Prevention of Significant Deterioration and Title V Greenhouse Gas Tailoring Rule

AGENCY:

Environmental Protection Agency (EPA).

ACTION:

Final rule.

SUMMARY:

EPA is tailoring the applicability criteria that determine which stationary sources and modification projects become subject to permitting requirements for greenhouse gas (GHG) emissions under the Prevention of Significant Deterioration (PSD) and title V programs of the Clean Air Act (CAA or Act). This rulemaking is necessary because without it PSD and title V requirements would apply, as of January 2, 2011, at the 100 or 250 tons per year (tpy) levels provided under the CAA, greatly increasing the number of required permits, imposing undue costs on small sources, overwhelming the resources of permitting authorities, and severely impairing the functioning of the programs. EPA is relieving these resource burdens by phasing in the applicability of these programs to GHG sources, starting with the largest GHG emitters. This rule establishes two initial steps of the phase-in. The rule also commits the agency to take certain actions on future steps addressing smaller sources, but excludes certain smaller sources from PSD and title V permitting for GHG emissions until at least April 30, 2016.

DATES:

This action is effective on August 2, 2010.

ADDRESSES:

EPA has established a docket for this rulemaking under Docket ID No. EPA-HQ-OAR-2009-0517. All documents in the docket are listed in the http://www.regulations.gov index. Although listed in the index, some information is not publicly available, e.g., Confidential Business Information or other information whose disclosure is restricted by statute. Certain other material, such as copyrighted material, is not placed on the Internet and will be publicly available only in hard copy. Publicly available docket materials are available either electronically in http://www.regulations.gov or in hard copy at the EPA Docket Center EPA/DC, EPA West, Room 3334, 1301 Constitution Avenue, Northwest, Washington, DC. The Public Reading Room is open from 8:30 a.m. to 4:30 p.m., Monday through Friday, excluding legal holidays. The telephone number for the Public Reading Room is (202) 566-1744, and the telephone number for the EPA Docket Center is (202) 566-1742.

FOR FURTHER INFORMATION CONTACT:

Mr. Joseph Mangino, Air Quality Policy Division, Office of Air Quality Planning and Standards (C504-03), Environmental Protection Agency, Research Triangle Park, NC 27711; telephone number: (919) 541-9778; fax number: (919) 541-5509; e-mail address: mangino.joseph@epa.gov.

SUPPLEMENTARY INFORMATION:

I. General Information

A. Does this action apply to me?

Entities affected by this action include sources in all sectors of the economy, including commercial and residential sources. Entities potentially affected by this action also include States, local permitting authorities, and tribal authorities. The majority of categories and entities potentially affected by this action are expected to be in the following groups:

B. How is this preamble organized?

The information presented in this preamble is organized as follows:

Outline

I. General Information

A. Does this action apply to me?

B. How is this preamble organized?

C. Preamble Acronyms and Abbreviations

II. Overview of the Final Rule

III. Background

A. What are GHGs and their sources?

B. Endangerment Finding and the LDVR

1. Endangerment Finding

2. Light-Duty Vehicle Rule

C. What are the general requirements of the PSD program?

1. Overview of the PSD Program

2. General Requirements for PSD

D. What are the general requirements of the Title V operating permits program?

1. Overview of Title V

2. Title V Permit Requirements

E. The Interpretive Memo

IV. Summary of Final Actions

A. How do you define the GHG pollutant for PSD and Title V purposes?

1. GHG Pollutant Defined as the Sum-of-Six Well-Mixed GHGs

2. What GWP values should be used for calculating CO₂ e?

B. When will PSD and Title V applicability begin for GHGs and emission sources?

1. What are the Step 1 thresholds, timing, and calculation methodology?

2. What are the Step 2 thresholds, timing, and calculation methodology?

3. What about Step 3?

4. What about the proposed 6-year exclusion for smaller sources?

5. When and how will EPA take further action on smaller sources?

C. How do state, local, and tribal area programs adopt the final GHG applicability thresholds?

D. How do you treat GHGs for purposes of Title V permit fees?

E. Other Actions and Issues

1. Timing for Permit Streamlining Techniques

2. Guidance for BACT Determinations

3. Requests for Higher Category-Specific Thresholds and Exemptions From Applicability

4. Transitional Issues Including Requests for Grandfathering

V. What is the legal and policy rationale for the final actions?

A. Rationale for Our Approach to Calculating GHG Emissions for PSD and Title V Applicability Purposes

1. Grouping of GHGs Into a Single Pollutant

2. Identifying Which GHGs Are Included in the Group

3. Use of GWP vs. Mass-Based GHG Thresholds

4. Determining What GWP Values Are To Be Used

5. Use of Short Tons vs. Metric Tons

B. Rationale for Thresholds and Timing for PSD and Title V Applicability to GHG Emissions Sources

1. Overview

2. Data Concerning Costs to Sources and Administrative Burdens to Permitting Authorities

3. “Absurd Results,” “Administrative Necessity,” and “One-Step-at-a-Time” Legal Doctrines

4. The PSD and Title V Programs

5. Application of the “Absurd Results” Doctrine for the PSD Program

6. Application of the “Absurd Results” Doctrine for the Title V Program

7. Additional Rulemaking for the PSD and Title V Programs

8. Rationale for the Phase-in Schedule for Applying PSD and Title V to GHG Sources

9. “Administrative Necessity” Basis for PSD and Title V Requirements in Tailoring Rule

10. “One-Step-at-a-Time” Basis for Tailoring Rule

C. Mechanisms for Implementing and Adopting the Tailoring Approach

1. PSD Approach: Background and Proposal

2. Rationale for Our Final Approach to Implementing PSD

3. Other Mechanisms

4. Codification of Interpretive Memo

5. Delaying Limited Approvals and Request for Submission of Information From States Implementing a SIP-Approved PSD Program

6. Title V Programs

D. Rationale for Treatment of GHGs for Title V Permit Fees

E. Other Actions and Issues

1. Permit Streamlining Techniques

2. Guidance for BACT Determinations

3. Requests for Higher Category-Specific Thresholds or Exemptions From Applicability

4. Transitional Issues Including Requests for Grandfathering

VI. What are the economic impacts of the final rule?

A. What entities are affected by this final rule?

B. What are the estimated annual benefits to sources due to regulatory relief from the statutory requirements?

1. What are annual estimated benefits or avoided burden costs for title V permits?

2. What are annual benefits or avoided costs associated with NSR permitting regulatory relief?

C. What are the economic impacts of this rulemaking?

D. What are the costs of the final rule for society?

E. What are the net benefits of this final rule?

VII. Comments on Statutory and Executive Order Reviews

A. Comments on Executive Order 12866—Regulatory Planning and Review

B. Comments on the Paperwork Reduction Act

C. Comments on the RFA

D. Comments on the Unfunded Mandates Reform Act

E. Comments on Executive Order 13132—Federalism

F. Comments on Executive Order 13175—Consultation and Coordination With Indian Tribal Governments

G. Comments on Executive Order 13211—Actions That Significantly Affect Energy Supply, Distribution, or Use

VIII. Statutory and Executive Order Reviews

A. Executive Order 12866—Regulatory Planning and Review

B. Paperwork Reduction Act

C. Regulatory Flexibility Act

D. Unfunded Mandates Reform Act

E. Executive Order 13132—Federalism

F. Executive Order 13175—Consultation and Coordination With Indian Tribal Governments

G. Executive Order 13045—Protection of Children From Environmental Health and Safety Risks

H. Executive Order 13211—Actions That Significantly Affect Energy Supply, Distribution, or Use

I. National Technology Transfer and Advancement Act

J. Executive Order 12898—Federal Actions To Address Environmental Justice in Minority Populations and Low-Income Populations

K. Congressional Review Act

L. Judicial Review

IX. Statutory Authority

C. Preamble Acronyms and Abbreviations

The following are abbreviations of terms used in this preamble.

ANPR Advance Notice of Proposed Rulemaking

APA Administrative Procedure Act

AQRVs Air Quality Related Values

BACT Best Available Control Technology

Btu British thermal units

Btu/hr British thermal units per hour

CAA or Act Clean Air Act

CAAAC Clean Air Act Advisory Committee

CAFE Corporate Average Fuel Economy

CH₄ Methane

CO Carbon Monoxide

CO₂ Carbon Dioxide

CO₂ e Carbon Dioxide Equivalent

EPA U.S. Environmental Protection Agency

FDA Food and Drug Administration

FIP Federal Implementation Plan

FTEs Full-Time Equivalents

GHG Greenhouse Gas

GHz Gigahertz

GWP Global Warming Potential

HAP Hazardous Air Pollutant

HFCs Hydrofluorocarbons

ICR Information Collection Request

IPCC Intergovernmental Panel on Climate Change

LDVR Light-Duty Vehicle Rule

MACT Maximum Achievable Control Technology

MCL Maximum Contaminant Level

N₂ O Nitrous Oxide

NAAQS National Ambient Air Quality Standard

NHTSA National Highway Traffic Safety Administration

NMOC Nonmethane Organic Compounds

NO_X Nitrogen Oxides

NPDES National Pollutant Discharge Elimination System

NSPS New Source Performance Standard

NSR New Source Review

NTAA National Tribal Air Association

NTTAA National Technology Transfer and Advancement Act

OMB  Office of Management and Budget

PFCs Perfluorocarbons

PM Particulate Matter

PSD Prevention of Significant Deterioration

PTE Potential to Emit

RFA Regulatory Flexibility Act

RIA Regulatory Impact Analysis

RTC Response to Comment

SBA Small Business Administration

SBAR Small Business Advocacy Review

SBREFA Small Business Regulatory Enforcement Fairness Act

SF₆ Sulfur Hexafluoride

SIP State Implementation Plan

SNPR Supplemental Notice of Proposed Rulemaking

TIP Tribal Implementation Plan

TRS Total Reduced Sulfur

TSD Technical Support Document

tpy Tons Per Year

UMRA Unfunded Mandates Reform Act

UNFCCC United Nations Framework Convention on Climate Change

VOC Volatile Organic Compound

II. Overview of the Final Rule

EPA is relieving overwhelming permitting burdens that would, in the absence of this rule, fall on permitting authorities and sources. We accomplish this by tailoring the applicability criteria that determine which GHG emission sources become subject to the PSD and title V programs of the CAA. In particular, EPA is establishing with this rulemaking a phase-in approach for PSD and title V applicability, and is establishing the first two steps of the phase-in for the largest emitters of GHGs. We also commit to certain follow-up actions regarding future steps beyond the first two, discussed in more detail later. Our legal basis for this rule is our interpretation of the PSD and title V applicability provisions under the familiar Chevron two-step framework for interpreting administrative statutes, taking account of three legal doctrines, both separately and interdependently: They are what we will call (1) The “absurd results” doctrine, which authorizes agencies to apply statutory requirements differently than a literal reading would indicate, as necessary to effectuate congressional intent and avoid absurd results, (2) the “administrative necessity” doctrine, which authorizes agencies to apply statutory requirements in a way that avoids impossible administrative burdens; and (3) the “one-step-at-a-time” doctrine, which authorizes agencies to implement statutory requirements a step at a time. This legal basis justifies each of the actions we take with this rule—e.g., each of the first two steps of the phase-in approach—both (1) as part of the overall tailoring approach, and (2) independently of each other action we take with this rule. EPA also has authority for this Tailoring Rule under CAA section 301(a)(1), which authorizes the Administrator “to prescribe such regulations as are necessary to carry out his functions under [the CAA].”

