenefit. The results indicate that the only net 
disbenefit predicted in any of the scenarios was in Cincinnati for the 
1-hour NAAQS. However, these disbenefits occurred only in the 0.25 and 
0.20 scenarios. In the 0.15t scenario, there is a net 32 percent 
benefit in Cincinnati with Metric 3 and a net benefit of 23 percent 
with Metric 4. There were no net Statewide 8-hour disbenefits in any of 
the scenarios examined by EPA.
    b. Impacts of Upwind Controls on Downwind Nonattainment. The 
impacts of the budgets applied in upwind States on downwind ozone in 
the (a) the Northeast, (b) Georgia, and (c) Illinois-Indiana-Wisconsin, 
were evaluated by comparing the 0.15nt scenario to the three downwind 
transport assessment scenarios listed in Table IV-1. In each of these 
three scenarios, EPA modeled the 0.15nt option in one of the downwind 
areas with the Base Case emissions applied in the rest of the OTAG 
region.62 The results of each downwind control run were 
compared to the Base Case in order to assess the benefits of the 
controls applied within those areas (i.e., the downwind areas). 
Similarly, the predictions for the 0.15nt regionwide scenario were 
compared to the Base Case to estimate the benefits in each area of the 
downwind plus upwind controls. The benefits of the upwind controls were 
determined by calculating the difference between the benefits of the 
downwind controls compared to the benefits of the downwind plus upwind 
controls. The results are provided in Table IV-7. The following is an 
example of how the benefits of upwind controls were calculated for 
Metric 1 (i.e., number of exceedances). In the Northeast, there were 
1052 grid-day exceedances of the 1-hour NAAQS predicted in the Base 
Case scenario. In the downwind control scenario (i.e., 0.15nt applied 
in the Northeast only), the number of exceedances declined to 827 grid-
days which represents a 21 percent reduction in exceedances from the 
Base Case due to controls in the Northeast. In the downwind plus upwind 
scenario, the number of 1-hour exceedances declined even further to 670 
grid-days which is a 36 percent reduction from the Base Case. 
Therefore, the upwind controls provide a 15 percent reduction in 1-hour 
exceedances in the Northeast (i.e., 36 percent versus 21 percent).
---------------------------------------------------------------------------

    \62\ As described in the Air Quality Modeling TSD, emissions 
from the intrastate trading scenario rather than the interstate 
trading scenario were used for the analysis of upwind controls in 
order to avoid any potentially confounding effects of small changes 
in the downwind emissions between the downwind control scenario and 
the downwind plus upwind control scenario due to interstate trading.
---------------------------------------------------------------------------

    For Metric 3 (i.e., Total ``ppb'' Reduced), the impact of upwind 
controls on downwind ozone was determined using two approaches. The 
first approach is similar to the procedures followed described above 
for exceedances. For example, in the Northeast the total ppb >=125 ppb 
(across all grids and days) in the Base Case was 14,724 ppb. In the 
downwind control scenario the total ppb reduced by these controls was 
3289 ppb which represents a 22 percent reduction (i.e., 3289 ppb 
divided by 14,724 ppb) in total ppb >=125 ppb. In the downwind plus 
upwind control scenario, the total ppb reduced was 5500 ppb which 
represents a 37 percent reduction in total ppb >=125 ppb in the Base 
Case. Therefore, the upwind controls provide a 15 percent reduction in 
total ppb >=125 ppb (i.e., 37 percent versus 22 percent). The results 
for Metric 3 calculated using this first approach are presented in 
Table IV-7.
    A second approach to analyze the benefits of upwind controls using 
Metric 3 is to determine the fraction or percentage of the total 
reduction from downwind plus upwind controls that comes from just the 
upwind controls. This is determined by first subtracting the ppb 
reduced by downwind controls from the ppb reduced by downwind plus 
upwind controls. This difference provides an estimate of the portion of 
the reduction due to upwind controls. Then, the portion of the 
reduction due to upwind controls is divided by the reduction from 
downwind plus upwind controls to estimate the percent of reduction due 
to the upwind controls only. For example, in the Northeast the 1-hour 
total ppb reduced by the downwind plus upwind controls is 5500 ppb and 
the total ppb reduced by the downwind controls is 3289 ppb. The 
difference (2211 ppb) is the estimated amount of reduction due to 
upwind controls. Thus, in this example, the upwind controls provide 40 
percent (i.e., 2211 ppb divided by 5500 ppb) of the total ppb reduction 
in the downwind plus upwind regionwide scenario. The results for Metric 
3 using this second approach for estimating the impacts of upwind 
controls are provided in Table IV-8.

--------------------------------------------------------------------------------------------------------------------------------------------------------
                                                                         1-hour daily max                                8-hour daily max
                                                         -----------------------------------------------------------------------------------------------
                                                              DW \1\        DW + UW \1\       UW \1\            DW            DW + UW           UW
--------------------------------------------------------------------------------------------------------------------------------------------------------
                                                            Percent Reduction in Exceedances
--------------------------------------------------------------------------------------------------------------------------------------------------------
Northeast...............................................              21              36              15              18              40              22
Lake MI.................................................              29              36               7              11              17               6
IL/IN/WI................................................              35              50              15              27              57              30
Atlanta.................................................              30              39               9          \2\ NA              NA              NA
Georgia \3\.............................................              30              39               9              15              27              12
--------------------------------------------------------------------------------------------------------------------------------------------------------
                                                   Percent Reduction in Total ``ppb'' Above the NAAQS
--------------------------------------------------------------------------------------------------------------------------------------------------------
Northeast...............................................              22              37              15              23              43              20
Lake MI.................................................              39              44               5              20              28               8
IL/IN/WI................................................              17              33              16              32              62              30
Atlanta.................................................              37              43               6              NA              NA              NA

[[Page 57446]]

Georgia.................................................              37              43               6              25              35              10
--------------------------------------------------------------------------------------------------------------------------------------------------------
\1\ ``DW'' denotes the reductions due to the downwind controls; ``DW + UW'' denotes the reductions due to controls applied regionwide in upwind plus
  downwind areas; and ``UW'' denotes the incremental additional reduction in exceedances.
\2\ NA: The metrics for the 8-hour NAAQS were not calculated for individual 1-hour nonattainment areas.
\3\ The 1-hour results for Georgia are the same as for Atlanta because Atlanta is the only 1-hour nonattainment area in that State.


  Table IV-8.--Percent of the Total ppb Above the NAAQS That Is Reduced
                         Due to Upwind Controls
------------------------------------------------------------------------
                                           1-hour daily    8-hour daily
                                           max (percent)   max (percent)
------------------------------------------------------------------------
Northeast...............................              40              48
Lake MI.................................              12              27
IL/IN/WI................................              49              48
Atlanta.................................              14              NA
Georgia.................................              14              28
------------------------------------------------------------------------

    In the following discussion of the impacts of upwind controls on 
ozone in the three downwind areas, the results for Metric 3 focus on 
the second approach for calculating upwind impacts using this metric 
since the results based on the first approach are similar to those for 
Metric 1, as indicated in Table IV-7.
    In the Northeast, the upwind controls provide a 15 percent 
reduction in 1-hour exceedances and a 22 percent reduction in 8-hour 
exceedances. The results in Table IV-8 indicate that upwind controls 
provide 40 percent or more of the total ppb reduction from the downwind 
plus upwind control scenario for both the 1-hour and 8-hour NAAQS. 
Considering the results for several 1-hour nonattainment areas in the 
Northeast, the upwind controls reduce the number of 1-hour exceedances 
by 21 percent in Baltimore, 12 percent in Philadelphia, 12 percent in 
New York City, 19 percent in Greater Connecticut, and 3 percent in 
Boston. The percent of the total ppb reduction from the downwind plus 
upwind controls that is due to the upwind controls alone is 48 percent 
in Baltimore, 29 percent in Philadelphia, 38 percent in New York City, 
47 percent in Connecticut, and 25 percent in Boston. The results for 
all of the Northeast 1-hour nonattainment areas are provided in the Air 
Quality Modeling TSD.
    The impacts of upwind controls on nonattainment in Georgia were 
examined using the 0.15nt scenario in Georgia versus the Base Case 
scenario and the scenario with 0.15nt applied regionwide. The results, 
as shown in Table IV-7, indicate that the upwind controls are predicted 
to reduce the number of 1-hour exceedances in Atlanta by 9 percent. 
Also, in Atlanta, 14 percent of the 1-hour total ppb above the NAAQS 
reduced by the downwind plus upwind regionwide scenario is due to the 
controls applied in upwind States. For the 8-hour NAAQS, the upwind 
controls provide a 12 percent reduction in 8-hour exceedances within 
the State of Georgia. The upwind controls provide 28 percent of the 
total ppb reduction in the downwind plus upwind regionwide control 
scenario.
    To assess the benefits in Illinois-Indiana-Wisconsin due to upwind 
controls, EPA examined the data for the Lake Michigan receptor area and 
for the three States, combined. The discussion of results focuses on 
the Lake Michigan receptor area. The data for this area and the three 
States are provided in Table IV-7. For the Lake Michigan receptor area, 
there is a 7 percent reduction in 1-hour exceedances and a 6 percent 
reduction in 8-hour exceedances due to upwind controls. The upwind 
controls provide 12 percent of the total 1-hour reduction and 27 
percent of the total 8-hour reduction that results from the downwind 
plus upwind regionwide controls. In Illinois, Indiana, and Wisconsin, 
the reduction in 1-hour and 8-hour exceedances due to upwind controls 
are larger than over Lake Michigan (i.e., 15 percent and 30 percent for 
1-hour and 8-hour exceedances, respectively). The upwind controls 
provide nearly 50 percent of the total ppb reductions associated with 
the downwind plus upwind regionwide control scenario for both the 1-
hour and 8-hour NAAQS.
    Based on the results discussed above, EPA believes that the 
controls in today's rulemaking applied in upwind areas will reduce the 
number of 1-hour and 8-hour exceedances in downwind nonattainment 
areas. The analysis indicates that in downwind areas, a substantial 
portion of the 1-hour and 8-hour ozone reductions provided by the 
regionwide application of these controls are due to those controls in 
upwind areas.
    c. Summary of Findings. The EPA has performed an air quality 
assessment to estimate the ozone benefits of the proposal and several 
alternative uniform regionwide control levels. In addition, EPA 
examined the overall benefits in several major downwind nonattainment 
areas of the application of the proposal in upwind States. The results 
of EPA's assessment corroborate and extend the findings presented in 
the SNPR. The major findings are as follows: (1) The NOX 
emissions reductions associated with the proposed Statewide budgets are 
predicted to produce large reductions in (a) 1-hour concentrations 
>=125 ppb in areas which are currently nonattainment for the 1-hour 
NAAQS and which would likely continue to have a 1-hour nonattainment 
problem in the future without the SIP call budget reductions, and (b) 
8-hour concentrations >=85 ppb in areas which currently violate the 8-
hour NAAQS and which would likely continue to have an 8-hour ozone 
problem in the future without the SIP call budget reductions.
    (2) The more NOX emissions are reduced, the greater the 
benefits in reducing ozone concentrations. There does not appear to be 
any ``leveling off'' of benefits within the range of NOX 
reductions associated with EPA's proposal. That is, NOX 
reductions at control levels less than EPA's proposal provide fewer air 
quality benefits than the proposal and NOX reduction greater 
than the proposal provide more air quality benefits.

[[Page 57447]]

    (3) Any disbenefits due to the NOX reductions associated 
with the budgets are expected to be very limited compared to the extent 
of the benefits expected from these budgets.
    (4) There are likely to be benefits in major nonattainment areas 
due to the downwind application of controls in the proposed budgets. 
Reductions in ozone transport associated with the collective 
application of the budgets in upwind States are expected to provide 
substantial ozone benefits in downwind areas, beyond what is provided 
by the budgets applied in the downwind areas alone. Together, the 
downwind reductions and transport reductions from upwind controls will 
provide significant progress toward attainment in major nonattainment 
areas within the OTAG region. However, even with the most stringent 
control option considered, nonattainment problems requiring additional 
local control measures may continue in some areas currently violating 
the NAAQS.

V. NOX Control Implementation and Budget Achievement 
Dates

A. NOX Control Implementation Date

    In the NPR, the EPA proposed to mandate NOX emissions 
decreases in each affected State leading to a budget based on 
reductions to be achieved from both Federal and State measures. The EPA 
further proposed that the required SIP revisions for achieving the 
portion of the NOX reduction from State measures be 
implemented by no later than September 2002. The EPA also requested 
comment on a range of compliance dates between September 2002 and 
September 2004.
    The EPA stated that this range of compliance dates is consistent 
with the requirement for severe 1-hour nonattainment areas to attain 
the standard no later than 2005 (for severe-15 areas) or 2007 (for 
severe-17 areas). With respect to the 8-hour ozone standard, EPA stated 
that the CAA provides for attainment within 5 years of designation as 
nonattainment, which must occur no later than July 2000, with a 
possible extension of up to 10 years following designation as 
nonattainment. The EPA stated that the range of implementation dates--
from September 2002 to September 2004--is consistent with the 
attainment time frames for the 8-hour standard (62 FR 60328-29). For 
the reasons described in Section III, below, the applicable attainment 
date for all affected downwind areas is ``as expeditiously as 
practicable,'' but no later than certain prescribed dates. In many 
cases, the date for achieving the upwind reductions will make the 
difference as to when downwind States will attain. Thus, it is 
appropriate for EPA to require the upwind reductions to be achieved as 
expeditiously as practicable. Subsection 1., below, analyzes the 
earliest date feasible for achieving the upwind reductions.
1. Practicability
    After reviewing the comments and analyzing the feasibility of 
implementing the NOX controls assumed for purposes of 
developing the State emissions budgets, as well as other measures which 
States may choose to rely on to meet the rule, the EPA is today 
determining that the required implementation date must be by no later 
than May 1, 2003. The Agency received many comments on the feasibility 
of installing appropriate control technology by 2003, and the 
succeeding paragraphs address many of the significant comments 
submitted on this topic.
    Some commenters asserted that a compliance deadline of September 
2002 is infeasible for completing the installation of the assumed 
NOX controls. Some of these commenters argued that there are 
not enough trained workers, engineering services or materials and 
equipment to install NOX controls by the September 2002 
deadline. Other commenters expressed concern that utilities will not 
have sufficient time to install NOX controls without causing 
electrical power outages; these commenters stated that such power 
outages would have adverse impacts on the reliability of the 
electricity supply. Commenters also expressed concern that retrofitting 
NOX controls would require increasing the operation of less 
efficient units, which would increase compliance costs.
    In response to these comments, the Agency has conducted a detailed 
examination of the feasibility of installing the NOX 
controls that EPA assumed in constructing the emissions budgets for the 
affected States (hereinafter, the ``assumed control strategy''). See 
the technical support document ``Feasibility of Installing 
NOX Control Technologies By May 2003,'' EPA, Office of 
Atmospheric Programs, September 1998. The Agency's findings are 
summarized below. Based on these findings, the EPA believes that the 
compliance date of May 1, 2003 for NOX controls to be 
installed to comply with the NOX SIP call is a feasible and 
reasonable deadline. The Agency is also providing some compliance 
flexibility to States for the 2003 and 2004 ozone seasons by 
establishing State compliance supplement pools as described above in 
Section III.F.6.
    The EPA's projections for the assumed control strategy include 
post-combustion controls (Selective Catalytic Reduction [SCR] and 
Selective Noncatalytic Reduction [SNCR]) and combustion controls (e.g., 
low NOX burners, overfire air, etc.)
    a. Combustion Controls. In general, the implementation of 
combustion controls should be readily accomplished by May 1, 2003 for 
the following reasons. First, there is considerable experience with 
implementing combustion controls. Combustion control retrofits on over 
230 utility boilers, accounting for over 75 GWe of capacity under the 
title IV NOX program, took place within 4 years (i.e., from 
1992 through 1995). Moreover, the combustion retrofits under Phase I of 
the Ozone Transport Commission's Memorandum of Understanding were 
completed in the same time frame. As a result of this experience, the 
sources and permitting agencies are familiar with the installation of 
combustion controls. This familiarity should result in relatively short 
time frames for completing technology installations and obtaining 
relevant permits.
    Second, combustion controls are constructed of commonly available 
materials such as steel, piping, etc., and do not require reagent 
during operation. Therefore, the EPA does not expect delays due to 
material shortages to occur at sites implementing these controls.
    Third, there are many vendors of combustion control technology. 
These vendors should have ample capacity to meet the NOX SIP 
call needs because they were able to satisfy significant installation 
needs during the period 1992 through 1995, as mentioned above. Since 
then these vendors have had relatively few installation needs to fill.
    Therefore, it is reasonable to expect that implementation of post-
combustion controls, not combustion controls, would determine the 
schedule for implementing all of the projected NOX controls.
    b. Post-Combustion Controls. Tables V-1 and V-2 present the Agency 
projections of how many electricity generating units and industrial 
sources, respectively, would need to be retrofitted with post-
combustion NOX controls under the assumed control strategy.

[[Page 57448]]



                Table. V-1.--Electricity Generating Units
------------------------------------------------------------------------
                                                           Projected No.
                       NOX Control                               of
                                                           installations
------------------------------------------------------------------------
Coal SCR.................................................           142
Coal SNCR................................................           482
Oil/gas SNCR.............................................            15
                                                          --------------
    Total................................................           639
------------------------------------------------------------------------


              Table. V-2.--Non-Electricity Generating Units
------------------------------------------------------------------------
                                                           Projected No.
                       NOX Control                               of
                                                           installations
------------------------------------------------------------------------
SCR on coal-fired sources................................            55
SCR on oil/gas-fired sources.............................           225
SCR on other sources.....................................             1
                                                          --------------
    Total................................................           281
                                                          ==============
SNCR on coal-fired sources...............................           195
SNCR on oil/gas-fired sources............................             0
SNCR on other sources....................................            40
                                                          --------------
    Total................................................           235
------------------------------------------------------------------------

    There are three basic considerations related to implementation of 
post-combustion controls (SCR and SNCR) by the compliance date: (1) 
Availability of materials and labor, (2) the time needed to implement 
controls at plants with single or multiple retrofit requirements, and 
(3) the potential for interruptions in power supply resulting from 
outages needed to complete installations.
    The EPA examined each of these considerations. An adequate supply 
of off-the-shelf hardware (such as steel, piping, nozzles, pumps, soot 
blowers, fans, and related equipment), reagent (ammonia and urea), and 
labor would be available to complete implementation of post-combustion 
controls projected under the assumed control strategy.
    However, the catalyst used in the SCR process is not an off-the-
shelf item and, therefore, requires additional consideration. Based on 
the projections shown in the tables above, the EPA estimates that about 
54,000 to 90,000 m\3\ of catalyst may be needed in SCR installations. 
The EPA has found that currently the catalyst suppliers can supply 
about 43,000 to 67,000 m\3\ of catalyst per year. However, of this 
supply about 5,000 to 8,000 m\3\ of catalyst per year is needed to meet 
the requirements of the existing worldwide SCR installations. Based on 
these estimates, the EPA conservatively concludes that adequate 
catalyst supply should be available if SCR installations were to occur 
over a period of two years or more.
    In addition, in comments to EPA's proposed NOX reduction 
program, the Institute of Clean Air Companies (ICAC) stated that more 
than sufficient vendor capacity existed to supply retrofit SCR catalyst 
to the sources that would be controlled by SCR under the assumed 
control strategy.
    Implementation of a NOX control technology on a 
combustion unit involves conducting facility engineering review, 
developing control technology specifications, awarding a procurement 
contract, obtaining a construction permit, completing control 
technology design, installation, testing, and obtaining an operating 
permit. The EPA evaluated the amount of time potentially needed to 
complete these activities for a single unit retrofit and found that 
about 21 months would be needed to implement SCR while about 19 months 
would be needed to implement SNCR.
    The EPA examined several particularly complicated implementation 
efforts to assure an accurate and realistic estimate of the time needed 
to install SCR and SNCR. The EPA examined the data and determined that 
the assumed control strategy might lead one plant to choose to install 
a maximum of 6 SCRs. In another instance, a different plant might 
choose to install a maximum of 10 SNCRs under the assumed control 
strategy. The estimated total time needed to complete these 
installations is 34 months for 6 SCR systems and 24 months for 10 SNCR 
systems.
    Finally, the EPA examined the impact(s) that outages required for 
connecting NOX post-combustion controls to EGUs could 
potentially have on the supply of electricity and on the cost of this 
rule. The EPA has found that, generally, connections between a 
NOX control system and a boiler can be completed in 5 weeks 
or less. This connection period has been accounted for in both the 
single and multi-unit implementation times presented in the previous 
paragraph. On an EGU, the connection would have to be completed during 
an outage period in which the unit is not operational. The EPA's 
research reveals that currently, on average, about 5 weeks of planned 
outage hours are taken every year at an electricity generating unit. 
Therefore, the EPA expects that connection between a NOX 
control system and such a unit would be completed during one of these 
planned outages.
    Results of EPA's analyses reflect that, even if all of the post-
combustion controls projected in Table V-1 for the EGUs were to be 
connected to these units in one single year, no disruption in the 
supply of electricity would occur. If each of these plants takes the 
five week outage in a single block of time, no cost increase is 
expected to occur. However, if a plant divides the five week outage 
into two or more periods, a cost increase of less than one-half of one 
percent may be expected. See the technical support document 
``Feasibility of Installing NOX Control technologies By May 
2003,'' EPA, Office of Atmospheric Programs, September 1998.
    Based on the estimated timelines for implementing NOX 
controls at a plant and availability of materials and labor, the EPA 
estimates that the NOX controls in the assumed control 
strategy (which is one available method for achieving the required 
NOX reductions in each covered State) could be readily 
implemented by September 2002, without causing an adverse impact on the 
electricity supply or on the cost of compliance. The EPA bases this 
conclusion on its analysis that the most complex and time-consuming 
implementation effort--one involving 6 SCR systems--would take 34 
months, and that all of the controls could be installed within this 
period without causing any disruptions in the supply of electricity.
    Further, the EPA notes that the September 27, 1994 OTC 
NOX Memorandum of Understanding (MOU) provides that large 
utility and nonutility NOX sources should comply with the 
Phase III controls by the year 2003. The levels of control in the MOU 
are 75 percent or 0.15 lb/106 btu in the inner and outer 
zones of the Northeast OTR, levels comparable to the controls assumed 
in setting the budget for today's rulemaking. Moreover, several States 
in the Northeast OTR have submitted SIP revisions implementing this 
level of emissions reductions from NOX sources in those 
States by May 1, 2003. This further supports the feasibility of the May 
1, 2003 implementation date for these controls.
    The EPA has determined that States would have sufficient time to 
implement other NOX control measures in lieu of the boiler 
controls described above. For example, vehicle I/M programs have 
historically required no more than two years to implement, including 
the time needed to pass enabling State legislation and to construct the 
necessary emission testing facilities. The time required to implement 
measures to reduce VMT depends on the nature of the measure, but many 
VMT reduction measures require no more than one or two years to 
implement. State opt-ins to the RFG program have generally required 
less

[[Page 57449]]

than one year to implement. Even if the EPA were to determine that 
supply considerations warranted a delay in implementing the opt-in 
request, the delay cannot exceed two years.
    States can also take advantage of the NOX-reducing 
benefits that energy efficiency and renewables projects provide, many 
of which could be developed in less than three years and incorporated 
into a SIP. Examples of efficiency/renewables projects that have been 
accomplished within a 3-year time frame and have resulted in 
significant NOX reductions include reducing boiler fuel use 
by utilizing waste heat, implementing short-term steam trap maintenance 
and inspection programs, and undertaking building upgrades using EPA's 
Energy Star Buildings approach.
2. Relationship to SIP Submittal Date
    Under this rule, as explained in Section B. below, States are 
required to submit revised SIPs by September 30, 1999. Commenters have 
suggested that based on the requirements of this rulemaking, sources in 
these States would need to begin early planning of compliance 
strategies before the September 30, 1999 date. The EPA disagrees. The 
EPA's technical analysis described above indicates that if these 
sources begin planning and specification of controls by even as late as 
April 2000, then they would be able to complete control technology 
implementation by May 1, 2003.
3. Rationale
    To assure adequate lead-time for implementation of controls, the 
EPA has moved the compliance deadline from the proposed date of 
September 2002 in the NPR to May 1, 2003. Since the ozone seasons in 
areas in the eastern U.S. end in the fall and begin in the spring, 
setting the implementation date for May 1, 2003 will provide sources 7-
8 additional months for implementing control requirements while not 
undermining the ability of areas to attain. The additional 
implementation time will occur during the cooler months of the year, a 
time when ozone exceedances generally do not occur. Thus, with either 
the September 2002 implementation date or the May 1, 2003 
implementation date, the 2003 ozone season would be the first to 
benefit from full implementation of the SIP call reductions.
    Several commenters contend that EPA does not have the authority to 
establish the compliance date. Since section 110(a)(2)(D)(i) is silent 
as to the implementation schedule for measures to prevent significant 
contribution, the EPA disagrees that the statute prohibits the EPA from 
establishing an implementation date for control measures that will 
achieve the reductions established by the SIP call. Thus, the EPA must 
look to the other provisions in the CAA, the legislative history, and 
the specific facts of today's rule to determine whether it is 
reasonable for the Agency to set the implementation date for the 
control measures. Furthermore, for the reasons provided in this 
Section, the EPA believes it is necessary to use its general rulemaking 
authority under section 301(a) to establish the latest date for 
implementation through a rule in order to ensure that downwind areas 
attain the standard as expeditiously as practicable and that areas 
continue to make progress toward attaining the NAAQS. See NRDC v. EPA, 
22 F.3d 1125, 1146-48 (D.C. Cir. 1994).
    With respect to the facts of this particular situation, this SIP 
call entails a complex analysis of the interstate transport of 
NOx and ozone and involves 23 jurisdictions. Although the 
States made significant progress through the OTAG process, they were 
unable to reach a final resolution on the emission reductions necessary 
or the schedule to achieve reductions to address upwind emissions. 
Thus, it would not be reasonable for EPA to leave open the issue of 
implementation in light of the need for downwind areas to rely on these 
reductions in order to demonstrate attainment by their attainment 
dates. See also the discussion in Section II.A.
    Furthermore, EPA believes that requiring implementation of the SIP-
required upwind controls, and thereby mandating those upwind 
reductions, by no later than May 1, 2003, is consistent with the 
purpose and structure of title I of the CAA. Under both section 
172(a)(2), which establishes attainment dates for areas designated 
nonattainment for the 8-hour standard, and section 181(a), which 
establishes attainment dates for nonattainment areas for the 1-hour 
standard, areas are required to attain ``as expeditiously as 
practicable'' but no later than the statutorily-prescribed (for section 
181(a)) or EPA-prescribed (for section 172(a)(2)) attainment dates. The 
implementation date of May 1, 2003 fits with both the more general 
requirement for areas to attain ``as expeditiously as practicable'' and 
the latest attainment dates that apply for purposes of the 1-hour 
standard and that EPA will establish for the 8-hour standard.
    The overarching requirement for attainment is that areas attain 
``as expeditiously as practicable.'' This requirement was established 
in the CAA in the 1970 Amendments and has been carried through in both 
the 1977 and 1990 Amendments. Thus, although Congress has provided 
outside attainment dates under the 1970, 1977, and 1990 Amendments, 
States have always been required to attain as expeditiously as 
practicable. Congress has furthered this concept of ensuring that 
emission reductions are achieved on an expeditious, yet practicable, 
schedule through its inclusion of other provisions in the CAA that rely 
on similar concepts. Most notably, under both subpart 1 and subpart 2 
of part D of title I of the CAA, areas are required to make reasonable 
further progress toward attainment and thus are not allowed to delay 
implementation of all measures until the attainment year.\63\ While the 
ROP requirements directly apply only to emission reductions that 
designated nonattainment areas need to achieve to address local 
violations of the standard, these provisions highlight congressional 
intent that--at a minimum--reasonably available or practicable measures 
should not be delayed if such measures are needed to attain the 
standard by the applicable attainment date. Thus, it is consistent for 
EPA to require upwind areas to adopt practicable control measures on a 
schedule that will help to ensure timely attainment of the standard in 
downwind areas.
---------------------------------------------------------------------------

    \63\ CAA sections 171(1) and 172(c)(2) (requiring that 
nonattainment area SIPs provide for reductions in emissions that may 
reasonably be required by the Administrator for the purpose of 
ensuring attainment of the applicable national ambient air quality 
standard by the applicable date; 182(b)(1) and (c)(2)(B) (requiring, 
respectively, 15 percent reductions between 1990 and 1996 and 
additional 3 percent average reductions per year until the 
attainment date, unless, among other things, the plan includes ``all 
measures that can be feasibly implemented in the area, in light of 
technological achievability'').
---------------------------------------------------------------------------

    In addition, the May 1, 2003 implementation date is consistent with 
the statutorily-prescribed ``outside'' 1-hour attainment dates for many 
of the areas that will benefit from the SIP call reductions.
    Currently, areas designated nonattainment for the 1-hour standard 
have attainment dates ranging from 1996 to 2010. For those with 
attainment dates in the years 1996-1999, EPA is analyzing whether such 
areas should receive an attainment date extension due to transported 
emissions or whether such areas should be reclassified, or ``bumped 
up,'' under section 181(b)(2), to the next higher classification and 
therefore be subject to additional control requirements and a later 
attainment

[[Page 57450]]

date.\64\ To the extent that an attainment date extension is 
appropriate, consistent with the general requirement of the CAA, it 
should be no later than the date by which the necessary reductions can 
practicably be achieved. Thus, it is appropriate for EPA to require 
upwind reductions by May 1, 2003--a date that EPA has determined can be 
practicably achieved--in order to allow these areas to attain as 
expeditiously as practicable. Additionally, there are areas with 
attainment dates of 2005 \65\ and 2007 \66\ that will benefit from the 
reductions upwind States will require in response to the SIP call. The 
May 1, 2003 compliance date is sensible in light of the requirement for 
these areas to make reasonable further progress toward attainment under 
section 182(c)(2)(B) and to attain as expeditiously as practicable but 
no later than 2005 or 2007.
---------------------------------------------------------------------------

    \64\ See Guidance on Extension of Attainment Dates for Downwind 
Transport Areas, Memorandum from Richard Wilson, dated July 17, 
1998.
    \65\ Severe-15 areas, such as Baltimore and Philadelphia, as 
well as any Serious areas that do not receive an attainment date 
extension and are bumped up due to a failure to attain, will need to 
attain no later than 2005.
    \66\ Severe-17 areas, such as New York City, Philadelphia, 
Chicago and Milwaukee, need to attain the standard no later than 
2007.
---------------------------------------------------------------------------

    The implementation date of May 1, 2003 is also consistent with the 
attainment date scheme for the 8-hour ozone NAAQS. The EPA is required 
to promulgate designations for areas under the 8-hour ozone NAAQS by 
July 2000. Pub. L. No. 105-178 section 6103 and CAA section 107(d)(1). 
In draft guidance EPA made available for comment in August 1998, the 
EPA indicated that most new areas that violate the 8-hour ozone NAAQS 
(but not the 1-hour ozone NAAQS) can achieve sufficient emissions 
reductions to produce one ozone season's clean air quality by the end 
of 2003 if EPA establishes May 1, 2003 as the compliance date for this 
rule.\67\ The EPA suggested that these areas would also be eligible for 
an ozone transitional classification, provided they submit a SIP by 
2000 (see the August 1998 proposed guidance). Therefore, in the 
proposed guidance, EPA has indicated that when the Agency reviews and 
approves ozone transitional area SIPs, the Agency anticipates 
establishing December 31, 2003 as the attainment date, for planning 
purposes, for almost all of the transitional areas. The EPA believes 
that establishing December 31, 2003 as the attainment date for these 
areas is consistent with the requirement of CAA section 172(a)(2)(A) 
that ``the attainment date for an area designated nonattainment with 
respect to a [NAAQS] shall be the date by which attainment can be 
achieved as expeditiously as practicable, but no later than 5 years 
from the date of designation.'' The EPA interprets this requirement to 
mandate that controls, either in the downwind nonattainment area or in 
upwind areas, should be implemented as expeditiously as practicable, 
when doing so would accelerate the date of attainment. For the reasons 
described elsewhere, the EPA believes it is practicable for States to 
implement the controls mandated under today's rulemaking by May 1, 
2003, and that doing so would ensure that areas subject to the 8-hour 
NAAQS will attain the standard as expeditiously as practicable. Doing 
so will be consistent with the requirement that downwind nonattainment 
areas make reasonable further progress toward attainment.
---------------------------------------------------------------------------

    \67\ ``Proposed Implementation Guidance for the Revised Ozone 
and Particulate Matter (PM) National Ambient Air Quality Standards 
(NAAQS) and the Regional Haze Program,'' John S. Seitz, Director, 
Office of Air Quality Planning and Standards, to Regional Office Air 
Division Directors, August 18, 1998. The guidance has been made 
available for 30-days public comment through a Federal Register 
Notice of Availability (63 FR 45060, August 24, 1998). The date of 
the notice is the official start date for the comment period.
---------------------------------------------------------------------------

B. Budget Achievement Date

    In the NPR, the EPA stated that although it would mandate the full 
implementation of the required SIP controls by an earlier date, it 
would require the affected States to demonstrate that they will achieve 
their NOX budgets as of the year 2007. The NPR explained 
that the 2007 date would allow EPA to make use of the substantial 
technical information collected by OTAG. The OTAG had selected the year 
2007, had collected inventory data geared towards this date, and had 
generated air quality modeling information geared towards this date. 
The NPR further stated that the EPA had doubts that there would be 
significant differences in amounts of emissions and impact on ambient 
air quality between an earlier date and 2007, in light of the fact that 
during this period, emissions would generally increase somewhat as a 
result of growth in activities that generate emissions, but would also 
decrease due to continued application of federally mandated controls.
    The EPA continues to believe that 2007 is an appropriate target 
date for the affected States to use in demonstrating whether their SIP 
will achieve the required emissions reductions, generally for the same 
reasons as expressed in the NPR. Based on the 2007 projections, States 
are expected to achieve their statewide emissions budgets (based on the 
required emissions reductions achieved by May 1, 2003) by September 30, 
2007 which is the end of the ozone season.
    Throughout this rulemaking process, the EPA has relied on technical 
data generated by OTAG geared towards the 2007 date, and it would be an 
ill-advised use of resources if EPA did not incorporate the emissions 
inventories and modeling results generated by the multi-stakeholder 
OTAG process, and instead developed comparable information for an 
earlier date. Such an effort would be time consuming and resource 
intensive. Furthermore, no State is disadvantaged by the requirement to 
demonstrate compliance with the budget later than the requirement to 
implement SIP controls because States may count both the growth in 
emissions and the reductions in emissions from Federal measures that 
would occur in the interim. Finally, the year 2007 is the latest 
attainment date under the 1-hour NAAQS for areas in States affected by 
today's rulemaking, i.e., the severe-17 areas of including Chicago, 
Milwaukee, and New York, so that this date is a sensible target date 
for affected States to use in projecting whether they will achieve the 
required emissions reductions.