For the first step of this Tailoring Rule, which will begin on January 2, 2011, PSD or title V requirements will apply to sources' GHG emissions only if the sources are subject to PSD or title V anyway due to their non-GHG pollutants. Therefore, EPA will not require sources or modifications to evaluate whether they are subject to PSD or title V requirements solely on account of their GHG emissions. Specifically, for PSD, Step 1 requires that as of January 2, 2011, the applicable requirements of PSD, most notably, the best available control technology (BACT) requirement, will apply to projects that increase net GHG emissions by at least 75,000 tpy carbon dioxide equivalent (CO₂ e), but only if the project also significantly increases emissions of at least one non-GHG pollutant. For the title V program, only existing sources with, or new sources obtaining, title V permits for non-GHG pollutants will be required to address GHGs during this first step.

The second step of the Tailoring Rule, beginning on July 1, 2011, will phase in additional large sources of GHG emissions. New sources as well as existing sources not already subject to title V that emit, or have the potential to emit, at least 100,000 tpy CO₂ e will become subject to the PSD and title V requirements. In addition, sources that emit or have the potential to emit at least 100,000 tpy CO₂ e and that undertake a modification that increases net emissions of GHGs by at least 75,000 tpy CO₂ e will also be subject to PSD requirements. For both steps, we also note that if sources or modifications exceed these CO₂ e-adjusted GHG triggers, they are not covered by permitting requirements unless their GHG emissions also exceed the corresponding mass-based triggers (i.e., unadjusted for CO₂ e.)

EPA believes that the costs to the sources and the administrative burdens to the permitting authorities of PSD and title V permitting will be manageable at the levels in these initial two steps, and that it would be administratively infeasible to subject additional sources to PSD and title V requirements at those times. However, we also intend to issue a supplemental notice of proposed rulemaking (SNPR) in 2011, in which we will propose or solicit comment on a third step of the phase-in that would include more sources, beginning by July 1, 2013. In the same rulemaking, we may propose or solicit comment on a permanent exclusion from permitting for some category of sources, based on the doctrine of “absurd results,” within the Chevron framework. We are establishing an enforceable commitment that we will complete this rulemaking by July 1, 2012, which will allow for 1 year's notice before Step 3 would take effect.

In addition, we commit to explore streamlining techniques that may well make the permitting programs much more efficient to administer for GHGs, and that therefore may allow their expansion to smaller sources. We expect that the initial streamlining techniques will take several years to develop and implement.

We are also including in this action a rule that no source with emissions below 50,000 tpy CO₂ e, and no modification resulting in net GHG increases of less than 50,000 tpy CO₂ e, will be subject to PSD or title V permitting before at least 6 years from now, April 30, 2016. This is because we are able to conclude at the present time that the administrative burdens that would accompany permitting sources below this level will be so great that even the streamlining actions that EPA may be able to develop and implement in the next several years, and even with the increases in permitting resources that we can reasonably expect the permitting authorities to acquire, it will be impossible to administer the permit programs for these sources until at least 2016.

Further, we are establishing an enforceable commitment that we will (1) Complete a study by April 30, 2015, to evaluate the status of PSD and title V permitting for GHG-emitting sources, including progress in developing streamlining techniques; and (2) complete further rulemaking based on that study by April 30, 2016, to address the permitting of smaller sources. That rulemaking may also consider additional permanent exclusions based on the “absurd results” doctrine, where applicable.

This Tailoring Rulemaking is necessary because without it, PSD and title V would apply to all stationary sources that emit or have the potential to emit more than 100 or 250 tons of GHGs per year beginning on January 2, 2011. This is the date when EPA's recently promulgated Light-Duty Vehicle Rule (LDVR) takes effect, imposing control requirements for the first time on carbon dioxide (CO₂) and other GHGs. If this January 2, 2011 date were to pass without this Tailoring Rule being in effect, PSD and title V requirements would apply at the 100/250 tpy applicability levels provided under a literal reading of the CAA as of that date. From that point forward, a source owner proposing to construct any new major source that emits at or higher than the applicability levels (and which therefore may be referred to as a “major” source) or modify any existing major source in a way that would increase GHG emissions would need to obtain a permit under the PSD program that addresses these emissions before construction or modification could begin. Similarly, title V would apply to a new or existing source exceeding the 100 tpy applicability level in the Act, if the source did not already have a title V permit.

Under these circumstances, many small sources would be burdened by the costs of the individualized PSD control technology requirements and permit applications that the PSD provisions, absent streamlining, require. Additionally, state and local permitting authorities would be burdened by the extraordinary number of these permit applications, which are orders of magnitude greater than the current inventory of permits and would vastly exceed the current administrative resources of the permitting authorities. Permit gridlock would result with the permitting authorities able to issue only a tiny fraction of the permits requested.

These impacts—the costs to sources and administrative burdens to permitting authorities—that would result from application of the PSD and title V programs for GHG emissions at the statutory levels as of January 2, 2011, are so severe that they bring the judicial doctrines of “absurd results,” “administrative necessity,” and “one-step-at-a-time” into the Chevron two-step analytical framework for statutes administered by agencies. Under the U.S. Supreme Court's decision in Chevron, the agency must, at Step 1, determine whether Congress's intent as to the specific matter at issue is clear, and, if so, the agency must give effect to that intent. If congressional intent is not clear, then, at Step 2, the agency has discretion to fashion an interpretation that is a reasonable construction of the statute.

To determine congressional intent, the agency must first consider the words of the statutory requirements, and if their literal meaning answers the question at hand, then, in most cases, the agency must implement those requirements by their terms. However, under the “absurd results” doctrine, the literal meaning of statutory requirements should not be considered to indicate congressional intent if that literal meaning would produce a result that is senseless or that is otherwise inconsistent with—and especially one that undermines—underlying congressional purpose. In these cases, if congressional intent for how the requirements apply to the question at hand is clear, the agency should implement the statutory requirements not in accordance with their literal meaning, but rather in a manner that most closely effectuates congressional intent. If congressional intent is not clear, then an agency may select an interpretation that is reasonable under the statute.

Under the “administrative necessity” doctrine, Congress is presumed, at Chevron Step 1, to intend that its statutory directives to agencies be administrable, and not to have intended to have written statutory requirements that are impossible to administer. Therefore, under this doctrine, an agency may depart from statutory requirements that, by their terms, are impossible to administer, but the agency may depart no more than necessary to render the requirements administrable. Under the “one-step-at-a-time” doctrine, Congress is presumed at Chevron Step 1 to have intended to allow the agency to administer the statutory requirements on a step-by-step basis, as appropriate, when the agency remains on track to implement the requirements as a whole. Each of these doctrines supports our action separately, but the three also are intertwined and support our action in a comprehensive manner.

Here, we have determined, through analysis of burden and emissions data as well as consideration of extensive public comment, that the costs to sources and administrative burdens to permitting authorities that would result from application of the PSD and title V programs for GHG emissions at the statutory levels as of January 2, 2011 should be considered “absurd results.” Therefore, we conclude that under the “absurd results” doctrine, Congress could not have intended that the PSD or title V applicability provisions—in particular, the threshold levels and timing requirements—apply literally to GHG sources as of that date.

Even so, the PSD and title V provisions and their legislative history do indicate a clear congressional intent, under Chevron Step 1, as to whether the two permitting programs applied to GHG sources, and that the intent was in the affirmative, that the permitting programs do apply to GHG sources. Our previous regulatory action defining the applicability provisions made this clear, and we do not reopen this issue in this rulemaking. Moreover, even if this long-established regulatory position were not justifiable based on Chevron Step 1—on the grounds that in fact, congressional intent on this point is not clear—then we believe that this position, that the statutory provisions to apply PSD and title V generally to GHG sources, was justified under Chevron step 2.

As to how to apply the PSD program to GHG sources, congressional intent, as expressed in the various statutory provisions and statements in the legislative history, is clear that PSD should apply at least to the largest sources initially, at least to as many more sources as possible and as promptly as possible over time—consistent with streamlining actions that we intend to consider coupled with increases in permitting authority resources—and at least to a certain point. This is the approach we take in this Tailoring Rule, and because it is consistent with congressional intent, we believe it is required under Chevron Step 1. Even if congressional intent were not clear as to how to apply the PSD requirements to GHG sources, we would have authority under Chevron Step 2 to establish a reasonable interpretation that is consistent with the PSD provisions, and we believe that the tailoring approach so qualifies.

As for title V, the statutory provisions and legislative history, which of course are different than those concerning the PSD program, do not express a clear intent as to how title V applies to GHG sources, which leads our analysis to Chevron Step 2, and here, again, we believe that the tailoring approach is a reasonable interpretation that is consistent with the title V provisions.

For both PSD and title V, we intend to use the tailoring approach to address smaller GHG sources over time, consistent with Congress's expectations that the programs would not impose undue costs to sources or undue administrative burdens to permitting authorities. However, we cannot say at this point how close to the statutory thresholds we will eventually reach. Because this rule establishes only the first two phases of the tailoring approach, we do not find it necessary to answer these questions in this rule, and instead we expect to resolve them through future rulemaking. We will remain mindful of the concerns that Congress expressed about including small sources in either program. We intend to consider the issue of the applicability of title V to GHG sources without applicable requirements (i.e., “empty permits”) in future steps of our tailoring approach. When we do so, we will further assess the potential for the approach of excluding empty permits from title V to relieve burden consistent with statutory requirements.

In addition, because Congress can be said to have intended the PSD and title V programs to apply to GHG sources, the Tailoring Rule is also justifiable under the “administrative necessity” and “one-step-at-a-time” doctrines.