VI. SIP Criteria and Emissions Reporting Requirements

A. SIP Criteria

    The NPR and SNPR discussed SIP revision approval criteria and the 
schedule for States' submission plans for meeting statewide emission 
budgets in response to this SIP call under section 110(a)(2)(D). The 
EPA received a number of comments related to the proposed SIP approval 
criteria. This section summarizes these comments on key issues and 
presents EPA responses.
1. Schedule for SIP Revision
    In the NPR, EPA proposed that each State must submit a 
demonstration that it will meet its assigned Statewide emission budget 
(including adopted rules needed to meet the emission budget) by 
September 30, 1999.68 The EPA received numerous comments 
concerning this proposed timeframe.
---------------------------------------------------------------------------

    \68\ In the NPR, EPA proposed the SIP submittal date to be 
within 12 months of the date of final promulgation of this 
rulemaking. Promulgation means signature so long as the rulemaking 
is made available to the public on the same day.
---------------------------------------------------------------------------

    Comments: The EPA received many comments on the practicality of 
allowing States 12 months to submit SIPs in response to this 
rulemaking. Some commenters articulated that some States anticipate 
administrative obstacles that could create problems in

[[Page 57451]]

submitting their SIP revisions by 1999. On the other hand, many 
commenters expressed concern about extending the SIP submittal deadline 
to 18 months based on the additional adverse impact that NOX 
emissions from upwind areas would have on downwind air quality if the 
schedule for reductions were extended. Arguing that the States would 
have ample time to formulate an approvable SIP, these commenters 
supported a 12-month SIP submission date.
    Response: After considering these comments, EPA is requiring that 
SIP revisions be submitted within 12 months after the date of signature 
of this final rule. This date is appropriate in light of the fact that 
States which are subject to today's rulemaking will need to achieve 
reductions in NOX emissions by May 1, 2003. Requiring States 
to submit SIP revisions within the 12-month timeframe will ensure that 
controls necessary to reduce these emissions will be in place on time.
    The Agency believes the health risks associated with ozone 
pollution require the NOX SIP call to proceed expeditiously. 
Delaying the SIP submission date by an additional 6 months would hinder 
downwind areas' efforts to improve air quality in a timely manner.
    Twelve months is adequate time to submit a NOX reduction 
SIP. States were involved in the OTAG for 2 years and, during that 
time, developed lists of feasible NOX control strategies and 
compiled information about control strategy costs. This groundwork will 
assist States in making decisions about their NOX reduction 
strategies and should expedite the SIP submittal process. Further, 
States developed NOX emission inventories for modeling 
purposes during the OTAG process. The States, therefore, have the 
information for the source categories on which to focus. As a result, 
many elements needed for putting together a NOX reduction 
strategy have already moved forward.
    Since OTAG concluded in June 1997, the States have had time for 
internal review of data, and refinement of their emission inventories. 
This SIP call rulemaking provides EPA's view of a reasonable cost-
effective strategy to reduce NOX in the 23 jurisdictions. 
The EPA's action provides a good starting point for State 
NOX reduction strategies; States can embrace the Agency's 
approach or use it as a basis for tailoring their own programs. If 
States elect to participate in EPA's model trading rule, the SIP 
process will be further simplified because States can adopt the entire 
package of recommended strategies.
    Therefore, under section 110(k)(5) for the 1-hour NAAQS and section 
110(a)(1) for the 8-hour NAAQS, a demonstration that each State will 
meet the assigned Statewide emission budget (including adopted rules 
needed to meet the emission budget) must be submitted to EPA in its SIP 
revision.
2. Approvability Criteria
    In the NPR, EPA described the elements listed below that States 
must include in their ozone transport SIP revisions (62 FR 60365).
    The EPA proposed that the approvability criteria for transport SIP 
submissions appear in 40 CFR 51.121. Most of the criteria are 
substantially identical to those that already apply to attainment SIPs, 
for example, a description of control measures that the State intends 
to use.
    The SNPR proposed additional SIP approvability criteria for control 
strategies that will help States meet their NOX budgets (63 
FR 25912-25914). The legal authority for these additional approvability 
criteria was articulated in the SNPR (63 FR 25913, footnote 5). The EPA 
received numerous comments related to these additional criteria.
    a. Source Categories Subject to Additional Approvability Criteria. 
In the SNPR, EPA proposed that, if a State should choose to meet this 
SIP call by regulating NOX sources (boilers, turbines and 
combined cycle units) serving electric generators with a nameplate 
capacity greater than 25 MWe and boilers with a maximum design heat 
input greater than 250 mmBtu/hr, the State would need to frame these 
control measures and monitoring requirements as either: (1) Mass 
emissions limits, (2) emissions rates assuming maximum utilization, or 
(3) an alternative approach, as described more fully in the next 
subsection. The EPA solicited comment on the reasonableness of 
extending these approvability criteria to additional NOX 
sources. The EPA explained that the ability to comply with a mass 
emissions limit using reasonably available technology and to accurately 
and consistently monitor mass emissions were key factors for coverage 
by the additional approval criteria.
    In the SNPR (63 FR 25923), EPA also outlined criteria for sources 
to participate in the NOX Budget Trading Program. The EPA 
explained that the ability to accurately and consistently monitor 
NOX mass emissions was a key factor for participation in the 
trading program. The EPA proposed that the trading program include the 
same sources listed above as well as other large steam-producing units 
(units above 250 mmBtu/hr) which would include combustion turbines or 
combined cycle systems, as well as boilers that do not serve electrical 
generators.
    The EPA now believes that the SIP approvability criteria should 
cover all NOX sources serving electric generators with a 
nameplate capacity greater than 25 Mwe and all boilers, combustion 
turbines and combined cycle units with a maximum design heat input 
greater than 250 mmBtu/hr. The Agency believes this group is 
appropriate because of the considerations set forth in the SNPR. For 
example, all of these sources can comply with a mass emissions limit 
using reasonably available technology and can accurately and 
consistently monitor mass emissions. In addition, EPA believes that 
mass emissions limits remain highly cost-effective for these sources, 
even when future growth is accommodated within the limits. Based on the 
analyses in the RIA, EPA projects that even if actual growth for this 
group of sources exceeds EPA's projected growth by over one-third, mass 
emission limits would remain highly cost-effective according to the 
criteria used for this rule. Therefore, in this final rule, EPA is 
requiring that the additional SIP approvability criteria outlined below 
apply to States that select regulatory requirements covering boilers, 
turbines and combined cycle units that are greater than 250 mmBtu/hr--
regardless of whether they are connected to an electrical generator of 
any size--or to boilers, turbines and combined cycle units that serve 
electrical generators greater than 25 Mwe, regardless of the heat input 
capacity of the unit.
    b. Pollution Abatement Requirements. The EPA proposed requiring 
States that choose to meet their budget through control requirements 
for such large NOX sources to express the requirements in 
one of three ways: (1) In terms of mass emissions, which would limit 
total emissions from a source or group of sources; (2) in terms of 
emissions rates that when multiplied by the affected source's maximum 
operating capacity would meet the tonnage component of the emissions 
budget for this source or for these sources; or (3) an alternative 
approach for expressing regulatory requirements, provided the State 
demonstrates to EPA that its alternative provides assurance equivalent 
to or greater than option (1) or (2) that seasonal emissions budgets 
will be attained and maintained.
    Comments: Seven commenters generally support the approach of

[[Page 57452]]

expressing regulatory requirements as mass emissions limitations. One 
of these commenters does not object to a mass limit provided that the 
limit covers a time period no shorter than the ozone season, and that 
sources should be allowed to maintain flexibility within the ozone 
season. Several commenters generally support a rate-based limit, one of 
which noted that EPA's own rule-effectiveness studies show that rate-
based limits can be very effective. Another commenter opposes the use 
of mass emission limits and urges EPA not to require monitoring 
procedures and data generation that are inconsistent with current 
requirements under the Acid Rain Program (namely the use of an 
emissions rate limit). Other commenters believe that States, not EPA, 
should decide the form of the limit. Finally, one commenter recommends 
both a cap on mass emissions and an emissions rate limitation.
    Response: As explained in the SNPR (63 FR 25912), EPA believes that 
regulatory requirements in the form of a maximum level of mass 
emissions for a source or group of sources have the greatest likelihood 
of achieving and maintaining the Statewide NOX emissions 
budget. As with the entire SIP call, the new approvability criteria are 
designed to apply to total emissions throughout the ozone season and 
are not intended to apply to shorter time periods within the ozone 
season. This, however, does not limit a State's ability to require 
emissions limitations for a shorter time period if deemed necessary in 
a specific ozone attainment plan.
    Although several commenters supported using rate-based limits, they 
did not provide evidence to refute EPA's belief that the proposed 
criteria would provide superior environmental results over rate-based 
limits alone. The EPA maintains that the proposed criteria provide the 
greatest assurance to downwind States that the air emissions from 
upwind States will be effectively managed over time. Regarding EPA's 
rule effectiveness studies, they do confirm that rate-based limits can 
be effective in achieving a specific emissions rate. However, the 
studies do not address the emissions variations that may take place at 
the regulated sources due to changes in utilization under rate-based 
limits, including the potential for significant increases, particularly 
in light of utility restructuring. Under the proposed criteria, mass 
emissions from the regulated sources would stay within a fixed tonnage 
amount despite shifts in utilization of the sources. Finally, EPA does 
not believe that the rate-based NOX emissions limits 
prescribed under title IV of the CAA are relevant to this rulemaking. 
Since the time of the 1990 CAA amendments, EPA, States, local 
governments, and the regulated community have all gained considerable 
experience with regulatory requirements expressed in terms of mass 
emissions limitations which demonstrates their feasibility and high 
degree of effectiveness. For these reasons and the reasons described in 
the SNPR, EPA is including these additional SIP approvability criteria 
in today's action.
    c. Monitoring Requirements. The Agency proposed requiring these 
large combustion NOX sources to use continuous emissions 
monitoring systems (CEMS), and requested comment on requiring the use 
of the NOX mass monitoring provisions in 40 CFR part 75 to 
demonstrate compliance with applicable emissions control requirements.
    Comments: Some commenters generally support the use of CEMS for 
large combustion sources. One commenter noted that while the preamble 
and the proposed revisions to part 51 would require CEMS on all 
sources, the requirements set forth in subpart H of part 75 allow for 
non-CEMS monitoring options for units that are infrequently operated or 
that have low mass emissions of NOX.
    Response: The EPA believes that programs like the Acid Rain Program 
and RECLAIM have shown that CEMS can be effectively used on boilers, 
turbines and combined cycle units to demonstrate compliance with a mass 
emissions limitation. The Agency also believes that, while CEMS provide 
more consistent and accurate data, allowing non-CEMS monitoring options 
for low-emitting or infrequently operated units greatly increases the 
cost effectiveness of these requirements without significantly 
jeopardizing the quality of the data used to ensure compliance with the 
requirements of the SIP call. Therefore, EPA agrees with the commenter 
that the part 75 provisions allowing non-CEMS monitoring options for 
low-emitting or infrequently operated units are reasonable. The EPA is 
requiring the use of the NOX mass monitoring provisions in 
40 CFR part 75 in the final SIP approval criteria.
    d. Approvability of Trading Program. In the SNPR, EPA expressed its 
intent to approve the portion of any State's SIP submission that adopts 
the model rule, provided: (1) The State has the legal authority to 
adopt the model rule and implement its responsibilities under the model 
rule, and (2) the SIP submission accurately reflects the NOX 
emissions reductions to be expected from the State's adoption of the 
model rule (63 FR 25913). The EPA also stated that a State could 
develop State regulations in accordance with the model rule. In Section 
VII.C.3 of this preamble, the Agency clarifies the extent to which a 
State's regulations may deviate from the model rule and still receive 
streamlined approval. Regulations providing for streamlined approval 
appear in paragraph (p) of 40 CFR 51.121.
3. Sanctions
    In the preamble to the proposed rule, EPA explained the mandatory 
sanctions process that is established in section 179(a) and (b) of the 
CAA (62 FR 60368). This process is triggered upon a finding by EPA that 
a State failed to submit a SIP in response to a SIP call. One 
sanction--either increased offsets for new or modified major stationary 
sources or restrictions on highway funding--is imposed 18 months after 
the finding is made and the second sanction 6 months later. The EPA 
requested comment on the order in which these two sanctions should be 
imposed in response to the SIP call. The EPA further requested comment 
on whether EPA should use its discretion under section 110(m) to expand 
the geographic scope of the highway funding sanction.
    Comment: One commenter specifically commented on the order in which 
the two sanctions should be imposed. The commenter recommended that the 
offset sanctions apply first--18 months after the finding--and the 
restrictions on highway funding apply second--6 months after the offset 
sanction.
    Response: This is the approach that EPA took in its final rule 
addressing the sequence of mandatory sanctions for State failures to 
respond to submittals required under part D of title I of the CAA. For 
the reasons stated in the preamble to that final rule (59 FR 39832), 
EPA is providing in the final SIP call rule that the offset sanction 
will apply 18 months after EPA makes a finding and the restrictions on 
highway funding will apply 6 months after the offset sanction applies.
    Comments: Several commenters generally commented that EPA should be 
fair and equitable in making findings and imposing sanctions. Other 
commenters suggested that to be fair and equitable--and because the 
sanctions are an important backstop to ensuring emission reduction are 
achieved--EPA should apply the same or similar sanctions to upwind 
attainment areas as to nonattainment areas that do not comply with the 
SIP call. Recognizing that the highway

[[Page 57453]]

sanction can apply to attainment areas only under section 110(m), one 
commenter encouraged EPA to develop a mandatory clock for the 
imposition of discretionary sanctions. Finally, one commenter stated 
that the nature and timing of sanctions should reflect a State's 
particular circumstances; however, this commenter also emphasized the 
need for parties to know the impact of sanctions ahead of time so that 
they can effectively react.
    Response: The EPA agrees that sanctions are an important backstop 
and plans to make timely findings where States fail to submit or submit 
an incomplete or disapprovable SIP in response to the SIP call. The EPA 
agrees that areas should be treated fairly and plans to ensure that 
areas with similar circumstances are not treated differently in making 
findings of failure to submit and incompleteness. However, at this 
time, EPA is not prepared to determine whether and when it is 
appropriate to use the discretion provided under section 110(m) in 
imposing sanctions. The EPA believes it is not appropriate to make a 
general determination regarding the application of sanctions under 
section 110(m); rather if circumstances warrant the use of sanctions 
under section 110(m), EPA may take future rulemaking action to use that 
authority. Before EPA uses the section 110(m) authority, EPA must go 
through notice-and-comment rulemaking, which should provide States 
adequate certainty about EPA's intentions on the use of discretionary 
sanctions and time to respond to any action that EPA may take.
    Comment: One commenter suggested that the timeframes for the 
imposition of sanctions are too short and will undermine States' 
efforts to comply with the SIP call. In addition, the commenter states 
that the imposition of sanctions serves no useful purpose in light of 
EPA's intent to promulgate a FIP.
    Response: The EPA did not propose imposing sanctions more 
expeditiously than the timeframes mandated by the CAA. If EPA makes a 
finding of failure to submit or incompleteness shortly after the SIP is 
due, the State will have 18 months in which to make a submission that 
EPA determines is complete before the first sanction would be imposed. 
Thus, the statute provides sufficient additional time for the State to 
correct the problem before any sanction would apply. Under the statute, 
sanctions apply independently of EPA's obligation to promulgate a FIP. 
Congress recognized that the most efficient and effective programs are 
those operated by the State; thus, the CAA provides for the continued 
imposition of sanctions as a means to encourage States to adopt a 
program to replace the FIP.
    Comment: One commenter opposes restrictions on highway funding 
imposed by any highway sanction in nonattainment areas and especially 
Statewide.
    Response: Under section 179(a) and (b), the highway funding 
sanction is one of two sanctions that must be imposed due to a 
continuing failure of a State to adopt a SIP program, including a SIP 
in response to a SIP call. Under section 179(b), the highway funding 
sanction can only apply in a nonattainment area. However, under the 
discretionary sanctions provision in section 110(m), EPA may impose the 
highway funding Statewide. (See 59 FR 1476, 1479-80 for a more detailed 
discussion.) The EPA would undertake notice-and-comment rulemaking 
before imposing sanctions beyond the nonattainment area pursuant to 
section 110(m).
    Comments: Finally, several commenters recommended that EPA not 
sanction serious areas for failing to demonstrate attainment by 1999 
where those areas are affected by transported emissions that will not 
be controlled until after the 1999 attainment date.
    Response: The EPA is not addressing in this rulemaking the process 
for imposing sanctions for areas that fail to submit or submit 
incomplete or unapprovable attainment demonstrations. The EPA recently 
issued a policy memorandum explaining how it anticipates addressing 
transport for serious areas through rulemaking actions on submitted 
attainment demonstrations. See memorandum from Richard D. Wilson, EPA 
Acting Assistant Administrator, to EPA Regional Administrators, dated 
July 16, 1998, ``Extension of Attainment Dates for Downwind Transport 
Areas.''
    In the preamble to the proposed rule, EPA indicated that if an area 
fails to implement an approved SIP, the Agency can make a finding that 
triggers the sanctions clock but does not trigger an obligation to 
promulgate a FIP. Compare sections 179(a)(1) and 110(c)(1). One 
commenter noted that EPA should take a forceful role in assuring 
implementation. Implementation of control measures to achieve the 
reductions required under the NOX SIP call is crucial in 
moving all areas to attainment of the ozone standards. The EPA intends 
to make findings of failure to implement where the circumstances 
warrant such a finding.
4. FIPs
    Comment: The EPA received several comments supporting the approach 
outlined in the NPR in which EPA would propose a FIP at the same time 
as taking final action on the SIP call. The comments noted that the 
FIPs may be necessary to enforce the SIP call budgets and to assure 
fair treatment of complying States and industry as compared to States 
that are not responsive to the SIP call. In addition, many comments 
were submitted urging EPA to delay proposal of FIPs until (1) after the 
States have had time to respond to the SIP call, (2) the need for the 
FIP is established, or (3) up to 2 years after the final SIP call.
    Response: Also signed today is a separate notice titled ``Federal 
Implementation Plans to Reduce the Regional Transport of Ozone,'' EPA 
is proposing FIPs for each of the jurisdictions affected by the final 
SIP call rulemaking. While EPA will have a non-discretionary duty to 
promulgate a FIP within 2 years of a finding that a State has failed to 
submit a complete SIP, EPA agrees with certain commenters that the 
timing of the FIP proposal should allow for promulgation in time to 
require NOX emissions reductions by sources at about the 
same time in States that comply with the SIP call and States that do 
not. Under a delayed FIP proposal approach, sources in the non-
complying States might experience an unfair competitive advantage over 
sources in States which elected to reduce their NOX 
emissions and reduce interstate transport of ozone and ozone precursors 
in an earlier timeframe, consistent with the SIP call rulemaking. More 
importantly, delaying the FIP proposal would potentially delay 
reductions of ozone pollution and NOX emissions in any non-
complying State which would unnecessarily jeopardize attainment and 
public health and welfare. Therefore, proposing a FIP today will ensure 
that EPA can promulgate a FIP very shortly after the time the SIPs are 
due, in the event of any State's failure to comply with today's final 
rule.

B. Emissions Reporting Requirements for States

    As stated in the November 7, 1997 NPR and the May 11, 1998 SNPR, 
the EPA believes it is essential that compliance with the regional 
control strategy be verified. Tracking emissions is the principal 
mechanism to ensure compliance with the SIP call and to assure the 
downwind affected States

[[Page 57454]]

and EPA that the ozone transport problem is being 
mitigated.69
---------------------------------------------------------------------------

    \69\ Legal authority for the reporting requirements was 
articulated in the supplemental notice of proposed rulemaking (63 FR 
25915-6).
---------------------------------------------------------------------------

1. Use of Inventory Data
    If tracking and periodic reports indicate that a State is not 
implementing all of its NOX control measures beginning on 
May 1, 2003 or is off track to meet its required reductions by 
September 30, 2007, EPA will work with the State to determine the 
reasons for noncompliance and what course of remedial action is needed. 
The EPA will expect the State to submit a plan showing what steps it 
will take to correct the problems. Noncompliance with the 
NOX transport SIP call may lead EPA to make a finding of 
failure to implement the SIP and potentially to implement sanctions, if 
the State does not take corrective action within a specified time 
period.
    The EPA will use 2007 data to assess how each State's SIP actually 
performed in meeting the statewide NOX emissions budget.
2. Response to Comments
    The EPA proposed reporting requirements in the May 11, 1998 SNPR. 
That proposal elicited several comments during the public comment 
period. Some of these comments resulted in changes to the final 
reporting requirements.
    Comment: One commenter asked that the EPA review the need for 
triennial collection of annual (i.e for the full year) emissions data 
for uncontrolled sources, as compared to collection of only ozone 
season data for uncontrolled sources.
    Response: The EPA has reviewed the need for reporting of full year 
emissions (as opposed to only ozone season emissions), and has revised 
the final rule to remove a requirement that full year emissions be 
reported. In the final rule, only ozone season emissions must be 
reported in the annual, triennial and 2007 reports. This NOX 
SIP call is aimed at controlling transport of emissions during the 
ozone season and reporting of full year emission for the purposes of 
this SIP call is not necessary.
    Comment: One commenter said that EPA should evaluate the reporting 
burden to entities other than the 22 States and the District of 
Columbia. These entities are likely to include owners/operators of 
facilities that will be required to report emissions data to States as 
part of this information collection. Another commenter said EPA should 
address the additional resource burden on States and facilities 
required to report.
    Response: Since the emissions reporting rule does not place 
requirements directly on any sources but only on the 23 jurisdictions 
which receive the SIP call, the EPA is under no legal obligation to 
evaluate the indirect burdens on sources that may result from the 
promulgation of this rule. However, based on EPA's assumed control 
strategy, EPA has performed an analysis of costs which could be 
incurred by facilities if States require facilities analyzed in EPA's 
assumed control strategy to report information to aid States in 
complying with the rule. This cost information includes both capital 
costs for monitoring equipment, such as continuous emission monitors, 
and labor costs for testing. These costs are included in the RIA for 
this rule which is located in the docket for the rulemaking (docket no. 
A-96-56).
    Comment: One commenter is concerned that the definition of point 
and area sources does not coincide with the definition of smaller point 
sources included in the inventory, nor with the definition of major 
sources in ozone nonattainment areas where the threshold is either 25 
or 50 tons per year. Another commenter stated that the definition of 
``point source'' should reach at least down to the 50 ton per year 
level, if not lower. This commenter also said that, for consistency, 
EPA should have a single definition of ``point source'' for the purpose 
of this rule.
    Response: All sources with NOX emissions equal to or 
greater than 100 tons per year will remain point sources. However, the 
EPA has revised its definition of point source for this final rule's 
reporting requirements to allow States the option of specifying a 
smaller threshold than 100 tons/year of NOX for defining 
point source. When a State chooses this option, non-mobile sources 
smaller than the State-defined threshold would be area sources in that 
State. This allows States to tailor their definition of point source to 
maintain consistency with their own current requirements.
    In the proposal, the EPA specifically solicited comments on whether 
the State reporting time for source emissions should be shortened to no 
later than 6 or 9 months after the end of the calendar year for which 
the data are collected. This would allow corrective actions, if needed, 
to be taken prior to the next ozone season. The EPA also solicited 
comments on whether different reporting schedules should be established 
for the different source categories, so that the data which can be 
obtained more readily would be submitted sooner. The EPA has received 
several comments on these topics, suggesting a variety of reporting 
times.
    Comment: A State recommended that since the performance of electric 
generating facilities is known promptly, EPA should shorten the 
reporting time to no later than 4 to 6 months after the end of the 
ozone season for which the data are collected. The comment did not 
specify whether this reporting period , which is shorter than the 
proposed 12 months, would apply only to electric generating facilities 
or should apply to all NOX emitting sources. Another State 
said the point source emissions reporting period can be shortened to 9 
months. Other commenters favored a 12 month or more reporting period. 
Several commenters did not believe that 12 months after the end of the 
calendar year is a reasonable time to submit reports and suggested 
periods ranging from 18 to 24 months. Some commenters thought the 
reporting time for area and mobile sources must be longer than for 
point sources; one commenter thought the reporting time for all source 
types should be uniform.
    Response: Many of the emissions from large electric generating 
facilities would be reported directly to EPA more rapidly than 12 
months, if States elect to adopt the model trading program; however, 
the EPA continues to believe that 12 months from the end of the 
calendar year for which the data is collected is a reasonable time to 
require a State to report all emissions from all types of sources. This 
12 month period is supported by the comments which say that 12 months, 
or even less in some situations, is a sufficient reporting time. The 
EPA believes that States can report emissions from area and mobile 
sources, as well as stationary sources, within the 12 month period. The 
uniform 12 month reporting period for all source types was chosen to 
simplify reporting requirements. However, a State has the option of 
collecting emissions from particular sectors more rapidly if it wishes. 
Therefore in the final rule, the EPA is requiring that States submit 
the required annual and triennial emissions inventory reports no later 
than 12 months after the end of the calendar year for which the data 
are collected. Because downwind nonattainment areas will be relying on 
the upwind NOX reductions to assist them in reaching 
attainment by the required dates, EPA believes it is important that 
data be submitted as soon as practicable to verify that the necessary 
emissions reductions are being achieved. Early reports will allow 
States to more quickly respond to implementation problems detected by 
the reports. States should formally notify the appropriate EPA

[[Page 57455]]

Regional Office when making the submittals.
3. Final Rule
    After taking into account the comments submitted in response to the 
May 11, 1998 proposal, EPA today is promulgating emission inventory 
reporting requirements for States subject to the NOX SIP 
call. The regulatory text appears in 40 CFR 51.122, and the main 
emission reporting requirements are summarized in Table VI-1 below.

           Table VI-1.--Summary of NOX Reporting Requirements
------------------------------------------------------------------------
                                                      then, your State
    If you own or operate              and           must report to EPA
                                                        the source's
------------------------------------------------------------------------
A point source..............  You are not subject   Ozone season2
                               to regulations        emissions.
                               relied on to
                               achieve the NOX
                               reductions required
                               in this SIP call 1.
                                                    1. triennially 3,5.
                                                    2. for 20075.
A point source..............  You are subject to    Ozone season
                               regulations relied    emissions.
                               on to achieve the
                               NOX reductions
                               required in this
                               SIP call 1.
                                                    1. annually 4.
                                                    2. triennially 5.
                                                    3. for 2007 5.
An area source..............  You are not subject   Ozone season
                               to regulations        emissions.
                               relied on to
                               achieve the NOX
                               reductions required
                               in this SIP call 1.
                                                    1. triennially.
                                                    2. for 2007.
An area source..............  You are subject to    Ozone season
                               regulations relied    emissions.
                               on to achieve the
                               NOX reductions
                               required in this
                               SIP call 1.
                                                    1. annually 6.
                                                    2. triennially.
                                                    3. for 2007.
A mobile source.............  You are not subject   Ozone season
                               to regulations        emissions.
                               relied on to
                               achieve the NOX
                               reductions required
                               in this SIP call 1.
                                                    1. triennially.
                                                    2. for 2007.
A mobile source.............  You are subject to    Ozone season
                               regulations relied    emissions.
                               on to achieve the
                               NOX reductions
                               required in this
                               SIP call 1.
                                                    1. annually 6.
                                                    2. triennially.
                                                    3. for 2007.
------------------------------------------------------------------------
 1The EPA considers the State to rely on regulations to achieve the NOX
  reductions required if those regulations require reductions beyond
  those reflected in the base case 2007 inventory.
2 Ozone season is May 1 through September 30.
3 Triennial reporting (which is every 3 years) starts with emissions
  occurring in 2002.
4 Annual reporting starts with emissions occurring in 2003.
5 Triennial and 2007 reports for point sources contain additional data
  elements not required in the annual reports.
6 The data elements in the annual report for area and mobile sources
  satisfy the reporting requirements for these source categories for the
  triennial and 2007 reports. However, the triennial reports start with
  emissions occurring in the year 2002 and the annual reports start with
  emissions occurring in the year 2003.

4. Data Elements to be Reported
    In addition to reporting the NOX emissions values shown 
in Table VI-1, the State must report other critical data necessary to 
generate and validate these values. This includes data used to identify 
source categories such as site name, location and (source 
classification code) SCC codes. It also includes data used to generate 
the NOX emissions values such as fuel heat content and 
activity level. The specific data elements required for each source 
category are further defined in 40 CFR 51.122.
5. 2007 Report
    The EPA is requiring that States submit to EPA for the year 2007 a 
special onetime statewide NOX emissions inventory from all 
NOX sources (point, area, and mobile) within the State. The 
data reporting requirements are identical to the reporting requirements 
for the triennial inventories, and this reporting requirement is being 
imposed to allow evaluation of whether the budget is met in 2007. This 
one-time special inventory is necessary because the ordinary 3-year 
reporting cycle does not fall in the year 2007.
    States which must submit the 2007 inventory may project incremental 
changes in emissions from 2007 to 2008 to allow the 2008 inventory 
requirement to be more easily met and to reduce the burden on States 
which must submit full NOX inventories for consecutive 
years, i.e., 2007 and 2008.
    The EPA received comments saying that EPA should not require the 
special report in 2007 due to increased resources required but rather 
should adjust the schedule of the triennial reports so that a triennial 
report year will fall on 2007. Alternatively, the EPA could eliminate 
the 2008 triennial report. The EPA has considered these alternatives, 
but believes that the schedule which was proposed is necessary to 
maintain consistency with

[[Page 57456]]

other EPA reporting requirements and is not unnecessarily burdensome.
6. Ozone Season Reporting
    The EPA is requiring that the States provide ozone-season (i.e., 
May 1 through September 30) inventories for the sources for which the 
State reports annual, triennial and 2007 emissions. The ozone season 
emissions may be calculated from annual data by prorating emissions 
from the ozone season by utilization factors that must be reported and 
that are further defined in 40 CFR 51.122. For the triennial and 2007 
reports, ozone season emissions from all NOX source 
categories within the State, controlled or uncontrolled, must be 
reported. The EPA is requiring that each State provide its ozone season 
calculation method to EPA for approval.
7. Data Reporting Procedures
    When submitting a formal NOX budget emissions report and 
associated data, the State should formally notify the appropriate EPA 
Regional Office of its activities. States are required to report 
emissions data in an electronic format to one of the locations given 
below. Several options are available for data reporting. The State may 
choose to continue reporting to the EPA Aerometric Information 
Retrieval System (AIRS) using the AIRS facility subsystem (AFS) format 
for point sources. (This option will continue for point sources for 
some period of time after AIRS is reengineered (before 2002), at which 
time this choice may be discontinued or modified.) A second option is 
for the State to convert its emissions data into the Emission Inventory 
Improvement Program/Electronic Data Interchange (EIIP/EDI) format. This 
file can then be made available to any requestor, either using E-mail, 
floppy disk, or value added network, or can be placed on a file 
transfer protocol (FTP) site. As a third option, the State may submit 
its emissions data in a proprietary format based on the EIIP data 
model. For the last two options, the terms ``submitting'' and 
``reporting'' data are defined as either providing the data in the 
EIIP/EDI format or the EIIP based data model proprietary format to EPA, 
Office of Air Quality Planning and Standards, Emission Factors and 
Inventory Group, directly or notifying that group that the data are 
available in the specified format and at a specific electronic location 
(e.g., FTP site). A fourth option for annual reporting (not for third 
year reports) is to have sources submit the data directly to EPA. This 
option will be available to any source in a State that is both 
participating in an approved trading program and that has agreed to 
submit data in this format. The EPA will make both the raw data 
submitted in this format and summary data available to any State that 
chooses this option.
    For the latest information on data reporting procedures, call the 
EPA Info Chief help desk at (919) 541-5285 or e-mail to 
[email protected].
8. Confidential Data
    Emissions data being requested in today's action are not considered 
confidential by the EPA (See 42 U.S.C. 7414). However, some States may 
restrict the release of certain types of data, such as process 
throughput data. Where Federal and State requirements are inconsistent, 
the EPA Regional Office should be consulted for final reconciliation.

C. Timeline

    The reporting requirements fit into the general time line 
summarized below:
    September 30, 1999--Deadline for SIP submissions in response to 
this SIP call.

2002--The first triennial emissions inventory report must be submitted 
for ozone season emissions for this year. States must collect emissions 
inventory information for all NOX sources in the State. This 
report must be submitted by December 31, 2003 (i.e., 12 months after 
the end of the calendar year for which the data are collected.)
May 1, 2003--The SIP measures required to achieve the NOX 
reductions must be implemented by this date.
2003--The first annual emissions inventory report must be submitted for 
certain ozone season NOX emissions for this year. 
Specifically, States must collect emissions information regarding all 
sources for which the State is relying on measures to meet its 
NOX budget (``SIP call sources''). This report is due 
December 31, 2004.
2004--The second annual emissions inventory report must be submitted 
for ozone season emissions from SIP call sources for this year. This 
report is due December 31, 2005.
2005--The second triennial report must be submitted for ozone season 
emissions from all NOX sources for this year. The report is 
due December 31, 2006.
2006--The third annual report must be submitted for ozone season 
emissions from SIP call sources in the State for this year. This report 
is due December 31, 2007.
2007--The special year 2007 emission inventory report for ozone season 
emissions from all NOX sources in the State must be 
submitted for this year. This report is due December 31, 2008. The EPA 
will assess whether States have met their budgets in the year 2007.
2008--The third triennial emissions inventory report must be submitted 
for ozone season emissions for this year. This report is due December 
31, 2009.

    Annual and triennial reports must continue to be submitted in 
future years beyond 2008 in order for the EPA to track compliance with 
the budget or any revisions to the budget that may occur after 2007.

VII. NOX Budget Trading Program

A. General Background

    In the November 7, 1997 proposed rulemaking, EPA offered to develop 
and administer a multi-state NOX trading program to assist 
States in the achievement of their budgets. Today's notice sets forth a 
model program on which States may choose to base their SIP submittal. 
The trading program employs a cap on total emissions in order to ensure 
that emissions reductions under the transport rulemaking are achieved 
and maintained, while providing the cost effectiveness of a market-
based system. States can voluntarily choose to participate in the 
NOX Budget Trading Program by adopting the final model rule, 
which is a fully approvable control strategy for achieving over 90 
percent of the emissions reductions required under the transport 
rulemaking.

B. NOX Budget Trading Program Rulemaking Overview

    Prior to publication of the proposed NOX Budget Trading 
Program, EPA held two public workshops to solicit comments and 
suggestions from States and other stakeholders on a NOX cap-
and-trade program. Over 150 people participated in each of the 
workshops. To facilitate meaningful comments from these participants, 
EPA developed papers on critical issues that were made available for 
review prior to each workshop. These papers discussed major issues 
relevant to developing a NOX Budget Trading Rule, delineated 
options and, in some cases, offered recommendations. The issues 
associated with each working paper were presented at the workshops, 
followed by open discussion periods allowing workshop participants to 
comment and discuss each issue. Input from workshop participants was 
extremely helpful in drafting the proposed NOX Budget 
Trading Program. In addition to

[[Page 57457]]

input gained from the workshop process, the NOX Budget 
Trading Program builds directly upon the Ozone Transport Commission's 
NOX Budget Program and recommendations from the OTAG's 
Trading and Incentives Workgroup. On May 11, 1998, EPA published the 
proposed NOX Budget Trading Program as a part of the 
supplemental notice for the proposed ozone transport rulemaking. The 
final NOX Budget Trading Rule published in today's notice 
reflects changes that have been made in response to comments received 
on the May 11, 1998 proposal.

C. General Design of NOX Budget Trading Program

1. Appropriateness of Trading Program
    The EPA proposed that a voluntary market-based program be 
established as one possible means for a State to meet its 
NOX emissions reduction obligations under the NOX 
SIP call. The vast majority of commenters, including States, industry, 
and environmental groups, supported a market approach over traditional 
``command and control'' mechanisms to fulfill reduction requirements. 
However, many commenters argued that the proposed State budgets, based 
on the cost-effectiveness of an emission limit of 0.15 lb/mmBtu for 
large combustion sources, are too stringent to provide sufficient 
surplus allowances to support a market. These commenters argued that 
cost and technological constraints would prevent regulated sources from 
over-controlling, thus reducing the pool of allowances and the cost 
savings EPA predicts would accompany trading. However, several other 
commenters stated that the trading program was the most cost-effective 
means to reduce emissions and would in fact generate sufficient 
allowances for trading. These commenters noted that all but the highest 
emitting coal-fired units can achieve this rate, and that many sources 
are able to achieve emission limits significantly below 0.15 lb/mmBtu. 
They also argued that, at least in the early years of the trading 
program, the growth factors used to determine the budgets will lead to 
a less stringent emission reduction requirement than 0.15 lb/mmBtu.
    The EPA notes that nothing requires a State to impose a 0.15 lb/
mmBtu limit on its large combustion sources. The States will select in 
their SIPs which sources to regulate and the type of regulation to 
impose in order to achieve their NOX budgets. The EPA 
believes that trading for large combustion sources under a budget based 
on 0.15 lb/mmBtu is a feasible, highly cost-effective means of meeting 
a State's budget. The Agency believes that 0.15 lb/mmBtu can easily be 
achieved by gas and oil-fired boilers. In fact, more than 50 percent of 
gas and oil-fired boilers already operate at NOX levels 
below 0.15 lb/mmBtu and should therefore easily be able to generate 
excess allowances if trading is allowed. The EPA recognizes that for 
coal-fired boilers to operate at or below a 0.15 lb/mmBtu emission 
limit, selective catalytic reduction (SCR) will generally be necessary. 
Under a trading scenario, however, if one coal-fired boiler is able to 
emit below 0.15 lb/mmBtu by installing SCR, it can provide excess 
allowance to another coal-fired boiler and obviate the need for that 
boiler to install SCR. (For further technical justification for the 
feasibility of 0.15 lb/mmBtu, see Section III.B.2 of this preamble.) In 
summary, EPA concludes that, should a State elect to control large 
combustion sources with a budget based on an emission rate of 0.15 lb/
mmBtu, ample allowances would exist to sustain a market under the 
NOX Budget Trading Program.
    Several of the commenters who did not support the trading program 
proposed by EPA were generally wary of the use of market approaches for 
environmental regulation, especially in the context of ozone attainment 
strategies, citing concerns that emissions in existing nonattainment 
areas may increase under such a program. The EPA, however, believes 
that a trading program is an appropriate mechanism to achieve the 
NOX reductions required under the SIP call. The EPA proposed 
the trading program in the SNPR based on recommendations from OTAG, 
experience from the Ozone Transport Commission, and EPA's public 
workshops held in November and December 1997. This trading program was 
designed to mitigate transport of ozone and its precursors to 
facilitate attainment and maintenance of the ozone NAAQS. Analyses in 
conjunction with the SIP call show that implementation of a trading 
program with a uniform control level results in no significant changes 
in the location of emissions reductions than would result from a non-
trading scenario (``Supplemental Ozone Transport Rulemaking Regulatory 
Analysis'', April 1998, page 2-19). The NOX reductions 
required by the SIP call will significantly lower background levels of 
ozone and can be coupled with local measures to achieve further 
NOX reductions, as well as VOC reductions, where necessary 
to reach attainment. States concerned with contribution by local 
sources in the trading program are free to limit emissions from 
particular sources by imposing source-specific emission limits where 
deemed necessary.
2. Alternative Market Mechanisms
    The SNPR proposed to establish a model cap-and-trade program for 
certain large combustion sources. This proposed program employs a cap 
on total emissions to ensure achievement and maintenance of the 
emissions reductions required under the NOX SIP call while 
providing the flexibility and cost effectiveness of a market-based 
system. Several commenters supported EPA's recommendation for a cap-
and-trade program. Several others complained that EPA's focus on a 
capped trading program was inappropriate, citing OTAG's recognition 
that NOX market systems could also be implemented without an 
emissions cap. As a result, these commenters felt that EPA could not 
make a cap a prerequisite to approval of a State trading program. They 
suggested that EPA recognize that a rate-based program can be part of a 
viable SIP, perhaps by outlining parameters of an acceptable 
alternative program or working with OTAG States to develop a rate-based 
program that would better accommodate future growth. Another issue 
raised by a few commenters was that the trading program would either 
conflict with or would ignore existing local or State-based trading 
programs.
    The EPA first reiterates that the model program is voluntary (63 FR 
25918). In providing a cap-and-trade program as a streamlined means by 
which to comply with the NOX SIP call, EPA does not preclude 
implementation of other solutions. The purpose of the trading program 
is to provide a compliance mechanism that capitalizes on a proven means 
of cost effectively meeting a specific emissions budget that the Agency 
will assist States in administering.
    As OTAG concluded, the procedures for a cap-and-trade program have 
already been developed and used successfully, whereas procedures for 
other types of multi-state trading programs have not been developed and 
implemented to the same degree. Therefore, EPA does not have the same 
level of experience or established protocols to follow in the design 
and administration of other types of trading programs. The OTAG did 
encourage development of provisions to implement other types of trading 
programs, and EPA recognizes that these alternative trading programs 
may be appropriate in some circumstances.

[[Page 57458]]

However, EPA recommends a cap-and-trade program for purposes of the 
NOX SIP call because, by limiting total NOX 
emissions to the level determined to address the interstate transport 
problem, a cap better ensures achievement and maintenance of the 
environmental goal articulated in the NOX SIP call. In 
contrast, under a non-cap trading program, the addition of new sources 
to the regulated sector or increased utilization of existing sources 
could increase total emissions above the level determined to address 
transport, even though a NOX rate limit is met.
    States, however, have the flexibility to respond as they see fit to 
meet their emissions budgets established under the NOX SIP 
call. States are free to pursue other regulatory mechanisms or include 
other types of trading programs in their SIPs, whether newly created or 
already existing, on the condition that they meet EPA's SIP approval 
criteria as delineated for the NOX SIP call. These criteria 
mandate that regulatory requirements for boilers, turbines and combined 
cycle units that are greater than 250 mmBtu or that serve electrical 
generators that are greater than 25 MWe be expressed in one of three 
ways: (1) In terms of mass emissions; (2) in terms of emissions rates 
that when multiplied by the affected sources' maximum operating 
capacity would meet the tonnage component of the emissions budget for 
these sources; or (3) an alternative approach for expressing regulatory 
requirements, provided the State demonstrates, to EPA's satisfaction, 
that its alternative provides equivalent or greater assurance than 
options (1) or (2) that seasonal emissions budgets will be attained and 
maintained. For further information regarding SIP approvability 
criteria, see Section VI.A.2.b of this preamble.
3. State Adoption of Model Rule
    In the SNPR, EPA proposed that States electing to participate in 
the NOX Budget Trading Program could either adopt the model 
rule by reference or develop State regulations in accordance with the 
model rule. The few commenters on this issue were primarily concerned 
about lack of guidance by EPA in this area for State adoption of the 
model rule and the potential for deviation from the model rule in the 
State-adopted rules. This section clarifies EPA's intent in issuing a 
model rule and distinguishes between sections of the model rule that 
State rules must mirror, and those that States may choose to alter or 
eliminate while maintaining a SIP that is approvable for purposes of 
joining the NOX Budget Trading Program.
    a. Process for Adoption. One commenter suggested that rather than 
adopting the NOX Budget Trading Program, it should be 
sufficient for each State to include a statement in its SIP declaring 
that the State will participate in the Federal program, along with a 
demonstration of the authority for the State to do so. This would leave 
the details in the Federal rule and avoid differences that could arise 
through each State adopting its own rule. However, EPA does not have 
the statutory authority under title I to promulgate a Federal cap-and-
trade program to achieve a State's SIP call budget unless the State 
fails to respond adequately to the SIP call. The EPA understands the 
commenter's concern regarding differences among State rules to 
implement the NOX Budget Trading Program, and intends to 
ensure consistency as explained in the following Section.
    The EPA's intent in issuing a model rule for the NOX 
Budget Trading Program is to provide States with a model program that 
serves as an approvable strategy for achieving more than 90 percent of 
the required reductions under the NOX SIP call. States 
choosing to participate in the program will be responsible for adopting 
State regulations to support the NOX Budget Trading Program, 
and submitting those rules as part of the SIP. As articulated in the 
proposed rulemaking (63 FR 25920), there are two legal alternatives for 
a State to use in joining the NOX Budget Trading Program: 
incorporate 40 CFR part 96 by reference into the State's regulations, 
or adopt State regulations that mirror 40 CFR part 96 but for the 
variations and omissions described below.
    b. Model Rule Variations. The EPA would like to clarify the 
variations and omissions from the model rule that are acceptable in a 
State rule, to provide States flexibility while still ensuring the 
environmental results and administrative feasibility of the program. 
More specifically, EPA will clarify those variations that maintain a 
State's eligibility for the streamlined SIP approval associated with 
adoption of the model rule, those changes that will require more 
extensive review by EPA prior to approval, and those changes that are 
not acceptable for incorporation into the NOX Budget Trading 
Program.
    In order for a SIP revision to be approved for State participation 
in the NOX Budget Trading Program, on a streamlined basis or 
otherwise, the State rule should not deviate from the model rule except 
in the areas of applicability, NOX allowance allocation 
methodology, and early reduction credit methodology (all of which are 
described briefly in the following paragraphs and in more detail in 
subsequent Sections of today's notice). Deviations from the model rule 
regarding allocation methodologies and early reduction credit 
methodologies as defined in this Section do not impact a State's 
eligibility for streamlined approval of its SIP with respect to the 
NOX Budget Trading Program. However, some deviations 
regarding applicability will require more extensive EPA review, as 
explained below. Changes to program applicability may render a State's 
rule ineligible for streamlined approval, though the rule would still 
be eligible for approval after a more thorough EPA review.
    State rules that deviate beyond the applicability, allocation, and 
early reduction credit flexibility provided in the model rule would not 
be approvable for inclusion in the NOX Budget Trading 
Program. SIPs incorporating a trading program that is not approved for 
inclusion in the broader NOX Budget Trading Program may 
still be acceptable for purposes of achieving some or all of a State's 
obligations under the NOX SIP call, provided the SIP 
criteria outlined in Section VI.A.2.b are met. However, only States 
participating in the NOX Budget Trading Program would be 
included in EPA's tracking systems for NOX emissions and 
allowances used to administer the multi-state trading program.
    For States participating in the NOX Budget Trading 
Program, applicability is one of the three main areas in which the 
State may deviate from the model rule. State rules need to include an 
applicability section that at least covers the core sources defined in 
the model rule, but States may allow additional stationary sources to 
participate in the trading program. These sources must be able to 
monitor and report emissions in accordance with the model rule, and 
identify an individual responsible for fulfilling program requirements 
to be eligible for inclusion. States have three options to expand 
applicability and one to limit it, as explained in the following 
paragraphs.
    States may choose to expand applicability either by: (1) Including 
smaller sources in the core source categories, (2) including additional 
source categories, or (3) providing individual sources the ability to 
opt in. Expansion of applicability to smaller core sources will 
maintain the State's eligibility for streamlined SIP approval with 
regard to the NOX Budget Trading Program. Including 
additional source categories beyond the core sources (e.g., municipal 
waste combustors), however, will require more careful review by EPA

[[Page 57459]]

in some cases to ensure that the trading program requirements can be 
met, and therefore preclude streamlined SIP approval otherwise 
associated with adoption of the model rule. Regarding individual source 
opt-ins, States have the discretion to determine whether or not to 
include this provision in their State rule. The opt-in provision is not 
a prerequisite to approval of a SIP incorporating the NOX 
Budget Trading Program. However, if a State does choose to include 
provisions for opt-in sources, these provisions must mirror those in 
the model rule. Providing the provisions do so, the SIP remains 
eligible for streamlined EPA approval.
    States may also choose to limit applicability of the trading 
program by allowing units with a low federally enforceable 
NOX emission limit (e.g. 25 tons per control period) to be 
exempt from trading program requirements. A State may include this 
exemption provision as it appears in the model rule to allow these 
sources not to participate in the trading program, or a State may omit 
the provision. Neither of these actions will interfere with streamlined 
SIP approval by EPA, provided the exemption provisions mirror the model 
rule if included in the State rule.
    In terms of allocations, States must include an allocation section 
in their rule, conform to the timing requirements for submission of 
allocations to EPA that are described in this preamble, and allocate an 
amount of allowances that does not exceed their State trading program 
budget. However, States may allocate NOX allowances to 
NOX budget sources according to whatever methodology they 
choose. The EPA has included an optional allocation methodology in 40 
CFR part 96, but States are free to allocate as they see fit within the 
bounds specified above, and still receive streamlined SIP approval for 
purposes of the NOX Budget Trading Program.
    Today's final rule also includes an optional methodology in 
Sec. 96.55(c) that States may use for issuing early reduction credits 
from the State compliance supplement pools. However, States may 
distribute the State compliance supplement pool to sources as they wish 
in accordance with the requirements set forth in 40 CFR 51.121(e)(3) 
and still receive streamlined SIP approval for purposes of the 
NOX Budget Trading Program.
    In summary, a State is eligible for streamlined approval of the 
portion of their SIP incorporating the NOX Budget Trading 
Program if the State adopts all the provisions of the model rule (e.g., 
banking and monitoring provisions) with variations incorporated only in 
the manner explained in this Section. Streamlined approval requires 
that applicability extends only to the core sources, or to core sources 
and smaller sources within the core source categories and that the opt-
in provision and the exemption option for sources with a low federally 
permitted emission limit, if included, mirror those in the model rule. 
Regarding allocations, eligibility for streamlined approval extends to 
those State rules whose allocations do not exceed the State trading 
program budget and are determined in accordance with the timing 
requirements delineated in the model rule. A State rule is still 
eligible for approval, but not streamlined approval, if the 
applicability determination for the NOX Budget Trading 
Program extends beyond the core sources to additional source 
categories, to allow for the additional review necessary to ensure such 
an extension of applicability is administratively feasible and 
environmentally sound. A State rule is also eligible for streamlined 
approval if it includes methodologies for issuing credit from the State 
compliance supplement pool in accordance with the provisions in 40 CFR 
51.121(e)(3). Differences among States in these areas will provide 
flexibility while not detracting from the operation or implementation 
of the multi-state trading program. Therefore, variations as explained 
in this section are acceptable to EPA with assurance that State rules 
will be sufficiently consistent. In addition, joint implementation of 
the program with EPA will ensure that once these consistent rules are 
established, they will be implemented consistently as well.
    Several commenters expressed concern that the lack of prohibitions 
on State-imposed trading restrictions in conjunction with the model 
rule would lead to variation between States and cripple the trading 
program. The EPA agrees with commenters that additional restrictions 
imposed on the trading program by individual States could increase 
economic costs without providing significant environmental benefit. 
Therefore, EPA does not believe that any restrictions on trading are 
necessary, and does not foresee approving State rules that include 
trading restrictions in SIPs incorporating the NOX Budget 
Trading Program. However, to address local air quality problems, a 
State participating in the NOX Budget Trading Program may 
establish permit limitations for specific sources participating in the 
trading program. The EPA considers such a limitation appropriate given 
local air quality concerns and does not consider it a trading 
restriction, and therefore the incorporation of such limitations will 
not preclude streamlined SIP approval. These sources would still 
participate in the NOX Budget Trading Program and the 
unconstrained market operating in the program, but could not use 
allowances to exceed their permit limitation; the source would be held 
to the permitted limit, regardless of how many allowances it holds for 
the purposes of the trading program. This topic is discussed in more 
detail in the next Section.
4. Unrestricted Trading Market
    a. Geographic Issues. For the NOX SIP call, EPA is 
basing the State budgets on the uniform application of reasonable, 
cost-effective NOX control measures for each State 
determined to contribute significantly to nonattainment in a downwind 
State. The EPA's analyses show that the collective reductions across 
the region will produce significant air quality benefits across the 
region. The development of and justification for the State budgets 
under the NOX SIP call is described in Section III, 
Determination of Budgets. Although the analyses in today's final action 
demonstrate that the collective emissions for the NOX SIP 
call region significantly contribute to nonattainment, the location of 
particular emissions does impact the effects that the emissions have on 
other areas within the region. Emissions in some locations may cause 
greater overall effects than emissions from other locations.
    In the SNPR, EPA proposed a single trading program allowing all 
emissions to be traded on a one-for-one basis without restrictions on 
trading allowances within the SIP call region. The EPA also solicited 
comment on whether the trading program should attempt to factor in 
differential effects of NOX emissions based on the location 
of the emissions. Possible options for factoring in the differential 
effects include defining exchange ratios for trades between areas based 
on the differential effects of emissions between areas, establishing 
subregions for trading, and/or prohibiting certain trades (63 FR 25902 
at 25919).
    The Agency received more than fifty comments on this issue from the 
regulated community, States, and environmental organizations. A number 
of commenters did support limiting trading by establishing smaller 
subregions within the SIP call region or

[[Page 57460]]

establishing trading ratios based on the idea that there are 
differential effects of NOX emissions based on the location 
of the emissions. However, none of these commenters included a complete 
proposal with a justification or description for the appropriate 
subregional boundaries or trading ratios. The majority of commenters on 
this subject favored unrestricted trading within areas having a uniform 
level of control. Most commenters supporting unrestricted trading 
stated that restrictions would result in fewer cost-savings without 
achieving any additional environmental benefit and would increase the 
administrative burden of implementing the program. They expressed 
concern that discounts or other adjustments or restrictions would 
unnecessarily complicate the trading program, and therefore reduce its 
effectiveness.
    Consistent with the proposal, the final model rule is designed to 
be a single jurisdiction trading program allowing all emissions to be 
traded on a one-for-one basis, without restrictions or limitations on 
trading allowances within the trading area. EPA has used the IPM to 
evaluate the emissions and cost impacts of alternative regulatory 
options under the SIP call for the electric power sector. These 
analyses can be found in the RIA. The model has been used to show the 
level and location of emissions if the SIP call were implemented under 
a number of different alternatives including unrestricted trading and 
command-and-control approaches. The results indicate that significant 
shifts in the location of emissions reductions would not occur with 
unrestricted trading compared to where the reductions would occur under 
command-and-control and intrastate only trading scenarios. Based upon 
the IPM results and EPA's air quality modeling, EPA has chosen a 
region-wide trading program allowing all emissions to be traded on a 
one-for-one basis without trading restrictions. EPA's analyses suggest 
that the net effect of all the trades is that the net emissions will 
not significantly shift within the region compared to a command-and-
control scenario. For this reason, EPA believes that the need for 
trading subregions or trading ratios that differ from one-for-one are 
unsubstantiated for the purposes of this SIP call and the 
NOX Budget Trading Program.
    Although the location of net emissions is not expected to 
significantly shift as a result of trading, it is possible that a State 
may identify a specific location (e.g., major NOX source 
adjacent to or within an urban center) where NOX reductions 
would be particularly beneficial for ozone mitigation. For these 
situations, a State may establish a specific permit limitation 
restricting the amount of NOX that may be emitted from the 
source. The source would still be included in the trading program but 
it would not be allowed to emit above the amount specified in the 
permit limitation regardless of the number of NOX allowances 
it may hold. The source would be allowed to trade the allowances it is 
unable to use. In this way, States will be able to tailor specific 
attainment strategies within the framework of the NOX Budget 
Trading Program without restricting the trading options for most 
sources included in the program.
    b. Episodic Issues. The EPA also received several comments 
addressing the episodic nature of ozone formation and whether this 
should be factored into the design of the trading program. Commenters 
noted that under the NOX SIP call, which is designed to 
reduce total NOX emissions from May through September of 
each year, it is still possible that NOX emissions may be 
relatively higher during ozone episodes compared with NOX 
emissions on other days between May and September. In addition, the 
effect of a unit of emissions may be higher during ozone episodes. To 
address this concern, the commenters stated that the trading program 
should provide incentives or safeguards to ensure that NOX 
emissions reductions are achieved specifically during ozone episodes. 
One commenter asserted that emissions could either be capped during 
ozone episodes or that the trading program could place a premium on the 
use of NOX allowances during ozone episodes. The commenter 
recommended the latter option. The premium would require that sources 
surrender NOX allowances at rates greater than 1-to-1 for 
each ton of NOX emitted during the ozone episodes.
    Consistent with the NOX SIP call, the NOX 
Budget Trading Program focuses on reducing total NOX 
emissions from May to September for the jurisdictions that are 
identified in the NOX SIP call and that choose to 
participate in the trading program. Proposals to address NOX 
emissions during specific episodes and in specific nonattainment areas 
are more closely tied to issues affecting individual attainment plans 
rather than the goal of the NOX SIP call which is to reduce 
transport. It would be very difficult to apply the appropriate premium 
to the individual sources that contribute NOX emissions 
affecting specific ozone episodes. The meteorology and source 
contribution for each ozone episode is different. And in some cases, 
NOX emissions and the resulting ozone may be transported for 
several days before contributing to an ozone violation.
    Provisions designed to ensure that NOX emissions 
reductions are achieved specifically during ozone episodes are more 
likely to be effective in controlling NOX emissions that are 
released adjacent to or within locations frequently affected with 
elevated ozone levels. Where a State identifies such a source, EPA 
believes specific permit limitations are an appropriate and effective 
method for controlling the source's emissions. As stated in the 
previous section, EPA believes that States may use permit limitations 
to tailor specific attainment strategies within the framework of the 
NOX Budget Trading Program without restricting the trading 
options for most sources included in the program. Furthermore, this 
provides each State more flexibility in establishing its attainment 
plan rather than applying one approach to address the episodic nature 
of ozone throughout the SIP call region. Therefore, EPA has not 
included additional trading restrictions to address ozone episodes in 
the design of the final NOX Budget Trading Program.

D. Applicability

1. Core Sources
    In the SNPR, EPA proposed that compliance with the emission 
limitation requirements of the NOX Budget Trading Rule, 
i.e., the requirement to hold sufficient NOX allowances to 
cover emissions, apply to a core group of large stationary sources that 
includes all fossil fuel-fired stationary boilers, combustion turbines, 
and combined cycle systems (i.e., units) that serve an electrical 
generator of capacity greater than 25 MWe and to any fossil fuel-fired 
stationary boilers, combustion turbines, and combined cycle systems not 
serving a generator that have a heat input capacity greater than 250 
mmBtu/hr. A unit was considered fossil fuel-fired if fossil fuels 
accounted for more than 50 percent of the unit's heat input on an 
annual basis. The EPA solicited comment on the appropriateness of the 
categories included in the core group, whether the size cut-offs should 
be higher or lower for the source categories, and the appropriateness 
of including other source categories in the core group. Comments on the 
concept of a core group fell into three broad categories:
     Those who agreed with the core group concept and who 
generally agreed

[[Page 57461]]

with EPA's proposed core group definition;
     Those who felt that the core group definition was too 
limiting; and
     Those who felt that the core group definition was too 
inclusive.
    a. Commenters Who Felt the Core Group Should Not Be Changed. 
Commenters who supported the concept of a core group generally and the 
cut-offs proposed by EPA specifically explained that the cut-offs are 
consistent with the Acid Rain Program and that the use of a core group 
will minimize inconsistencies that could impede establishment of 
interstate trading. Commenters also added that the program should 
provide the flexibility to allow additional sources to opt-in on an 
individual basis or for States to bring in additional sources on a 
categorical basis. Some of these commenters added that the timing for 
bringing in these sources or source categories should be dependent upon 
the ability of the source or source category to accurately monitor 
emissions. For some source categories it might be appropriate to bring 
them in at the start of the program; for others, it might be necessary 
to wait until their ability to quantify emissions has improved.
    Commenters who generally supported the concept of a core group of 
sources as it was defined in the SNPR did have several specific 
concerns. One commenter noted that while the SNPR preamble clearly 
explained that the rule only included fossil-fuel-fired units, the rule 
itself was not clear on this issue. Another commenter suggested that 
because the proposed definition differentiated between electrical 
generating units and non-electrical generating units it excluded 
sources that should be in the trading program such as cogeneration 
facilities that consisted of boilers greater than 250 mmBtu/hr that 
served electric generating units with a rating of less than 25 MWe.
    The EPA agrees that the establishment of a core group will help 
facilitate interstate trading as well as compliance with the emissions 
budget. If there is not some minimum group of trading participants, 
sources that are in the program will have less of an opportunity to 
trade allowances and realize the economic benefits of trading. In 
addition, by ensuring that most of the emissions from industries 
covered by the trading program are included in a capped system, the 
trading program can be simplified because concerns about load shifting 
to uncapped sources is minimized. The EPA also agrees that making the 
cut-offs consistent with existing regulatory programs helps to minimize 
conflicts with existing regulatory programs. The EPA also agrees with 
both of the concerns raised by the commenters. Therefore the regulatory 
definition of unit has been clarified to make it clear that a unit must 
be fossil-fuel fired. The EPA has also added a clarification to the 
definition of fossil-fuel fired. This clarification is intended to 
define a baseline period for determining if a unit is fossil-fuel 
fired. The revised definition states that fossil-fuel fired means the 
combustion of fossil fuel, alone or in combination with any other fuel, 
where the fossil fuel comprises more than 50 percent of the annual heat 
input on a Btu basis. An existing unit is considered fossil-fuel fired 
if it meets this criterion for any year since 1990 (or if not operating 
since 1990 during the last year of operation). A new unit is considered 
fossil-fuel fired if it is projected to meet this criterion or, if 
after operation begins, it does meet this criterion.
    In addition, to address the concern about excluding cogeneration 
facilities that are greater than 250 mmBtu/hr that serve electric 
generating units with a rating of less than 25 MWe, the applicability 
has been changed to include all units greater than 250 mmBtu/hr, 
regardless of how much electricity they generate.
    b. Commenters Who Felt the Core Group Should Be Expanded. 
Commenters who felt the trading program should be expanded focused on a 
number of areas. Several commenters argued generally that the program 
should allow any source to participate if the source can document that 
emissions reductions have been achieved. A number of commenters 
mentioned as examples the inclusion of medium-sized and smaller 
stationary sources in the RECLAIM program. A few commenters argued that 
the addition of certain sources is needed for consistency with the OTC 
NOX Budget Rule. Other commenters opposed the core group 
concept because they believe that regulation of low-level and local 
sources in the Northeast is an essential step in solving the ozone 
problem. Others argued that excluding non-utility sources from the 
trading program unfairly excludes these sources from least-cost 
compliance options. Some commenters suggested specific categories of 
units that should be allowed to, but not required to, participate in 
the trading program. These included:

(1) Municipal waste combustors;
(2) Internal combustion engines;
(3) Process units;
(4) Units for which the output product is not comparable to other 
units on which the allocations are based, such as process heaters, 
hazardous waste incinerators, process vents and nitric acid plants.

    The EPA believes that many of the concerns about the core source 
definition stem from a misunderstanding of its purpose. The core 
sources definition was intended to indicate the minimum applicability 
requirements that a State rule would have to include to participate in 
a larger multi-state program that EPA would help to administer. It was 
not intended to limit individual States from including more sources (as 
long as the sources meet certain criteria further explained below) in 
the larger multi-state program (63 FR 25924). Nor was it intended to 
prohibit a State (or group of States) from developing its own trading 
program with a more limited applicability.
    If, however, a State or group of States developed a trading program 
that did not meet the minimum requirements set forth in the model 
NOX Budget Trading Program, such as minimum core source 
applicability, EPA would not participate in the administration of such 
a trading program. This is because it would not be administratively 
cost-efficient for EPA to manage multiple trading programs with a 
variety of applicability and other requirements designed to address the 
same issue.
    The EPA is not expanding the core source group to include any 
additional sources because EPA believes that this decision is better 
left to the states. Therefore the model rule will allow a State to 
expand the applicability of the trading program to include additional 
stationary sources if the sources meet certain criteria. These criteria 
include the ability to accurately and consistently monitor and report 
emissions and the ability to identify a party responsible for ensuring 
that monitoring and reporting requirements are met, for authorizing 
allowance transfers and for ensuring compliance. The EPA's rationale 
for setting these minimum criteria are set forth in the preamble to the 
SNPR (63 FR 25923). Also, EPA addresses issues specifically related to 
the monitoring requirements for these sources in Section D.3 of today's 
preamble.
    There are two mechanisms that can be used to include more sources 
in the program. One is for a State to expand the applicability criteria 
to include other source categories; the other is to give individual 
sources the ability to opt-in.
    States that choose to expand the applicability criteria can do so 
(1) by lowering the applicability threshold for source categories that 
are already part of

[[Page 57462]]

the core group in order to include smaller sources or (2) by including 
additional source categories that are not included in the core group. 
For instance a State in the OTC might choose to lower the applicability 
cut-off for electrical generating units to 15 MWe to make the program 
more consistent with the existing OTC NOX Budget Program. If 
a State chose to expand the applicability criteria for source 
categories already included in the core group this would not affect 
EPA's streamlined approval of the NOX Budget Trading program 
component of the State's SIP.
    A State might choose to lower the applicability cut-off for sources 
in the core group to create different applicability cut-offs for new 
and existing units. This could help to better facilitate integration 
with a State's new source review program. The EPA took comment on this 
concept in the SNPR and received comments both for and against this 
proposal. Commenters who opposed it suggested that it would be a 
disincentive to replace old units with new cleaner units. Some of these 
commenters also noted that expanding the applicability cut-off for all 
units would provide an incentive to replace these older units. 
Commenters who favored it suggested that it would be an incentive to 
make new units as clean as possible. The EPA believes that it is 
appropriate for States to determine how best to handle the issue of 
small new units.
    Another reason to allow smaller sources to opt-in is to simplify 
monitoring for situations in which a common stack is shared by a number 
of units, some of which are affected and some which are not. In this 
situation the owner or operator would have to either install monitors 
at each of the affected units, or install monitors at the common stack 
and at all of the non-affected units, so that the emissions from these 
units could be deducted from the emissions from the affected units. If 
the owner or operator is allowed to opt-in the nonaffected unit, they 
will be able to install one set of monitors at the common stack 
accounting for the emissions from all of the units.
    If a State chose to include additional source categories, EPA would 
have to review the SIP submittal to ensure that those additional source 
categories met the minimum criteria for monitoring and reporting 
emissions and for having a responsible official. As further explained 
in the SNPR (63 FR 25924), EPA would also have to determine if it could 
successfully administer a regional trading program with the inclusion 
of these additional source categories.
    In the SNPR, EPA proposed developing a list of specific additional 
source categories beyond the core group which a State could bring into 
the trading program without affecting EPA's streamlined approval of the 
trading component of the SIP. While this concept received general 
support, none of the commenters provided enough specific support to 
demonstrate that all of the sources in a given source category could 
meet the criteria to accurately and consistently monitor emissions. 
These comments are discussed in Section D.3.
    The EPA believes that the opportunity for States to expand the 
applicability to include additional sources addresses concerns about 
incompatibility with the applicability requirements of existing 
programs, such as the OTC Trading Program, as well as concerns that an 
individual State might want to expand the program to address local 
ozone problems.
    The other mechanism that can be used to broaden the applicability 
of the program is the individual opt-in procedures in subpart I of part 
96. These provisions allow a source to opt-in, if it can meet the 
monitoring and reporting requirements of part 75. The EPA received a 
number of comments about the monitoring requirements of part 75 as they 
related to opt-ins. These comments are addressed in Section D.3 of 
today's preamble.
    In the SNPR (62 FR 25940-25942 and 62 FR 25991-25994), EPA proposed 
that the individual opt-in provisions would only be applicable to 
fossil-fuel-fired, stationary boilers, combustion turbines, and 
combined cycle systems smaller than the applicability cut-offs of 25 
MWe or 250 mmBtu/hr. The EPA agrees that the RECLAIM program has 
demonstrated that many combustion sources that are not included in the 
core applicability criteria can accurately and consistently monitor 
NOX mass emissions using CEM (or other alternative protocols 
for units with low mass emissions) that are very similar to the 
provisions in subpart H of part 75. Therefore, in today's action EPA is 
allowing States to expand the opt-in provisions to include any 
stationary combustion source that emits to a stack and can meet the 
monitoring and reporting requirements of subpart H of part 75.
    States that choose to add other combustion sources that are not 
part of the core group would also have to address issues related to 
allocating allowances for those types of sources. Allocation 
methodologies that may be appropriate for source categories covered in 
the core group may not be as applicable for other source categories. 
For instance, as one commenter noted, an output based allocation 
methodology might not make as much sense for a municipal waste 
combustor, since the primary purpose of a municipal waste combustor is 
to combust waste, not to generate usable output.
    c. Commenters Who Felt the Core Group Is Overly Inclusive. A number 
of commenters argued that the burdens associated with including certain 
source categories would outweigh the benefits and that particular types 
of sources should therefore be excluded from the core group. Many of 
these commenters stated that individual sources in these groups should 
be allowed to opt in where there is a net economic benefit to them to 
participate rather than mandating inclusion of the source category. 
Specific categories include: non-utility boilers generally; generators 
of power for on-site use; combustion turbines exempt from Title IV; 
small cyclone boilers; combustion turbines below 100 MWe; small, 
particularly municipal, electric generating units (e.g., those under 25 
MWe); and units with low potential to emit as defined by enforceable 
limits (e.g., peaking units with potential to emit less than 100 tons 
per year).
    The EPA does not believe there is a great distinction between 
similarly sized utility and non-utility boilers. Both categories of 
boilers are similar in design, have similar control options and have 
similar control costs. Therefore, EPA is not excluding large non-
utility boilers from the trading program. The EPA believes the same 
arguments that apply to utility and non-utility boilers also apply to 
generators of power for on-site use and generators of power for resale. 
In light of the fact that utility restructuring will provide more 
opportunities for generators of power for on-site use to resell the 
power they produce in the future, EPA believes that this distinction is 
even harder to make. Therefore, EPA is not excluding large generators 
of power for on-site use from the trading program.
    In accordance with title IV of the CAA, the Acid Rain Program 
exempts simple combustion turbines that commenced commercial operation 
before November 15, 1990. These units were exempted from the Acid Rain 
Program because the SO2 emissions from these units were 
extremely low. The NOX emissions from these units are 
potentially higher; therefore, EPA is not adding a specific exemption 
for these types of units. However, many of these units are small and/or 
infrequently operated, so their actual NOX emissions may be 
quite low; therefore, some of these units may qualify for the

[[Page 57463]]

alternative compliance options for units with low NOX mass 
emissions, explained below. Combustion turbines smaller than 100 MWe 
are also likely candidates to qualify for the alternative compliance 
option explained below.
    The Acid Rain Program exempts cyclone boilers with a maximum 
continuous steam flow at 100 percent load of greater than 1060 thousand 
lb/hr from NOX control requirements under part 76. These 
units were exempted because one of the primary criteria in title IV of 
the CAA for setting emissions limitations under part 76 was 
comparability of cost with low NOX emission controls on 
boilers categorized as group 1 boilers under Title IV (large 
tangentially fired and dry bottom, wall fired). There is no such 
criterion in the CAA applicable to this rulemaking. Also, since the 
emission reductions required by this rulemaking are more substantial 
than the emission reductions required under part 76 70, the 
cost per ton of reducing NOX emission reductions is 
correspondingly higher. Therefore, applicability cutoffs that were 
relevant in the part 76 rulemaking are not relevant in this rulemaking.
---------------------------------------------------------------------------

    \70\ The lowest emission rate required under part 76 is 0.40 
lbs/mmBtu.
---------------------------------------------------------------------------

    In response to the comment that small electrical generators less 
than 25 MWe should be exempt from the NOX Budget Trading 
Program, they were proposed to be exempt and will be exempt under the 
final model rule. They do still have the option of opting into the 
program if they choose to do so.
    In the SNPR (63 FR 25926), EPA took comment on allowing units with 
a low federally enforceable NOX emission limit (e.g. 25 tons 
per ozone season), that because of their size would be included in the 
trading program, to be exempt from the requirements of the trading 
program. In general commenters supported this concept. One commenter 
who supported the concept also added that it would be important to 
ensure that there were adequate requirements to assure that the 
individual sources who took advantage of this option demonstrated 
compliance with their unit-specific caps. The commenters who disagreed 
with this option expressed concern that a State's budget could be 
exceeded if emissions from these units were not accounted for.
    Based on the comments received EPA continues to believe that it is 
appropriate to offer States the option of providing units that are 
above the applicability threshold but that have a very low potential to 
emit an alternative compliance option. This option would allow units 
that meet the requirements described below to be exempt from the 
requirements to hold allowances, and to comply with quarterly reporting 
requirements. In order to address the concern that sources must 
demonstrate compliance with their individual cap, EPA has added 
specific requirements that sources must meet in order to use this 
alternative compliance option.
    Units that use this option would be required to:
    (1) have a federally enforceable permit restricting ozone season 
emissions to less than 25 tons;
    (2) keep on site records demonstrating that the conditions of the 
permit were met, including restrictions on operating time;
    (3) report hours of operation during the ozone season to the 
permitting authority on an annual basis.
    A unit choosing to use this compliance option would be required to 
determine the appropriate restrictions on its operating time by 
dividing 25 tons by the unit's maximum potential hourly NOX 
mass emissions. The unit's maximum potential hourly NOX mass 
emissions would be determined by multiplying the highest default 
emission rate for any fuel that the unit burned (using the default 
emission rates, in part 75.19 of this chapter) by the maximum rated 
hourly heat input of the unit (as defined in part 72 of this chapter).
    States would be allowed, but not required, to incorporate this 
alternative compliance option into their SIPs. The EPA does agree that 
if a State does incorporate this option into the SIP, it would have to 
account for the emissions under its budget. Thus a State that chose to 
use this option would have to either:
    (1) Subtract the total amount of potential emissions permitted to 
be emitted using this approach from the trading portion of the budget 
before the remaining portion of the trading budget is allocated to the 
trading participants; or (2) Offset the difference between total amount 
of potential emissions permitted to be emitted using this approach and 
the 2007 base year inventory emissions for these same sources with 
additional reductions outside of the trading portion of the budget.
    If States choose not to incorporate this alternative compliance 
option into their SIPs, or if they choose to incorporate it exactly as 
it is set forth in the model rule, it will not affect the streamlined 
approval of the trading rule portion of the SIP. A State may choose to 
require an alternative means of ensuring that the potential to emit for 
units utilizing the alternative means of compliance is limited to less 
than 25 tons, however if a State deviates from the model rule in this 
way, the SIP will no longer receive streamlined approval.
2. Mobile/Area Sources
    The proposed rule did not include mobile or area sources in the 
trading program, but solicited comment on expanding applicability to 
include these sources, or to include credits generated by these 
sources, in the trading program. Mobile and area sources were not 
included in the proposed trading rule due to EPA's concerns related to 
ensuring that reductions were real, developing and implementing 
procedures for monitoring emissions, and identifying responsible 
parties for the implementation of the program and associated emissions 
reductions.
    The EPA received comment from State and local government, industry 
and coalitions of industry, and environmental groups regarding the 
inclusion of mobile and area sources in the program. Comments focused 
on the following main areas: inclusion or exclusion of mobile and area 
sources, subcategories of mobile sources for inclusion, and the use of 
pilot programs to foster innovation.
    Some commenters urged EPA to include mobile and area sources with 
as few restrictions as possible in the trading program, primarily on an 
opt-in or voluntary basis. These commenters argued that excluding 
mobile sources would reduce the potential scope and benefits of the 
trading by placing a large portion of States' NOX inventory 
outside the scope of the trading program. They noted that the existence 
of RECLAIM protocols for mobile and area source credit generation 
demonstrated that EPA's quantification, verification, and 
administration concerns were misplaced.
    The majority of commenters, however, indicated that mobile sources 
should not be included at this time and that the model rule should not 
be delayed to address concerns related to inclusion of these sources. 
Some commenters argued against ever including mobile and area sources 
in the program. One State argued that inclusion of mobile and area 
sources would destroy the integrity of the program since mobile and 
area source reductions are not necessarily real, verifiable and 
quantifiable, failing to display a level of certainty comparable to 
those sources included in the trading program. A few commenters 
indicated that mobile sources were inherently unsuited to a capped 
system, since the difficulties of measuring emissions from these 
sources precludes their inclusion in a budget.