The legal analysis just described justifies each of the actions in this rule. The first two steps that we promulgate in this rule, which take effect on January 2, 2011 and July 1, 2011, constitute the most that permitting authorities can reasonably be expected to do by those times. Similarly, the 50,000 tpy floor that we promulgate through at least April 30, 2016 is reasonable because the information we have available now shows that it constitutes the most that permitting authorities can reasonably be expected to do by that date. Finally, the study and two additional rulemakings—to take effect by July 1, 2013 and April 30, 2016—to which we commit in this rule establish a track for acquiring additional information and for taking further steps to address the application of PSD and title V more closely to the literal statutory levels. We intend to apply them as closely to those levels as is consistent with congressional intent and administrative imperatives, in light of the “absurd results,” “administrative necessity,” and “one-step-at-a-time” doctrines, although, as noted previously, we will consider in future rulemaking how closely to the statutory thresholds we will be able to implement the PSD and title V programs as well as what to require with respect to a potentially large number of sources with empty title V permits.

In this rule, we are adopting regulatory language codifying our phase-in approach. As we will explain, many state, local and tribal area programs will likely be able to immediately implement our approach without rule or statutory changes by, for example, interpreting the term “subject to regulation” that is part of the applicability provisions for PSD and title V. We ask permitting authorities to confirm that they will follow this implementation approach for their programs, and if they cannot, then we ask them to notify us so that we can take appropriate follow-up action to narrow our federal approval of their programs before GHGs become subject to regulation for PSD and title V programs on January 2, 2011. Narrowing our approval will ensure that for federal purposes, GHG sources below the size thresholds we establish in this Tailoring Rule are not obligated to hold PSD or title V permits until the states develop and submit revised PSD and title V programs that EPA approves, either because they adopt our tailoring approach or because, if they continue to cover smaller GHG sources, the states have demonstrated that they have adequate resources to administer those programs.

The thresholds we are establishing are based on CO₂ e for the aggregate sum of six greenhouse gases that constitute the pollutant that will be subject to regulation, which we refer to as GHGs. These gases are: CO₂, methane (CH₄), nitrous oxide (N₂ O), hydrofluorocarbons (HFCs), perfluorocarbons (PFCs), and sulfur hexafluoride (SF₆). Thus, in this rule, we provide that PSD and title V applicability is based on the quantity that results when the mass emissions of each of these gases is multiplied by the Global Warming Potential (GWP) of that gas, and then summed for all six gases. However, we further provide that in order for a source's GHG emissions to trigger PSD or title V requirements, the quantity of the GHGs must equal or exceed both the applicability thresholds established in this rulemaking on a CO₂ e basis and the statutory thresholds of 100 or 250 tpy on a mass basis. Similarly, in order for a source to be subject to the PSD modification requirements, the source's net GHG emissions increase must exceed the applicable significance level on a CO₂ e basis and must also result in a net mass increase of the constituent gases combined.

We are adopting this rule after careful consideration of numerous public comments. On October 27, 2009 (74 FR 55292), EPA proposed the GHG Tailoring Rule. EPA held two public hearings on the proposed rule, and received over 400,000 written public comments. The public comment period ended on December 28, 2009. The comments have provided detailed information that has helped EPA understand better the issues and potential impacts of this rule, and the final rule described in this preamble incorporates many of the suggestions we received. We respond to many of these comments in explaining our rationale for the final rule, which is described in section V. The final rule adopts many elements of the proposal but differs from the proposal in several important respects. We proposed to apply PSD and title V to GHG sources that emit or have the potential to emit at least 25,000 tpy CO₂ e, and we proposed a PSD significance level in a range between 10,000 and 25,000 tpy CO₂ e, but based on consideration of the additional information we received and our further analysis, we are finalizing the threshold levels in the amounts and on the schedule described previously. In addition, the mechanism for state, local, and tribal program implementation has been significantly changed to reflect the comments received that we needed to develop an implementation approach that states could adopt under state law more expeditiously.

The remainder of this notice describes our approach and rationale in more detail. Following this overview, section III of this preamble provides background information on the nature of GHG emissions, recent regulatory developments that affect when and how GHG emissions are subject to stationary source permitting, and the general requirements of the PSD and title V programs. Section IV describes in detail the summary of the key actions being taken in this rule, including the determination of emissions, the thresholds and timing for the phase-in, our approach to implementing the phase-in, and the additional future actions we will take. Section V provides a more detailed description of each action, explaining the policy and legal rationale and responding to comments received. Section V begins with our decisions on how to calculate the mass-based and CO₂ e-based emissions used in the phase-in. Section V then turns to our legal and policy rationale for the first two steps of the phase-in, the 50,000 tpy floor, and the subsequent study and rulemakings to determine whether and how smaller sources should be subject to permitting. This section then describes key implementation issues including the approach to state adoption. After describing our plans for follow-up on title V fee programs, the section concludes by describing permit streamlining techniques; guidance on BACT for the GHG sources that are affected under the first two steps of the Tailoring Rule phase-in; requests for exemptions; and transitional issues, including grandfathering. Finally, section VI describes the expected impacts that will result from the phase-in approach (i.e., the narrower application of PSD and title V requirements during the phase-in period) and sections VII and VIII address administrative requirements.

III. Background

A. What are GHGs and their sources?

Greenhouse gases trap the Earth's heat that would otherwise escape from the atmosphere into space, and form the greenhouse effect that helps keep the Earth warm enough for life. Greenhouse gases are naturally present in the atmosphere and are also emitted by human activities. Human activities are intensifying the naturally occurring greenhouse effect by increasing the amount of GHGs in the atmosphere, which is changing the climate in a way that endangers human health, society, and the natural environment.

Some GHGs, such as CO₂, are emitted to the atmosphere through natural processes as well as human activities. Other gases, such as fluorinated gases, are created and emitted solely through human activities. As previously noted, the well-mixed GHGs of concern directly emitted by human activities include CO₂, CH₄, N₂ O, HFCs, PFCs, and SF₆. These six GHGs will, for the purposes of this final rule, be referred to collectively as “the six well-mixed GHGs,” or, simply, GHGs, and together constitute the “air pollutant” upon which the GHG thresholds in this action are based. These six gases remain in the atmosphere for decades to centuries where they become well-mixed globally in the atmosphere. When they are emitted more quickly than natural processes can remove them from the atmosphere, their concentrations increase, thus increasing the greenhouse effect. The heating effect caused by the human-induced buildup of GHGs in the atmosphere is very likely the cause of most of the observed global warming over the last 50 years. A detailed explanation of greenhouse gases, climate change and its impact on health, society, and the environment is included in EPA's technical support document (TSD) for the endangerment finding final rule (Docket ID No. EPA-HQ-OAR-2009-0472-11292).

In the United States, the combustion of fossil fuels (e.g., coal, oil, gas) is the largest source of CO₂ emissions and accounts for 80 percent of the total GHG emissions. Anthropogenic CO₂ emissions released from a variety of sources, including through the use of fossil fuel combustion and cement production from geologically stored carbon (e.g., coal, oil, and natural gas) that is hundreds of millions of years old, as well as anthropogenic CO₂ emissions from land-use changes such as deforestation, perturb the atmospheric concentration of CO₂ and the distribution of carbon within different reservoirs readjusts. More than half of the energy related emissions come from large stationary sources such as power plants, while about a third comes from transportation. Of the six well-mixed GHGs, four (CO₂, CH₄, N₂ O, and HFCs) are emitted by motor vehicles. In the United States industrial processes (such as the production of cement, steel, and aluminum), agriculture, forestry, other land use, and waste management are also important sources of GHGs.

Different GHGs have different heat-trapping capacities. The concept of GWP was developed to compare the heat-trapping capacity and atmospheric lifetime of one GHG to another. The definition of a GWP for a particular GHG is the ratio of heat trapped by one unit mass of the GHG to that of one unit mass of CO₂ over a specified time period. When quantities of the different GHGs are multiplied by their GWPs, the different GHGs can be summed and compared on a CO₂ e basis. For example, CH₄ has a GWP of 21, meaning each ton of CH₄ emissions would have 21 times as much impact on global warming over a 100-year time horizon as 1 ton of CO₂ emissions. Thus, on the basis of heat-trapping capability, 1 ton of CH₄ would equal 21 tons of CO₂ e. The GWPs of the non-CO₂ GHGs range from 21 (for CH₄) up to 23,900 (for SF₆). Aggregating all GHGs on a CO₂ e basis at the source level allows a facility to evaluate its total GHG emissions contribution based on a single metric.

B. Endangerment Finding and the LDVR

1. Endangerment Finding

On April 2, 2007, the U.S. Supreme Court found that GHGs are air pollutants under CAA section 302(g). Massachusetts v. EPA, 549 U.S. 497 (2007). As a result, the Supreme Court found that EPA was required to determine, under CAA section 202(a), whether (1) GHGs from new motor vehicles cause or contribute to air pollution which may reasonably be anticipated to endanger public health or welfare, or (2) the science is too uncertain to make a reasoned decision. After issuing a proposal and receiving comment, on December 7, 2009, the Administrator signed two distinct findings regarding GHGs under CAA section 202(a):

• Endangerment Finding: The Administrator found that the current and projected atmospheric concentrations of the mix of six long-lived and directly emitted GHGs—CO₂, CH₄, N₂ O, HFCs, PFCs, and SF₆ (referred to as “well-mixed greenhouse gases” in the endangerment finding)—are reasonably anticipated to endanger the public health and welfare of current and future generations.
• Cause or Contribute Finding: The Administrator found that the emissions of the single air pollutant defined as the aggregate group of six well-mixed greenhouses gases from new motor vehicles and new motor vehicle engines contributes to the GHG air pollution that threatens public health and welfare.

These findings, which were published December 15, 2009 (74 FR 66496), do not themselves impose any requirements on industry or other entities. However, they were a prerequisite to finalizing the GHG standards for light-duty vehicles, described next.

2. Light-Duty Vehicle Rule

The LDVR, 75 FR 25324 (May 7, 2010), is a joint rule between EPA and the Department of Transportation's National Highway Traffic Safety Administration (NHTSA) that establishes a national program consisting of new standards for light-duty vehicles that will reduce GHG emissions and improve fuel economy. EPA finalized the national GHG emissions standards under the Act, and NHTSA finalized Corporate Average Fuel Economy (CAFE) standards under the Energy Policy and Conservation Act, as amended. The new standards apply to new passenger cars, light-duty trucks, and medium-duty passenger vehicles, starting with model year 2012. The EPA GHG standards are projected to result in an estimated combined average emissions level of 250 grams of CO₂ per mile for model year 2016 vehicles. The standards begin with the 2012 model year, with standards increasing in stringency through model year 2016. The standards are a fleet average for each manufacturer, based on a footprint attribute curve, meaning that the actual target for a vehicle will vary depending on the size of the vehicle. Under the footprint-based standards, each manufacturer will have a GHG standard unique to its fleet, depending on the footprints of the vehicle models produced by that manufacturer. A manufacturer will have separate footprint-based standards for cars and for trucks.