[[Page 57464]]

    Several commenters suggested that some categories of mobile sources 
should be included while other categories should not. Commenters 
indicated, for example, that it is not feasible to have individual 
motorists participate in the cap-and-trade program due to the burdens 
and administrative complexity associated with such a vast number of 
sources and responsible parties in a trading system. Alternatively, 
commenters argued that manufacturers, fuel distributors, and fleet 
owners could be included if they were able to generate surplus emission 
reductions by going beyond the requirements established by some Federal 
measures. These commenters specifically cited the low-RVP regulations, 
the vehicle scrappage guidance, and the locomotive regulations as 
examples of such Federal measures.
    Several commenters who recommended that mobile sources not be 
included in the program at this time also recommended that EPA sponsor 
pilot programs in States to study the feasibility of inter-sector 
trading and to develop mechanisms to address the specific concerns 
mentioned regarding the inclusion of mobile and area sources. Along 
similar lines, one industry commenter stated that mobile sources may be 
appropriate candidates for participation in the trading program only if 
adequate emission reduction measurement protocols can be developed. 
Foreseeing this occurrence, some commenters felt that EPA should leave 
a placeholder in the rule or add a provision that would include mobile 
and area sources once the mechanisms to address the specific concerns 
of EPA and others have been developed.
    The model trading program that EPA is finalizing today will not 
include mobile and area sources for the reasons outlined in the SNPR. 
The EPA concurs with the concerns raised by commenters against the 
inclusion of mobile and area sources, regarding program integrity, 
emissions monitoring, and accountability. Most of the proponents of 
including mobile or area sources listed general reasons for including 
them such as increasing market efficiency, lowering costs, or simply 
the existence of RECLAIM protocols to do so. However, these commenters 
did not provide sufficient information or documentation to support the 
validity of these assertions, and several acknowledged that the 
potential for improvement in market efficiency or lower compliance 
costs was difficult to ascertain. Further, one proponent acknowledged 
that the RECLAIM protocols are new and not yet extensively utilized.
    In fact, a recent audit of the RECLAIM program indicates that the 
volume of mobile source credits used under the program is very small 
(only 99 NOX tons have been converted from mobile source 
reductions in the last five years). Only 5 requests for conversion of 
mobile source emission reduction credits to RECLAIM trading credits 
were approved in 1994, and no further requests had been received as of 
May 1998. The small amount of credits relative to the significant 
resource expenditure for the conversion of mobile source credits under 
the RECLAIM program (i.e., the need for case-by-case review given the 
variability and complexity of the petitions) suggests that the RECLAIM 
mobile source protocols and strategy are not yet a cost-effective 
option for the trading program.
    The EPA remains willing to consider adding mobile or area sources 
to the trading program in the future. Most commenters recommended that 
the program be opened to mobile or area sources once adequate 
mechanisms are developed for addressing related concerns. In response 
to these comments, and those recommending that EPA support pilot 
programs in States in order to facilitate resolution of the areas of 
concern for mobile and area sources, EPA will investigate how grant 
funding may be used for such pilots. Additionally, EPA is pursuing 
possible ways to incorporate mobile and area source strategies into 
other trading and incentive programs. Through these efforts, EPA will 
work with States in finding solutions to adequately address concerns 
such as emissions variability, difficulty in controlling emissions 
growth, difficulty in monitoring emissions levels, and difficulty in 
establishing emissions baselines. Through this process, EPA and States 
will explore and develop the necessary protocols that could eventually 
allow the inclusion of mobile and area sources in some capacity in the 
NOX Budget Trading Program. Anticipating that the 
quantification, verification, and administration concerns regarding 
expansion of the trading program to include mobile and area sources may 
be sufficiently resolved in the future, EPA is reserving in this 
rulemaking a section in part 96 for future inclusion of mobile or area 
sources in the NOX Budget Trading Program.
    The EPA is aware of other concerns on which the Agency did not 
receive comment, including the adequacy of some of the existing mobile 
source protocols and the enforcement of mobile source credit generation 
strategies. These emerging issues, coupled with past experience, and 
the issues raised by commenters lead EPA to conclude that it is not 
appropriate to include mobile and area sources in the NOX 
Budget Trading Program at this time.
3. Monitoring
    For the reasons set forth in the SNPR (63 FR 25938-40), EPA 
proposed that sources in the NOX Budget Trading Program use 
the monitoring methodologies in proposed subpart H of part 75 to 
quantify their NOX mass emissions (63 FR 28032). The 
comments that EPA has received can be classified into three main 
categories:
     Support for requiring the use of part 75 to demonstrate 
compliance with the trading program,
     Support for using CEMS on large units, but concerns about 
using part 75 as the monitoring protocol, and
     Concerns about requiring CEMS.
    Some of the commenters concerned about requiring CEMS focused on 
units of any size that are not subject to the provisions of the Acid 
Rain Program. Others focused on smaller units.
    The EPA proposed revisions to part 75 (63 FR 28032) for a number of 
reasons, one of which was to add procedures for monitoring 
NOX mass emissions (subpart H). These procedures could be 
used by sources to comply with any State or Federal program requiring 
measurement and reporting of NOX mass emissions. In 
particular, subpart H would be used by sources to meet the monitoring 
and reporting requirements of the NOX Budget Trading Rule 
(part 96) and the monitoring and reporting requirements of the SIP call 
for (1) combustion units (boilers, turbines and combined cycle units) 
which serve electric generators greater than 25 MWe and (2) combustion 
units greater than 250 mmBtu/hr, regardless of whether they serve a 
generator.
    The part 75 revisions also proposed to make a number of other 
changes that would affect units using part 75 to comply either with the 
requirements of title IV or the requirements of a NOX mass 
emissions program that incorporated or adopted the requirements of part 
75. These included a number of minor changes to simplify and streamline 
the rule to make it more efficient for both affected facilities and 
EPA, a new excepted monitoring methodology that would reduce monitoring 
burdens for affected facility units with low mass emissions, new 
quality assurance requirements based on gaps identified by EPA during 
evaluation of the initial implementation of part 75, and several minor 
technical

[[Page 57465]]

changes to maintain uniformity within part 75 and to clarify various 
provisions.
    The following discussion addresses comments received in the SNPR 
docket (A-96-56) that are related to the general requirement to monitor 
emissions, the requirement to monitor emissions using CEMS, and the 
requirement to monitor using part 75. Although EPA had requested that 
all comments related to the use of part 75 for monitoring 
NOX mass be submitted to the part 75 docket (A-97-35), some 
comments also dealt with the specific requirements set forth in part 
75.
    In today's rulemaking, EPA is finalizing sections of part 75 
related to monitoring NOX mass emissions as well as those 
which address the excepted monitoring methodology for units with low 
mass emissions of NOX and SO2 that combust oil or 
natural gas. Units using this methodology to comply with the 
requirements of part 96 would be subject only to the NOX 
mass emission requirements and not to the SO2 mass emission 
requirements. For a more complete discussion of the NOX mass 
monitoring and reporting provisions in part 75, see the Amendments to 
Part 75 Section below and Appendix A of this preamble. These Sections 
discuss both the comments received in the part 75 docket as well as the 
comments received in the SNPR docket that address the specific 
requirements of part 75.
    a. Use of Part 75 to Ensure Compliance with the NOX 
Budget Trading Program. Several commenters supported the idea of 
requiring all sources in the trading program to meet the monitoring 
provisions of part 75. Some of these commenters noted that part 75 
provides the consistent and accurate monitoring requirements necessary 
to ensure the integrity of a cap and trade program. They also noted 
that the proposed revisions offered the flexibility needed for sources 
to be able to reasonably comply.
    Several commenters supported the concept of trying to consolidate 
the monitoring and reporting requirements for units in the 
NOX Budget Trading Program already subject to part 75 under 
the Acid Rain Program.
    Response: The EPA agrees that accurate and consistent data are 
important to ensure the integrity of a trading program and that the 
protocols in part 75 provide for such accurate and consistent data from 
stationary combustion sources. Today's final model rule would require 
all sources in the trading program (including sources currently subject 
to part 75) to use the monitoring and reporting procedures set forth in 
subpart H of part 75.
    b. Use of CEMS on Large Units. A number of commenters expressed 
support for the requirement that large units should use CEMS to 
quantify NOX mass emissions. Many of these commenters did, 
however, have concerns about using part 75 as the basis for this 
monitoring. Some of these commenters elaborated that part 75 was 
specifically developed for utility units and that it might not be 
applicable to other types of units. Commenters also expressed concerns 
about costs associated with upgrading existing CEM systems to meet the 
part 75 requirements. The main alternatives they suggested were either 
using existing State monitoring and reporting requirements or allowing 
States the discretion to create or approve new monitoring and reporting 
requirements.
    Response: For reasons set forth in the preamble to the SNPR, EPA 
believes that the use of CEMS, in general, and the protocols in part 
75, more specifically, are the most effective way to ensure that 
NOX mass emissions from large combustion sources are 
quantified in an accurate and consistent manner from source to source 
and are reported in a consistent and cost-efficient way. This is 
important to maintain the integrity and efficiency of the trading 
system.
    The EPA believes that the protocols in part 75 can appropriately be 
applied to all of the core sources (fossil fuel-fired electric 
generating units and industrial boilers). The issues associated with 
monitoring NOX mass emissions from a stack attached to a 
boiler, turbine, or combined cycle unit are the same regardless of 
whether that boiler, turbine, or combined cycle unit is owned or 
operated by a utility, by an independent power producer, or by a 
manufacturer. The EPA does acknowledge that there may be additional 
issues associated with monitoring NOX mass from units such 
as process heaters or cement kilns.
    The RECLAIM program uses very similar protocols to the ones in part 
75 to quantify NOX mass emissions. Both RECLAIM and part 75 
require the use of NOX CEMS and flow CEMS to quantify 
NOX mass emissions from large sources combusting solid fuel. 
Both RECLAIM and part 75 also offer large oil and gas units an 
additional option for monitoring. This option involves the use of a 
fuel flowmeter and fuel sampling and analysis. The RECLAIM program 
requires monitoring of source categories that are in the NOX 
Budget Trading Program core group, such as boilers and turbines, but 
also requires monitoring of source categories that are not in the core 
group, such as process heaters and cement kilns.
    RECLAIM needed to establish a standing working group to resolve 
issues related to monitoring NOX mass from such a wide range 
of source categories (See South Coast Air Quality Management District, 
RECLAIM Program Three Year Audit and Progress Report, May 8, 1998). EPA 
does not believe that the problems that RECLAIM has had with monitoring 
are related to the protocols that program uses. Rather, EPA believes 
these problems are due to the limited experience that both States and 
sources have with monitoring such a wide range of source categories.
    The EPA believes that regardless of what protocols are used, if 
States opt to bring additional source categories into the trading 
program, issues related to monitoring at specific source categories 
will arise. These issues will need to be resolved, thus improving State 
and EPA experience with those source categories. If a State wants to 
include additional sources beyond those included in the core group, 
then EPA would resolve issues through the initial certification process 
for opt-in units. The EPA will also provide additional guidance on 
specific source categories, sharing the experiences gained with 
individual opt-in units.
    Using one basic set of protocols will make it easier for states, 
sources and EPA to work together while gaining more experience with 
these sources and resolving the issues in a cooperative and consistent 
manner.
    The EPA believes that the most significant costs associated with 
upgrading from an existing NOX emission rate monitoring 
system to a part 75 NOX mass monitoring system are 
associated with the need to monitor NOX mass and would be 
incurred regardless of the specific monitoring protocol that was 
required. Many existing CEM rules other than part 75 require sources to 
monitor NOX emission rate (in lbs/mmBtu) or NOX 
concentration corrected for oxygen (in ppm)(e.g. monitoring 
requirements under Subpart D, Da, Db of part 60). In order to meet 
these requirements, a NOX monitoring system must consist of 
a NOX concentration CEM, a diluent CEM and a data 
acquisition and handling system (DAHS). The DAHS is the part of the 
system that collects raw monitor data, performs calculations, and 
generates reports.
    In order to upgrade an existing system so that it can monitor 
NOX mass, a source must install a flow CEMS, if it burns 
solid fuels, or must install either a flow CEMS or a fuel flow meter if 
it burns a homogeneous oil or gas. In addition, the source would have 
to

[[Page 57466]]

upgrade its DAHS to reflect the reporting of NOX mass rather 
than NOX emission rate or NOX concentration. 
These costs must be incurred, regardless of the protocol that a source 
used to monitor NOX mass.
    The EPA believes that a single monitoring and reporting protocol 
for the NOX Budget Trading Program will keep the costs of 
upgrading systems to a minimum. This is because equipment vendors will 
be able to create standardized systems that will be applicable to all 
sources in the program, rather than having to create many different 
State- and source-specific systems. A single monitoring and reporting 
protocol will also help ensure a level playing field for all affected 
sources.
    For these reasons, part 96 requires all large units to monitor 
NOX mass emissions using CEMS in accordance with part 75. 
However, as explained below, part 75 does offer various monitoring 
options for low-emitting or infrequently operated oil- and gas-fired 
units, in addition to CEMS.
    c. Commenters Who Do Not Believe That CEMS Are Necessary. Some 
commenters expressed concerns about requiring CEMS on any unit that 
does not currently have a CEMS monitoring requirement. Suggested 
alternatives included the use of stack test data and emission factors. 
Some commenters also suggested the testing and monitoring provisions of 
a source's title V permit.
    Response: For large sources, EPA does not believe that stack test 
data and emission factors provide the consistent and accurate data 
needed to facilitate a trading program. Stack test data provide a one-
time assessment of a source's emission rate. Emission factors at best 
are based on a series of stack tests at similar units. A unit's actual 
emission rate may fluctuate greatly over time due to factors such as 
the way the unit and/or its associated control equipment is operated 
and maintained and the quality of fuel that the unit burns. An emission 
factor or stack test will often not be representative of that unit's 
actual normal emissions. Continuous monitoring of actual emissions will 
ensure that fluctuations in emission rates are accounted for. Because 
CEMS provide continuous monitoring, they can also indicate when 
emission control equipment is malfunctioning, thus, helping to ensure 
that the owners of units continue to properly operate and maintain any 
installed emission control equipment.
    Title V permits incorporate all of the monitoring requirements to 
which a source is subject in order to demonstrate compliance with its 
current regulatory requirements. In addition, where a source is not 
subject to any other monitoring requirements, it sets forth minimum 
monitoring requirements. In many cases the current regulatory 
requirements do not require compliance with a mass emissions 
limitation. Therefore, the monitoring requirements are not designed to 
demonstrate compliance with a mass emission limitation.
    Even when a source may have monitoring requirements designed to 
demonstrate compliance with a mass emissions limitation, the stringency 
of these requirements often varies from source to source and from State 
to State. These variations in turn lead to inconsistencies in sources' 
accounting of mass emissions. This both creates an uneven playing field 
for sources and undermines the integrity of the trading program.
    The EPA believes that it is necessary for all sources in the 
trading program to be subject to accurate and consistent monitoring 
requirements designed to demonstrate compliance with a mass emission 
limitation. This will ensure compliance with the requirements of the 
SIP Call and will ensure the integrity of the trading program.
    The EPA does believe that it is appropriate to provide lower cost 
monitoring options for units with low NOX mass emissions. 
Part 75 allows non-CEMS alternatives to quantify NOX mass 
emissions for gas and oil fired units that have low NOX mass 
emissions and/or that operate infrequently.
    In contrast, EPA does not believe that the types of protocols set 
forth in the Compliance Assurance Monitoring (CAM) rule, part 64, are 
appropriate for a trading program because they were not designed to 
quantify mass emissions. The preamble to the CAM rule further 
elaborates why these protocols are not appropriate for a trading 
program (62 FR 54915, 54916, 54922).
    The EPA believes that the types of protocols in RECLAIM and the 
Ozone Transport Commission's NOX Budget Trading Program 
(``OTC Program'') are more appropriate for a trading program because 
they were specifically designed to quantify NOX mass 
emissions. The EPA also believes that the flexible monitoring options 
offered by part 75 are consistent with the type of flexibilities 
offered in RECLAIM and the OTC Program. RECLAIM requires CEMS on all 
units that burn solid fuels and all units that emit more than 10 tons 
per year, regardless of the type of fuel they burn.. The OTC Program 
requires CEMS on all units that burn solid fuels and all units that do 
not qualify as peaking units, that are larger than 250 mmBtu/hr or that 
serve generators greater than 25 MW. Like RECLAIM and the OTC Program, 
part 75 requires CEMS on all units that burn solid fuel. Part 75 also 
requires the use of CEMS on oil and gas fired units that emit more than 
50 tons of NOX annually (or for units that only report 
during the ozone season, 25 tons of NOX during the ozone 
season), or that don't qualify as peaking units. In both the OTC 
Program and part 75, a peaking unit is defined as a unit that has a 
capacity factor of no more than 10 percent per year averaged over a 
three year period and no more than 20 percent in any one year.
    The EPA believes that these exceptions in part 75 provide cost-
effective monitoring alternatives to CEMs for small, low mass emitting, 
or infrequently used units, and therefore, it is appropriate that part 
96 require all units to use part 75.
     d. Issues Related to Monitoring and Reporting Needed to Support a 
Heat Input Allocation Methodology. For monitoring and reporting 
NOX mass emissions, subpart H of part 75 requires the use a 
NOX concentration CEM and a flow CEM. Since the methodology 
does not require the use of heat input, EPA would not require sources 
to monitor or report heat input or NOX emission rate for a 
NOX mass emission reduction program. If a State elects to 
use a periodically updating allocation methodology that utilizes heat 
input, it may need to require sources using this methodology to monitor 
and report heat input also.

e. Amendments to Part 75 (1) Summary of Part 75 Rulemaking. Title IV of 
the CAA requires the EPA to promulgate regulations for continuous 
emissions monitoring (CEM). On January 11, 1993, final rules (40 CFR 
part 75) were published (58 FR 3590). Technical corrections were 
published on June 23, 1993 (58 FR 34126) and July 30, 1993 (58 FR 
40746). A notice of direct final rulemaking and a notice of interim 
final rulemaking making further changes to the regulations were 
published on May 17, 1995 (60 FR 26510 and 60 FR 26560, respectively). 
Subsequently, on November 20, 1996, a final rule was published in 
response to public comments received on the direct final and interim 
rules (61 FR 59142).

    The EPA proposed further revisions to part 75 on May 21, 1998 (63 
FR 28032). These revisions included a new subpart H which sets forth 
procedures for monitoring NOX mass emissions, which could be 
used by sources to comply with any State or Federal program requiring 
measurement of NOX mass emissions, including the 
requirements

[[Page 57467]]

of the NOX Budget Trading Rule (part 96). The May 21, 1998 
proposed revisions also proposed to make a number of other changes that 
would affect units that were using part 75 to comply either with the 
requirements of title IV or the requirements of a NOX mass 
trading program under title I that incorporated or adopted the 
requirements of part 75. These included a number of minor changes to 
simplify and streamline the rule to make it more efficient for both 
affected facilities and EPA; a new excepted monitoring methodology that 
would reduce monitoring burdens for affected facility units with low 
mass emissions; and new quality assurance requirements to fill in gaps 
identified by EPA during evaluation of the initial implementation of 
Part 75.
    (2) Schedule For Part 75 Final Rulemaking. The comment period for 
the proposed revisions to part 75 ended on July 20, 1998. EPA 
anticipates completing rulemaking on all of proposed revisions to part 
75 by the end of the year. However, because the revisions to subpart H 
of part 75 relating to the monitoring and reporting of NOX 
mass emissions are integral requirements of the SIP Call, EPA is 
finalizing most of the requirements of subpart H of part 75 with 
today's action.
    The EPA is also finalizing a new excepted monitoring methodology 
for units that combust natural gas and or fuel oil with low mass 
emissions of NOX and SO2. These provisions are 
being finalized because they are one of the methodologies that certain 
gas and oil units can use to quantify NOX mass under the new 
subpart H of part 75.
    The EPA is not finalizing the rest of the proposed revisions to 
Part 75 at this time because EPA is still evaluating the comments 
received on the proposed rulemaking. Many of these remaining provisions 
will be applicable to any unit that must use the requirements of part 
75 in order to meet the requirements of title IV or to meet the 
requirements of a State or Federal NOX reduction program 
that adopts the part 75 requirements. For example, the proposed 
revisions would allow a unit with CEMS to be exempt from the 
requirement to perform a linearity test in any quarter that the 
combustion unit for which the CEMs is installed operates for less than 
168 hours. If EPA ultimately finalizes this proposed flexibility, it 
will become available both to units using part 75 to comply with title 
IV and to units using it to comply with the part 96 model trading rule. 
As another example, EPA proposed quality assurance requirements for 
moisture monitors that would be needed if pollutant concentration 
(NOX, SO2 or CO2) were measured on a 
dry basis and needed to be converted to a wet basis so that mass 
emissions could be determined using a stack flow meter. If EPA 
ultimately finalizes this proposed requirement it will affect both 
units using part 75 to comply with title IV and units using it to 
comply with part 96 (or a State or Federal NOX mass 
reduction program that adopts part 75).
    The EPA is also not yet finalizing the recordkeeping and reporting 
requirements associated with either the NOX mass monitoring 
provisions in subpart H or the low mass emitter monitoring methodology 
because EPA believes that these reporting requirements should be 
coordinated with any changes in the reporting requirements that result 
from the finalization of the rest of proposed revisions to part 75.
    Therefore, EPA has closed the part 75 docket (A-97-35, with respect 
to the provisions that are being finalized in today's rulemaking: 
section 75.19, a new excepted methodology for estimating emissions for 
units with low mass emissions; and subpart H, a new subpart setting 
forth provisions for monitoring, recording and reporting NOX 
mass emissions, except where EPA has reserved final action on related 
aspects of these provisions. EPA has not closed the docket with respect 
to the other provisions that were the subject of EPA's, May 21, 1998 
proposal (63 FR 28032).
    (3) Summary of Major Differences Between Proposed and Final 
Revisions to Part 75. The final rule contains two main differences to 
the NOX mass monitoring and reporting provisions from what 
was proposed. The first is that a new methodology for calculating 
NOX mass emissions is included. This methodology utilizes a 
NOX concentration CEM and a flow CEM to calculate 
NOX mass emissions. The second is that sources that are not 
subject to title IV are not required to monitor and report data outside 
of the ozone season unless otherwise required to do so by the 
Administrator or the permitting authority administering the 
NOX mass trading program.
    The final rule also contains two main differences from the proposal 
with regard to the new excepted monitoring methodology for low mass 
emitters. The first is that the methodology is applicable to units with 
calculated NOX mass emissions of up to 50 tons, rather than 
25 tons as proposed. The second is that in lieu of using default rates 
for NOX set forth in the rule, the owner or operator of a 
unit using this methodology may instead elect to determine a unit 
specific rate by conducting stack testing. All of these changes are 
discussed in greater detail in Appendix A of this notice. At this time 
EPA is only addressing the comments dealing with the two main issues 
for which EPA is finalizing revisions to part 75, the reporting of 
NOX Mass (subpart H) and a new excepted monitoring 
methodology for low emitters (Sec. 75.19). The EPA intends to address 
the rest of the comments on the part 75 rulemaking in a separate, 
future rulemaking. The discussions in Appendix A also address comments 
received in the SNPR docket (A-96-56) that related specifically to the 
monitoring requirements set forth in part 75.

E. Emission Limitations/Allowance Allocations

    Each State has the ultimate responsibility for determining the size 
of its trading program budget and its individual source allocations as 
long as the trading budget plus emissions from all other sources do not 
exceed the State's SIP Call budget. The proposed rule published on May 
11, 1998 set timing requirements identifying when the allocations 
should be completed by each State and submitted to EPA for inclusion in 
the NOX Allowance Tracking System (NATS) and provided an 
option specifying how a State might allocate NOX allowances 
to the NOX budget units. Today's final model rule clarifies 
the timing requirements for submission of allowance allocations to EPA 
and provides an optional allocation approach. Each State remains free 
to adopt the Model Rule's allocation approach or adopt an allocation 
scheme of its own provided it meets the specified timing requirements, 
requires new sources to hold allowances, and does not allocate more 
allowances than are available in the State trading budget.
1. Timing Requirements
    In the SNPR, EPA set timing requirements identifying when a State 
would finalize NOX allowance allocations for each control 
period in the NOX Budget Trading Program and submit them to 
EPA for inclusion into the NATS. In developing the proposal, the Agency 
reasoned that uniform timing requirements would be important to ensure 
that all NOX budget units in the trading program would have 
sufficient time and the same amount of time to plan for compliance for 
each control period, and sufficient time and the same amount of time to 
trade NOX allowances. After considering a range of timing 
requirements, EPA proposed options that allocated NOX 
allowances 5

[[Page 57468]]

to 10 years in advance of the applicable control period. The proposal 
attempted to strike a balance between systems that change the 
allocations on an annual basis and systems that establish a single, 
permanent allocation.
    The proposed rule included the following timing requirements for 
the allocation of NOX allowances: by September 30, 1999, 
each participating State would submit NOX allowance 
allocations to EPA for the control periods in the years 2003, 2004, 
2005, 2006, and 2007. After the initial allocation, two timing 
requirements were proposed for allocations following the year 2007. The 
option set forth in the proposed Model Rule would require a State to 
submit allocations to EPA for the control period in the year that is 5 
years after the applicable submission deadline. For example, by January 
1, 2003 each State participating in the trading program would issue its 
allocations for the control period in 2008. The State would issue 
allocations for the 2009 summer season by January 1, 2004. The second 
option, discussed in the preamble of the supplemental notice, would 
require the State to submit five years' worth of allowance allocations 
at a time, every five years, starting in 2003. For example, by January 
1 , 2003, each State participating in the trading program would issue 
allocations for the control periods in the years 2008 through 2012. The 
supplemental notice solicited comment on these timing options as well 
as the full range of possible timing requirements (including a single, 
permanent allocation system and an annually changing allocation 
system). The supplemental notice also solicited comment on a provision 
requiring EPA to allocate NOX allowances to NOX 
budget units if a State were to fail to meet the timing requirements.
    Comments: Although comments covered the entire range of possible 
timing requirements, commenters generally supported striving for 
administrative simplicity and ensuring sufficient planning horizons for 
affected sources, while still addressing the needs of a changing 
marketplace. Most comments fell into one of five categories.
    First, a few commenters favored the option set forth in the 
proposed Model Rule that would update the allocations each year, five 
years in advance of the applicable control period. However, most of 
these commenters also supported a system which would update the 
allocations less than five years prior to the applicable control period 
as that would allow more recent data to be used in the allocations. One 
commenter advocated allocating for the previous season based on current 
year data (i.e., allocations would be issued at the end of the season 
for the preceding control period).
    Approximately ten commenters favored the approach which would issue 
allowances five to ten years in advance. This group found that five to 
ten years of allocations satisfies the desire to have a sufficient 
planning horizon while still ensuring responsiveness to changing market 
conditions. Utilities generally opposed allocating single year 
allowances as it might be disruptive to utility planning.
    The third category of commenters advocated longer term or permanent 
allocations. Most utility and business commenters favored allocations 
that were issued in ten year blocks at a minimum to provide sufficient 
time to plan future activities and amortize investments. A report 
submitted by a State proposed that allocations extend over the capital 
life of equipment, which was at least ten years.
    A fourth set of commenters, which included three States, favored 
shorter term allocations. These States commented that they may want to 
base their allocations on more recent data than that proposed by the 
Model Rule and suggested that three years would provide sufficient 
planning time for sources. One State suggested tying allocations to the 
submission of triennial inventories.
    A final group of commenters suggested that no timing requirement 
was necessary. They suggested that just as sources may participate in 
an interstate trading program with allocations based upon different 
methodologies, those same sources may participate in such a program 
even if they receive their allowances at different times or for 
different periods.
    Several State commenters asserted that September 1999 was too early 
to have allocations set. These States suggested that the allocation 
process is difficult and takes longer than one year. One State 
suggested that the early allocation deadline would effectively prevent 
States from issuing allowances based upon output for the first period 
because an output approach could not be developed in time.
    Response: Most commenters supported issuing allowances at least a 
couple of years prior to the season in which they would be used. The 
commenters generally cited the goal of balancing changing market 
conditions with providing sufficient planning horizons, as had the 
Agency in the proposal. The EPA agrees that the certainty in having 
allowances at least a couple of years into the future would provide 
some predictability for sources in their control planning and build 
confidence in the market. Most of the State commenters suggested three 
years prior to the control season as an adequate length of time for 
sources to know their allocations. The Agency agrees that a trading 
system could work with sources knowing their allocations three years 
prior to the control season. Therefore, EPA has modified its original 
proposal to ensure that sources would always have allowances at least 
three years in advance of the use date.
    In addition to addressing how many years in advance the allocations 
are determined, the Agency has also considered whether allocations 
should be issued one control period at a time or for multiple control 
periods at a time (e.g., five to ten control periods). In response to 
the comments received, the Agency has determined that it would be 
appropriate to set minimum timing requirements rather than prescribing 
a set length of time for all States. Therefore, the Agency is now 
requiring States choosing to participate in the NOX Budget 
Trading Program to allocate a minimum of one summer season of 
allowances at a time (at least three years in advance of the applicable 
control period).
    Moving from requiring five summer seasons of allocations (three 
years in advance of the first season) to one summer season of 
allocations (three years in advance) has the advantage of allowing the 
allocation system to be updated sooner with more recent data. This 
would provide those States that want to use updating systems to more 
fully avail themselves of an updating system. The system could also 
incorporate new sources more quickly, thus reducing the need for larger 
new source set-asides.
    However, the Agency has determined that a State may decide to issue 
allowances further into the future than the one-season minimum period 
required by this final rule and still receive streamlined EPA review of 
its trading program. The NOX Allowance Tracking System will 
be able to handle allocations for longer periods. Therefore, this Final 
Rule sets out minimum timing requirements of one season (three years in 
advance), but States may issue allocations in larger blocks for as many 
as 30 seasons into the future and still receive streamlined EPA review. 
However, in determining the length of time for which a State issues 
allocations, a State should consider any potential adjustments that may 
occur to its future State budgets. For example, as stated in Section 
III.B.5.

[[Page 57469]]

of this preamble, the Agency may establish new budget levels for the 
post-2007 timeframe. States issuing long-term allocations should 
address how the allocations would be adjusted if new budget levels are 
established in the future. The Agency does believe that having 
allocations three years prior to the relevant control period would be 
the minimum needed to support an active multi-state trading market 
intended to reduce compliance costs for all States involved.
    The three-year minimum timing requirement also is compatible with 
beginning the program in 2003, with at least the first year's 
allocations submitted to EPA by September 30, 1999. Sources will know 
their first year's allocations three years prior to the start of the 
program, and by April 1, 2003, all sources will have allocations for at 
least four seasons--2003, 2004, 2005 and 2006. The Agency maintains 
that the first year's allowances should be issued by September 30, 1999 
to provide some predictability for sources in their control planning 
and build confidence in the market. It also ties in with the State's 
SIP submittal deadlines. For States participating in the trading 
program, the allowances are an integral part of the State's plan to 
satisfy the requirements of this SIP call. For sources in the Trading 
Program, the allowances are the mechanism by which State budget 
requirements are translated into source-specific limitations, and 
therefore the allocations should be submitted with the SIP submittals. 
In response to States who are worried about completing allocations in 
this time frame, EPA notes that one State in the OTC resolved its 
allocations in six weeks, demonstrating that it is possible to 
establish allocations in less than one year.
    Requiring only one year's worth of allowances at a time has the 
added benefit of being able to more quickly accommodate States that 
want to switch allocation methodologies after the start of the program. 
For example, a State may decide to issue its initial allocations based 
on heat input data because it has not yet finalized an approach to 
issuing output-based allocations. The State could take a few additional 
years to refine the alternative approach to issuing allowances. When 
the State is ready to adopt the output approach, the State would be 
able to start using the new approach much sooner than it would be able 
to under a system that issued allocations in larger blocks.
    Therefore, this preamble sets the following timing requirements for 
the allocation of NOX allowances which will be able to 
accommodate States that want to issue allocations one year at a time as 
well as States that would like to issue allocations in larger blocks: 
by September 30, 1999, the State would submit NOX allowance 
allocations to EPA for at least the control period of 2003. After this 
initial allocation, by April 1 of every year starting in 2001, the 
State must, at a minimum, submit allowance allocations to EPA for the 
control period in the year that is three years after the applicable 
submission deadline. For example, by April 1, 2001, a State would 
submit allocations for the control period in 2004. By April 1, 2002, a 
State would submit allocations for the control period in 2005. This 
minimum requirement would allow a State to submit blocks of allowances 
that represent any number of years should the State prefer to do so. 
For example, by the September 30, 1999 deadline, a State could submit 
allocations for only the 2003 control period or for multiple control 
periods (e.g., the five control periods of 2003-2007). The SIP would 
provide that if the State fails to submit allocations by the required 
date, EPA would allocate allowances based on the previous year's 
allocation within 60 days of the applicable deadline. This approach 
would ensure that starting in 2003, all sources would always have at 
least three years of allowances in their accounts.
    Today's Model Rule presents an allocation approach that satisfies 
the minimum timing requirements. However, the initial allocation is for 
three control periods (2003-2005) because this would avoid updating 
allocations on an input basis. Any variation on the following approach 
would be acceptable providing it satisfies the minimum requirements 
specified in the previous paragraph. After this initial allocation, the 
model rule would have the State submit allowance allocations to EPA for 
the control period in the year that is three years after the applicable 
submission deadline. By April 1, 2003, a State would submit allocations 
for the control period in 2006. By April 1, 2004, a State would submit 
allocations for the control period in 2007, and so forth.
2. Options for NOX Allowance Allocation Methodology
    The Agency proposed that the NOX Budget Trading Rule 
include a recommended NOX allowance allocation methodology. 
The proposed Model Rule laid out an example of an allocation 
methodology using heat input data for source allocations. The preamble 
to the proposed Model Rule solicited comment on this methodology as 
well as two additional options using either input or output data for 
determining allocations. The first alternative to using heat input 
would base the allocation recommendation on heat input data for the 
first five control periods of the trading program and then convert the 
allocations to an output basis for the control periods after 2007. The 
final option would base the allocation recommendation on output data 
for all NOX Budget units from the start of the trading 
program. The Agency also solicited comment on a suggested schedule for 
establishing a method for output-based allocations, and on any 
technical or data issues relevant to output-based allocations, as well 
as on the use of a fuel-neutral or output-neutral calculation to 
determine allocations for NOX Budget units.
    Comments: The Agency received numerous comments on the issue of 
whether to suggest an allocation recommendation to States. 
Approximately 25 commenters suggested that no recommendation is 
necessary. Many of these commenters emphasized that EPA had no 
authority to prescribe an allowance allocation methodology and a 
recommendation could be misinterpreted as a requirement for SIP 
approval. Several commenters requested that EPA clarify that the SIP 
approval process will be consistently applied to all States regardless 
of the allocation method chosen by a State, as long as the total 
allocation does not exceed a State's trading budget. Approximately half 
of the commenters who stated that no recommendation was necessary 
suggested that if EPA were going to make a recommendation, the 
recommendation should be a heat input approach.
    Close to fifty commenters suggested that an Agency recommendation 
was a good idea, but they were divided on the appropriate methodology. 
This group included all the State commenters who suggested that a 
recommended approach was appropriate for use as a default allocation 
mechanism by States that did not determine their own allocations.
    Many commenters supported the heat input approach used in the 
example in the supplemental notice. Two State commenters said that the 
proposed example approach was a useful default for States that did not 
come up with their own allocations. Other commenters suggested that 
heat input is an easily understood metric for all sources and the data 
is readily available.
    However, many suggested that EPA should recommend an output method 
because they believe output-based allocations tend to reward more 
efficient

[[Page 57470]]

fuels over fuels that require a higher heat input to generate the same 
amount of electricity. Other reasons cited for output-based allocations 
include the incentive that updating output allocations provides for 
reducing emissions of pollutants such as CO2 and mercury. 
Several commenters suggested that output-based allocations would allow 
the environmental goals of the program to be achieved more cost-
effectively; their arguments rested upon assertions that issuing 
allowances to non-NOX emitting units in an output-based 
system would reduce the need for NOX controls over time. One 
State commenter said that an output approach was the consensus of 
participants at EPA Workshops held prior to drafting of the 
Supplemental Notice and therefore should be the recommended approach 
suggested by EPA.
    One commenter had a specific recommendation for an updating output-
based allocation system which would issue allowances each year for the 
current control period. Administrative simplicity, economic efficiency, 
incentives for innovation, and lower consumer impact were cited as 
reasons supporting that position.
    Additional commenters favored the output-based approach but only 
for fossil-fuel fired sources and renewables. Several commenters 
submitted letters opposing a ``fuel-neutral'' policy and objected to 
including nuclear sources in an output allocation to sources. They 
stated that a fuel neutral policy would provide incentives for nuclear 
generation which has the potential to release small amounts of 
radiation to the environment as well as the potential for generation of 
high-and low-level radioactive waste.
    Response: As was stated in the SNPR, EPA believes that it is 
important for as many States as possible to participate in the 
NOX Budget Trading Program. The Agency recognizes that 
States have unanimously favored flexibility in developing their own 
allocation methodologies. Further, the comments that EPA received in 
response to the SNPR (as well as in response to the workshops held 
prior to publication of the SNPR) provided no clear consensus for one 
methodology over another.
    However, the Agency believes it is important to provide a model 
allocation methodology that States may choose to use as a guide for 
their own allocation process. Several States have commented that 
including an example method in the Model Rule would be useful as a 
backup for States who do not come up with an alternative method of 
allocation. An outlined approach in the Model Rule may also facilitate 
the regulatory process within a State that wants to quickly adopt the 
Model Rule.
    Therefore, today's Model Rule includes an optional allocation 
methodology. The Agency has carefully considered arguments for 
alternative allocation methods. The EPA would support a decision by a 
State to use either heat input or output data as a basis for source 
allocations or for the State to auction some or all of its allocation. 
In determining the basis for the methodology presented in today's Model 
Rule, EPA has decided to use the heat input approach because it is 
concerned that an output-based approach has not been fully developed or 
made available for public comment. Further, before issuing a model 
output-based allocation approach, the Agency would need to make several 
revisions to current reporting and monitoring provisions. EPA would 
have to revise part 75 to monitor and report temperature, pressure, and 
steam heat output (mmBtu) for units with some or all of their output as 
heated steam. EPA would also need to put in place procedures which take 
advantage of the most accurate data possible. For example, the Energy 
Information Administration (EIA) solicited comment in a July 17, 1998 
Federal Register Notice on a proposal to make electricity generating 
data non-confidential and publicly available from non-utility 
electricity generators (63 FR 38620). EPA will not know if this 
information is available to the Agency or to States through EIA for 
some time. If EIA were to decide that this information should remain 
confidential in the future, then EPA and States would need to collect 
their own data from sources. Additionally, the Agency is currently 
unaware of any public databases of output information besides those for 
electrical generation output for certain electrical generating units. 
Output information would only become available if sources report it 
directly to the Agency or to States.
    While today's final Model Rule includes a heat input approach, the 
Agency is continuing to work on developing an updating output approach 
to source allocations. For States that wish to use output in developing 
their source allocations and are willing to wait for EPA to finalize 
such an approach, EPA plans to issue a proposed system for output-based 
allocations in 1999 and finalize an output-based option in 2000. 
However, the Agency's ability to issue an output-based approach on this 
schedule is contingent upon resolving the issues and promulgating the 
necessary rule changes mentioned in the previous paragraph.
    Assuming EPA finalizes an output-based option in early 2000, States 
wishing to use this output-based system could adopt the necessary 
rules, and output data could be measured and collected at 
NOX budget units during the control periods in the years 
2001 and 2002. Output data could then be available for States to 
calculate allocations for the control periods starting in 2006. Heat-
input-based allocations could be used for the 2003 through 2005 control 
seasons.
    However, this does not prohibit a State from developing its own 
output-based system on a faster timeline. For example, if a State has 
developed an output-based approach for use in its initial allocations, 
it may use that approach. Or, the State may issue its initial 
allocation for 2003 using heat input data and then by April 1, 2001 
issue output allocations for the control periods starting in 2004.
    The Agency recognizes that a State's choice of when and for what 
blocks of time it issues allocations is intertwined with the choice of 
allocation methodology. Several commenters suggested that more 
incentives for generation efficiency and therefore ancillary 
environmental benefits (CO2 and mercury reductions) are 
provided in an output system with periodic updates, and those 
incentives are lost in an heat input system that is periodically 
updated. These commenters suggested that with a heat-input-based 
system, States should issue permanent allocations rather than updating 
the allocations. An allocation system that issues permanent streams of 
allowances (using either a heat input or an output methodology) would 
still provide an incentive for generation efficiency although perhaps 
not to the extent that an updating output system might. However, if a 
State issues a permanent stream of allowances to existing sources, that 
State would have to decide how to address new sources (options include 
establishing an allocation set aside or an auction, or requiring new 
sources to obtain allowances from existing sources).
3. New Source Set-Aside
    The Agency proposed an allocation set-aside account equaling 2 
percent of the State trading program budget for each control period for 
new NOx Budget units as part of its recommended allocation 
approach. The concept and size of the set-aside is included only as an 
optional feature of the Model Rule; however, the Model Rule requires 
new sources to hold allowances to cover

[[Page 57471]]

their emissions. The supplemental notice proposed that allowances from 
the set-aside be given out on a first-come, first-served basis at an 
emission rate of 0.15 lb/mmBtu multiplied by a budget unit's maximum 
design heat input. The source would then be subject to a reduced 
utilization calculation so that a reduction in the emission rate below 
0.15 lb/mmBtu would be rewarded, but a reduction in utilization would 
not. In other words, EPA would deduct NOx allowances 
following each control period based on the unit's actual utilization 
for the control period. After the deduction, the allocation that had 
been granted to the new unit from the set-aside would equal the product 
of 0.15 lb/mmBtu and the budget unit's actual heat input for the 
season. EPA solicited comments on the use of a set-aside as part of the 
recommended allocation methodology as well as the proposed size and 
operation of the set-aside.
    Comments: The Agency received many comments regarding the proposal 
for a new source set-aside. While several commenters were opposed to a 
new source set-aside because it might bias control decisions in favor 
of adding new sources relative to controlling existing sources, 
numerous other commenters expressed general support for accommodating 
new sources with allowances.
    Several of these commenters offered suggestions for how the set-
aside should be designed. A few commenters stated that the size of the 
set-aside should be related to the timing requirements and noted that 
shorter timing requirements make it easier to accommodate new growth. 
One commenter who advocated annually updating the allocation system 
noted that its proposal would eliminate the need for a new source set-
aside. Some commenters supported the set-aside concept but asserted 
that States should be able to decide the correct size. Other commenters 
agreed with the set-aside concept in theory but did not think the 
allowances should come from existing sources.
    Additional commenters had specific proposals for the size of the 
set-aside. One commenter suggested that the size of the set-aside 
should reflect the actual growth projected in budget calculations and 
that the unused portion of the set-aside should be retired. A few 
commenters agreed with the proposed 2 percent size.
    Several commenters offered suggestions on how to issue the set-
aside allowances to new sources. One commenter suggested that the 
allowances should be given to new sources at the actual emission rate 
if it was below the proposed 0.15 lb/mmBtu level.
    Finally, several commenters suggested that the concept of a set-
aside was an issue that should be left completely up to the States.
    Response: The Agency believes that a new source set-aside should be 
large enough to provide all new units entering the trading program with 
allocations. The Agency maintains that as much as possible within the 
context of the overall trading budget, allocations should be provided 
to new sources on the same basis as that used for existing units until 
the time when the new sources receive an allocation as part of an 
updating allocation system. Therefore, the Agency continues to include 
a new source set-aside as part of its optional allocation methodology 
described in the Model Rule. The EPA proposed the 2 percent set-aside 
in the SNPR after looking at the amount of growth from new sources 
projected by the Integrated Planning Model (and used in the budget 
determinations) and estimating how much growth could be expected over 
the five year period that new sources might have to wait before 
receiving an allocation. In light of the allocation methodology and 
timing specified in today's Model Rule as well as revisions made to the 
growth factors used in State budget determinations since the SNPR, the 
Agency has re-evaluated the size of the new source set-aside proposal. 
The revised Integrated Planning Model projects approximately \1/2\ 
percent annual growth in capacity utilization for new sources. Given 
the timing and optional allocation methodology specified in today's 
Model Rule, the 2003, 2004, and 2005 set-aside would need to 
accommodate any source that started operating after May 1, 1995. 
Assuming the \1/2\ percent growth rate projected by IPM, the Agency 
finds that a 5 percent set-aside should be large enough to accommodate 
all new sources for the 2003, 2004, and 2005 control seasons.
    After 2005, the new source set-aside would need to accommodate any 
source that commenced operation after May 1 of the control period three 
years prior to the control period in which the set-aside would be 
available. For example, in 2006, the set-aside should be large enough 
to accommodate any source that commenced operation after May 1, 2003. 
Assuming the growth rates predicted by the IPM, the Agency finds that a 
2 percent set-aside should be large enough to accommodate new source 
growth after May 1, 2003.
    A 5 percent set-aside provision for the first three control seasons 
and 2 percent for the control periods starting in 2006 is incorporated 
into today's Model Rule as an option States may adopt. However, States 
may choose to handle new sources in any way as long as the emissions 
from new sources are subject to the overall State budget. For example, 
some States may choose to issue allowances for longer periods of time 
than that outlined as the minimum requirement in today's Model Rule. 
These States may find that a 5 percent set-aside is not sufficient to 
accommodate all their new source growth, and may want to consider a 
larger set-aside or alternative means to accommodate new sources. Or, 
States may decide to allocate allowances based on a new source's 
permitted or actual emissions, which may be lower than 0.15 lb/mmBtu. 
This would require a smaller set-aside.
    In the model rule set-aside provision, allowances will be issued to 
new sources on a first-come, first-served basis. Allowances that are 
not issued to new sources in the applicable control period will be 
returned to the existing sources in the State on a pro-rata basis to 
guard against the possibility of a disproportionately large set-aside.
    The EPA maintains its position that new sources should receive 
allowances at the same rate as that applied to existing sources (i.e., 
large electric generating units would receive allowances at a 0.15 lb/
mmBtu rate, large non-electric generating units would receive 
allowances at the average emission rate for existing large non-electric 
generating units after controls are in place, as explained in section 4 
below). However, to reinforce the flexibility available on these 
issues, as long as a State requires new sources to hold allowances, the 
Agency reiterates that States may have any size set-aside (including 
zero), may allocate the set-aside in whatever manner they choose, and 
may carry over from one year to the next any amount of allowances 
(subject to the banking provisions on this SIP call). If a State 
decides to return unused allowances from a new source set-aside to 
existing sources, the State would indicate to EPA (as the administrator 
of the allowance tracking system) what number of allowances should be 
returned to which existing units.
4. Optional NOX Allocation Methodology in Model Rule
    While specific source allocations are required for States 
participating in the NOX Budget Trading Program, the 
allocation methodology presented here is an optional approach that may 
be adopted by States. As long as a State (1) does not allocate more 
allowances than are available in the State NOX trading

[[Page 57472]]

budget, (2) requires new sources to hold allowances, and (3) issues 
allocations on a schedule that meets the minimum timing requirements, 
the State may adopt whatever methodology it finds the most appropriate 
and still qualify for inclusion in the NOX Budget Trading 
Program.
    The Model Rule contains the following optional allocation 
methodology. It differs from the approach presented in the proposed 
rule on the timing provisions, the allocation methodology for non-
electric generating units, and the size of the optional new source set-
aside. As proposed in the SNPR, initial unadjusted allocations to 
existing NOX Budget units serving electric generators would 
be based on actual heat input data (in mmBtu) for the units multiplied 
by an emission rate of 0.15 lb/mmBtu. For the control periods in 2003, 
2004, and 2005, the heat input used in the allocation calculation for 
large electric generating units equals the average of the heat input 
for the two highest control periods for the years 1995, 1996, and 1997. 
Once the State completes the initial allocation calculation for all the 
existing NOX budget units serving electric generators for 
2003, 2004, and 2005, the State would adjust the allocation for each 
unit upward or downward so that the total allocations match the 
aggregate emission levels apportioned by an approved SIP to the State's 
NOX Budget units serving electric generators. Then, the 
State would adjust the allocation for each unit proportionately so that 
the total allocation equals 95 percent of the aggregate emission levels 
apportioned to the State's NOX Budget units serving electric 
generators (to provide for the 5 percent new source set-aside). A State 
would submit the 2003, 2004, and 2005 allocations to EPA by September 
30, 1999.
    For the control periods starting in 2006, the heat input used in 
the allocation calculation for large electric generating units equals 
the heat input measured during the control period of the year that is 
four years before the year for which the allocations are being 
calculated. Once the State completes the initial allocation calculation 
for all existing budget units, and the State adjusts the allocations to 
match the aggregate emission levels apportioned to NOX 
Budget units serving electric generators, the State would adjust the 
allocation for each unit proportionately so that the total allocation 
equals 98 percent of the aggregate emission levels apportioned to 
NOX Budget units serving electric generators (to provide for 
the 2 percent new source set-aside).
    For reasons explained elsewhere in today's rulemaking, EPA 
determined the aggregate emission levels for large non-electric 
generating units in each State budget based upon a 60 percent reduction 
rather than the 70 percent proposed in the SNPR. The 60 percent 
reduction results in an average emission rate across the region of 0.17 
lbs/mmBtu for large non-electric generating units. Therefore, initial 
unadjusted allocations to existing large non-electric generating units 
would be based on actual heat input data (in mmBtu) for the units 
multiplied by an emission rate of 0.17 lb/mmBtu. For non-electric 
generating units subject to the trading program, 1995 heat input data 
is used in the allocation calculation for the control periods 2003, 
2004, and 2005 (1995 is the most recent data the Agency knows is 
currently available for non-electric generating units). Once the State 
completes the initial allocation calculation for all the existing large 
non-electric generating units for 2003, 2004, and 2005, the State would 
adjust the allocation for each unit upward or downward so that the 
total allocations match the aggregate emission levels apportioned by an 
approved SIP to the State's large non-electric generating units. Then, 
the State would adjust the allocation for each unit proportionately so 
that the total allocation equals 95 percent of the aggregate emission 
levels apportioned to the State's large non-electric generating units 
(to provide for the 5% new source set-aside). A State would submit the 
2003, 2004, and 2005 allocations to EPA by September 30, 1999.
    For the control periods starting in 2006, the heat input used in 
the allocation calculation equals the heat input measured during the 
control period of the year that is four years before the year for which 
the allocations are being calculated. Once the State completes the 
initial allocation calculation for all existing budget units, and the 
State adjusts the allocations to match the aggregate emission levels 
apportioned to large non-electric generating units, the State would 
adjust the allocation for each unit proportionately so that the total 
allocation equals 98 percent of the aggregate emission levels 
apportioned to large non-electric generating units (to provide for the 
2% new source set-aside).
    A State would establish a separate allocation set-aside for new 
units each control period. Five percent of the seasonal trading budget 
will be held in a set-aside account for the control periods in 2003, 
2004, and 2005. At the end of the relevant control period, the State 
would submit a NOX allowance transfer request to EPA to 
return any allowances remaining in the account to the existing sources 
in the State on a pro-rata basis.
    The allowances would be issued to new sources on a first-come 
first-served basis at a rate of 0.15 lb/mmBtu for NOX Budget 
units serving electric generators and 0.17 lb/mmBtu for large non-
electric generating units multiplied by the budget unit's maximum 
design heat input. Following each control period, the source would be 
subject to a reduced utilization calculation, in which EPA would deduct 
NOX allowances based on the unit's actual utilization. 
Because the allocation for a new unit from the set-aside is based on 
maximum design heat input, this procedure adjusts the allocation by 
actual heat input for the control period of the allocation. This 
adjustment is a surrogate for the use of actual utilization in a prior 
baseline period which is the approach used for allocating 
NOX allowances to existing units.

F. Banking Provisions

    As explained in Section III.F.7., EPA requested comment in the SNPR 
on whether and how banking should be incorporated into the design of 
the NOX Budget Trading Program. Banking may generally be 
defined as allowing sources that make emissions reductions beyond 
current requirements to save and use these excess reductions to exceed 
requirements in a later time period. Options ranged from a program 
without banking to several variations of a program with banking, prior 
to and/or following the start of the program. The EPA also requested 
comment on options for managing the use of banked allowances in order 
to limit the emissions variability associated with banking. The EPA 
specifically proposed using a ``flow control'' mechanism in cases where 
the potential exists for a large amount of banked allowances to be 
available.
    This section addresses how banking has been incorporated into the 
NOX Budget Trading Program based on the criteria set forth 
in the NOX SIP call.
1. Banking Starting in 2003
    In accordance with the provisions discussed in III.F.7.a., trading 
programs used to comply with the NOX SIP call may allow 
banking to start in the first control period of the program, the 2003 
ozone season. The majority of commenters supported banking in the 
context of the NOX Budget Trading Program. Based on the 
advantages that banking can provide, as discussed in the SNPR and the 
comments, the NOX

[[Page 57473]]

Budget Trading Program has been designed to allow banking starting in 
the first control period of the trading program. NOX Budget 
units that hold additional NOX allowances beyond what is 
required to demonstrate compliance for a given control period may 
carry-over those allowances to the next control period. These banked 
allowances may be used or sold for compliance in future control 
periods.
2. Management of Banked Allowances
    The NOX SIP call establishes that a flow control 
mechanism be paired with any banking provisions to limit the potential 
for emissions to be significantly higher than budgeted levels because 
of banking. This mechanism allows unlimited banking of allowances saved 
through emissions reductions by sources, but discourages the 
``excessive use'' of banked allowances by establishing either an 
absolute limit on the number of banked allowances that can be used each 
season or a rate discounting the use of banked allowances over a given 
level. In the SNPR, EPA solicited comment on the application of flow 
control in the NOX Budget Trading Program. Although many 
commenters were opposed to any restrictions on the use of banked 
allowances, several commenters stated that if restrictions were to be 
imposed, they would favor flow control as the most cost-effective, 
least rigid means of management. A few commenters added that, if 
implemented, flow control should be applied on a source-by-source basis 
so as to avoid penalizing all of the participants in the trading 
program for the excess banking of individual participants. One 
commenter stated that if EPA concludes that there is an adequate basis 
for imposing some type of restriction, it should avoid placing any 
absolute limit on the amount of banked allowances that can be used in a 
given season.
    The NOX SIP call established that flow control should be 
set at the 10 percent level. The effect of setting flow control at 10 
percent of the trading program budget is that on a season-by-season 
basis, sources may use banked allowances or credits for compliance 
without restrictions in an amount up to 10 percent of the 
NOX budget for those sources in the trading program. Banked 
allowances or credits that are used in an amount greater than 10 
percent of the NOX budget for those sources will have 
restrictions on their use.
    The following provides a brief description of exactly how the flow 
control mechanism will operate in the NOX Budget Trading 
Program. The number of banked allowances held by all participants in 
the multi-state trading program will be tabulated each year following 
the compliance certification process to determine what percentage 
banked allowances are of the overall multi-state trading budget for the 
next year. If this percentage is equal to or below 10 percent, all 
banked allowances may be used in the upcoming control season on a one 
allowance for one ton basis. If this percentage is greater than 10 
percent, flow control will be triggered. In years when flow control is 
triggered, a withdrawal ratio will be established prior to the control 
period for which it would apply. The withdrawal ratio will be 
calculated by dividing 10 percent of the total trading program budget 
by the total number of banked allowances. This ratio will be applied to 
each compliance or overdraft account (only accounts used for 
compliance) holding banked allowances as of the allowance transfer 
deadline at the end of the control period for which it applies. Banked 
allowances in each account may be used for compliance on a one-for-one 
basis in an amount not exceeding the amount established by the 
withdrawal ratio. Banked allowances used in an amount exceeding that 
established by the withdrawal ratio must be used on a two-for-one 
basis. By setting the withdrawal ratio prior to the applicable control 
period (in years flow control is triggered) and applying it at the time 
of compliance certification at the end of the applicable control 
period, sources have one full control period to incorporate the value 
of using banked allowances into their operations.
    As described above, the NOX Budget Trading Program 
applies the flow control mechanism on a regional basis and establishes 
a 2-for-1 discount for banked allowances that are used in an amount 
greater than the flow control limit. The regional approach for applying 
flow control was selected over the source-by-source approach for the 
following reasons:
     EPA believes this option provides more flexibility to 
individual sources than the source-by-source approach. If the 10 
percent limit were placed on each source based on the source's 
allocation, the limit would be in effect every year for every source, 
even when the amount of banked allowances throughout the entire trading 
region was below 10 percent of the regional trading budget. In 
contrast, the regional approach only applies flow control when the 
amount of banked allowances throughout the region (entire multi-state 
trading area) exceeds the 10 percent limit. In response to the 
commenter suggesting that the regional approach penalizes all 
participants in the trading program for the excess banking of 
individual participants, EPA notes that it would be difficult for a few 
sources to cause the entire regional bank to exceed 10 percent of the 
budget. In addition, based on the analyses presented in the RIA, EPA 
does not anticipate that flow control is likely to be triggered. 
Consequently, flow control is more of an insurance policy, rather than 
a provision that is routinely expected to be operational.
     The regional approach also provides flexibility to sources 
if and when it is triggered. Because the withdrawal ratio is set before 
the applicable control period but not applied until the control 
period's allowance transfer deadline, sources have over seven months to 
manage the amount of banked allowances they use on a 1-for-1 basis 
versus a 2-for-1 basis.
     EPA believes the regional approach is also a more 
universal approach than the source-by-source approach under a variety 
of allocation programs that States may use in the NOX Budget 
Trading Program. To apply the flow control mechanism on a source-by-
source basis, the 10 percent limit would be applied to each source's 
allocation. In this way, a source could use an amount of banked 
allowances up to 10 percent of it's allocation without restrictions. 
Restrictions would be placed on banked allowances that the source uses 
in an amount greater than 10 percent of its allocation. Under certain 
allocation programs, States may choose not to allocate NOX 
allowances to new sources and require that these sources obtain the 
necessary amount of NOX allowances for compliance from the 
market. By not having an allocation of NOX allowances, new 
sources would be prevented from using banked allowances under the 
source-by-source approach. EPA believes that approaches to accommodate 
sources without a fixed allocation under the source-by-source flow 
control approach would overly complicate the system.
     The regional approach for applying flow control is also 
the approach used in the Ozone Transport Commission's (OTC) trading 
program. Because the NOX Budget Trading Program is designed 
to include States currently operating in the OTC program, using the 
same approach for flow control will minimize the disruption for these 
sources to convert to the NOX Budget Trading Program.
    The other issue for flow control is the type of restriction to 
place on banked allowances used in an amount greater than the 10 
percent limit. The NOX Budget Trading Program includes the 
2-for-1 discount as the applicable

[[Page 57474]]

restriction. EPA agrees with the commenters that favored this approach 
over using an absolute limit. The EPA believes the 2-for-1 discount 
provides more flexibility for sources to achieve compliance than is 
offered by the absolute limit. The discount is also beneficial to the 
environment, when triggered, by allowing only one ton of NOX 
emissions for every two tons removed. Additionally, the OTC program 
uses the 2-for-1 discount.
    The following example illustrates how flow control will be used. 
For the year 2006, assume the total trading program budget across all 
States equals 300,000 allowances and 35,000 allowances are banked from 
control periods prior to the 2006 control period. Since more than 10 
percent (35,000/300,000 = 11.7%) of the total trading program budget is 
banked, a withdrawal ratio will be established prior to the 2006 
control period and will apply to all compliance and overdraft accounts 
(only accounts that may be used for compliance) holding banked 
allowances at the end of the 2006 control period. In this case, the 
withdrawal ratio would be 0.86 (determined by dividing 10 percent of 
the total trading program budget by the total number of banked 
allowances, or 30,000/35,000). Thus if a source holds 1,000 banked 
allowances at the end of the 2006 control period, it will be able to 
use 860 on a 1-for-1 basis, but will have to use the remaining 140, if 
necessary, on a 2-for-1 basis. As a result, if the source used all its 
banked allowances for compliance in the 2006 control period, the 1,000 
banked allowances could be used to cover only 930 tons of 
NOX emissions (860 + 140/2). Of course, a source could buy 
additional current year allowances to cover emissions on a 1-for-1 
basis or buy additional banked allowances (allowances not needed by 
other sources for compliance) to increase the amount of banked 
allowances it may use on a 1-for-1 basis.
3. Early Reduction Credits
    As described in section III.F.7.c., the majority of commenters 
generally supported the option of awarding early reduction credits. EPA 
is allowing, but not requiring, States to grant early reduction credits 
to sources for reductions in ozone season NOX emissions 
prior to the 2003 ozone season. States may issue early reduction 
credits in an amount not exceeding the State's compliance supplement 
pool. The compliance supplement pool is further explained in section 
III.F.6.
    Based on the support the commenters on the NOX Budget 
Trading Program expressed for early reduction credits, EPA is including 
optional provisions in the trading program that States may use for 
issuing credits. States participating in the NOX Budget 
Trading Program that choose to issue early reduction credits may follow 
the methodology included in part 96 or may develop their own 
methodology, provided the State's program meets the following 
requirements. The State program must ensure that early reduction 
credits will not be issued in an amount exceeding the State's 
compliance supplement pool. The State program must also meet the 
criteria for early reduction credits discussed in section III.F.7.c. 
Finally, the State should notify EPA of the amount of credits issued to 
particular NOX Budget units by no later than May 1, 2003. 
Early reduction credits shall be issued to units as allowances for the 
2003 control period. For purposes of the banking provisions, the 
allowances will not be considered banked in the 2003 control period. 
However, any unused allowances carried from the 2003 control period to 
the 2004 control period shall be considered banked as will be the case 
for all unused allowances carried over to the next control period. Per 
the requirements discussed in section III.F.7.c., allowances issued for 
early reduction credits may be used for compliance by sources in the 
2003 and 2004 control periods. Any of these allowances that are not 
used for compliance in the 2003 or 2004 control periods shall be 
retired by EPA from the account in which they are held.
    As discussed in Section III.F.6.b.ii., States also have the option 
of issuing some or all of the State's compliance supplement pool 
directly to sources according to the criteria for direct distribution. 
Consequently, States participating in the NOX Budget Trading 
Program may also use the direct distribution option for issuing the 
compliance supplement pool. In this case, the State must notify EPA by 
May 1, 2003 of the specific NOX Budget units that will be 
receiving the direct distribution.
4. Optional Methodology for Issuing Early Reduction Credits
    The methodology described below is an optional methodology included 
in part 96 that States participating in the NOX budget 
Trading Program and choosing to issue early reduction credits may 
follow. States participating in the NOX Budget Trading 
Program may also choose to develop their own methodology as discussed 
above. The following methodology is designed to meet the criteria for 
issuing early reduction credits discussed in section III.F.7.c. and to 
provide incentives for a State's NOX budget units to 
generate early credits in an amount no greater than the size of the 
State's compliance supplement pool. The State may choose to issue the 
entire compliance supplement pool as early reduction credits through 
this methodology, or the State may choose to reserve some of the 
compliance supplement pool to be issued to sources according to the 
direct distribution criteria as described above.
    This methodology is applicable for reductions made during the 2001 
and 2002 ozone seasons. NOX budget units that request early 
reduction credits will be required to monitor ozone season 
NOX emissions according to the monitoring provisions of part 
75, subpart H by the 2000 ozone season. The information from the 2000 
ozone season shall be used to establish a baseline emission rate for 
the NOX budget unit. To be eligible for early reduction 
credits, a NOX budget unit shall reduce its emissions rate 
in the 2001 and/or 2002 control period(s) no less than 20 percent below 
its baseline emissions rate established for the 2000 ozone season. The 
size of the early reduction credit request shall equal the difference 
between 0.25 lb/mmBtu and the unit's actual emissions rate multiplied 
by the unit's actual heat input for the applicable control period. 
NOX Budget units requesting early reduction credits should 
submit the request to the State by no later than October 30 of the year 
for which the early reductions were generated.
    The methodology conforms with the NOX SIP call's 
criteria for early reduction credits. By requiring that the reductions 
be measured using provisions in part 75, the reductions will be 
verified as having actually occurred and will be quantified according 
to the same procedures as required for compliance with the general 
requirements of the NOX Budget Trading Program. The 
procedure for calculating the credit request is intended to ensure that 
the reductions are surplus. Phase II of the title IV NOX 
emissions limits are required to be installed at specific coal-fired 
boilers by January 1, 2000. By requiring that an early reduction credit 
must be generated by no less than a 20 percent reduction below the 2000 
baseline emission rate, credits will only be issued for reductions that 
go below emissions levels achieved for compliance with title IV 
requirements. This provision ensures that the early reduction credits 
are only issued for reductions below existing requirements (i.e., 
surplus).
    Calculating the early credit based on the difference between 0.25 
lb/mmBtu

[[Page 57475]]

and the unit's actual emissions rate establishes a standard emissions 
rate from which all early reduction credits are calculated. This 
approach ensures that sources with higher NOX emissions 
rates prior to the 2001 ozone season are not provided an opportunity to 
generate more early reduction credits than relatively cleaner sources. 
In this way, all sources have an equal opportunity to generate early 
reduction credits below a standard emissions rate.
    According to the requirements in the NOX SIP call, 
States may not issue early reduction credits in an amount greater than 
the State's compliance supplement pool. To ensure this provision is 
met, the optional methodology is designed for States to issue all early 
reduction credits following the 2002 ozone season. By October 30, 2002, 
a State will have received all early reduction requests for both the 
2001 and 2002 ozone seasons. After review of the requests, the State 
would issue credit to all valid requests according to the following 
procedure. If the amount of valid requests is less than the size of the 
State's compliance supplement pool, the State would issue one allowance 
for each ton of early reduction credit requested. If the amount of 
valid requests is more than the size of the State's pool, the State 
would reduce the amount in the credit requests on a pro-rata basis so 
that the requests equal the size of the State's pool. After the 
requests have been reduced, the State would then issue allowances based 
on the remaining size of each credit request. States would complete the 
issuance of allowances for the early reduction credit requests as soon 
as possible following October 30, 2002, but no later than May 1, 2003.
5. Integrating the OTC Program With the NOX Budget Trading 
Program's Banking Provisions
    The OTC NOX Budget Program is a multi-state, capped 
NOX trading program that begins in 1999 and includes many 
States subject to today's action. By the start of the NOX 
Budget Trading Program under the NOX SIP call, sources in 
the OTC program will potentially hold banked NOX allowances 
resulting from early reductions and/or overcontrol with program 
requirements. At issue is the ability of OTC sources to use these 
banked allowances in the NOX Budget Trading Program.
    Commenters have supported allowing OTC sources to use banked 
allowances (i.e., early reductions from the 1997 and 1998 ozone seasons 
and unused allowances from the 1999 through 2002 ozone seasons) from 
the OTC program for compliance in the NOX Budget Trading 
Program. Commenters have stated that because OTC sources will be 
subject to a market-based cap-and-trade program prior to the 2003 ozone 
season, it is important to create a smooth transition from the OTC 
program to the NOX Budget Trading Program. They have 
suggested discounting OTC Phase II allowances to make them equivalent 
to those achieved under the NOX SIP call. One OTC State 
suggested accomplishing this by adjusting the OTC banked allowances by 
a ratio of the Phase II OTC control requirement to the Phase III OTC 
control requirement, working with EPA to determine the exact ratio. A 
few OTC States suggested that OTC allowances banked in Phase II could 
be used as early reduction credits in the NOX Budget Trading 
Program. A commenter from outside the OTC voiced concern that the use 
of OTC allowances banked by sources for the years 1999 through 2002 
could distort the larger trading market established under the SIP call.
    The EPA believes that the compliance supplement pool provides the 
opportunity to integrate the OTC program into the NOX Budget 
Trading Program by allowing OTC States to bring their banked allowances 
into the NOX Budget Trading Program as early reduction 
credits after the 2002 ozone season. The EPA established two primary 
criteria for the generation of early reduction credits in III.F.7.c.: 
first, the credits must be surplus, verifiable, and quantifiable; and 
second, a State may not grant an amount of early reduction credits in 
excess of a State's compliance supplement pool. EPA believes that 
banked allowances held by sources in the OTC program would qualify as 
being surplus, verifiable, and quantifiable. The banked allowances 
would be surplus because they would represent emissions reductions that 
go beyond what is required by the emissions limitations established by 
the OTC program in the applicable ozone seasons. The banked allowances 
would also be verified and quantified according to the procedures in 
the OTC program which are essentially identical to the requirements 
that will be in place under the NOX Budget Trading Program.
    As for the second criterion that a State issue no more early 
reduction credits than provided through the compliance supplement pool, 
EPA believes this could be addressed according to the following 
procedure. If the number of banked allowances held by an OTC State's 
NOX Budget units, after the compliance certification process 
for the 2002 ozone season, is less than the number of credits available 
in the pool for that State, the NOX budget units in that 
State may carry all of their banked allowances from the OTC program 
into the NOX Budget Trading Program. The banked allowances 
brought in from the OTC program would be subtracted from the State's 
compliance supplement pool. Any remaining credits in the compliance 
supplement pool could be distributed by the OTC State through the 
direct distribution option, if necessary. If, on the other hand, an OTC 
State's NOX Budget units hold banked allowances from the OTC 
program in excess of the amount of credits in the State's pool, after 
the compliance certification process for the 2002 ozone season, the 
State would need to reduce the amount of allowances eligible for being 
carried into the NOX Budget Trading Program. This could be 
achieved by reducing the amount of banked allowances held by the units 
on a pro rata basis so that the number of allowances carried into the 
NOX Budget Trading Program is less than or equal to the size 
of the State's compliance supplement pool.
    The process described above provides a mechanism for OTC States to 
use the compliance supplement pool to carry banked allowances from the 
OTC program as of the end of the compliance period in 2002 over into 
the NOX Budget Trading Program. The EPA believes this 
integration acknowledges the important reductions made in the OTC 
program prior to 2003 while providing similar opportunities for sources 
outside the OTC to generate credits for early reductions. Since all 
States in the NOX Budget Trading Program will have an 
opportunity to receive credit for early reductions, EPA does not 
believe any market distortion will occur.

G. New Source Review

    Under the New Source Review (NSR) provisions of section 173 of the 
CAA, a new major source or a major modification to an existing major 
source of a particular pollutant that proposes to locate in an area 
designated nonattainment for that pollutant must offset its new 
emissions. In the SNPR, the EPA solicited comment on whether and how 
the offset requirement could be met by sources' participation in the 
NOX Budget Trading Program. The Agency stated its belief 
that sources obligated to obtain NOX offsets under the NSR 
program should be able to do so by acquiring NOX allowances 
through the trading program. In essence, the EPA reasoned that, where a 
trading program is a capped system, a new source's acquisition of 
allowances to cover its increased emissions would necessarily

[[Page 57476]]

result in actual emissions reductions elsewhere in the system.
    The EPA continues to believe that nonattainment NSR offset 
requirements of the CAA can be met using the mechanism of the 
NOX Budget Trading Program. However, there are a number of 
complex issues involved with integrating these programs, for example, 
the statutory requirements to obtain offsets from certain geographic 
areas and, depending on the classification of the 1-hour ozone 
nonattainment area, at certain offset ratios. Because the Agency is 
continuing to evaluate these issues, it will not be providing guidance 
at this time on integrating these programs; however, the EPA intends to 
provide such guidance as soon as possible. At that time, the EPA will 
respond to the comments received on this topic in the course of this 
rulemaking.