The endangerment and contribution findings described previously require EPA to issue standards under section 202(a) “applicable to emission” of the air pollutant that EPA found causes or contributes to the air pollution that endangers public health and welfare. The final emissions standards satisfy this requirement for GHGs from light-duty vehicles. Under section 202(a), the Administrator has significant discretion in how to structure the standards that apply to the emission of the air pollutant at issue here, the aggregate group of six GHGs. EPA has the discretion under section 202(a) to adopt separate standards for each gas, a single composite standard covering various gases, or any combination of these. In the LDVR, EPA finalized separate standards for N₂ O and CH₄, and a CO₂ standard that provides for credits based on reductions of HFCs, as the appropriate way to issue standards applicable to emission of the single air pollutant, the aggregate group of six GHGs. EPA did not set any standards for PFCs or SF₆, as they are not emitted by motor vehicles.

C. What are the general requirements of the PSD program?

1. Overview of the PSD Program

The PSD program is a preconstruction review and permitting program applicable to new major stationary sources and major modifications at existing major stationary sources. The PSD program applies in areas that are designated “attainment” or “unclassifiable” for a National Ambient Air Quality Standard (NAAQS). The PSD program is contained in part C of title I of the CAA. The “nonattainment new source review (NSR)” program applies in areas not in attainment of a NAAQS or in the Ozone Transport Region and is implemented under the requirements of part D of title I of the CAA. Collectively, we commonly refer to these two programs as the major NSR program. The governing EPA rules are contained in 40 CFR 51.165, 51.166, 52.21, 52.24, and part 51, Appendices S and W. There is no NAAQS for CO₂ or any of the other well-mixed GHGs, nor has EPA proposed any such NAAQS; therefore, unless and until we take further such action, we do not anticipate that the nonattainment NSR program will apply to GHGs.

The applicability of PSD to a particular source must be determined in advance of construction or modification and is pollutant-specific. The primary criterion in determining PSD applicability for a proposed source is whether the source is a “major emitting facility,” based on its predicted potential emissions of regulated pollutants, within the meaning of CAA section 169(1) and either constructs or undertakes a modification. EPA has implemented these requirements in its regulations, which use somewhat different terminology for determining PSD applicability, which is whether the source is a “major stationary source” or whether the proposed project is a “major modification.”

a. Major Stationary Source

Under PSD, a “major stationary source” is any source belonging to a specified list of 28 source categories which emits or has the potential to emit 100 tpy or more of any pollutant subject to regulation under the CAA, or any other source type which emits or has the potential to emit such pollutants in amounts equal to or greater than 250 tpy. We refer to these levels as the 100/250-tpy thresholds. A new source with a potential to emit (PTE) at or above the applicable “major stationary source threshold” is subject to major source NSR. These limits originate from section 169 of the CAA, which applies PSD to any “major emitting facility” and defines the term to include any source that emits or has a PTE of 100 or 250 tpy, depending on the source category. Note that the major source definition incorporates the phrase “subject to regulation,” which, as described later, will begin to include GHGs on January 2, 2011, under our interpretation of that phrase discussed in the recent Interpretive Memo notice. 75 FR 17004, April 2, 2010.

b. Major Modifications

PSD also applies to existing sources that undertake a “major modification,” which occurs: (1) When there is a physical change in, or change in the method of operation of, a “major stationary source;” (2) the change results in a “significant” emission increase of a pollutant subject to regulation (equal to or above the significance level that EPA has set for the pollutant in 40 CFR 52.21(b)(23)); and (3) there is a “significant net emissions increase” of a pollutant subject to regulation that is equal to or above the significance level (defined in 40 CFR 52.21(b)(23)). Significance levels, which EPA has promulgated for criteria pollutants and certain other pollutants, represent a de minimis contribution to air quality problems. When EPA has not set a significance level for a regulated NSR pollutant, PSD applies to an increase of the pollutant in any amount (that is, in effect, the significance level is treated as zero).

2. General Requirements for PSD

This section provides a very brief summary of the main requirements of the PSD program. One principal requirement is that a new major source or major modification must apply BACT, which is determined on a case-by-case basis taking into account, among other factors, the cost effectiveness of the control and energy and environmental impacts. EPA has developed a “top-down” approach for BACT review, which involves a decision process that includes identification of all available control technologies, elimination of technically infeasible options, ranking of remaining options by control and cost effectiveness, and then selection of BACT. Under PSD, once a source is determined to be major for any regulated NSR pollutant, a BACT review is performed for each attainment pollutant that exceeds its PSD significance level as part of new construction or for modification projects at the source, where there is a significant increase and a significant net emissions increase of such pollutant.

In addition to performing BACT, the source must analyze impacts on ambient air quality to assure that no violation of any NAAQS or PSD increments will result, and must analyze impacts on soil, vegetation, and visibility. In addition, sources or modifications that would impact Class I areas (e.g., national parks) may be subject to additional requirements to protect air quality related values (AQRVs) that have been identified for such areas. Under PSD, if a source's proposed project may impact a Class I area, the Federal Land Manager is notified and is responsible for evaluating a source's projected impact on the AQRVs and recommending either approval or disapproval of the source's permit application based on anticipated impacts. There are currently no NAAQS or PSD increments established for GHGs, and therefore these PSD requirements would not apply for GHGs, even when PSD is triggered for GHGs. However, if PSD is triggered for a GHG emissions source, all regulated NSR pollutants which the new source emits in significant amounts would be subject to PSD requirements. Therefore, if a facility triggers review for regulated NSR pollutants that are non-GHG pollutants for which there are established NAAQS or increments, the air quality, additional impacts, and Class I requirements would apply to those pollutants.

The permitting authority must provide notice of its preliminary decision on a source's application for a PSD permit, and must provide an opportunity for comment by the public, industry, and other interested persons. After considering and responding to comments, the permitting authority must issue a final determination on the construction permit. Usually NSR permits are issued by state or local air pollution control agencies, which have their own permit programs approved by EPA in their State Implementation Plans (SIPs). In some cases, EPA has delegated its authority to issue PSD permits to the state or local agency. In other areas, EPA issues the permits under its own authority.

D. What are the general requirements of the title V operating permits program?

1. Overview of Title V

The operating permit requirements under title V are intended to improve sources' compliance with other CAA requirements. The title V program is implemented through regulations promulgated by EPA, 40 CFR part 70, for programs implemented by state and local agencies and tribes, and 40 CFR part 71, for programs generally implemented by EPA.

In summary, the title V program requires major sources (defined and interpreted by EPA to include sources that emit or have a PTE of 100 tpy of any pollutant subject to regulation) and certain other sources to apply for operating permits. Under EPA's long-standing interpretation, a pollutant, such as a GHG, is “subject to regulation” when it is subject to a CAA requirement establishing actual control of emissions. Title V generally does not add new pollution control requirements, but it does require that each permit contain all pollution control requirements or “applicable requirements” required by the CAA (e.g., New Source Performance Standard (NSPS), and SIP requirements, including PSD), and it requires that certain procedural requirements be followed, especially with respect to compliance with these requirements. “Applicable requirements” for title V purposes include stationary source requirements, but do not include mobile source requirements. Other procedural requirements include providing review of permits by EPA, states, and the public, and requiring permit holders to track, report, and annually certify their compliance status with respect to their permit requirements.

2. Title V Permit Requirements

This section provides a brief summary of the requirements of the title V program that are most relevant to this action. A source generally must apply for a title V permit within 1 year of first becoming subject to permitting—for new sources, this is usually within 1 year of commencing operation. The application must include, among other things, identifying information, a description of emissions and other information necessary to determine applicability of requirements and information concerning compliance with those requirements. The permitting authority uses this information to develop the source's operating permit.

Title V permits generally contain the following elements: (1) Emissions limitations and standards to assure compliance with all applicable requirements; (2) monitoring, recordkeeping, and reporting requirements, including submittal of a semiannual monitoring report and prompt reporting of deviations from permit terms; (3) fee payment; and (4) an annual certification of certification by a responsible official. The detailed requirements are set forth at 40 CFR 70.6.

In addition to the permit content requirements, there are procedural requirements that must be followed in issuing title V permits, including (1) Application completeness determination; (2) public notice and a 30-day public comment period, including an opportunity for a public hearing, on draft permits; (3) EPA and affected state review; and (4) a statement of the legal and factual basis of the draft permit. The permitting authority must take final action (issue or deny) on the permit applications within 18 months of receipt. EPA also has 45 days from receipt of a proposed permit to object to its issuance, and citizens have 60 days after that to petition EPA to object to a permit. Permits may also need to be revised or reopened if new requirements come into effect during the permit terms or if the source makes changes that conflict with, or necessitate changes to, the current permit. Permit revisions and re-openings follow procedural requirements which vary depending on the nature of the necessary change to the permit.

E. The Interpretive Memo

On December 18, 2008, EPA issued a memorandum, “EPA's Interpretation of Regulations that Determine Pollutants Covered by Federal Prevention of Significant Deterioration (PSD) Permit Program” (known as the “Johnson Memo” or the “PSD Interpretive Memo,” and referred to in this preamble as the “Interpretive Memo”) that set forth EPA's interpretation regarding which EPA and state actions, with respect to a previously unregulated pollutant, cause that pollutant to become “subject to regulation” under the Act. Whether a pollutant is “subject to regulation” is important for the purposes of determining whether it is covered under the federal PSD and title V permitting programs. The Interpretive Memo established that a pollutant is “subject to regulation” only if it is subject to either a provision in the CAA or regulation adopted by EPA under the CAA that requires actual control of emissions of that pollutant (referred to as the “actual control interpretation”). On February 17, 2009, EPA granted a petition for reconsideration on the Interpretive Memo, and announced its intent to conduct a rulemaking to allow for public comment on the issues raised in the memorandum and on related issues. EPA also clarified that the Interpretive Memo would remain in effect pending reconsideration.