VIII. Interaction With Title IV NOX Rule

    The EPA proposed, in the May 11, 1998 supplemental notice, to add a 
new Sec. 76.16 to part 76, the Acid Rain NOX Emission 
Reduction Program regulations. The purpose of the proposed Sec. 76.16 
was to increase utilities' flexibility in situations where units owned 
or operated by a utility were subject to both a NOX cap-and-
trade program and the Phase II NOX emission limitations 
under the Acid Rain NOX Emission Reduction Program. Under 
proposed Sec. 76.16, a State or group of States could request that the 
Administrator relieve all units located in the State or States and 
otherwise subject to the Phase II NOX emission limitations 
(under Secs. 76.6 and 76.7) of the requirement to comply with such 
emission limitations. The Administrator could also take this action on 
his or her own motion. All Group 1 boilers (i.e., tangentially fired or 
dry bottom wall fired boilers) would remain subject to the Phase I 
NOX emission limitations (under Sec. 76.5), while Group 2 
boilers (i.e., cell burner boilers, cyclones, wet bottom boilers, and 
vertically fired boilers) would have no NOX limits under the 
Acid Rain Program. This relief would be available if all such units 
were subject, under a SIP or a FIP, to a NOX cap-and-trade 
program meeting certain requirements. The NOX cap-and-trade 
program had to include, inter alia, either an annual cap or seasonal 
caps that together limited total annual emissions and a requirement 
that each unit use authorizations to emit (or allowances) to account 
for all NOX emissions. In addition, there had to be a 
demonstration that total annual NOX emissions from all units 
otherwise subject to the Acid Rain NOX emission limitations 
and located in the State or group of States would, under the 
NOX cap-and-trade program, be equal to or lower than the 
total number of annual NOX emissions if the units remained 
subject to the Acid Rain NOX emission limitations. 
Alternative emission limitations and NOX averaging plans 
under part 76 would not be taken into account in such a demonstration.
    Although the purpose of proposed Sec. 76.16 was to provide more 
flexibility to utilities consistent with the requirements of section 
407, almost all utility commenters and many State and State agency 
commenters opposed the proposal. Many commenters argued that relieving 
a utility's units in one State of the applicability of the Phase II 
NOX emission limitation would prevent the utility from using 
those units, along with units that the utility owns or operates in 
other States, in an interstate averaging plan under the Acid Rain 
Nitrogen Oxides Emission Reduction Program. Under section 407(e) of the 
CAA, as implemented under Sec. 76.11, a utility may comply with the 
Acid Rain NOX emission limitations by averaging the 
emissions of units that the utility owns or operates in the same State 
or other States. Many utilities have complied, or plan to comply, with 
the Acid Rain NOX Emission Reduction Program by using 
averaging plans, including some interstate averaging plans. However, a 
unit that has no Acid Rain emission limitation obviously cannot be 
included in an averaging plan since EPA would have no authority under 
title IV to limit the unit's emissions, whether on an individual-unit 
or a group-average basis. Further, as a practical matter, the group 
average limit for any given year, which must be calculated based on the 
limit applicable to each individual unit in the averaging plan, could 
not reflect any limit for such a unit. See 40 CFR 76.11(a) (1) and (2) 
(allowing only units with Acid Rain NOX emission limitations 
in effect to participate in an averaging plan) and (d)(1)(ii)(A) 
(showing calculation of the group average limit using each unit's Acid 
Rain NOX emission limitation).
    In the proposal, EPA attempted to address the issue of the 
potential impact of proposed Sec. 76.16 on averaging plans. Proposed 
Sec. 76.16(b)(1)(ii) required that, in determining whether a 
NOX cap-and-trade program met the requirements for granting 
units relief from the Phase II NOX emission limitations, the 
Administrator must consider ``whether the cost savings from trading 
will be offset by elimination of the ability of an owner or operator of 
a unit in the State or the group of States to use a NOX 
averaging plan under Sec. 76.11.'' 63 FR 25974. However, commenters 
were still concerned that the Administrator could, even after taking 
this into consideration, grant the relief over a utility's objections 
and prevent the utility from using an averaging plan that included the 
units for which the Administrator made the Phase II NOX 
emission limitations inapplicable. In light of the utilities' concerns 
that proposed Sec. 76.16 would actually reduce utilities' compliance 
flexibility, albeit under title IV, and prevent the use of averaging 
plans authorized under section 407(e), EPA has decided not to revise 
part 76 as proposed and is not adopting proposed Sec. 76.16 as a final 
rule.
    Suggestions by some commenters that, instead of adopting proposed 
Sec. 76.16, EPA extend the compliance date under the Acid Rain Program 
for the Phase II NOX emission limitations are rejected as 
outside the scope of this rulemaking. As acknowledged by commenters, 
that issue was raised in the rulemaking adopting the Phase II 
NOX emission limitations, and the compliance deadline of 
January 1, 2000 set in that rulemaking was recently upheld by the 
courts in Appalachian Power v. EPA, 135 F.3d 791 (D.C. Cir. 1998). The 
SIP call rulemaking did not include any proposal to alter that date. On 
the contrary, EPA stated in the SIP call:
    Obviously, in proposing a new 40 CFR 76.16, EPA is not requesting 
comment on any aspect of the December 19, 1996 final rule [i.e., the 
rule that set the Phase II NOX emission limitations and that 
included an earlier, proposed version of Sec. 76.16], including any 
issues addressed by the Court in Appalachian Power. 63 FR 25951.
    Similarly, commenters' suggestions concerning other revisions to 
the Acid Rain NOX Emission Reduction Program regulations 
(e.g., revisions to change the averaging provisions in the Acid Rain 
regulations to allow averaging among units that lack common owners or 
operators) are rejected as outside the scope of this rulemaking.

IX. Non-Ozone Benefits of NOX Emissions Decreases

A. Summary of Comments

    One commenter suggested that drinking water nitrate is not affected 
by atmospheric emissions and that the impacts of eutrophication are 
unknown, although no evidence was presented. Another commenter stated 
that EPA should estimate in the RIA the benefits of the SIP call with 
respect to the non-ozone impacts. One comment was received stating that 
EPA should not consider non-ozone benefits as

[[Page 57477]]

justification for the proposed emission reductions.

B. Response to Comments and Conclusion

1. Drinking Water Nitrate
    There is no disagreement that high levels of nitrate in drinking 
water is a health hazard, especially for infants. The contribution of 
atmospheric nitrogen (N) deposition to elevated levels of nitrate in 
drinking water supplies can be described as an evolving impact area. 
The Ecological Society of America has included discussion of this 
impact in a recent major review of causes and consequences of human 
alteration of the global N cycle in its Issues in Ecology series 
(Vitousek, Peter M., John Aber, Robert W. Howarth, Gene E. Likens, et 
al. 1997. Human Alteration of the Global Nitrogen Cycle: Causes and 
Consequences. Issues in Ecology. Published by Ecological Society of 
America, Number 1, Spring 1997). For decades, N concentrations in major 
rivers and drinking water supplies have been monitored in the United 
States, Europe, and other developed regions of the world. Analysis of 
these data confirms a substantial rise of N levels in surface waters, 
which are highly correlated with human-generated inputs of N to their 
watersheds. These N inputs are dominated by fertilizers and atmospheric 
deposition.
    Increases in atmospheric N deposition to sensitive forested 
watersheds approaching N saturation would be expected to result in 
increased nitrate concentrations in stream water. This phenomenon has 
been documented in the Los Angeles, California area and has been well-
established for areas in Germany and the Netherlands (Riggan, P.J., 
R.N. Lockwood, and E.N. Lopez, ``Deposition and Processing of Airborne 
Nitrogen Pollutants in Mediterranean-Type Ecosystems of Southern 
California'' Environmental Science and Technology, vol. 19, 1985). 
Stream water nitrate concentrations in watersheds subject to chronic 
air pollution in the Los Angeles area were two to three orders of 
magnitude greater than in chaparral regions outside the air basin.
2. Eutrophication
    The EPA believes that the eutrophication problem associated with 
atmospheric nitrogen deposition is well established. The National 
Research Council recently identified eutrophication as the most serious 
pollution problem facing the estuarine waters of the United States 
(NRC, 1993). NOX emissions contribute directly to the 
widespread accelerated eutrophication of United States coastal waters 
and estuaries. Atmospheric nitrogen deposition onto surface waters and 
deposition to watershed and subsequent transport into the tidal waters 
has been documented to contribute from 12 to 44 percent of the total 
nitrogen loadings to United States coastal water bodies. Nitrogen is 
the nutrient limiting growth of algae in most coastal waters and 
estuaries. Thus, addition of nitrogen results in accelerated algae and 
aquatic plant growth causing adverse ecological effects and economic 
impacts that range from nuisance algal blooms to oxygen depletion and 
fish kills.
3. Regulatory Impact Analysis
    The EPA believes it is important to note the potential impacts of 
the rulemaking, including the substantial benefits to the environment 
of several non-ozone impacts. As described in the November 7 proposal, 
in addition to contributing to attainment of the ozone NAAQS, decreases 
of NOX emissions will also likely help improve the 
environment in several important ways: (1) On a national scale, 
decreases in NOX emissions will also decrease acid 
deposition, nitrates in drinking water, excessive nitrogen loadings to 
aquatic and terrestrial ecosystems, and ambient concentrations of 
nitrogen dioxide, particulate matter and toxics; and (2), on a global 
scale, decreases in NOX emissions will, to some degree, 
reduce greenhouse gases and stratospheric ozone depletion. These 
benefits were also specifically recognized by OTAG, which in its July 
8, 1997 final recommendations, stated that it ``recognizes that 
NOX controls for ozone reductions purposes have collateral 
public health and environmental benefits, including reductions in acid 
deposition, eutrophication, nitrification, fine particle pollution, and 
regional haze.'' However, the benefits of some of these impacts are 
very difficult to estimate. Where possible, EPA provides estimates of 
the impacts of the rulemaking--both ozone and non-ozone--in the RIA.
4. Justification for Rulemaking
    While EPA believes this information is important for the public to 
understand and, thus, needs to be described as part of the rulemaking 
and RIA, there should be no misunderstanding as to the legal basis for 
the rulemaking, which is described in Section I, Background, of this 
notice and does not depend on the non-ozone benefits. The non-ozone 
benefits did not affect the method in which EPA determined significant 
contribution nor the calculation of the emissions budgets.

X. Administrative Requirements

A. Executive Order 12866: Regulatory Impacts Analysis

    Under Executive Order 12866 (58 FR 51735, October 4, 1993), the 
Agency must determine whether a regulatory action is ``significant'' 
and therefore subject to Office of Management and Budget (OMB) review 
and the requirements of the Executive Order. The Order defines 
``significant regulatory action'' as one that is likely to result in a 
rule that may:
    1. Have an annual effect on the economy of $100 million or more or 
adversely affect in a material way the economy, a sector of the 
economy, productivity, competition, jobs, the environment, public 
health or safety, or State, local, or tribal governments or 
communities;
    2. Create a serious inconsistency or otherwise interfere with an 
action taken or planned by another agency;
    3. Materially alter the budgetary impact of entitlements, grants, 
user fees, or loan programs or the rights and obligations of recipients 
thereof; or
    4. Raise novel legal or policy issues arising out of legal 
mandates, the President's priorities, or the principles set forth in 
the Executive Order.
    In view of its important policy implications and potential effect 
on the economy of over $100 million, this action has been judged to be 
a ``significant regulatory action'' within the meaning of the Executive 
Order. As a result, the final rulemaking was submitted to OMB for 
review, and EPA has prepared a Regulatory Impact Analysis (RIA) 
entitled ``Regulatory Impact Analysis for the Regional NOX 
SIP Call (September 1998).''
    This RIA assesses the costs, benefits, and economic impacts 
associated with potential State implementation strategies for complying 
with this rulemaking. Any written comments from OMB to EPA and any 
written EPA response to those comments are included in the docket. The 
docket is available for public inspection at the EPA's Air Docket 
Section, which is listed in the ADDRESSES Section of this preamble. The 
RIA is available in hard copy by contacting the EPA Library at the 
address under ``Availability of Related Information'' and in electronic 
form as discussed above under ``Availability of Related Information.''
    The RIA attempts to simulate a possible set of State implementation 
strategies and estimates the costs and benefits associated with that 
set of

[[Page 57478]]

strategies. The RIA concludes that the national annual cost of possible 
State actions to comply with the SIP call are approximately $1.7 
billion (1990 dollars). The associated benefits, in terms of 
improvements in health, crop yields, visibility, and ecosystem 
protection, that EPA has quantified and monetized range from $1.1 
billion to $4.2 billion. Due to practical analytical limitations, the 
EPA is not able to quantify and/or monetize all potential benefits of 
this action.

B. Regulatory Flexibility Act: Small Entity Impacts

    The Regulatory Flexibility Act (5 U.S.C. 601 et seq.) (RFA), as 
amended by the Small Business Regulatory Enforcement Fairness Act (Pub. 
L. No. 104-121) (SBREFA), provides that whenever an agency is required 
to publish a general notice of proposed rulemaking, it must prepare and 
make available an initial regulatory flexibility analysis, unless it 
certifies that the proposed rule, if promulgated, will not have ``a 
significant economic impact on a substantial number of small 
entities.'' 5 U.S.C. 605(b). Courts have interpreted the RFA to require 
a regulatory flexibility analysis only when small entities will be 
subject to the requirements of the rule. See, Motor and Equip. Mfrs. 
Ass'n v. Nichols, 142 F.3d 449 (D.C. Cir. 1998); United Distribution 
Cos. v. FERC, 88 F.3d 1105, 1170 (D.C. Cir. 1996); Mid-Tex Elec. Co-op, 
Inc. v. FERC, 773 F.2d 327, 342 (D.C. Cir. 1985) (agency's 
certification need only consider the rule's impact on entities subject 
to the rule).
    The NOX SIP Call would not establish requirements 
applicable to small entities. Instead, it would require States to 
develop, adopt, and submit SIP revisions that would achieve the 
necessary NOX emissions reductions, and would leave to the 
States the task of determining how to obtain those reductions, 
including which entities to regulate. Moreover, because affected States 
would have discretion to choose which sources to regulate and how much 
emissions reductions each selected source would have to achieve, EPA 
could not predict the effect of the rule on small entities.
    For these reasons, EPA appropriately certified that the rule would 
not have a significant impact on a substantial number of small 
entities. Accordingly, the Agency did not prepare an initial RFA for 
the proposed rule.
    For the final rule, EPA is confirming its initial certification. 
However, the Agency did conduct a more general analysis of the 
potential impact on small entities of possible State implementation 
strategies. This analysis is documented in the RIA. The EPA did receive 
comments regarding the impact on small entities. These comments will be 
addressed in the Response to Comment document.
    This final rule will not have a significant impact on a substantial 
number of small entities because the rule does not establish 
requirements applicable to small entities. Therefore, I certify that 
this action will not have a significant impact on a substantial number 
of small entities.

C. Unfunded Mandates Reform Act

    Title II of the Unfunded Mandates Reform Act of 1995 (Pub. L. 104-
4) (UMRA), establishes requirements for Federal agencies to assess the 
effects of their regulatory actions on State, local, and tribal 
governments and the private sector. Under section 202 of the UMRA, 2 
U.S.C. 1532, EPA generally must prepare a written statement, including 
a cost-benefit analysis, for any proposed or final rule that ``includes 
any Federal mandate that may result in the expenditure by State, local, 
and tribal governments, in the aggregate, or by the private sector, of 
$100,000,000 or more * * * in any one year.'' A ``Federal mandate'' is 
defined under section 421(6), 2 U.S.C. 658(6), to include a ``Federal 
intergovernmental mandate'' and a ``Federal private sector mandate.'' A 
``Federal intergovernmental mandate,'' in turn, is defined to include a 
regulation that ``would impose an enforceable duty upon State, local, 
or tribal governments,'' section 421(5)(A)(i), 2 U.S.C. 658(5)(A)(i), 
except for, among other things, a duty that is ``a condition of Federal 
assistance,'' section 421(5)(A)(i)(I). A ``Federal private sector 
mandate'' includes a regulation that ``would impose an enforceable duty 
upon the private sector,'' with certain exceptions, section 421(7)(A), 
2 U.S.C. 658(7)(A).
    Before promulgating an EPA rule for which a written statement is 
needed under section 202 of the UMRA, section 205, 2 U.S.C. 1535, of 
the UMRA generally requires EPA to identify and consider a reasonable 
number of regulatory alternatives and adopt the least costly, most 
cost-effective, or least burdensome alternative that achieves the 
objectives of the rule.
    The EPA has prepared a written statement consistent with the 
requirements of section 202 of the UMRA and placed that statement in 
the docket for this rulemaking. Furthermore, as EPA stated in the 
proposal, EPA is not directly establishing any regulatory requirements 
that may significantly or uniquely affect small governments, including 
tribal governments. Thus, EPA is not obligated to develop under section 
203 of the UMRA a small government agency plan. Furthermore, as 
described in the proposal, in a manner consistent with the 
intergovernmental consultation provisions of section 204 of the UMRA 
and Executive Order 12875, EPA carried out consultations with the 
governmental entities affected by this rule. Finally, the written 
statement placed in the docket also contains a discussion consistent 
with the requirements of section 205 of the UMRA.
    For several reasons, however, EPA is not reaching a final 
conclusion as to the applicability of the requirements of UMRA to this 
rulemaking action. First, it is questionable whether a requirement to 
submit a SIP revision would constitute a federal mandate in any case. 
The obligation for a state to revise its SIP that arises out of 
sections 110(a) and 110(k)(5) of the CAA is not legally enforceable by 
a court of law, and at most is a condition for continued receipt of 
highway funds. Therefore, it is possible to view an action requiring 
such a submittal as not creating any enforceable duty within the 
meaning of section 421(5)(9a)(I) of UMRA (2 U.S.C. 658 (a)(I)). Even if 
it did, the duty could be viewed as falling within the exception for a 
condition of Federal assistance under section 421(5)(a)(i)(I) of UMRA 
(2 U.S.C. 658(5)(a)(i)(I)).
    As noted earlier, however, notwithstanding these issues EPA has 
prepared the statement that would be required by UMRA if its statutory 
provisions applied and has consulted with governmental entities as 
would be required by UMRA. Consequently, it is not necessary for EPA to 
reach a conclusion as to the applicability of the UMRA requirements. 
The analysis assumes that states would adopt the control strategies 
that EPA assumed in its analyses underlying this action. The EPA 
further notes that in two related proposals also signed today--one 
concerning federal implementation plans if States do not comply with 
the SIP call and one concerning the petitions submitted to the Agency 
under section 126 of the CAA--EPA is taking the position that the 
requirements of UMRA apply because both of those actions could result 
in the establishment of enforceable mandates directly applicable to 
sources (including sources owned by state and local governments).

D. Paperwork Reduction Act

    The information collection requirements in this rule have been 
submitted for approval to the Office of

[[Page 57479]]

Management and Budget (OMB) under the Paperwork Reduction Act, 44 
U.S.C. 3501 et seq. An Information Collection Request (ICR) document 
has been prepared by EPA (ICR No. 1857.02) and a copy may be obtained 
from Sandy Farmer by mail at Regulatory Information Division; U.S. 
Environmental Protection Agency (2137); 401 M St., SW., Washington, DC 
20460, by email at [email protected], or by calling (202) 260-2740. 
A copy may also be downloaded from the internet at http://www.epa.gov/
icr. The information requirements are not effective until OMB approves 
them.
    The EPA believes that it is essential that compliance with the 
regional control strategy be verified. Tracking emissions is the 
principal mechanism to ensure compliance with the budget and to assure 
the downwind affected States and EPA that the ozone transport problem 
is being mitigated. If tracking and periodic reports indicate that a 
State is not implementing all of its NOX control measures 
beginning with the compliance date for NOX controls or is 
off track to meet its statewide budget by September 30, 2007, EPA will 
work with the State to determine the reasons for noncompliance and what 
course of remedial action is needed.
    The reporting requirements are mandatory and the legal authority 
for the reporting requirements resides in section 110(a) and 301(a) of 
the CAA. Emissions data being requested in today's rule is not be 
considered confidential by EPA. Certain process data may be identified 
as sensitive by a State and are then treated as ``State-sensitive'' by 
EPA.
    The reporting and record keeping burden for this collection of 
information is described below:
    Respondents/Affected Entities: States, along with the District of 
Columbia, which are included in the NOX SIP call.
    Number of Respondents: 23.
    Frequency of Response: annually, triennially.
    Estimated Annual Hour Burden per Respondent: 269.
    Estimated Annual Cost per Respondent: $7,140.00.
     Estimated Total Annual Hour Burden: 6,197.
    Estimated Total Annualized Cost: $164,190.00.
    There are no additional capital or operating and maintenance costs 
for the States, along with the District of Columbia, associated with 
the reporting requirements of this rule. During the 1980s, an EPA 
initiative established electronic communication with each State 
environmental agency. This included a computer terminal for any States 
needing one in order to communicate with the EPA's national data base 
systems. Costs associated with replacing and maintaining these 
terminals, as well as storage of data files, have been accounted for in 
the ICR for the existing annual inventory reporting requirements (OMB # 
2060-0088).
    Burden means the total time, effort, or financial resources 
expended by persons to generate, maintain, retain, or disclose or 
provide information to or for a Federal agency. This includes the time 
needed to review instructions; develop, acquire, install, and utilize 
technology and systems for the purposes of collecting, validating, and 
verifying information, processing and maintaining information, and 
disclosing and providing information; adjust the existing ways to 
comply with any previously applicable instructions and requirements; 
train personnel to be able to respond to a collection of information; 
search data sources; complete and review the collection of information; 
and transmit or otherwise disclose the information.
    An Agency may not conduct or sponsor, and a person is not required 
to respond to a collection of information unless it displays a 
currently valid OMB control number. The OMB control numbers for EPA's 
regulations are listed in 40 CFR Part 9 and 48 CFR Chapter 15.
    Send comments on the Agency's need for this information, the 
accuracy of the provided burden estimates, and any suggested methods 
for minimizing respondent burden, including through the use of 
automated collection techniques to the Director, Office of Policy, 
Regulatory Information Division; U.S. Environmental Protection Agency 
(2137); 401 M St., SW.; Washington, DC 20460; and to the Office of 
Information and Regulatory Affairs, Office of Management and Budget, 
725 17th St., NW., Washington, DC 20503, marked ``Attention: Desk 
Officer for EPA.'' Comments are requested by November 27, 1998. Include 
the ICR number in any correspondence.

E. Executive Order 13045: Protection of Children From Environmental 
Health Risks and Safety Risks

1. Applicability of E.O. 13045
    The Executive Order 13045 applies to any rule that EPA determines 
(1) ``economically significant'' as defined under Executive Order 
12866, and (2) the environmental health or safety risk addressed by the 
rule has a disproportionate effect on children. If the regulatory 
action meets both criteria, the Agency must evaluate the environmental 
health or safety effects of the planned rule on children; and explain 
why the planned regulation is preferable to other potentially effective 
and reasonably feasible alternatives considered by the Agency. This 
proposed rule is not subject to E.O. 13045, entitled ``Protection of 
Children from Environmental Health Risks and Safety Risks (62 FR 19885, 
April 23, 1997), because it does not involve decisions on environmental 
health risks or safety risks that may disproportionately affect 
children.
2. Children's Health Protection
    In accordance with section 5(501), the Agency has evaluated the 
environmental health or safety effects of the rule on children, and 
found that the rule does not separately address any age groups. 
However, the Agency has conducted a general analysis of the potential 
changes in ozone and particulate matter levels experienced by children 
as a result of the NOX SIP call; these findings are 
presented in the Regulatory Impact Analysis. The findings include 
population-weighted exposure characterizations for projected 2007 ozone 
and PM concentrations. The population includes a census-derived 
subdivision for the under 18 group.

F. Executive Order 12898: Environmental Justice

    Executive Order 12898 requires that each Federal agency make 
achieving environmental justice part of its mission by identifying and 
addressing, as appropriate, disproportionately high and adverse human 
health or environmental effects of its programs, policies, and 
activities on minorities and low-income populations. The Agency has 
conducted a general analysis of the potential changes in ozone and 
particulate matter levels that may be experienced by minority and low-
income populations as a result of the NOX SIP call; these 
findings are presented in the Regulatory Impact Analysis. The findings 
include population-weighted exposure characterizations for projected 
ozone concentrations and PM concentrations. The population includes 
census-derived subdivisions for whites and non-whites, and for low-
income groups.

G. Executive Order 12875: Enhancing the Intergovernmental Partnerships

    Under Executive Order 12875, EPA may not issue a regulation that is 
not required by statute and that creates a mandate upon a State, local 
or tribal government, unless the Federal

[[Page 57480]]

government provides the funds necessary to pay the direct compliance 
costs incurred by those governments. If the mandate is unfunded, EPA 
must provide to the Office of Management and Budget a description of 
the extent of EPA's prior consultation with representatives of affected 
State, local and tribal governments, the nature of their concerns, 
copies of any written communications from the governments, and a 
statement supporting the need to issue the regulation. In addition, 
Executive Order 12875 requires EPA to develop an effective process 
permitting elected officials and other representatives of State, local 
and tribal governments ``to provide meaningful and timely input in the 
development of regulatory proposals containing significant unfunded 
mandates.''
    Today's rule does not create a mandate on State, local or tribal 
governments. As explained in the discussion of UMRA (Section X.C), this 
rule does not impose an enforceable duty on these entities. 
Accordingly, the requirements of section 1(a) of Executive Order 12875 
do not apply to this rule.

H. Executive Order 13084: Consultation and Coordination With Indian 
Tribal Governments

    Under Executive Order 13084, EPA may not issue a regulation that is 
not required by statute, that significantly or uniquely affects the 
communities of Indian tribal governments, and that imposes substantial 
direct compliance costs on those communities, unless the government 
provides the funds necessary to pay the direct compliance costs 
incurred by the tribal governments. If the mandate is unfunded, EPA 
must provide to the Office of Management and Budget, in a separately 
identified section of the preamble to the rule, a description of the 
extent of EPA's prior consultation with representatives of affected 
tribal governments, a summary of the nature of their concerns, and a 
statement supporting the need to issue the regulation. In addition, 
Executive Order 13084 requires EPA to develop an effective process 
permitting elected and other representatives of Indian tribal 
governments ``to provide meaningful and timely input in the development 
of regulatory policies on matters that significantly or uniquely affect 
their communities.''
    Today's rule does not significantly or uniquely affect the 
communities of Indian tribal governments. The rule applies only to 
certain States, and does not require Indian tribal governments to take 
any action. Moreover, EPA does not, by today's rule, call on States to 
regulate NOX sources located on tribal lands. Accordingly, 
the requirements of section 3(b) of Executive Order 13084 do not apply 
to this rule.
    The only circumstance in which the rule might even indirectly 
affect sources on tribal lands would be if the budget set for one or 
more of the 23 jurisdictions reflects assumed emissions reductions from 
NOX sources on tribal lands located within the exterior 
boundaries of those States. The EPA is not aware of any such sources. 
However, to address the possibility that one or more of the State 
budgets reflects reductions from such sources, and because any such 
State generally would not have jurisdiction over such sources (see 
EPA's rule promulgated under CAA section 301(d), 63 FR 7254, February 
12, 1998), EPA will consider any request to revise as appropriate the 
budget and base year 2007 emissions inventory for such a State, based 
on a demonstration that the State does not have authority to regulate 
those sources.

I. Judicial Review

    Section 307(b)(1) of the CAA indicates which Federal Courts of 
Appeal have venue for petitions of review of final actions by EPA. This 
Section provides, in part, that petitions for review must be filed in 
the Court of Appeals for the District of Columbia Circuit if (i) the 
agency action consists of ``nationally applicable regulations 
promulgated, or final action taken, by the Administrator,'' or (ii) 
such action is locally or regionally applicable, if ``such action is 
based on a determination of nationwide scope or effect and if in taking 
such action the Administrator finds and publishes that such action is 
based on such a determination.''
    Any final action related to the NOX SIP call is 
``nationally applicable'' within the meaning of section 307(b)(1). As 
an initial matter, through this rule, EPA interprets section 110 of the 
CAA in a way that could affect future actions regulating the transport 
of pollutants. In addition, the NOX SIP call, as proposed, 
would require 22 States and the District of Columbia to decrease 
emissions of NOX. The NOX SIP call also is based 
on a common core of factual findings and analyses concerning the 
transport of ozone and its precursors between the different States 
subject to the NOX SIP call. Finally, EPA has established 
uniform approvability criteria that would be applied to all States 
subject to the NOX SIP call. For these reasons, the 
Administrator also is determining that any final action regarding the 
NOX SIP call is of nationwide scope and effect for purposes 
of section 307(b)(1). Thus, any petitions for review of final actions 
regarding the NOX SIP call must be filed in the Court of 
Appeals for the District of Columbia Circuit within 60 days from the 
date final action is published in the Federal Register.

J. Congressional Review Act

    The Congressional Review Act, 5 U.S.C. 801 et seq., as added by the 
Small Business Regulatory Enforcement Fairness Act of 1996, generally 
provides that before a rule may take effect, the agency promulgating 
the rule must submit a rule report, which includes a copy of the rule, 
to each House of the Congress and to the Comptroller General of the 
United States. The EPA will submit a report containing this rule and 
other required information to the U.S. Senate, the U.S. House of 
Representatives, and the Comptroller General of the United States prior 
to publication of the rule in the Federal Register. A ``major rule'' 
cannot take effect until 60 days after it is published in the Federal 
Register. This action is a ``major rule'' as defined by 5 U.S.C. 
Sec. 804(2). This rule will be effective December 28, 1998.

K. National Technology Transfer and Advancement Act

    Section 12(d) of the National Technology Transfer and Advancement 
Act of 1995 (NTTAA), Pub. L. No. 104-113, section 12(d) (15 U.S.C. 272 
note) directs EPA to use voluntary consensus standards in its 
regulatory activities unless to do so would be inconsistent with 
applicable law or otherwise impractical. Voluntary consensus standards 
are technical standards (e.g., materials specifications, test methods, 
sampling procedures, and business practices) that are developed or 
adopted by voluntary consensus standards bodies. The NTTAA directs EPA 
to provide Congress, through OMB, explanations when the Agency decides 
not to use available and applicable voluntary consensus standards.
    This final rulemaking sets forth a model trading program including 
environmental monitoring and measurement provisions that States are 
encouraged to adopt as part of their SIPs. If States adopt those 
provisions, sources that participate in the trading program would be 
required to meet the applicable monitoring requirements of part 75. In 
addition, this final rulemaking requires States that choose to regulate 
certain large stationary sources to meet the requirements of the SIP 
call to use part 75 to ensure compliance with their regulations. Part 
75 already incorporates a number of voluntary consensus standards. In

[[Page 57481]]

addition, EPA's proposed revisions to part 75 proposed to add two more 
voluntary consensus standards to the rule (see 63 FR at 28116-17, 
discussing ASTM D5373-93 ``Standard Methods for Instrumental 
Determination of Carbon, Hydrogen and Nitrogen in laboratory samples of 
Coal and Coke,'' and API Section 2 ``Conventional Pipe Provers'' from 
Chapter 4 of the Manual for Petroleum Measurement Standards, October 
1988 edition). The EPA's proposed revisions to part 75 also requested 
comments on the inclusion of additional voluntary consensus standards. 
The EPA is finalizing some revisions to part 75 now, including the 
incorporation of two voluntary consensus standards, in response to 
comments submitted on the proposed part 75 rulemaking:
    (1) American Petroleum Institute (API) Petroleum Measurement 
Standards, Chapter 3, Tank Gauging: Section 1A, Standard Practice for 
the Manual Gauging of Petroleum and Petroleum Products, December 1994; 
Section 1B, Standard Practice for Level Measurement of Liquid 
Hydrocarbons in Stationary Tanks by Automatic Tank Gauging, April 1992 
(reaffirmed January 1997); Section 2, Standard Practice for Gauging 
Petroleum and Petroleum Products in Tank Cars, September 1995; Section 
3, Standard Practice for Level Measurement of Liquid Hydrocarbons in 
Stationary Pressurized Storage Tanks by Automatic Tank Gauging, June 
1996; Section 4, Standard Practice for Level Measurement of Liquid 
Hydrocarbons on Marine Vessels by Automatic Tank Gauging, April 1995; 
and Section 5, Standard Practice for Level Measurement of Light 
Hydrocarbon Liquids Onboard Marine Vessels by Automatic Tank Gauging, 
March 1997; for Sec. 75.19 and,
    (2) Shop Testing of Automatic Liquid Level Gages, Bulletin 2509 B, 
December 1961 (Reaffirmed October 1992), for Sec. 75.19.
    These materials are available for purchase from the following 
address: American Petroleum Institute, Publications Department, 1220 L 
Street NW, Washington, DC 20005-4070.
    These standards are used to quantify fuel use from units that have 
low emissions of NOX and SOX.
    The EPA intends to finalize other revisions to part 75 in the near 
future and address comments related to the proposed voluntary consensus 
standards and to additional voluntary consensus standards at that time.
    Consistent with the Agency's Performance Based Measurement System, 
part 75 sets forth performance criteria that allow the use of 
alternative methods to the ones set forth in part 75. The PBMS approach 
is intended to be more flexible and cost effective for the regulated 
community; it is also intended to encourage innovation in analytical 
technology and improved data quality. The EPA is not precluding the use 
of any method, whether it constitutes a voluntary consensus standard or 
not, as long as it meets the performance criteria specified, however 
any alternative methods must be approved in advance before they may be 
used under part 75.

List of Subjects

40 CFR Part 51

    Air pollution control, Administrative practice and procedure, 
Carbon monoxide, Environmental protection, Intergovernmental relations, 
Nitrogen dioxide, Ozone, Particulate matter, Reporting and 
recordkeeping requirements, Sulfur oxides, Transportation, Volatile 
organic compounds.

40 CFR Parts 72 and 75

    Air pollution control, Carbon dioxide, Continuous emissions 
monitors, Electric utilities, Environmental protection, Incorporation 
by reference, Nitrogen oxides, Reporting and recordkeeping 
requirements, Sulfur dioxide.

40 CFR Part 96

    Environmental protection, Administrative practice and procedure, 
Air pollution control, Nitrogen dioxide, Reporting and recordkeeping 
requirements.

    Dated: September 24, 1998.
Carol M. Browner,
Administrator.

Appendix A to the Preamble--Detailed Discussion of Changes to Part 
75

    The following discussion addresses the comments received both on 
the SNPR (68 FR 25902) and the proposed part 75 revisions (68 FR 28032) 
that relate to the monitoring of NOX mass emissions. In 
addition, it addresses the comments received on the excepted monitoring 
methodology for low mass emitting units that would apply to both units 
affected by title IV of the CAA and to units affected by a State or 
Federal NOX mass reduction program that adopted or 
incorporated the requirements of this part.

I. NOX Mass Monitoring and Reporting Provisions

    Commenters raised four main issues with the proposed NOX 
mass monitoring and reporting provisions in subpart H. The first issue 
has to do with the appropriate monitoring requirements necessary to 
support a NOX mass monitoring program, particularly in light 
of the fact that many of the units that would be subject to a program 
based on Part 96 are not currently monitoring NOX mass 
emissions. The second has to do with using a NOX 
concentration CEMS and a flow CEMS to calculate NOX mass. 
The third has to do with the requirement to report NOX mass 
emissions year round even though the ozone season is only 5 months 
long. The final issue has to do with the requirement to have petitions 
for alternatives to part 75 be approved by both the state permitting 
authority and by EPA.

A. Background on Use of Part 75 to Monitor and Report NOX 
Mass Emissions

    Subpart H of the proposed part 75 rule set forth general monitoring 
and reporting requirements that sources subject to a State or Federal 
NOX mass emission reduction program could incorporate or 
adopt into that program. Several commenters argued that it was 
inappropriate to require sources, who were not already required to meet 
the requirements of part 75, to meet those requirements for purposes of 
a state program.
    Commenters who suggested that it was inappropriate to require a 
source that is not already subject to part 75 to meet the requirements 
of part 75 for purposes of a state program suggested that the State 
should decide what requirements the source needs to meet. The EPA 
agrees that this would be appropriate in the case of a program that 
only affected that state. For instance, if a State was developing a 
NOX reduction program to address its own non-attainment 
problem, it would not be necessary to adopt requirements that were 
consistent across a larger geographic area. However, in a multi-state 
program, particularly a multi-state trading program which engages in 
interstate commerce like the one set forth in part 96, EPA believes it 
is necessary to account for emissions in a consistent manner across the 
whole region. This ensures that all sources that participate in the 
trading program account for their emissions in a consistent manner, 
ensuring both integrity in the trading program and a level playing 
field for all program participants. Therefore, EPA believes that it is 
necessary to create one set of consistent monitoring and reporting 
requirements that can be used for such a program. This is consistent 
with the way the Act mandated that a multi-state trading program be 
implemented under Title IV. It is also consistent with the

[[Page 57482]]

approach taken in implementing other emissions standards, such as the 
new source performance standards that affect many states. This approach 
also makes it easier for states designing their programs since they 
would not have to reinvent the monitoring requirements in each case.
    Commenters who suggested that part 75 did not provide enough 
flexibility focused on three areas: they suggested that other programs 
such as RECLAIM or the OTC trading program provided more flexible non-
CEMS options for units that operated infrequently or had low 
NOX mass emissions; they suggested that sources should be 
allowed to use predictive emissions monitoring systems (PEMS); and they 
suggested that sources should be allowed to use coal sampling and 
weighting to determine heat input.
    The EPA believes that the flexibilities offered by part 75 are 
consistent with the type of flexibilities offered in RECLAIM and the 
OTC Program. RECLAIM requires CEMS on all units that emit more than 10 
tons of any individual pollutant per year. The OTC Program requires 
CEMS on all units that do not qualify as peaking units that are larger 
than 250 mmBtu or serve generators greater than 25 MWs. Subpart H of 
part 75 allows non-CEMS alternatives for units that have emissions less 
than 50 tons per year of NOX. If a unit is not required to 
report SO2 and CO2 for Acid Rain compliance, then 
the unit may use the low mass emissions provisions of Part 75 if its 
NOX emissions are less than 50 tons per year. Part 75 also 
allows non-CEMS alternatives for units that qualify as peaking units. 
In both the OTC Program and part 75, a peaking unit is defined as a 
unit that has a capacity factor of no more than 10 percent per year 
averaged over a three year period and no more than 20 percent in any 
one year. The EPA believes that these options provide cost effective 
monitoring methodologies for small or infrequently used units.
    While commenters who supported the use of PEMS and the use of coal 
sampling and weighting asserted that these methodologies would provide 
data equivalent to that provided by the methodologies in Part 75, none 
of the commenters provided any data to justify this claim. Therefore 
EPA is not adding specific requirements that would allow either of 
these methodologies. It should be noted that subpart E of part 75 does 
provide a means for a source to demonstrate that an alternative 
methodology such as PEMS or coal sampling and weighting is equivalent 
to CEMS. Subpart E of part 75 is consistent with Performance Based 
Measurement Systems criteria. Any source wishing to use an alterative 
methodology may petition the agency under subpart E of part 75.

B. Background on Use of a NOX Concentration CEMS and a Flow 
CEMS to Calculate NOX Mass

    Subpart H of the proposed part 75 rule called for sources in the 
NOX Budget Program to monitor NOX emission rate 
in lb/mmBtu using a NOX concentration monitor and a diluent 
monitor, and then to multiply this by heat input, calculated using a 
flow monitor and a diluent monitor. Under this proposal, sources would 
then calculate NOX mass emissions by multiplying the hourly 
NOX emission rate by the hourly heat input to obtain the 
pounds of NOX emitted during the hour. The EPA also 
requested comment on whether it would be appropriate for sources in the 
NOX Budget Program to use the NOX concentration 
monitor and flow monitor without a diluent monitor to calculate 
NOX mass emissions. This is analogous to the Acid Rain 
Program's current approach to monitoring SO2 mass emissions.
    Commenters recommended that the Agency require sources to determine 
NOX mass emissions from pollutant concentration and stack 
gas volumetric flow. The commenters stated that this approach would be 
more accurate, more familiar to sources, and more consistent with the 
SO2 mass emissions monitoring in the existing part 75.
    The Agency agrees that using NOX pollutant concentration 
and volumetric flow is an appropriate method for monitoring 
NOX mass emissions. Today's final rule includes provisions 
in Subpart H and Section 8 of Appendix F of part 75 to allow sources to 
choose one of several options for monitoring and calculating 
NOX mass emissions. Sources may monitor NOX mass 
emissions by using either:

All Units

     A NOX pollutant concentration monitor and a 
volumetric flow monitor, or a NOX concentration monitor and 
a diluent monitor to calculate NOX emission rate in lb/
mmBtu, and a flow monitor and a diluent monitor to calculate heat 
input; or
     A NOX concentration monitor and a diluent 
monitor to calculate NOX emission rate in lb/mmBtu, and a 
fuel flow meter and oil or gas sampling and analysis to calculate heat 
input; or

Oil/Natural Gas Fired Units

     Peaking units may use NOX to load correlation 
procedures from Appendix E of part 75 for NOX emission rate, 
and a fuel flow meter and oil or gas sampling and analysis to calculate 
heat input; or
     Units with less than 50 tons of Nox and 25 tons of 
SO2 may use emission rates multiplied by either the maximum 
rated heat input capacity of the unit or by the actual heat input of 
the unit which may be determined on a longer term basis than a single 
hour.
    The EPA decided to allow sources several options so that they could 
use monitoring equipment that is already installed under part 75 to the 
greatest extent possible.
    In implementing these options, a source would need to designate a 
primary approach to calculating NOX mass emissions. For 
example, the designated representative of a coal-fired unit could 
choose to designate a primary monitoring approach under Option 1 
(pollutant concentration monitor and diluent monitor, and diluent 
monitor and flow monitor). The designated representative could then use 
a (pollutant concentration monitor and flow monitor) as a backup 
monitoring approach. This would be useful for periods when the diluent 
monitor is not operating properly, where NOX emission rate 
data in lb/mmBtu would not be available, but NOX mass 
emission data in lb could still be available. The OTC NOX 
Budget Program allows this approach (see docket A-97-35 item II-I-7).
    In order to make monitoring as consistent as possible between the 
first two approaches for monitoring NOX mass emissions using 
continuous emission monitoring systems (CEMS), EPA is making additional 
changes to part 75. First, the Agency is adding language in Section 8 
of Appendix F that specifies the calculations for NOX mass 
emissions using either approach. Second, EPA is requiring sources that 
use a NOX pollutant concentration monitor and a flow monitor 
as the primary method for calculating NOX mass emissions to 
substitute for missing NOX pollutant concentration data 
using the same missing data procedures as for NOX CEMS (lb/
mmBtu) under Secs. 75.31(c), 75.33(c) and Appendix C. Third, the Agency 
is establishing a relative accuracy testing requirement for 
NOX pollutant concentration monitors that are used to 
calculate NOX mass emissions independently of a 
NOX CEMS (lb/mmBtu). The NOX pollutant 
concentration monitors will need to meet a relative accuracy of 10.0 
percent to pass the relative accuracy test audit (RATA). They will need 
to meet a relative accuracy of 7.5 percent to perform a RATA on an 
annual basis instead of a semi-annual basis. Because the vast majority 
of NOX CEMS (lb/

[[Page 57483]]

mmBtu) and SO2 pollutant concentration monitors routinely 
meet a relative accuracy of 7.5 percent or less, the Agency concludes 
that it will also be possible for a NOX pollutant 
concentration monitor, which is part of a NOX CEMS, to meet 
this standard. Fourth, EPA requires these sources to test their 
NOX pollutant concentration monitor and flow monitor for 
bias. If the monitor is found to be biased low, then the source must 
either fix the monitor and retest it to show it is not biased, or apply 
a bias adjustment factor to hourly data. These changes to part 75 make 
monitoring consistent between the different monitoring approaches using 
CEMS, prevent underestimation of emissions, preserve monitoring 
accuracy, and take advantage of approaches already developed for other 
monitoring systems that will be familiar to sources.
    The EPA decided to allow sources to calculate NOX mass 
emissions using NOX concentration and flow rate for several 
reasons:
     This approach would allow sources to remove bias due to 
the diluent monitor from calculations of NOX mass emissions.
     Sources affected by the NOX Budget Program, but 
not by the Acid Rain Program, such as industrial boilers, may be able 
to simplify their recordkeeping and reporting because they will not 
need to calculate or report NOX emission rate in lb/mmBtu 
for each hour for the trading program.
     Sources will be able to maintain higher availability of 
quality-assured NOX mass emission data, because they will 
not need to substitute missing data for purposes of NOX mass 
emissions when data are not available from the diluent monitor.
     As the commenters suggested, this approach is more 
analogous to monitoring for SO2 mass emissions in the Acid 
Rain Program.
    Because this approach is already allowed under the OTC 
NOX Budget Program, EPA already has accounted for this 
possibility in the electronic data reporting format and in its 
computerized Emission Tracking System.
    For these reasons, the Agency believes that it is appropriate to 
allow sources the option of monitoring and calculating NOX 
mass emissions using NOX pollutant concentration and flow 
monitors.
    Sources using this approach may still be required to install 
maintain and operate a diluent monitor to calculate heat input if 
required to do so by their state for purposes of obtaining data needed 
to support allocation of NOX allowances.

C. Background on Year Round Reporting of NOX Mass Emissions

    The proposal would have required all units to report NOX 
mass emissions on an annual basis rather than on an ozone season basis. 
One commenter noted that since the proposed SIP call would not require 
emission reductions outside of the ozone season it is not necessary to 
report NOX mass emissions outside of the ozone season. The 
EPA agrees that solely for the purposes of an ozone program, it may not 
be necessary to report NOX mass emissions outside of the 
ozone season except if a source wants to qualify for the low mass 
emissions provision. However the requirements of subpart H could be 
used to support NOX mass emission reduction programs where 
reductions would be required annually. In addition, the monitoring and 
reporting requirements could be used to help consolidate other State or 
Federal reporting that would be required on an annual basis. Therefore 
in the final rule the requirements of subpart H have been modified so 
that they no longer require annual reporting of NOX mass 
emissions, but rather defer to the State or Federal rule that is 
incorporating these requirements to define the applicable time period 
for reporting.
    In addition a new section has been added to subpart H that details 
how the requirements of part 75, which are designed to be used 
annually, should be used if monitoring and reporting is being done for 
only part of the year.
    Some of the most significant differences include:
     Owners and operators of units using the fuel sampling 
procedures in Appendix D must ensure that they have accurate fuel 
sampling information at the beginning of the ozone season. This 
requires either sampling the fuel tank itself before the start of the 
ozone season or meeting the requirements to sample fuel deliveries on a 
year round basis.
     Historical lookback periods for missing data periods only 
need to include data from the ozone season. However, if a monitor is 
out of control at the beginning of the season, historical data from 
seven months ago may represent significantly different operating 
conditions (e.g. fuel burned or use of control equipment). Therefore 
the AAR would have to certify that the operating conditions are 
representative of the previous years operating conditions. If the 
conditions are not representative, the standard missing data procedures 
could not be used. In this case maximum potential NOX mass 
emissions would have to be substituted.
     The owner or operator of a unit must ensure that the 
monitors used for monitoring and reporting are in control. Since CEMS 
require ongoing quality assurance to ensure that they are operating 
properly, owners and operators of units that do not meet this 
requirement during the non-ozone season will have to recertify their 
monitors before the start of the ozone season.

D. Background on Requiring EPA and the State Permitting Authority to 
Approve Alternatives to Part 75

    The proposal would have required owners and operators of units that 
are not subject to the requirements of title IV of the CAA that wish to 
petition for an alternative to any of the requirements of part 75 to 
petition both the state permitting authority and the Administrator. 
Several commenters suggested that approval of one or the other should 
suffice. Some of the commenters also noted that the requirements were 
different for units affected by title IV, who are only required to 
petition the Administrator.
    The EPA agrees that the requirements for units affected by title IV 
and units not affected by title IV are inconsistent. Because of 
different requirements of the Act this inconsistency is necessary. The 
EPA has the sole authority to grant petitions to units affected by 
title IV under Sec. 75.66 of part 75. If a State incorporates those 
monitoring requirements into its State rules, this still does not give 
it the authority to change or waive the monitoring requirements for a 
unit subject to title IV. However, recognizing that granting a petition 
affects the accounting of NOX mass emissions for a State 
program, EPA does intend to work cooperatively with State agencies on 
petition requests that could affect monitoring and reporting of 
NOX mass emissions.
    For sources not affected by title IV that are complying with the 
requirements of subpart H because they have been adopted or 
incorporated into a State SIP, neither EPA nor the State has sole 
authority to approve a petition for an alternative. While the State 
does have the authority to set forth specific monitoring and reporting 
requirements in a SIP and submit those requirements for EPA approval, a 
State does not have the discretion to modify the SIP by changing or 
waiving those monitoring and reporting requirements without obtaining 
EPA approval. Likewise, EPA does not have sole authority to revise a 
SIP since the primary responsibility to develop and implement a SIP is 
granted

[[Page 57484]]

to the States under the CAA. The EPA is however required by the CAA to 
review and approve or disapprove SIP revisions. Since a petition to 
change or waive unspecified requirements related to monitoring and 
reporting can not be approved as part of the original SIP approval 
process, EPA must be involved in any approvals of alternatives to the 
SIP.
    In addition to the title I requirements for EPA to be involved in 
approval of petitions for alternatives to part 75, there are several 
other reasons that EPA needs to be involved. The first is that since 
EPA is administering the emissions data collection system under part 
75, EPA must ensure that any changes to the reporting requirements can 
be handled by the emissions tracking system that EPA maintains. 
Secondly, in order to ensure the integrity of a multi-state market 
based system and to ensure that participants in the system are treated 
equitably, it is important to ensure that sources are treated equitably 
from State to State. Therefore, if interstate trading is taking place 
EPA clearly has a role in approving petitions for alternatives to 
ensure that sources are treated consistently from state to state when 
engaging in such interstate commerce.

II. Low Mass Emissions Excepted Monitoring Methodology

A. Background

    In the January 11, 1993 Acid Rain permitting rule, EPA provided for 
a conditional exemption from the emissions reduction, permitting, and 
emissions monitoring requirements of the Acid Rain Program for new 
units having a nameplate capacity of 25 MWe or less that burn fuels 
with a sulfur content no greater than 0.05 percent by weight, because 
of the de minimis nature of their potential SO2, 
CO2 and NOX emissions (see 58 FR 3593-94 and 
3645-46). Moreover, in the January 11, 1993 monitoring rule, EPA 
allowed gas-fired and oil-fired peaking units to use the provisions of 
Appendix E, instead of CEMS, to determine the NOX emission 
rate, stating that this was a de minimis exception. The EPA allowed 
this exception from the requirements of section 412 of the CAA because 
the NOX emissions from these units would be extremely low, 
both collectively and individually (see 58 FR 3644-45). One utility 
wrote to the Agency, suggesting that the Agency consider further 
regulatory relief for other units with extremely low emissions that do 
not fall under the categories of small new units burning fuels with a 
sulfur content less than or equal to 0.05 percent by weight or gas-
fired and oil-fired peaking units (see Docket A-97-35, Item II-D-31). 
The utility specifically suggested that the Agency consider an 
exemption, the ability to use Appendix E, or some other simplified 
methods which are more cost effective.
    In the process of implementing part 75, other utilities also have 
suggested to EPA that it provide regulatory relief to low mass emitting 
units (see Docket A-97-35, Items II-D-29, II-E-25). These units might 
be low mass emitting because they use a clean fuel, such as natural 
gas, and/or because they operate relatively infrequently. Some 
utilities stated that they spend a great deal of time reviewing the 
emissions data when preparing quarterly reports for these units. Others 
argued that it would be important to reduce monitoring and quality 
assurance (QA) requirements in order to save time and money currently 
devoted to units with minimal emissions (see Docket A-97-35, Item II-E-
25).
    In response to the requests for simplified monitoring and 
recordkeeping requirements for units which both operate infrequently 
and have low mass emissions on May 21, 1998 the Agency proposed, under 
Sec. 75.19 of part 75, changes to the monitoring requirements that 
would allow a new excepted methodology for low mass emission units. The 
proposed low mass emissions methodology would have allowed units which 
have emissions less than 25 tons of both NOX and 
SO2 to use a methodology with reduced monitoring, reporting 
and quality assurance requirements than the use of CEMS or either 
appendix D or E methodologies. The methodology proposed used a unit's 
maximum rated hourly heat input and generic defaults for 
SO2, NOX and CO2 mass emissions. The 
proposed methodology was a less accurate methodology for determining 
emissions for SO2, NOX and CO2 but 
would significantly reduce the burden on industry for these sources. 
The allowance of this methodology was justified using the de minimis 
individual and aggregate emissions represented by the units who would 
qualify for the methodology.
    While the proposed methodology did not contain an explicit cutoff 
for CO2, EPA believes that the limited applicability of the 
proposal ensured that emissions of CO2 from units that would 
qualify to use the proposal was also de minimis. This is important, 
because under section 821 of the Act, the agency is also required to 
collect CO2 emissions data from sources subject to title IV. 
This data is required to be collected ``in the same manner and to the 
same extent'' as required under title IV.
    The Agency solicited comments on both the proposed methodology for 
determining emissions and the proposed applicability limits of 25 tons 
for both NOX and SO2 as well as any other 
comments related to the proposed low mass emission methodology. In 
reviewing the comments submitted on the proposal, the Agency noted that 
several commenters suggested the methodology was too restrictive and 
would only allow reduced monitoring to a limited number of units. The 
commenters suggested various methods for expanding applicability to the 
low mass emission methodology the most common which are; (i) remove the 
requirement for units to have both SO2 and NOX 
emissions of less than 25 tons and instead to allow units to use the 
methodology on a pollutant specific basis; (ii) increase the 25 ton 
limit for NOX and SO2 to 50, 100 or 250 tons; 
(iii) allow additional methods for calculating heat input; and (iv) 
allow the use of unit-specific NOX emission rates. One other 
significant comment was received which indicated that the default 
values for NOX emission rate in table 1b of proposed 
Sec. 75.19 (c) could significantly underestimate emissions from certain 
types of units.
    In response to the comments, which generally advocating the 
applicability of the low mass emissions methodology to more units, the 
Agency is adopting the proposed low mass emissions methodology with the 
following changes: (1) the NOX applicability limit is being 
raised to 50 tons which will increase the number of units that can use 
the methodology; (2) units are being allowed an optional procedure for 
heat input which will increase the number of units that can use the 
methodology and provide more accurate emission estimates; (3) units are 
being allowed to use unit-specific NOX emission rates 
determined through testing which will allow increased applicability and 
more accurate emissions estimates for NOX; and (4) the 
values for NOX emission rate in table 1b of proposed 75.19 
(c) are being changed to prevent underestimation of emissions using the 
methodology.

B. Discussion of Low Mass Emissions Methodology

    Today's new Low Mass Emissions methodology incorporates optional 
reduced monitoring, quality assurance, and reporting requirements into 
part 75 for units that burn only natural gas or fuel oil, emit no more 
than 25 tons of SO2 and no more than 50 tons of 
NOX annually, and have calculated annual

[[Page 57485]]

SO2 and NOX emissions that do not exceed such 
limits. Units that are not subject to Title IV of the Act and that are 
only subject to subpart H of part 75 are not required to meet the 
SO2 limit to qualify to use the methodology. In addition, if 
allowed by their State, they may qualify as low mass emission units 
during the ozone season if they emit less than 25 tons of 
NOX per ozone season.
    A unit may initially qualify for the reduced requirements by 
demonstrating to the Administrator's satisfaction that the unit meets 
the applicability criteria in Sec. 75.19(a). Section 75.19(a) requires 
facilities to submit historical actual (or projections, as described 
below) and calculated emissions data from the previous three calendar 
years demonstrating that a unit falls below the 25-ton cutoff for 
SO2 and the 50 ton cutoff for NOX. The calculated 
SO2 mass emissions data for the previous three calendar 
years will be determined by choosing one of the two heat input options 
in Sec. 75.19(c) and the appropriate emission rate from table 1a in 
Sec. 75.19(c). The calculated NOX mass emissions data for 
the previous three calendar years will be determined by choosing one of 
the two heat input options in Sec. 75.19(c) and either the appropriate 
emission rate from table 1b in Sec. 75.19(c) or a unit-specific 
NOX emission rate as allowed under Sec. 75.19(c). The data 
demonstrating that a unit meets the applicability requirements of 
Sec. 75.19(a) will be submitted in a certification application for 
approval by the Administrator to use the low mass emissions excepted 
methodology.
    For units that lack historical data for one or more of the previous 
three calendar years (including new units that lack any historical 
data), Sec. 75.19(a) will require the facility to provide (1) any 
historical emissions and operating data, beginning with the unit's 
first calendar year of commercial operation, that demonstrates that the 
unit falls under the 25-ton cutoffs for SO2 and the 50 ton 
cutoff for NOX, both with actual emissions and with 
calculated emissions using the proposed methodology, as described 
below; and (2) a demonstration satisfactory to the Administrator that 
the unit will continue to emit below the tonnage cutoffs (e.g., for a 
new unit, applying the applicable emission rates and applicable hourly 
heat input, under Sec. 75.19(c), to a projection of annual operation 
and fuel usage to determine the projected mass emissions).
    For units with historical actual (or projections, as described 
above) emissions and calculated emissions falling below the tonnage 
cutoffs, facilities allowed to use the optional methodology in 
Sec. 75.19(c) in lieu of either CEMS or, where applicable, in lieu of 
the excepted methods under Appendix D, E, or G for the purpose of 
determining and reporting heat input, NOX emission rate, and 
NOX, SO2, and CO2 mass emissions. The 
facility will no longer be required to keep monitoring equipment 
installed on low mass emissions units, nor will it be required to meet 
the quality assurance test requirements or QA/QC program requirements 
of Appendix B to part 75. Moreover, emissions reporting requirements 
will be reduced by requiring only that the facility report the unit's 
hourly mass emissions of SO2, CO2, and 
NOX, the fuel type(s) burned for each hour of operation, and 
report the quarterly total and year-to-date cumulative mass emissions, 
heat input, and operating time, in addition to the unit's quarterly 
average and year-to-date average NOX emission rate for each 
quarter. Owners and operators may also choose to report partial hour 
operating time and use the operating time to obtain a more accurate 
estimate of heat input determined using the maximum hourly heat input 
option. For units which use the optional long term fuel flow 
methodology for heat input the source will report hourly and cumulative 
quarterly and yearly output in either megawatts electrical output or 
thousands of pounds of steam. For units which use unit-specific 
NOX emission rates determined through testing, reporting of 
the Part 75 Appendix E test results will be required. For units that 
have NOX controls, data demonstrating that these controls 
are operating properly will have to be kept on site. Facilities will 
continue to be required to monitor, record, and report opacity data for 
oil-fired units, as specified under Secs. 75.14(a), 75.57(f), and 
75.64(a)(iii) respectively. Under Sec. 75.14(c) and (d), however, gas-
fired, diesel-fired, and dual-fuel reciprocating engine units will 
continue to be exempt from opacity monitoring requirements.
    If an initially qualified unit subsequently burns fuel other than 
natural gas or fuel oil, the unit will be disqualified from using the 
reduced requirements starting the first date on which the fuel (other 
than natural gas or fuel oil) burned.
    In addition, if an initially qualified unit subsequently exceeds 
the 25-ton cutoff for either SO2 or the 50 ton cutoff for 
NOX while using the adopted methodology, the facility will 
no longer be allowed to use the reduced requirements in Sec. 75.19(c) 
for determining the affected unit's heat input, NOX emission 
rate, or SO2, CO2, and NOX mass 
emissions (unless at a future time the unit can again meet the 
applicability requirements based on the recent three years of data). 
Adopted Sec. 75.19(b) allows the facility two quarters from the end of 
the quarter in which the exceedance of the relevant ton cutoff(s) 
occurred to install, certify, and report SO2, 
CO2, and NOX data from a monitoring system that 
meets the requirements of Secs. 75.11, 75.12, and 75.13, respectively.
    Under the low mass emission excepted methodologies in 
Sec. 75.19(c), a facility will calculate and report hourly 
SO2, NOX and CO2 mass emissions by 
multiplying hourly unit heat input by an appropriate emission rate. 
Unit heat input is determined using one of two heat input 
methodologies, maximum rated hourly heat input or long term fuel flow; 
unit SO2 and CO2 emission rates are determined 
using generic defaults; and unit NOX emission rate is 
determined using one of two methodologies, generic defaults or unit-
specific NOX emission rate testing.
    Commenters raised three major issues, which have led EPA to modify 
its proposal. The three major issues raised were: (i) Should the 
proposed initial and ongoing applicability criteria of 25 tons of both 
NOX and SO2 be modified; (ii) was the proposed 
methodology for estimating emissions appropriate and, should other 
options for calculating emissions be allowed; and (iii) what should the 
reduced monitoring and quality assurance requirements be for these 
units?
1. Applicability Criteria
    a. Approach. Based on the rationale described in the preamble to 
the May 12, 1998 proposal (63 FR 28037) and in the absence of 
significant adverse comment, the Agency is using both actual and 
calculated emissions as the basis for determining initial 
applicability.
    b. Cutoff Limit for Applicability. Several commenters requested 
that the cutoff limit for applicability of the low mass emission 
provision be increased. These comments fell into two broad categories: 
(1) decouple the NOX and SO2 requirements and 
allow units which qualify as a low mass emissions unit for only one 
pollutant to monitor that pollutant using the low mass emissions 
methodology (see Docket A-97-35, Items, IV-D-24, IV-D-11, IV-D-23, IV-
G-03, IV-D-20); and (2) raise the tonnage cutoff for NOX and 
SO2 (see Docket A-97-35, Items, IV-G-03, IV-D-24, IV-D-22, 
IV-D-23, IV-D-07, IV-G-02).
    c. Determining the Criteria for Low Mass Emitters. Based on 
comments received the Agency believes that the

[[Page 57486]]

low mass emission provision is appropriate for units which have low 
mass emissions because: (i) a unit has a low capacity factor usage or 
operates infrequently; or (ii) a unit has low mass emissions despite a 
relatively high capacity factor due to the small size of the unit. For 
these units, the cost of installing and maintaining CEMS would 
represent a relatively large portion of the total value of the 
electricity or steam produced by the unit. The Agency, also reasoned 
that the types of units identified above can use the excepted 
methodology without any significant risk to the environment or 
impairment of the Agency's ability to meet its obligations under the 
CAA.
    The Agency also determined the types of units which were not 
appropriate candidates for use of the low mass emissions excepted 
methodology. In particular, the Agency has concerns about allowing 
large numbers of controlled units to use an estimation methodology such 
as the low mass emission methodology. Because many of these units have 
low mass emissions not because they operate infrequently, but rather 
because they have controls which reduce their emission rates, their 
continued low mass emissions is dependent on continued proper operation 
of the controls on the unit. The EPA believes that monitoring actual 
emission rates is necessary to ensure that installed emission controls 
are operating properly and that actual emissions remain low. On the 
other hand, EPA believes that it is appropriate to allow small or 
infrequently operated units with controls, such as peaking turbines 
with water or fuel injection, to use the low mass emissions provision. 
This is appropriate because as long as these units continue to limit 
their operation, their potential to emit still remains low, even if 
their controls are not working. Therefore, while EPA believes it is 
appropriate to allow small infrequently operated units with controls 
that have both low actual emissions and a low potential to emit (as 
long as they continue to operate at low levels), EPA does not believe 
that it is appropriate to allow controlled units that have large 
potential to emit if their controls are not operating properly to use 
this methodology.
    The low mass emission excepted methodology is a new exception, in 
addition to the exceptions in the existing rule, from the requirement 
for a NOX CEMS. The determination of whether individual and 
collective emissions covered by the exceptions from CEMS are de minimis 
must include consideration of emissions from both new and existing 
units that will qualify to use the new low mass emissions excepted 
methodology and also new and existing units that will qualify to use 
other exceptions from the NOX CEM requirement, i.e. units 
using the existing appendix E excepted methodology and units with new 
unit exemptions under Sec. 72.7.
    The EPA has first considered the level of projected aggregate 
emissions determined to be de minimis for purposes of developing the 
new unit exemption promulgated in the January 11, 1993 Acid Rain 
permitting rule (58 FR 3593-94 and 3645-46). Aggregate emissions 
projected for units under the exemption were approximately 138 
cumulative tons of SO2 and 1934 cumulative tons of 
NOX emitted per year from an estimated 170 new units which 
might qualify for the exception before the year 2000. As of September 
of 1998, 278 exemptions have actually been granted under the new unit 
exemption. The Agency estimates that the level of SO2 and 
NOX mass emissions from these units is 226 tons of 
NOX and 3163 tons of SO2. The Agency further 
believes that this group of excepted units will continue to increase at 
the current rate.
    The EPA has also considered the level of emissions projected to be 
covered by appendix E. The EPA, in the January 11, 1993 Acid Rain 
monitoring rule, allowed gas-fired and oil-fired peaking units to use 
the provisions of appendix E, instead of CEMS, to determine the 
NOX emission rate. The Agency stated that, even though this 
method was less accurate than CEMS, this was a de minimis exception 
because emissions from all units that qualify to use the appendix E 
reporting methodology were projected to be extremely low, the units did 
not have a NOX compliance obligation, and the cost of 
installing and operating CEMS for these units would be high (see 58 FR 
3644-45). The preamble to the January 11, 1993 rule estimated the 
emissions from oil and gas units which operated with a capacity factor 
of less than 10 percent to be 40,000 tons of NOX per year. 
The Agency has analyzed existing appendix E units to determine the 
actual NOX mass emissions reported by these units in 1997. 
This analysis indicates that in 1997 approximately 235 units used the 
appendix E methodology and had total emissions of approximately 11,000 
tons of NOX in 1997. (see Docket A-97-35, Items, IV-A-1).
    The Agency has then considered what level of total NOX 
emissions would be de minimis for all units that may be covered by de 
minimis exceptions from the requirement to use CEMS i.e. all units 
using the new unit exemption, appendix E, and the new low mass 
emissions methodology. The Agency maintains that a de minimis level of 
total NOX emissions should not be more than one percent of 
the total NOX emission inventory currently or in the future 
for all units. This approach is supported by the treatment of 40,000 
tons of NOX as de minimis in the January 11, 1993 rule 
preamble concerning appendix E, which is somewhat less than 1 percent 
of the total NOX emissions estimated for 1993. However, the 
40,000 tons of NOX determined to be de minimis emissions in 
1993 is not an appropriate de minimis level with regard to current and 
future levels of NOX emissions. Several factors have 
increased the importance of monitoring lower levels of NOX 
emissions including: (i) The new more stringent NAAQS for ozone 
(NOX is an ozone precursor); (ii) title IV Phase II 
NOX reductions which will reduce the total NOX 
inventory; (iii) today's NOX SIP call which may result in 
NOX compliance obligations for gas-and oil-fired units and 
will reduce the NOX emission inventory; and (iv) State and 
regional NOX reduction programs, such as the OTC program, 
State RACT rules and the RECLAIM program in California, which result in 
NOX compliance obligations for gas-and oil-fired units and 
reduced NOX emission inventory. As a result, EPA views about 
20,000 tons (close to 1 percent of projected NOX emission 
inventory) as the de minimis level of NOX emissions for the 
present and foreseeable future. Given that appendix E units and new 
unit exemption units currently account for about 14,100 tons of 
NOX there is not a large margin left for establishing 
additional exception to the CEM requirements. The Agency has considered 
potential future growth in the number of units using the new unit 
exemption or appendix E in order to estimate what level of additional 
NOX, SO2 and CO2 emissions might be 
appropriate to allow under the low mass emissions methodology. Taking 
account of the uncertainty inherent in such estimates EPA has set the 
applicability criteria for the low mass emission methodology so that 
the NOX emissions covered by the methodology plus future 
growth in NOX emissions covered by the other current de 
minimis exceptions (appendix E and the new unit exemption) will not 
exceed 5000 tons of NOX per year in the future.
    The Agency has analyzed SO2, NOX and 
CO2 emissions and determined that, as long as the cutoffs 
for NOX and SO2 are coupled so that a unit must 
meet both the 50 tons of NOX and 25 tons of

[[Page 57487]]

SO2 limits, that SO2, NOX and 
CO2 emissions under all exceptions from CEMS requirements 
will remain de minimus. Additionally decoupling the NOX and 
SO2 tons would allow only marginal simplification in 
monitoring while significantly complicating the low mass emissions 
methodology.
    d. Determining the Tonnage Cutoffs for SO2 and 
NOX. The Agency has conducted a study of actual emissions 
data from 1997 quarterly reports under part 75 and evaluated potential 
tonnage cutoffs for SOX and NOX (see Docket A-97-
35, Item IV-A-1). The analysis was based on the assumption that 
reported 1997 emissions of NOX and SO2 will be 
more representative of calculated emissions under the final low mass 
emissions methodology than they would have been under the proposed 
methodology. The assumption is considered valid because the final low 
mass emissions methodology allows more accurate heat input 
determination using long term fuel flow and the use of fuel and unit 
specific NOX emission rates. These options allow more 
accurate emissions estimates than the proposed methodology would have. 
This differs from the analysis performed for the proposed low mass 
emission methodology which calculated emissions based on operating 
hours and maximum rated heat input.
    Based on this analysis, EPA estimates that the existing Acid Rain 
affected sources that would qualify for the low mass emissions excepted 
methodology using a coupled 50 tons NOX and 25 tons 
SO2 limit would represent aggregate emissions of 
approximately 3100 tons of NOX and approximately 260 tons of 
SO2 in 1997 from 224 units. The analysis indicates that the 
applicability has been substantially increased in response to the 
comments received.
    For the proposed 25 ton NOX cutoff , which is the 
limiting factor for applicability in nearly all instances, the Agency 
has considered increasing the tons of NOX to 50 tons, 75 
tons, 100 tons, and 250 tons as suggested by various commenters. In its 
analysis, the Agency kept SO2 at 25 tons, as discussed 
above.
    The analysis showed that by increasing the NOX limit to 
250 tons coupled to 25 tons of SO2, the aggregate tons of 
NOX and SO2 emitted by units which could 
currently qualify for the low mass emissions methodology increased to 
approximately 23124 tons NOX and 4503 tons of 
SO2; this is without considering potential future growth in 
the number of units that could qualify to use this exemption. 
Increasing the cutoff for NOX to 250 tons could also allow 
many units with highly effective NOX controls to use the low 
mass emissions provision. As explained previously, units with effective 
NOX controls and high operating capacity should not use the 
low mass emission provision. The EPA concludes that with a 250 ton 
NOX mass emissions applicability cutoff, the aggregate 
NOX tons and percentage of inventory potentially covered by 
all the exceptions encompassed would easily exceed the de minimis level 
of emissions. The EPA has therefore, not adopted an increased cutoff 
limit for NOX of 250 tons. Similarly, EPA concludes that an 
increased cutoff of 100 tons of NOX would not be consistent 
with the type of source which the Agency has identified for use of the 
low mass emission excepted methodology or fit under the de minimis 
level of emissions defined for NOX by the Agency. At the 100 
ton cutoff for NOX coupled to a 25 ton cutoff for 
SO2 the aggregate NOX emissions are 8841 tons of 
NOX and 540 tons of SO2 from 408 qualifying 
units. The analysis performed by the Agency indicates that 50 tons of 
NOX coupled to 25 tons of SO2 is the appropriate 
cutoff limit for applicability to the low mass emissions excepted 
methodology. The approximate aggregate emissions of 3600 tons of 
NOX and 250 tons of SO2 from 240 sources allows 
the appropriate type of units to use the provisions without great 
potential of exceeding a de-minimus level of NOX emissions. 
In choosing the 50 ton NOX mass emission cutoff limit over 
other limits, the Agency evaluated the available data and applied the 
following criteria: (1) The NOX tons limit should allow 
reduced monitoring for the units which EPA determined were appropriate 
candidates for the low mass emissions provisions during the rulemaking 
process, namely units with low mass emissions both collectively and 
individually due to low operating levels or small size but not highly 
controlled units which operate at higher levels; (2) the NOX 
tons limit should allow reduced monitoring for a group of units 
consistent with the level of de minimis emissions inventory for all 
exceptions for the CEMS requirement; and (3) the limit should not 
jeopardize the Agency's ability to effectively fulfill its obligations 
under of the CAA.
    From the analysis performed, the Agency has demonstrated that 
increasing the 25 ton limit for SO2 would result in allowing 
few additional sources the option to use the low mass emissions 
methodology. For example at a coupled 50 tons of NOX and 25 
tons of SO2 increasing the SO2 tonnage cutoff to 
50 tons would allow only 7 additional units to use the methodology. The 
additional units identified all combusted oil as the primary fuel which 
has a very high sulfur content in comparison to natural gas. While 
natural gas fired units could easily increase operations without 
substantial increases in SO2 emissions oil fired units could 
not. The additional units which burn oil and qualify are considered 
inappropriate candidates for use of the low mass emission provision. 
Therefore, the Agency has chosen to leave the tonnage limit at the 
proposed level of 25 tons for SO2. Leaving the cutoff for 
applicability for SO2 at 25 tons also reflected the opinion 
of commenters who suggested raising only the NOX tonnage.
    When considering the size cutoffs, EPA also took into account both 
the effect that the use of this methodology could have on other 
regulatory actions and the effect that other regulatory actions could 
have on the number of units and percentage of emissions that could be 
covered by units using this methodology. In particular, EPA was 
concerned about the SIP call. Units that could qualify to use the low 
mass emission methodology do not have a NOX emission limit 
under title IV. However, under the SIP call, units that are using the 
monitoring requirements of part 75 to comply with the requirements of 
the SIP call, including units that could qualify to use the low mass 
emitter methodology, would have an emission limit. As explained in 
Section VI.A.2.c and VII.D.3 of today's preamble, EPA believes that it 
is important that large sources of NOX mass emissions 
accurately account for their emissions. Because EPA is expecting 
substantial reductions in NOX emissions from the title IV 
phase II NOX emission rate limits, the SIP call and other 
similar programs, EPA believes that even if the total NOX 
emissions coming from units that could qualify for the low mass emitter 
methodology does not increase, the percentage of emissions coming from 
these units will increase. The EPA also believes that the incentives 
provided under a trading program could encourage smaller oil and gas 
fired units that may not currently qualify under the low mass emission 
methodology to install controls. As a result, this could increase the 
number of units, the amount of emissions and the percentage of 
emissions that could be accounted for by units using this methodology. 
EPA believes that the 50 ton cutoff is adequate to ensure that 
emissions from units that qualify for the low mass