On March 29, 2010, EPA signed a notice conveying its decision to continue applying (with one limited refinement) the Interpretive Memo's interpretation of “subject to regulation” (“Interpretation of Regulations that Determine Pollutants Covered by Clean Air Act Permitting Programs”). See 75 FR 17004. EPA concluded that the “actual control interpretation” is the most appropriate interpretation to apply given the policy implications. However, we refined our interpretation in one respect: we established that PSD permitting requirements apply to a newly regulated pollutant at the time a regulatory requirement to control emissions of that pollutant “takes effect” (rather than upon promulgation or the legal effective date of the regulation containing such a requirement). In addition, based on the anticipated promulgation of the LDVR, we stated that the GHG requirements of the vehicle rule would take effect on January 2, 2011, because that is the earliest date that a 2012 model year vehicle may be introduced into commerce. In other words, the compliance obligation under the LDVR does not occur until a manufacturer may introduce into commerce vehicles that are required to comply with GHG standards, which will begin with model year 2012 and will not occur before January 2, 2011. We also reiterated EPA's interpretation that the 100 tpy major source threshold for title V is triggered only by pollutants “subject to regulation” under the Act, and we defined and applied that term for title V purposes in the same way that we did for PSD purposes. That is, we stated that a pollutant is “subject to regulation” if it is subject to a CAA requirement establishing “actual control of emissions;” that a pollutant is considered “subject to regulation” for title V purposes when such a requirement “takes effect”; and, based on the anticipated promulgation of the LDVR, that the GHG requirements of the vehicle rule would take effect on January 2, 2011.

On April 1, 2010, we finalized the LDVR as anticipated, confirming that manufacturer certification can occur no earlier than January 2, 2011. Thus, under the terms of the final notice for the Interpretive Memo, GHGs become subject to regulation on that date, and PSD and title V program requirements will also begin to apply upon that date.

IV. Summary of Final Actions

This section describes the specific actions we are taking in this final rule. It describes the overall tailoring approach for NSR and title V applicability, the steps we are taking to put it into place, and future actions that we commit to take. The next section, V, provides the legal and policy rationale for these actions. In that section, we provide a description of our rationale and response to comments for each action, presented in the same order as we describe the actions here.

A. How do you define the GHG pollutant for PSD and title V purposes?

1. GHG Pollutant Defined as the Sum-of-Six Well-Mixed GHGs

We are identifying the air pollutant for purposes of PSD and title V applicability to be the pollutant subject to regulation, which is the air pollutant for GHGs identified in EPA's LDVR, as well as EPA's endangerment and contribution findings. In the LDVR, EPA set emissions standards under section 202(a) that were “applicable to emission” of a single air pollutant defined as the aggregate sum of six GHGs. The six GHGs, which are well-mixed gases in the atmosphere, are CO₂, CH₄, N₂ O, HFCs, PFCs, and SF₆. Earlier, EPA made the contribution finding for this single air pollutant.

Furthermore, as proposed, we are using an emissions threshold that allows all six constituent gases to be evaluated using a common metric—CO₂ e. Thus, to determine applicability, a source's GHG emissions are calculated on a CO₂ e basis by multiplying the mass emissions of any of the six GHGs that the source emits by that gas's GWP and then summing the CO₂ e for each GHG emitted by the source. This sum, expressed in terms of tpy CO₂ e, is then compared to the applicable CO₂ e-based permitting threshold to determine whether the source is subject to PSD and title V requirements.

In addition, because we are implementing this phase-in through the term “subject to regulation,” the regulatory language is structured such that the statutory mass-based thresholds (i.e., for PSD, 100/250 tpy for new construction and zero tpy for modifications at a major stationary source, and for title V, 100 tpy) continue to apply. As a result, stationary source apply and stationary sources or modifications that do not meet these thresholds are not subject to permitting requirements. While technically evaluation of the mass-based thresholds is the second step in the applicability analysis, from a practical standpoint most sources are likely to treat this as an initial screen, so that if they would not trigger PSD or title V on a mass basis, they would not proceed to evaluate emissions on a CO₂ e basis. We have treated evaluation of mass-based thresholds as the initial step in our descriptions. As applicable, a source would evaluate these mass-based thresholds by summing each of the six GHGs it emits on a mass basis (i.e., before applying GWP). We expect that it will be very rare for a new stationary source or modification to trigger permitting based on CO₂ e and not also trigger based on mass alone.

Determining permit program applicability for the GHG “air pollutant” by using the sum-of-six GHGs is based on EPA's interpretation that the PSD and title V requirements apply to each “air pollutant” that is “subject to regulation” under another provision of the CAA. As discussed previously, the final LDVR for GHGs makes it clear that the emissions standards EPA adopted are standards applicable to emission of the single air pollutant defined as the aggregate mix of these six well-mixed GHGs. See LDVR, May 7, 2010, 75 FR 25398-99, section III.A.2.c, and 40 CFR 86.1818-12. For reasons explained in more detail in section V, we have determined it is legally required, and preferable from a policy standpoint, for EPA to use the same definition of the air pollutant for permitting purposes as that used in the rule that establishes the control requirements for the pollutant. We also believe there are implementation advantages for applying PSD and title V in this way. Thus, this rule establishes that a stationary source will use the group of six constituent gases for permitting applicability, rather than treating each gas individually. Similarly, you will include all six constituent gases because that is how the air pollutant is defined, even though motor vehicles only emit four of the six.

2. What GWP values should be used for calculating CO₂ e?

We are requiring that wherever you perform an emissions calculations involving CO₂ e for the purposes of determining the applicability of PSD or title V requirements, you use the GWP values codified in the EPA's mandatory GHG reporting rule. This approach will assure consistency between the values required for calculations under the reporting rule and for PSD or title V. In addition, because any changes to Table A-1 of the mandatory GHG reporting rule regulatory text must go through a rulemaking, this approach will assure that the values used for the permitting programs will reflect the latest values adopted for usage by EPA after notice and comment.

B. When will PSD and title V applicability begin for GHGs and emission sources?

Overview

In this action, we establish the first two phases of our phase-in approach, which we refer to as Steps 1 and 2. We also commit to a subsequent rulemaking in which we will propose or solicit comment on establishing a further phase-in, that is, a Step 3, that would apply PSD and title V to additional sources, effective July 1, 2013, and on which we commit to take final action, as supported by the record, by no later than July 1, 2012.

We also commit to undertaking an assessment of sources' and permitting authorities' progress in implementing PSD and title V for GHG sources, and to complete this assessment by 2015. We further commit to completing another round of rulemaking addressing smaller sources by April 30, 2016. Our action in that rulemaking would address permitting requirements for smaller sources, taking into account the remaining problems concerning costs to sources and burdens to permitting authorities. Finally, we determine in this action that we will apply PSD or title V requirements to sources that emit GHGs, or that conduct modifications that result in increases in emissions of GHGs, in amounts of less than 50,000 tpy CO₂ e any earlier than when we take the required further action to address smaller sources by April 30, 2016.

Through this process, we will implement the phase-in approach by applying PSD and title V at threshold levels that are as close to the statutory levels as possible, and do so as quickly as possible, at least to a certain point. The level and timing of the thresholds that we promulgate in future actions will be based on our assessment of the resulting costs to sources and burdens to permitting authorities, and that, in turn, will depend on such variables as our progress in developing streamlining approaches and on permitting authorities' progress in developing permitting expertise and acquiring more resources. At this time, we cannot foresee exactly when or in what manner those developments will occur. Therefore, we cannot promulgate more components of the tailoring approach beyond what we promulgate in this action. We can say only that we may continue the phase-in process with further rulemaking after 2016. Alternatively, we may make a definitive determination in one of the future rulemaking actions that, under the “absurd results” doctrine, PSD or title V applies only to certain GHG sources, and does not apply to the remaining GHG sources, and with that rulemaking, bring this tailoring process to a close.

1. What are the Step 1 thresholds, timing, and calculation methodology?

a. PSD Permitting

Step 1 of the Tailoring Rule phase-in will begin on January 2, 2011. With respect to the PSD program, GHG sources will become subject to PSD for their GHG emissions if they undergo PSD permitting anyway, either for new construction or for modification projects, based on emissions of non-GHG pollutants, in which case they will be subject to the PSD requirements for GHG if they increase GHG emissions by 75,000 tpy CO₂ e or more. Under this step, only these sources, which we refer to as “anyway” PSD sources, will become subject to PSD; no sources will become major sources for PSD purposes or be treated as undertaking modifications that trigger PSD based solely on their GHG emissions. As a result, no additional PSD permitting actions will be necessary solely due to GHG emissions. However, existing or newly-constructed sources that are determined to be major sources based on non-GHG emissions are required to conduct a BACT review for their GHG emissions (from new construction) or emissions increases (from modifications), if they are subject to PSD due to their non-GHG emissions from construction or modification actions and each of the following conditions is met:

(1) The GHG emissions (or net emissions increase) due to the new construction (or modification) project, calculated as the sum of the six well-mixed GHGs on a mass basis (no GWPs applied) exceed a value of 0 tpy; and

(2) The GHG emissions (or net emissions increase) due to the new construction (or modification) project, calculated as the sum of the six well-mixed GHGs on a CO₂ e basis (GWPs applied) equal or exceed a value of 75,000 tpy CO₂ e.

The purpose of the first condition is to determine whether the GHG emissions or net emissions increase has resulted in an “increase in the amount” of an air pollutant as required by the Act. Because EPA has not defined a mass-based regulatory significance level for GHGs, that level, in effect, is treated as zero. See 40 CFR 52.21(b)(23)(ii) and 51.166(b)(23)(ii). In practice, this means any amount of new emissions or an emission increase will exceed the mass-based limit. We are not, at this time, establishing a significance level based on mass emissions, and instead we are establishing one based on CO₂ e that addresses permitting burdens. The zero mass-based amount applies, but only as an initial screen to exclude sources or changes that have no mass increase of GHGs.

b. Title V Permitting

Under Step 1, only sources required to have title V permits for non-GHG pollutants (i.e., “anyway” title V sources) will be required to address GHGs as part of their title V permitting. That is, no sources will become major for title V based solely on their GHG emissions. Note further, however, that the 75,000 tpy CO₂ e limit does not apply to title V, so that anyway title V sources must apply any title V requirements to their GHG emissions. Sources with title V permits must address GHG requirements when they apply for, renew, or revise their permits. These requirements will include any GHG applicable requirements (e.g., GHG BACT requirements from a PSD process) and associated monitoring, record-keeping and reporting. When a permit application is otherwise required, they will also need to identify GHG emissions and other information in that application to the extent required under 40 CFR 70.5(c) and 71.5(c), including information necessary to determine applicable requirements.

2. What are the Step 2 thresholds, timing, and calculation methodology?

a. PSD Permitting

Step 2 will begin July 1, 2011. Under Step 2, anyway PSD sources—that is, sources already subject to PSD based on non-GHGs and covered under Step 1 previously—will remain subject to PSD. In addition, sources with the potential to emit 100,000 tpy CO₂ e or more of GHG will be considered major sources for PSD permitting purposes (provided that they also emit GHGs or some other regulated NSR pollutant above the 100/250 tpy (mass based) statutory thresholds. Additionally, any physical change or change in the method of operation at a major source (including one that is only major due to GHGs) resulting in a net GHG emissions increase of 75,000 tpy CO₂ e or more will be subject to PSD review and requirements with respect to GHGs (provided that it also results in an increase of GHG emissions on a mass basis).

Specifically, for purposes of determining whether a GHG emission source, resulting from either new construction or a physical or operational change at an existing source, is considered a major source under PSD, both of the following conditions must be met:

(1) The GHG emission source, which is not major for another pollutant, emits or has the potential to emit GHG in amounts that equal or exceed the following, calculated as the sum-of-six well-mixed GHGs on a mass basis (no GWPs applied):

• 100 tpy for sources in any of the 28 major emitting facility source categories listed under PSD, or
• 250 tpy for any other stationary source.

(2) The GHG emission source emits or has the potential to emit GHGs in amounts that equal or exceed 100,000 tpy CO₂ e basis.

For determining whether a modification project at a major stationary source is subject to PSD review, both of the following conditions must be met:

(1) The net GHG emissions increase resulting from the project, calculated as the sum-of-six well-mixed GHGs on a mass basis (no GWPs applied) equals or exceeds 0 tpy.

(2) The net GHG emissions increase resulting from the project, calculated as the sum-of-six well-mixed GHGs on a CO₂ e basis (GWPs applied) equals or exceeds 75,000 tpy CO₂ e.

The purpose of the first condition in both of these determinations is to confirm whether the GHG emissions or emissions increase have exceeded, on a mass-basis, the statutory major source thresholds (where the source is not otherwise major) and mass-based statutory significance level for GHGs, which, as noted previously, is 0 tpy. See 40 CFR 52.21(b)(23)(ii) and 51.166(b)(23)(ii).

As an example of how the mass-based test would apply, consider a modification project that results in a 5 tpy increase of GHG emissions on a mass basis, associated with a high-GWP GHG gas (for example, SF₆, with a GWP value of 23,900), but also results in a 100 tpy reduction in CO₂ emissions (assume no other contemporaneous increases or decreases of GHG). In this example, there would be a net decrease of GHG emissions on a mass basis (5 tpy−100 tpy = −95 tpy). Because there is no mass-based increase of GHG, this project does not trigger PSD, despite the fact that the net GWP-adjusted emissions increase of SF₆ in this example would equal 119,500 tpy of CO₂ e and the project would thus exceed 75,000 tpy CO₂ e.

b. Title V Permitting

Under Step 2, “anyway” title V sources—that is, sources already subject to title V based on non-GHGs and that are covered under Step 1 previously—will continue to be subject to title V. In addition, GHG emission sources that equal or exceed the 100,000 tpy CO₂ e threshold will be required to obtain a title V permit if they do not already have one. It is important to note that the requirement to obtain a title V permit will not, by itself, result in the triggering of additional substantive requirements for control of GHG. Rather, these new title V permits will simply incorporate whatever applicable CAA requirements, if any, apply to the source being permitted. Both of the following conditions need to be met in order for title V to apply under Step 2 to a GHG emission source:

(1) An existing or newly constructed source emits or has the potential to emit GHGs in amounts that equal or exceed 100 tpy calculated as the sum of the six well-mixed GHGs on a mass basis (no GWPs applied).

(2) An existing or newly constructed source emits or has the potential to emit GHGs in amounts that equal or exceed 100,000 tpy calculated as the sum of the six well-mixed GHGs on a CO₂ e basis (GWPs applied).

3. What about Step 3?

In this rule, EPA establishes an enforceable commitment to complete another rulemaking no later than July 1, 2012, in which we will propose or solicit comment on a Step 3 of the phase-in and may also consider other approaches that may result in the permanent exclusion of a category of sources from PSD or title V requirements, under the Chevron framework, taking account of the “absurd results” doctrine.

Consistent with our phase-in approach, it is important for us to consider whether, at some point during the implementation of Step 2, it will become possible to administer GHG permitting programs for additional sources. For example, if EPA is able to promulgate measures that streamline programs to at least some extent, if permitting authorities increase their resources, or if implementation experience and more seasoned staff results in more effective use of scarce permitting resources, then we expect that we will be able to phase in the application of PSD and title V to more sources by establishing Step 3. We do not have enough information now to establish a final Step 3, particularly because there will be significant transition occurring in the GHG permitting programs during Steps 1 and 2. However, we believe that it will be possible to develop a record on which to base Step 3 sometime soon after we begin to implement Step 2.

Therefore, we plan to propose a rule in which we solicit comment on or propose lower thresholds for PSD and title V applicability, and we establish an enforceable commitment to finalize a rule in which we address those matters by July 1, 2012. In order to provide a year for permitting authorities and sources to prepare for any additional GHG permitting action in Step 3, we will establish that Step 3 would take effect on July 1, 2013. We also commit to explore, between now and the Step 3 proposal, a wide range of streamlining options. In the proposal, we will take comment on streamlining approaches we think may be viable (except to the extent we will have already issued guidance documents concerning streamlining approaches), and we will address those options in the final rule.

In addition, as part of the Step 3 action, we may solicit comment on a permanent exclusion of certain sources from PSD, title V or both, based on an “absurd results” rationale. For example, we may make a final determination that under the “absurd results” doctrine, PSD and/or title V do not apply to a set of GHG sources that, although above the statutory thresholds for those programs, are too small and relatively inconsequential in terms of GHG contribution. Another type of such exclusion for the title V program could be for sources that would otherwise be required to obtain an “empty permit,” that is, for example, one that would not contain any applicable requirements because there are none that apply to the source. If we promulgate a permanent exclusion, we may conclude that by that time, we will have brought into the PSD and title V programs the full set of sources that would be consistent with congressional intent (or, if congressional intent on that point is unclear, with a reasonable policy consistent with statutory requirements) and, under those circumstances, we would find that such a rule brings the tailoring process to a close. The application of the “absurd results” rationale for a permanent exclusion is discussed in more detail in section V.B, later in this preamble.

4. What about the proposed 6-year exclusion for smaller sources?

The tailoring proposal contemplated at least a 6-year exclusion from permitting for small sources. This proposed exclusion was based on the overwhelming numbers of permitting actions at small sources and the need for time for permitting authorities to secure resources, hire and train staff, and gain experience with GHG permitting for new types of sources and technologies. It was also based on the time needed for EPA to develop, and for states to adopt, streamlining measures to reduce the permitting burden (e.g., concerning PTE, presumptive BACT, or general permits). We therefore proposed such an exclusion, and proposed that it would last 6 years—5 years to complete a required study evaluating permitting burden and assessing the effect of streamlining measures or techniques in reducing this burden, plus an additional year to complete a final rulemaking that would phase in additional sources as appropriate based on the study.

We are finalizing the 6-year exclusion, and for reasons described later, are establishing that in no event will sources below 50,000 tpy CO₂ e be subject to PSD or title V permitting during the 6-year period, nor will modifications be subject to PSD unless they increase emissions by 50,000 tpy CO₂ e or more. The exclusion will last until we take the action described later to address smaller sources, which is required by April 30, 2016. The exclusion provides certainty that, before this date, EPA will not act to cover sources and modifications below these thresholds, including during the required Step 3 rulemaking that will occur in 2012. In effect, this means that Step 3 will establish a major source threshold and significance level no lower than 50,000 tpy CO₂ e. This does not necessarily mean we will cover sources below this level on April 30, 2016. It simply means that the provision we are adopting would assure that EPA does not cover such sources any sooner than that.

5. When and how will EPA take further action on smaller sources?

As we proposed, we are establishing an enforceable commitment to act within 5 years to complete a study projecting the administrative burdens that remain for small sources after permitting authorities have had time to secure resources, hire and train staff, and gain experience with GHG permitting for new types of sources and technologies, and after EPA has had time to develop (and states have had time to adopt) streamlining measures to reduce the permitting burden for such sources. We will use the results of this study to serve as the basis for an additional rulemaking that would take further action to address small sources. Similar to the enforceable commitment to act on Step 3, we are making an enforceable commitment to complete this rulemaking by April 30, 2016.

We cannot predict at this time what form that final action will take. It could function as a Step 4, bringing in additional sources based on, for example, streamlining actions, increased permitting authority resources, and experienced and more efficient permitting staff; and it could further indicate that we intend to follow-up with a Step 5 to bring in more sources. Alternatively, it could also function as a final step excluding certain sources permanently based on our application of the Chevron framework, taking account of the “absurd results” doctrine, and subjecting the remaining sources to permitting. However, whatever final action we take would explain any necessary changes to the Step 3 thresholds and would supersede the 6-year exclusion for sources and modifications below 50,000 tpy CO₂ e.

C. How do state, local and tribal area programs adopt the final GHG applicability thresholds?

We are finalizing our proposed approach to change the definition of “major stationary source” in the PSD implementing regulations, and the “major source” definition in the title V implementing regulations to tailor the application of these permitting programs to GHG emissions. We are also finalizing a significance level for GHG emissions for purposes of defining a major modification under the PSD program, and add an exclusion from PSD and title V permitting for GHG emissions, until we complete a rulemaking required by April 30, 2016, for any sources that are not already subject to PSD and title V permitting and that emit less than 50,000 tpy of CO₂ e.

As explained earlier, we are adopting thresholds that phase in the applicability of GHG permitting over a specified time period. In adopting regulatory changes to implement these thresholds, we follow an approach that is substantively the same as the approach proposed, but takes a different form for purposes of revisions to our PSD and title V regulations. Specifically, in this final rule, for our regulations, in conjunction with the definitions of “major stationary source” and “major modification” (for PSD) and “major source” (for title V), we are adopting a definition of the term “subject to regulation.” Moreover, we are defining this term so that GHG emissions from sources above the threshold are treated as subject to regulation, and therefore the sources that emit them are subject to PSD and title V. We are not finalizing the approach we proposed, which was to revise the numerical thresholds in the definitions so that GHG sources would have a higher threshold. Although we are defining the term “subject to regulation,” we recognize that from a substantive standpoint, our tailoring approach entails interpreting the definitions of “major emitting facility,” “major modification,” and “major source” to phase in the applicability of PSD and title V, as applicable, to GHG sources, and it makes no difference whether we interpret those definitions through a definition of the term “subject to regulation,” revising the numerical thresholds, or revising other terms in those definitions.