[[Page 57488]]

emitter methodology are de-minimis today. In the future however, growth 
in the number of units may cause the level of NOX, 
SO2 or CO2 emissions from units qualifying for 
and using the new unit exemption, appendix E, the low mass emitter 
provision and other programs such as the SIP call to exceed a de-
minimis level and the agency reserves the right to re-assess any and 
all of these exceptions in the future if the need arises.
    e. Decoupling NOX and SO2. In order to 
qualify for the low mass emissions excepted methodology, the 
applicability criteria require a unit to meet annual tonnage cutoffs of 
25 tons for SO2 and 50 tons for NOX. The EPA has 
considered whether the excepted methodology should be available on a 
pollutant specific level so that, for example, a unit which falls below 
the tonnage cutoff for SO2 but not for NOX could 
use the excepted methodology under Sec. 75.19 to measure SO2 
emissions but use a NOX CEM or the excepted methodology 
under appendix E, where applicable, to measure NOX 
emissions. All analysis the Agency has done indicates that the 
NOX tonnage is the limiting factor for greater than 90 
percent of all units when applicability is for units to meet a coupled 
50 ton NOX and 25 ton SO2 limit (see Docket A-97-
35, Items, II-A-10, IV-A-1) For example, approximately 20 units were 
identified which would potentially be qualified to use the low mass 
emission methodology for a 50 tons of NOX cutoff who would 
not meet the 25 tons of SO2 cutoff and therefore be 
disqualified from using the methodology. Conversely, the agency's 
analysis indicated that leaving the tonnage cutoff for SO2 
mass emissions at 25 tons and decoupling NOX and 
SO2 would potentially allow approximately 650 units in the 
program to use the low mass emissions methodology for SO2 
(see Docket A-97-35, Items, II-A-10, IV-A-1). In particular allowing 
decoupling could impair the Agency's ability to collect data on 
CO2 emissions as required under the CAA section 821. The 
analysis performed by the Agency indicates, that even with a 25 ton 
limit on SO2, 652 units could qualify for the use of the low 
mass emissions methodology for SO2 only. The 652 units 
identified represent approximately 10 percent of the total program heat 
input and greater than 6 percent of the total program CO2 
emissions. If a unit which qualified for the use of only SO2 
were allowed to use the low mass emissions methodology for 
CO2 the result could be overestimation of CO2 
emissions from a sizeable percentage of the total CO2 
inventory. Future decisions based on such data might draw incorrect 
conclusions.
    For the reason stated above, if a unit were allowed to qualify for 
a single pollutant the unit would be allowed to use the low mass 
emissions methodology for that pollutant only and not for 
CO2 or heat input estimations. Therefore, no practical 
benefit for industry would result from decoupling SO2 and 
NOX. Decoupling would not be particularly beneficial because 
qualifying for one pollutant only allows only minimal monitoring 
reductions when CO2 and heat input are not simplified. In 
addition decoupling would dramatically increase the complexity of the 
low mass emissions methodology. The added complications which would 
benefit a limited number of sources in only a limited way would 
increase the time and effort needed for all other sources in 
understanding and implementing the methodology. The agency concludes 
that the burden from the increased rule complexity outweighs the 
benefit from decoupling SO2 and NOX.
    The following discussions further explain the Agencies position.
    One of the prime benefits of the low mass emissions excepted 
methodology will be the simplified reporting which will require less 
time and a less sophisticated Data Acquisition and Handling System 
(DAHS). In particular, the need for a DAHS that could calculate 
substitute data using the current missing data algorithms will be 
removed because there are no missing data algorithms for the low mass 
emissions excepted methodology. If the excepted methodology is only 
applied to one of the pollutants, much of the benefit would be negated 
because the DAHS will still need to be capable of calculating 
substitute data for the measured pollutant and close to the full 
quarterly report would still be required.
    Another prime benefit of the low mass emissions excepted 
methodology will be the reduction of monitoring and quality assurance 
requirements. A unit which would qualify for SO2 only would 
still need to determine CO2 mass emissions using a fuel flow 
meter. Additionally the units which would qualify are primarily gas 
fired units which would be allowed to use appendix D for 
SO2. In this case no benefit is allowed by using the low 
mass emissions methodology. A limited number of oil fired units would 
be granted some reduced sampling requirements.
    The agency's analysis indicates that most units which would qualify 
for NOX only can use the excepted methodology under appendix 
E.
    As stated before the analysis indicates that the benefits of 
decoupling are outweighed by the complications of allowing decoupling.
    f. The use of the Low Mass Emitter Methodology with fuels other 
than oil and natural gas. One commenter suggested that the 
applicability should be expanded to include other fuels including low 
sulfur solid fuels such as wood. EPA disagrees with the commenter who 
claims that the methodology should be irrespective of fuel type. The 
fuel type is an integral part of the emissions calculations and insures 
that emissions are not underestimated. The Agency does not have, and 
the commenter did not provide, sufficient data to justify including 
wood fired solid fuel units into the low mass emission methodology. The 
limited data EPA has does not provide assurance that wood is always low 
in sulfur or that it results in low mass emissions of NOX. 
The use of AP 42 emission factors was considered but rejected based on 
the possibility of underestimation of NOX emissions using 
the AP 42 factors, as stated in the January 11, 1993 rule preamble at 
58 FR 364445. If EPA is provided with information addressing this issue 
in the future, EPA will consider expanding the applicability to units 
that burn wood in the future.
2. Method for Determining Emissions
    On May 21, 1998 the Agency proposed a low mass emissions 
methodology which used maximum rated heat input as the only heat input 
option and default emission rates for SO2, NOX, 
and CO2. The Agency requested comment on whether this 
methodology was appropriate or whether an alternate approach should be 
adopted for low mass emitting units. In response, several commenters 
suggested changing the method for determining emissions. One commenter 
suggested allowing the use of unit-specific NOX testing (see 
Docket A-97-35, Item IV-D-20). Another commenter suggested that long 
term fuel flow heat input be allowed as an alternative to the proposed 
maximum rated heat input (see Docket A-97-35, Item IV-D-13). Two other 
commenters suggested that further unspecified options be allowed for 
determining heat input (see Docket A-97-35, Items, IV-D-03, IV-G-02). 
Additionally several commenters suggested that the reduced monitoring 
under the low mass emission methodology was being limited to too few 
sources (see Docket A-97-35, Items, IV-D-07, IV-D-22, IV-D-23, IV-D-24, 
IV-G-03). Other commenters made the general suggestion that part 75 
should

[[Page 57489]]

be more consistent with the monitoring requirements of the OTC 
NOX Budget Program. Finally the Agency received both 
comments and data which indicated that for uncontrolled gas fired 
turbines combusting both oil and gas the default emission rates for 
NOX in proposed table 1b of Sec. 75.19 (c) were potentially 
substantial underestimations of actual emission from these types of 
units (see Docket A-97-35, Item IV-D-22). Further analysis by the 
Agency provided supporting evidence that the emission rates in proposed 
75.19 (c), table 1b, might underestimate emissions significantly for 
gas and oil fired turbines (see Docket A-97-35, Item IV-A-1). In 
response to these comments which reflected a general desire to expand 
the applicability of the low mass emission methodology through changes 
in both the heat input and NOX emissions methodology, and in 
light of no negative comments reflecting opposition to allowing the low 
mass emission methodology, the Agency began analysis of what changes in 
the methods for determining heat input and NOX emissions 
could be allowed without risk of underestimation of emissions, or 
negative environmental consequences. The Agency received no comments on 
changing either the SO2 or CO2 methods for 
determining emissions and therefore did not attempt to change these 
methodologies.
    a. Adoption of the Proposed Methodology. In the proposal, the 
Agency considered several methods for determining the estimated 
emissions as the basis for applicability of the reduced monitoring and 
reporting excepted methodology. For each of the methods considered, 
rather than using actual measured sulfur and carbon values, 
CO2, SO2, and flow CEM readings, NOX 
CEM readings, or NOX values from an Appendix E 
NOX-versus-heat input correlation, a facility will calculate 
the unit's emissions based on an emission rate factor and one of two 
heat input methodologies. Since the units that will qualify for the 
excepted methodology will still be accountable for reporting emissions 
to the Agency and surrendering allowances based on those emissions, 
where applicable, the emissions estimations will not just be used to 
determine if the unit qualifies under the exception; the reported 
estimations will also be used to determine compliance. Prior to the 
proposal, some industry representatives suggested that facilities would 
be willing to use a conservative emission estimate, such as a maximum 
potential emission rate times the maximum heat input, if it would allow 
them to save time and money currently spent on monitoring and quality 
assurance (see Docket A-97-35, Items II-D-30, II-D-43, II-D-45, II-E-
13, and II-E-25). The Agency decided it was appropriate to retain the 
proposed methodologies of maximum rated heat input and default 
SO2, NOX and CO2 emission rates for 
the final rule. It was also decided to allow increased applicability of 
the low mass emissions methodology through optional unit-specific 
NOX emission rate determinations and the use of an optional 
heat input methodology (e.g., long term fuel flow).
    b. Change in Table 1b, Default NOX Emission Rates. In 
deciding to retain the proposed low mass emission methodology as part 
of the final rule the Agency had to consider that some values for 
NOX emission rate in proposed table 1b of Sec. 75.19 (c) had 
a high potential for underestimating emissions in at least some cases. 
The Agency acknowledged that increasing the default NOX 
emission rates in table 1b of Sec. 75.19 (c) will reduce the number of 
units allowed to use the low mass emissions methodology. Based on the 
comments received (see Docket A-97-35, Item IV-D-20) and to both allow 
increased applicability and increase the default rates to an 
appropriate level, the use of NOX testing to determine 
units-specific NOX emission rates will be allowed as an 
alternative option to using the default NOX emission rates 
in table 1b of Sec. 75.19 (c). Allowing the option of unit-specific 
NOX emission rates will generate more realistic 
NOX emission rates than the default NOX emission 
rates in table 1b of Sec. 75.19 (c) and will maintain some of the 
simplicity of the NOX mass methodology from the low mass 
emissions methodology proposal.
    The next issue was deciding which default NOX emission 
rates in table 1b of Sec. 75.19 (c) to raise and what level to raise 
the defaults to. As a first consideration the Agency noted that the 
default NOX emission rates in table 1b of proposed 
Sec. 75.19 (c) should be increased to the level at which it will be 
highly unlikely that any unit that performed testing will have a higher 
emission rate than the default. In this case, a source might opt to use 
a default which would knowingly underestimate emissions under certain 
operating conditions. Since all of the defaults used in table 1b of 
proposed Sec. 75.19 (c) were based on the 90th percentile it is very 
likely that some units would have a higher emission rate than the 
NOX emission rates in table 1b of proposed 75.19 (c). For 
this reason, all of the NOX emission rate values in proposed 
table 1b were increased to a level which will ensure that units will 
not have higher tested emission rates than the default rates in Table 
1b. A commenter suggested that these provisions be more consistent with 
the provisions for the Ozone Transport Commission (OTC), NOX 
Budget Program (see Docket A-97-35, Item IV-D-13). The default emission 
rates the Agency decided to adopt are the default rates used in the OTC 
NOX Budget Program (see Docket A-97-35, Item II-I-7). In the 
OTC NOX Budget Program, units similar in emission 
characteristics to those who will qualify as low mass emission units 
under today's rule have the option of unit specific testing or unit 
generic default OTC NOX emission rates. In the OTC 
NOX Budget Program units have chosen both options based on 
owner or operator preference. Finally, adopting the NOX 
Budget Program defaults creates consistency among programs which is a 
supplementary benefit.
    c. Unit-Specific NOX Emission Rate Testing. In 
considering the options for unit-specific NOX emission rate 
testing the Agency had to address several concerns, including the 
following: (1) Units with NOX controls who performed unit 
specific testing with the controls operating might have the potential 
to grossly underestimate emissions if the controls failed; (2) what 
sort of test would be appropriate for determining the low mass 
emissions methodology fuel -and-unit-specific NOX emission 
rate; (3) how long a period should a source be allowed to use the unit-
specific NOX rate once determined through testing; (4) under 
what conditions should a source be required to retest for a new unit-
specific NOX emission rate; (5) for sources with historical 
reported emissions data using CEMS under part 75, what historical 
NOX emission rate value might be appropriate for use in lieu 
of an initial test; and (6) if a source owns multiple identical units, 
should representative testing be allowed at some of the units to 
represent all units.
    The first issue resolved was the use of Appendix E of Part 75 
procedures for determination of a unit-specific NOX emission 
rate for each fuel combusted by the unit. The unit-specific 
NOX emission rate selected, for each fuel tested, will be 
the highest recorded NOX emission rate from the test at any 
test load or operating condition multiplied by 1.15. Units which 
combust multiple fuels can use, for different fuels, either a unit-
specific NOX rate determined through testing or use the 
default NOX emission rates listed in table 1b of Sec. 75.19 
(c). For example, a unit which primarily combusts oil but occasionally 
combusts natural gas could determine a unit-specific NOX 
emission rate for oil

[[Page 57490]]

through Appendix E testing and use the default NOX emission 
rate from table 1b of Sec. 75.19 (c) for gas. For hours in which a unit 
combusts multiple fuels in one hour, the unit must use the highest 
emission rate for that hour for all fuels combusted. In conducting the 
Appendix E test, the requirement for monitoring heat input to the unit 
during the test is removed as it is an unnecessary burden. The 
multiplier of 1.15 is required because of Agency analysis which 
indicates that appendix E testing is not representative of emissions at 
a given load at all times. In particular, the analysis of units with 
NOX emission rate CEMS indicated that the NOX 
emission rate can vary an average of 15 percent at a given load during 
different periods of operation. The most probable cause of the 
difference noted is variations in atmospheric moisture content. The 
agency notes that units which do appendix E testing during hot humid 
conditions would likely underestimate emissions during cooler less 
humid conditions. The Appendix E test was chosen for several reasons 
including: (1) many current Acid Rain sources which might qualify for 
the low mass emissions methodology already have performed Appendix E 
testing and will be allowed to use their historical Appendix E test 
data to determine a unit-specific NOX emission rate without 
further requirements; (2) the requirements of Appendix E testing are 
already familiar to sources and contractors who may perform the 
testing, thus reducing further burden imposed by requiring new testing 
methodologies; (3) The use of the Appendix E test and the multiplier of 
1.15 ensures that a unit uses a NOX emission rate which will 
not underestimate emissions at any normal operating condition.
    Once the Appendix E test was chosen, the use of a five year testing 
frequency was deemed appropriate as it matched the current Appendix E 
test period and matches the current permit renewal cycle.
    A special provision was included in the low mass emission 
methodology to allow units with historical CEMS NOX emission 
rate data to determine a unit-specific NOX emission rate 
from historical certified CEMS data. Under this provision a unit will 
analyze historical data from hours in which a unit combusted a 
particular fuel. The analysis will determine the unit-specific 
NOX emission rate which will yield a 95 percent confidence 
that the unit will not emit at a higher NOX emission rate 
while combusting the fuel being analyzed. The Agency also considered 
using the highest NOX rate from historical data but reasoned 
that the large data sets used to generate the unit-and fuel-specific 
emission rate would contain outliers which would make the procedure 
unfeasible for most units. The Agency considered several options for 
units which used NOX controls and wished to use unit-
specific NOX emission rates determined through Appendix E 
testing. One option was to allow units to test with the NOX 
control devices not operating or minimized. This option was rejected 
for the following two reasons: (1) the Agency does not support adopting 
a rule which would require sources to operate in a manner that would 
increase emissions; and (2) some sources which have controls are not 
allowed to operate when the controls are not operating by permit 
restrictions and these units would be disallowed from using the low 
mass emission methodology unfairly. The Agency also considered not 
allowing units with NOX emission controls to use the low 
mass emission methodology. While the Agency does believe that it is not 
appropriate to include large controlled units, the Agency does feel it 
is appropriate to allow infrequently used controlled units, such as 
peaking turbines with steam or water injection to benefit from the 
reduced requirements of this methodology (as further explained above). 
Therefore this solution was rejected as excluding many units for which 
the Agency believes it is appropriate to allow reduced monitoring from 
more accurate and more costly monitoring requirements.
    The Agency also considered allowing only units with certain types 
of controls to use the low mass emission methodology. This approach was 
rejected because the Agency does not, at this time, have the necessary 
information or expertise to make an appropriate determination on this 
approach.
    The Agency also considered allowing units to determine a unit-
specific NOX emission rate using NOX controls 
with no restriction. In analyzing this option, the Agency identified 
several units which would qualify for the low mass emission methodology 
based on the applicability criteria of 50 tons of NOX and 25 
tons of SO2 which the Agency did not believe were 
appropriate to use the low mass emission methodology. The units 
identified had advanced control technologies such as selective 
catalytic reduction (SCR) and burned low sulfur fuels such as natural 
gas. The units identified consistently reported hourly emission rates 
as low as 0.01 lb/mmBtu as compared to uncontrolled rates which are 
generally 10 to 100 times higher for these units. The best method of 
continued assurance that a unit's NOX controls are operating 
is monitoring with a NOX CEMS. These units also operated 
during more than half the hours of a year at an average heat input of 
greater than 1000 mmBtu/hr. While, for these units, the potential to 
underestimate SO2 emissions was low, the potential to 
grossly underestimate NOX mass emissions using the low mass 
emission methodology was much greater. For this reason, the Agency 
rejected allowing a controlled unit to use a single emission rate 
determined through Appendix E testing once every five years while 
NOX controls were operating.
    The methodology the Agency adopted in this rule was the use of a 
lower limit of 0.15 lb/mmBtu for a unit-specific NOX 
emission rate for units which opt to perform unit-and fuel-specific 
Appendix E testing while controls are operating. For units with 
NOX emission controls, which perform unit-specific 
NOX emission rate testing and whose test results in a 
NOX emission rate of less than 0.15 lb/mmBtu, the source 
will use the NOX emission rate limit of 0.15 lb/mmBtu for 
the unit-specific NOX emission rate instead of the lower 
tested NOX emission rate. Units with NOX emission 
controls who perform unit-specific NOX emission rate testing 
and whose results from the testing indicate a NOX emission 
rate of higher than 0.15 lb/mmBtu will be required to use the higher 
NOX emission rate as the fuel-and unit-specific 
NOX emission rate. In considering this approach the Agency 
considered using the lowest NOX emission rate proposed in 
75.19 (c), Table 1b, of 0.172 lb/mmBtu, as well as 0.15 lb/mmBtu, 0.1 
lb/mmBtu and 0.05 lb/mmBtu as lower limits for NOX emission 
rate. The proposed gas fired turbine emission rate was 0.172 lb/mmBtu. 
Using 0.172 lb/mmBtu as the lower limit for controlled units was 
rejected as being an arbitrary choice based on a number representative 
of only a single class of units and not representative of the 
difference between controlled and uncontrolled units. An analysis was 
performed to determine a reasonable lower cutoff between controlled and 
uncontrolled units which would allow controlled units to qualify for 
the reduced monitoring provisions of the excepted low mass emission 
methodology without serious risk of underestimation of emissions. The 
analysis indicated that a minimum allowable emission rate of 0.15 lb/
mmBtu for controlled units best allowed for fairness between controlled 
and uncontrolled units and insured that very

[[Page 57491]]

large units with high operating hours and extremely low NOX 
emission rates will not be allowed to use the low mass emission 
excepted methodology. The Agency's decision was also heavily influenced 
by the desire to insure that overall, the emission rate chosen would 
insure that aggregate emissions of controlled units were indeed de 
minimis. The Agency notes that the lower limit of 0.15 lb/mmBtu 
NOX emission rate, when coupled with the annual limit of 50 
tons of NOX, effectively limits the annual heat input of 
units using the methodology to 666,666 mmBtu annual heat input. 
Analysis done by EPA found this to be an appropriate limit on heat 
input for the low mass emission excepted methodology (see Docket A-97-
35, Item IV-D-20). In general, the lower emission rate limit for 
controlled units, and uncontrolled units inability to achieve such low 
rates, combines to limit the low mass emission methodology to the 
infrequently operated low mass emitting units the Agency was targeting 
for use of the provision in today's new rule.
    Controlled units that use this methodology are also subject to 
additional requirements. The owner or operator of the unit must ensure 
that the controls are being operated in the same manner that they were 
operated during the unit specific testing. Documentation of this must 
be kept on site. Any hour that the controls are not operating properly, 
the owner or operator must use the default emission rates for 
NOX in table 1.b of Sec. 75.19 (c), rather than the emission 
rate determined through unit specific testing.
    Based on experience gained working with the OTC in the 
implementation of the OTC NOX budget program, EPA believes 
that many of the units that may benefit from this new excepted 
monitoring methodology are banks of identical small emission turbines. 
The OTC has allowed these units to do representative sampling at a 
number of units rather than requiring testing at all of the units. 
While none of the commenters mentioned this specific flexibility of the 
OTC NOX Budget program, EPA believes that this is one of the 
flexibilities that commenters who suggested adopting some of the 
methodologies that the OTC has allowed for smaller units were referring 
to. Therefore this final rule contains a similar allowance for 
identical units. If the owner or operator of a number of units that are 
located at one facility can demonstrate that those units are identical, 
this final rule will allow emission rate testing to be done at a 
representative number of units.
     d. The Adoption of Maximum Rated Heat Input as Proposed. While 
several commenters suggested allowing alternative methods for 
determining heat input, none directly suggested replacing or altering 
the basic heat input approach as an option (as described in 68 FR 
28037-8). For this reason the maximum rated hourly heat input option 
from the proposal was retained as a less accurate but acceptable 
approach.
    e. Long Term Fuel Flow for Heat Input Determination. To allow 
greater flexibility to units under the low mass emissions methodology 
and to allow more realistic estimations of heat input as suggested by 
several commenters the Agency is allowing the use of long term fuel 
flow measurements to determine heat input to low mass emitting units as 
described earlier. The Agency chose to adopt this methodology for the 
following reasons: (1) The methodology allows more accurate 
measurements of total heat input into a unit over the reporting period 
than the use of maximum rated hourly heat input; (2) the methodology 
has proven to be usable by sources who have chosen to use a similar 
method in the Ozone Transport Commission, NOX Budget 
Program; and (3) the methodology is straightforward and is optional for 
sources which might be excluded from using the low mass emissions 
methodology if allowed to use maximum rated hourly heat input only.
    3. Reduced Monitoring and Quality Assurance Requirements. As 
discussed above, today's rule allows facilities to use a maximum rated 
hourly heat input value and an emission rate factor to determine the 
mass emissions from a low-emitting unit for each hour of actual 
operation. This approach involves no actual emissions monitoring and 
minimal quality assurance activities. Instead, the facility will only 
need to keep track of whether the unit combusted any fuel for a 
particular hour and what type of fuel was combusted. In this way, the 
revised rule significantly reduces the burden on affected facilities, 
while still ensuring that emissions are not under reported.
    For owners or operators which opt to use either the long term fuel 
flow methodology or a fuel-and unit-specific NOX emission 
rate, some additional quality assurance will be required. As these two 
options under the low mass emission methodology are not required and 
will allow units which would not otherwise qualify to use the low mass 
emission methodology, the additional quality assurance requirements are 
not burdensome to the sources using either long term fuel flow or unit-
specific NOX emission rates.
    For the reasons set forth in the preamble, parts 51, 72, 75, and 96 
of chapter I of title 40 of the Code of Federal Regulations are amended 
as follows:

PART 51--REQUIREMENTS FOR PREPARATION, ADOPTION, AND SUBMITTAL OF 
IMPLEMENTATION PLANS

    1. The authority citation for part 51 continues to read as follows:

    Authority: 42 U.S.C. 7401-7671q.

Subpart G--Control Strategy

    2. Subpart G is amended to add Secs. 51.121 and 51.122 to read as 
follows:


Sec. 51.121  Findings and requirements for submission of State 
implementation plan revisions relating to emissions of oxides of 
nitrogen.

    (a)(1) The Administrator finds that the State implementation plan 
(SIP) for each jurisdiction listed in paragraph (c) of this section is 
substantially inadequate to comply with the requirements of section 
110(a)(2)(D)(i)(I) of the Clean Air Act (CAA), 42 U.S.C. 
7410(a)(2)(D)(i)(I), because the SIP does not include adequate 
provisions to prohibit sources and other activities from emitting 
nitrogen oxides (``NOX'') in amounts that will contribute 
significantly to nonattainment in one or more other States with respect 
to the 1-hour ozone national ambient air quality standards (NAAQS). 
Each of the jurisdictions listed in paragraph (c) of this section must 
submit to EPA a SIP revision that cures the inadequacy.
    (2) Under section 110(a)(1) of the CAA, 42 U.S.C. 7410(a)(1), the 
Administrator determines that each jurisdiction listed in paragraph (c) 
of this section must submit a SIP revision to comply with the 
requirements of section 110(a)(2)(D)(i)(I), 42 U.S.C. 
7410(a)(2)(D)(i)(I), through the adoption of adequate provisions 
prohibiting sources and other activities from emitting NOX 
in amounts that will contribute significantly to nonattainment in, or 
interfere with maintenance by, one or more other States with respect to 
the 8-hour ozone NAAQS.
    (b)(1) For each jurisdiction listed in paragraph (c) of this 
section, the SIP revision required under paragraph (a) of this section 
will contain adequate provisions, for purposes of complying with 
section 110(a)(2)(D)(i)(I) of the CAA, 42 U.S.C. 7410(a)(2)(D)(i)(I), 
only if the SIP revision:

[[Page 57492]]

    (i) Contains control measures adequate to prohibit emissions of 
NOX that would otherwise be projected, in accordance with 
paragraph (g) of this section, to cause the jurisdiction's overall 
NOX emissions to be in excess of the budget for that 
jurisdiction described in paragraph (e) of this section (except as 
provided in paragraph (b)(2) of this section),
    (ii) Requires full implementation of all such control measures by 
no later than May 1, 2003, and
    (iii) Meets the other requirements of this section. The SIP 
revision's compliance with the requirement of paragraph (b)(1)(i) of 
this section shall be considered compliance with the jurisdiction's 
budget for purposes of this section.
    (2) The requirements of paragraph (b)(1)(i) of this section shall 
be deemed satisfied, for the portion of the budget covered by an 
interstate trading program, if the SIP revision:
    (i) Contains provisions for an interstate trading program that EPA 
determines will, in conjunction with interstate trading programs for 
one or more other jurisdictions, prohibit NOX emissions in 
excess of the sum of the portion of the budgets covered by the trading 
programs for those jurisdictions; and
    (ii) Conforms to the following criteria:
    (A) Emissions reductions used to demonstrate compliance with the 
revision must occur during the ozone season.
    (B) Emissions reductions occurring prior to the year 2003 may be 
used by a source to demonstrate compliance with the SIP revision for 
the 2003 and 2004 ozone seasons, provided the SIP's provisions 
regarding such use comply with the requirements of paragraph (e)(3) of 
this section.
    (C) Emissions reduction credits or emissions allowances held by a 
source or other person following the 2003 ozone season or any ozone 
season thereafter that are not required to demonstrate compliance with 
the SIP for the relevant ozone season may be banked and used to 
demonstrate compliance with the SIP in a subsequent ozone season.
    (D) Early reductions created according to the provisions in 
paragraph (b)(2)(ii)(B) of this section and used in the 2003 ozone 
season are not subject to the flow control provisions set forth in 
paragraph (b)(2)(ii)(E) of this section.
    (E) Starting with the 2004 ozone season, the SIP shall include 
provisions to limit the use of banked emissions reduction credits or 
emissions allowances beyond a predetermined amount as calculated by one 
of the following approaches:
    (1) Following the determination of compliance after each ozone 
season, if the total number of emissions reduction credits or banked 
allowances held by sources or other persons subject to the trading 
program exceeds 10 percent of the sum of the allowable ozone season 
NOX emissions for all sources subject to the trading 
program, then all banked allowances used for compliance for the 
following ozone season shall be subject to the following:
    (i) A ratio will be established according to the following formula: 
(0.10)  x  (the sum of the allowable ozone season NOX 
emissions for all sources subject to the trading program)  (the 
total number of banked emissions reduction credits or emissions 
allowances held by all sources or other persons subject to the trading 
program).
    (ii) The ratio, determined using the formula specified in paragraph 
(b)(2)(ii)(E)(1)(i) of this section, will be multiplied by the number 
of banked emissions reduction credits or emissions allowances held in 
each account at the time of compliance determination. The resulting 
product is the number of banked emissions reduction credits or 
emissions allowances in the account which can be used in the current 
year's ozone season at a rate of 1 credit or allowance for every 1 ton 
of emissions. The SIP shall specify that banked emissions reduction 
credits or emissions allowances in excess of the resulting product 
either may not be used for compliance, or may only be used for 
compliance at a rate no less than 2 credits or allowances for every 1 
ton of emissions.
    (2) At the time of compliance determination for each ozone season, 
if the total number of banked emissions reduction credits or emissions 
allowances held by a source subject to the trading program exceeds 10 
percent of the source's allowable ozone season NOX 
emissions, all banked emissions reduction credits or emissions 
allowances used for compliance in such ozone season by the source shall 
be subject to the following:
    (i) The source may use an amount of banked emissions reduction 
credits or emissions allowances not greater than 10 percent of the 
source's allowable ozone season NOX emissions for compliance 
at a rate of 1 credit or allowance for every 1 ton of emissions.
    (ii) The SIP shall specify that banked emissions reduction credits 
or emissions allowances in excess of 10 percent of the source's 
allowable ozone season NOX emissions may not be used for 
compliance, or may only be used for compliance at a rate no less than 2 
credits or allowances for every 1 ton of emissions.
    (c) The following jurisdictions (hereinafter referred to as 
``States'') are subject to the requirements of this section: Alabama, 
Connecticut, Delaware, Georgia, Illinois, Indiana, Kentucky, Maryland, 
Massachusetts, Michigan, Missouri, New Jersey, New York, North 
Carolina, Ohio, Pennsylvania, Rhode Island, South Carolina, Tennessee, 
Virginia, West Virginia, Wisconsin, and the District of Columbia.
    (d)(1) The SIP submissions required under paragraph (a) of this 
section must be submitted to EPA by no later than September 30, 1999.
    (2) The State makes an official submission of its SIP revision to 
EPA only when:
    (i) The submission conforms to the requirements of appendix V to 
this part; and
    (ii) The State delivers five copies of the plan to the appropriate 
Regional Office, with a letter giving notice of such action.
    (e)(1) The NOX budget for a State listed in paragraph 
(c) of this section is defined as the total amount of NOX 
emissions from all sources in that State, as indicated in paragraph 
(e)(2) of this section with respect to that State, which the State must 
demonstrate that it will not exceed in the 2007 ozone season pursuant 
to paragraph (g)(1) of this section.
    (2) The State-by-State amounts of the NOX budget, 
expressed in tons, are as follows:

------------------------------------------------------------------------
                           State                                Budget
------------------------------------------------------------------------
Alabama....................................................      158,677
Connecticut................................................       40,573
Delaware...................................................       18,523
District of Columbia.......................................        6,792
Georgia....................................................      177,381
Illinois...................................................      210,210
Indiana....................................................      202,584
Kentucky...................................................      155,698
Maryland...................................................       71,388
Massachusetts..............................................       78,168
Michigan...................................................      212,199
Missouri...................................................      114,532
New Jersey.................................................       97,034
New York...................................................      179,769
North Carolina.............................................      151,847
Ohio.......................................................      239,898
Pennsylvania...............................................      252,447
Rhode Island...............................................        8,313
South Carolina.............................................      109,425
Tennessee..................................................      182,476
Virginia...................................................      155,718
West Virginia..............................................       92,920
Wisconsin..................................................      106,540
                                                            ------------
    Total..................................................    3,023,113
------------------------------------------------------------------------


[[Page 57493]]

    (3)(i) Notwithstanding the State's obligation to comply with the 
budgets set forth in paragraph (e)(2) of this section, a SIP revision 
may allow sources required by the revision to implement NOX 
emission control measures by May 1, 2003 to demonstrate compliance in 
the 2003 and 2004 ozone seasons using credit issued from the State's 
compliance supplement pool, as set forth in paragraph (e)(3)(iii) of 
this section.
    (ii) A source may not use credit from the compliance supplement 
pool to demonstrate compliance after the 2004 ozone season.
    (iii) The State-by-State amounts of the compliance supplement pool 
are as follows:

------------------------------------------------------------------------
                                                              Compliance
                                                              supplement
                           State                              pool (tons
                                                               of NOX)
------------------------------------------------------------------------
Alabama....................................................       10,361
Connecticut................................................          559
Delaware...................................................          417
District of Columbia.......................................            0
Georgia....................................................       10,919
Illinois...................................................       17,455
Indiana....................................................       19,738
Kentucky...................................................       13,018
Maryland...................................................        3,662
Massachusetts..............................................          285
Michigan...................................................       15,359
Missouri...................................................       10,469
New Jersey.................................................        1,722
New York...................................................        1,831
North Carolina.............................................       10,624
Ohio.......................................................       22,947
Pennsylvania...............................................       13,716
Rhode Island...............................................            0
South Carolina.............................................        5,062
Tennessee..................................................       12,093
Virginia...................................................        6,108
West Virginia..............................................       16,937
Wisconsin..................................................        6,717
                                                            ------------
    Total..................................................      200,000
------------------------------------------------------------------------

    (iv) The SIP revision may provide for the distribution of the 
compliance supplement pool to sources that are required to implement 
control measures using one or both of the following two mechanisms:
    (A) The State may issue some or all of the compliance supplement 
pool to sources that implement emissions reductions during the ozone 
season beyond all applicable requirements in years prior to the year 
2003 according to the following provisions:
    (1) The State shall complete the issuance process by no later than 
May 1, 2003.
    (2) The emissions reduction may not be required by the State's SIP 
or be otherwise required by the CAA.
    (3) The emissions reduction must be verified by the source as 
actually having occurred during an ozone season between September 30, 
1999 and May 1, 2003.
    (4) The emissions reduction must be quantified according to 
procedures set forth in the SIP revision and approved by EPA. Emissions 
reductions implemented by sources serving electric generators with a 
nameplate capacity greater than 25 MWe, or boilers, combustion turbines 
or combined cycle units with a maximum design heat input greater than 
250 mmBtu/hr, must be quantified according to the requirements in 
paragraph (i)(4) of this section.
    (5) If the SIP revision contains approved provisions for an 
emissions trading program, sources that receive credit according to the 
requirements of this paragraph may trade the credit to other sources or 
persons according to the provisions in the trad