We are adopting definitions of the term “subject to regulation” to implement the tailoring approach because that will facilitate rapid implementation of the final rules by states. Under this approach, states may not need to undertake a regulatory or legislative action before implementing the final rule. These states would be able to establish their interpretations of the term “subject to regulation” used in existing state rules before January 2, 2011, which is the date that the LDVR and permitting requirements would take effect, and thereby exempt sources below the threshold from PSD and title V as a matter of both federal and state law. We are also codifying in this definition EPA interpretations discussed in our recent action “Reconsideration of Interpretation of Regulations that Determine Pollutants Covered by Clean Air Act Permitting Programs” (75 FR 17704) to provide a complete picture of the meaning of this phrase as it applies to all air pollutants.

Because we are finalizing the rule in a manner that will allow most states to rapidly implement the final rule, and because our recent action on the Interpretive Memo allowed for a longer transition time than we anticipated at proposal, we are delaying final action on our proposal to issue limited approvals for SIP-approved PSD programs and part 70 operating permit programs. Instead, we are requesting that states submit information to the appropriate EPA Regional Administrator by August 2, 2010 so that we may determine whether it is still necessary to finalize any of our proposed limited approvals for any SIP-approved PSD and part 70 title V state programs. In that letter, states should explain whether they will apply the meaning of the term “subject to regulation” established by EPA in this action in implementing both their PSD and part 70 title V permitting programs, and if so, whether the state intends to do so without undertaking a regulatory or legislative process. If a state must revise its statutes or regulations to implement this rule, we ask that it provide an estimate of the time to adopt final rules in its letter to the Regional Administrator. If a state chooses not to apply the approach reflected in this rule, the letter should address whether the state has alternative authority to implement the final rule's tailoring approach or some other approach that is at least as stringent, but which also addresses the expected shortfalls in personnel and funding that would exist if the state carried out permitting at thresholds lower than those in the final rules. For any state that is unable or unwilling to apply the permitting thresholds in the final rules, and otherwise is unable to demonstrate adequate personnel and funding, or alternate authority to permit GHG emissions sources consistent with the final rules, EPA will move forward with finalizing a limited approval of the state's permitting program. By the same token, if we do not receive a letter from a state in response to this request by August 2, 2010, we will be obliged to move forward with finalizing a narrowing of our approval of the existing SIP or title V program.

We also ask any state that currently lacks authority to issue PSD or title V permits to any GHG emissions sources to notify the EPA Regional Administrator by letter as to whether the state intends to undertake rulemaking to revise its rules consistent with these applicability thresholds. For any state that lacks the ability to issue PSD or title V permits for GHG emissions sources consistent with the final rule, we intend to undertake a separate action to call for revisions to these programs. We also intend to move quickly to impose a Federal Implementation Plan (FIP) for PSD through 40 CFR 52.21, and use our federal title V authority to ensure that GHG sources will be permitted consistent with the final rules. Our request for information from states is discussed further in section V.C.

D. How do you treat GHGs for purposes of title V permit fees?

We are not amending the title V regulations for fees at this time, including any of the provisions specifying the presumptive minimum fee. We are also not, at this time, calling for each state, local or tribal program to submit new fee adequacy demonstrations as a result of increased GHG permitting workload during Steps 1 and 2. However, as described in section VI.D the statutory and regulatory requirement to collect fees sufficient to cover all reasonable (direct and indirect) costs required to develop and administer title V programs still applies. Therefore, we are recommending that each program review its resource needs for GHG-emitting sources and determine if the existing fee approaches will be adequate. If those approaches will not be adequate, we suggest that state, local and tribal agencies should be proactive in raising fees to cover the direct and indirect costs of the program or develop other alternative approaches to meet the shortfall. We will closely monitor approved title V programs during implementation of the first two steps of the Tailoring Rule to ensure that the added workload from incorporating GHGs into the permit program does not result in fee shortfalls that imperil operating permit program implementation and enforcement. In developing alternative approaches, we note the value of approaches that do not require a per-ton fee for GHG and therefore do not require a GHG inventory to develop. Finally, we offer to work with permitting authorities that request our assistance with developing fee approaches.

E. Other Actions and Issues

This section describes other actions we intend to take in the future related to GHG permitting in addition to the actions that we are promulgating with this final rule. This section also responds to commenters' suggestions that we undertake certain additional actions in this rule, which we decline to do.

1. Timing for Permit Streamlining Techniques

As described at proposal, we intend to develop a series of streamlining approaches as an integral part of our phase-in approach. The approaches we described at proposal included: (1) Defining PTE for various source categories, (2) establishing emission limits for various source categories that constitute presumptive BACT, (3) establishing procedures for use of general permits and permits-by-rule, (4) establishing procedures for electronic permitting, and (5) applying lean techniques to establish more efficient permitting processes. Taken as a whole, these techniques have the potential to obviate the applicability of PSD and title V requirements for some GHG-emitting sources; promote more efficient treatment of GHG-emitting sources that will already be subject to PSD and title V; and allow the expeditious expansion of PSD and title V applicability to more GHG-emitting sources while protecting those sources and the permitting authorities from undue expenses.

As a result, we fully intend to move forward expeditiously with developing streamlining approaches. However, for reasons discussed in section V.E, we do not expect to develop and implement any of these approaches before Step 2 begins. Moreover, we generally expect that each of the first three—which are the most far-reaching—will take several years to implement because we will need to undertake notice and comment rulemaking to develop them, and then the permitting authorities will need to adopt them through the appropriate state or local processes. We commit to explore a wide range of possible approaches before the Step 3 rulemaking, and, in that rulemaking, to propose those that we think may be viable once we have had time to gather and review key supporting data, and once the states and we have key implementation experience that can inform our thinking. Because the streamlining approaches generally carry uncertainty—as demonstrated by comments we received raising legal and policy concerns, as discussed later, that we will have to address—we cannot commit with this action to adopt any streamlining actions in particular, nor to adopting them on any particular schedule. However, we intend to pursue streamlining options as expeditiously as possible, beginning immediately and proceeding throughout the phase-in period, and we encourage permitting authorities to do the same.

2. Guidance for BACT Determinations

Through this final rule we are not amending our regulations or issuing guidance on BACT for GHGs. As described in our proposal, we recognize the need to develop and issue technical and policy guidance for permitting of GHGs, and we plan to accomplish it through a separate effort that will involve stakeholder input. This effort is already underway; in addition to comments EPA received on the proposed Tailoring Rule related to GHG BACT guidance and information needs, EPA received a suite of recommendations from the Clean Air Act Advisory Committee (CAAAC) to which EPA is actively responding. This includes technical guidance and database tools that EPA anticipates issuing by June 2010, and policy guidance that will be issued by the end of 2010. Thus, this important information will be available to support permitting agencies in their BACT determinations at the time that the GHGs become a regulated NSR pollutant, once the LDVR takes effect in January 2011. EPA is confident that these measures will help support a smooth transition to permitting emissions of GHGs.

3. Requests for Higher Category-Specific Thresholds and Exemptions From Applicability

EPA has decided not to provide exemptions from applicability determinations (major source and major modification) under title V and PSD for certain GHG emission sources, emission activities, or types of emissions at this time. Commenters requested several applicability exemptions with respect to GHGs from, for example, agricultural sources, residential sources, small businesses, energy-intensive industrial processes (e.g., aluminum, steel, cement, glass, and pulp and paper manufacturers), lime production, semiconductor production, poultry production, solid waste landfills, biomass combustion/biogenic emissions, fugitive emissions, and pollution control projects. For reasons explained in section V.E, we have decided to address the need for tailoring through a uniform threshold-based approach, rather than through a collection of various specific exclusions.

4. Transitional Issues Including Requests for Grandfathering

For reasons explained in section V.E, EPA has determined that transitional issues for pending applications and permitted sources are adequately addressed by existing requirements and the amount of lead time provided before permitting requirements apply to GHGs under this rule and the March 29, 2010 final action regarding the Interpretive memo. This rule does not contain any additional exemptions or grandfathering provisions addressing the transition to PSD and title V permitting for GHGs.

We are not promulgating an exemption for PSD permit applications that are pending when Step 1 of the permitting phase-in begins for those sources that would otherwise need to obtain a PSD permit based on emissions of pollutants other than GHGs. Any PSD permits issued to such Step 1 sources on or after January 2, 2011 will need to address GHGs. This action makes no change to the position we expressed on this issue on April 2, 2010.

Final PSD permits issued before January 2, 2011 need not be reopened or amended to incorporate requirements for GHGs that take effect after the permit is issued. A source that is authorized to construct under a PSD permit but has not yet begun actual construction on January 2, 2011 may begin actual construction after that date without having to amend the previously-issued PSD permit to incorporate GHG requirements, provided the permit has not expired.

Sources that are not subject to PSD permitting requirements until Step 2 need not obtain a PSD permit addressing GHGs in order to continue any actual construction that begins before July 1, 2011, when such a source was not a major stationary source required to obtain a PSD permit. However, Step 2 sources that begin actual construction in Step 2 may do so only after obtaining a PSD permit.

The title V permitting regulations already include a robust set of provisions to address the incorporation of new applicable requirements and other transitional considerations. A title V source applying for the first time must submit its permit application within 12 months after the source becomes subject to the operating permit program or an earlier time at the discretion of the permitting authority. Where a source is required to obtain a PSD permit, the source must apply for a title V permit or permit revision within 12 months of commencing operation or on or before such earlier date as the permitting authority may establish. Where additional applicable requirements become applicable to a source after it submits its permit application, but prior to release of a draft permit, the source is obligated to supplement its application. Permitting authorities may also ask for additional information during the processing of an application. In addition, where a source that already has a title V permit becomes subject to additional applicable requirements, the permitting authority is required to reopen the permit to add those applicable requirements if the permit term has 3 or more years remaining and the applicable requirements will be in effect prior to the date the permit is due to expire.

V. What Is the Legal and Policy Rationale for the Final Actions?

In this section, we describe the legal and policy rationale for our action, including our rationale for the following: (1) Our approach to calculating GHG emissions for PSD and title V applicability purposes, (2) our approach to establishing the thresholds and timing of PSD and title V applicability to GHG emissions sources; (3) how state, local, and tribal area programs adopt the final GHG applicability thresholds; (4) treatment of GHGs for title V permit fees; (5) future activities, including streamlining actions. We present the rationale description in the following five subsections, corresponding to the basic presentation of the approach in section IV.

A. Rationale for Our Approach to Calculating GHG Emissions for PSD and Title V Applicability Purposes

1. Grouping of GHGs Into a Single Pollutant

In this section, we explain our treatment of the air pollutant at issue for purposes of PSD and title V, such that sources that emit that pollutant in the requisite quantities become subject to PSD and/or title V requirements. We explain our rationale for treating the GHG air pollutant as a combined group of six GHGs instead of six separate air pollutants defined by each individual GHG, and our rationale for including all six of the GHGs in that group. We also define the GHG metric to use for comparison to the applicability thresholds.

We proposed to identify the air pollutant as the aggregate group of the six GHGs that comprise the GHG pollutant, and to use a GHG metric for the applicability thresholds based on CO₂ e. The summed CO₂ e emissions would then be compared to the applicable permitting threshold to determine whether the source is subject to PSD and title V requirements. Historically, the PSD and title V regulatory provisions do not, in the first instance, define the “air pollutant” to which they apply, but rather rely for the definition of the pollutant on a cross-reference to the regulatory provision under another part of the Act that establishes the emission standards or limits for that pollutant that in turn causes the pollutant to be subject to regulation under PSD and title V permitting. As an example, the pollutant “total reduced sulfur” (TRS) is a pollutant comprised of the sum of multiple compounds that was originally defined under the NSPS, subpart BB, Standards of Performance for Kraft Pulp Mills, which then caused it to be subject to regulation under the PSD program. The actual compounds that define the pollutant TRS are identified in the NSPS. The PSD program regulations did not introduce its own independent definition of TRS, but instead relied on the definition as contained in the Kraft Pulp Mills NSPS.

However, at the time of our proposal, the endangerment and cause or contribute findings had not been completed and the LDVR for GHGs had not been finalized. Thus, there was no final agency action defining the “air pollutant” consisting of GHGs to be considered “subject to regulation.” Absent a definition of “greenhouse gases” under another regulatory provision that we could cross-reference, we proposed to define “greenhouse gases” for permitting purposes as “the single air pollutant that is comprised of the group of six GHGs, as proposed in the [CAA] section 202(a) endangerment and contribution findings.” 74 FR 55329, col. 1. The six well-mixed GHGs identified in the proposed contribution finding were: CO₂, CH₄, N₂ O, SF₆, HFCs, and PFCs.

In the proposal, we further recognized that the LDVR for GHGs, as it was proposed, would result in reductions of only four of the gases, not all six, because only four are emitted by vehicles. However, we concluded that if the LDVR were finalized as proposed, then the air pollutant for purposes of PSD and title V applicability would be a single air pollutant that is the aggregate mix of the group of six GHGs because—

[t]hese six GHGs as a class comprise the air pollutant that is the subject of the endangerment finding and companion contribution finding and constitute the air pollutant that is regulated by the light-duty vehicle rule through measures that address the components of that air pollutant that are emitted from the mobile sources. Thus, although the CAA section 202(a) proposal establishes controls only with respect to four GHGs, as a legal matter, the proposal covers the entire set of GHGs that as a class are the single “air pollutant” in the proposed endangerment and contribution findings.

74 FR 55329 col. 1.

We also solicited comment on whether we should identify the GHG metric in a different way, such as addressing each GHG constituent compound individually or including (whether individually or as a group) only those four GHG constituent compounds for which reductions would occur through the emission standards or limits proposed in the LDVR.

A minority of the comments on our proposal addressed this issue. Some commenters supported combining the individual GHGs as one pollutant for purposes of determining permitting applicability, and stated that it is not uncommon for EPA to recognize “collective” air pollutants comprised of many individual compounds based upon shared threats to health and welfare, including such EPA-created group pollutants as sulfur oxides, nitrogen oxides, volatile organic compounds (VOCs), and particulate matter (PM).

On the other hand, a significant number of commenters also raised concerns about grouping the individual GHGs into one metric. Some of these commenters argued that grouping GHGs is not appropriate because GHGs are not like other air pollutants that are comprised of numerous substances of concern (e.g., VOCs and PM), individual GHGs do not interact or combine to create a pollutant of concern, and EPA has not established a “GHG” NAAQS that supports the definition of the pollutant as a group. Some were concerned that regulating the GHGs as a group would increase the likelihood that a source will trigger permitting requirements, adding that this is unnecessary and would conflict with the “absurd results” and “administrative necessity” doctrines because it would lead to larger numbers of sources becoming subject to permitting. Some commenters opposing grouping suggested that we should explore regulating each of the GHG pollutants on an individual mass basis rather than collectively because in their view, it is reasonable and feasible to regulate and control emissions of each of the listed pollutants, other than CO₂, at the 100/250 tpy thresholds, or less if deemed necessary, in accordance with the established mechanisms of the Act and doing so would lead to a better environmental result. Finally, some commenters argued that disaggregating the pollutants would also allow for more appropriate technology review.

After considering these comments, and taking into account other related actions that have occurred since proposal, we have determined that PSD and title V permitting program requirements will apply, as proposed, to the “single air pollutant that is comprised of the group of six GHGs.” 74 FR 55329, col. 1. We believe that this approach is both compelled by the statute and reflects the preferable policy approach.

As more fully discussed elsewhere in this rulemaking, the PSD requirements apply to a “major emitting facility” that undertakes construction or “modification.” CAA sections 165(a), 169 (2)(C). The term “major emitting facility” is defined as, in general, a source that emits 100 or 250 tons of “any air pollutant,” CAA section 169(1), and, similarly, the term “modification” is defined as a physical or operational change that results in the increased or new emissions of “any air pollutant.” CAA sections 169(2)(C), 111(a)(4). Through regulation, we have interpreted the term “any air pollutant,” as found in both the terms “major emitting facility” and “modification,” more narrowly to mean any “regulated NSR pollutant,” and we further define this term to include any pollutant that is “subject to regulation under the Act.” 40 CFR 52.21(b)(50)(iv), 52.21.(b)(2).

Similarly, as discussed elsewhere, the title V requirements apply to a “major source,” which is defined, in general, as any source that emits at least 100 tpy of “any air pollutant.” CAA sections 502(a), 501(2)(B), 302(j). EPA has interpreted the term “any air pollutant” narrowly so that applies only with respect to air pollutants that are subject to regulation under the CAA. Memorandum from Lydia N. Wegman, Deputy Director, Office of Air Quality Planning and Standards, U.S. EPA, “Definition of Regulated Air Pollutant for Purposes of Title V” (Apr. 26, 1993).

Based on these provisions, the key issue for present purposes in determining whether a source is subject to PSD (because it qualifies as a major emitting facility that undertakes construction or modification) or title V is whether the pollutant or pollutants that the source emits comprise the “air pollutant” that is “subject to regulation” under the Act.

The phrase “subject to regulation under the Act,” by its terms, identifies the air pollutant that is subject to PSD and title V as the same air pollutant that is identified in the regulatory action under another provision of the Act. The term is a simple cross-reference. It carries no implication that EPA, in identifying the pollutant to which PSD or title V apply, may redefine the pollutant that is regulated elsewhere in the Act. Whatever the pollutant is that is regulated elsewhere, it is that pollutant to which PSD and title V apply.

Since the time of our proposal, we have finalized both the contribution finding and the LDVR for GHGs. The final LDVR for GHGs specifies, in the rule's applicability provisions, the air pollutant subject to control as the aggregate group of the six GHGs, including CO₂, CH₄, N₂ O, SF₆, HFCs, and PFCs. Because it is this pollutant that is regulated under the LDVR, it is this pollutant to which PSD and title V apply. Specifically, the applicability provision in the LDVR provides a clear reference to the definition of the single pollutant comprised of the aggregate group of the six well-mixed GHGs, which makes clear PSD and title V applicability depends on the same sum-of-six GHG construct. We must follow this construct of the aggregate group of the six gases and do not have discretion to interpret the GHG “air pollutant” differently for the purposes of PSD or title V.

This construct of the pollutant as the aggregate group of the six gases is also consistent with the definition of the air pollutant in the final contribution finding for GHGs [see 74 FR 66496, 66499, 66536-7 (December 15, 2009)]. There, the Administrator defined the air pollutant as the “aggregate group of the same six * * * greenhouse gases,” (74 FR 66536), and these well-mixed GHGs are defined to include CO_2, CH₄, N₂ O, SF₆, HFCs, and PFCs.

Moreover, even if we had discretion to identify the GHGs air pollutant differently in the permitting programs than in the LDVR, we believe it is reasonable to identify the GHGs air pollutant through the sum-of-six construct for the same reasons why we adopted that definition in the contribution finding and for additional reasons noted below specific to the permit programs. The term “air pollutant” is defined under CAA section 302(g) as “any air pollution agent or combination of such agents, including any physical, chemical, biological, radioactive * * * substance or matter which is emitted into or otherwise enters the ambient air.” Under this definition, EPA has broad discretion to identify an air pollutant, including, as appropriate, treating a combination of air pollutant agents as a single air pollutant. Here, we think that the six well-mixed gases are appropriately combined into a single air pollutant because, as noted in the contribution findings, they share several important attributes: Each of the six gases:

• Is directly emitted (and is not formed by secondary processes in the atmosphere);
• Is long-lived in the atmosphere after it is emitted;
• Is sufficiently long-lived that it becomes “well-mixed,” which means that its concentration is essentially uniform in the atmosphere (as opposed to having significant local/regional variation); and
• Has well understood atmospheric properties (e.g., radiative forcing).

See 74 FR 66516-66518.

In addition, treating the six GHGs as a single air pollutant is consistent with the actions of international scientific bodies. For example, the Intergovernmental Panel on Climate Change (IPCC) considers in various reports how the six gases drive human-induced climate change and how that affects health, society, and the environment. Similarly, the United Nations Framework Convention on Climate Change (UNFCCC) requires reporting of these six gases and the commitments under the UNFCCC and Kyoto Protocol are based on the combined emissions of these six gases. Finally, as discussed later, it is standard practice to compute the “CO₂ equivalency” of aggregate emissions using GWP.

We disagree with commenters who argued that grouping all six GHGs is not appropriate because GHGs are not like other air pollutants that are comprised of numerous substances of concern (e.g., VOCs and PM). First, as noted previously, we are following the approach to a single air pollutant comprised of the aggregate of the six GHGs initially adopted in the contribution finding and followed in the LDVR. Many of these same comments have already been addressed in the contribution finding and Response to Comment (RTC) document for that action, and those responses apply equally here.

In addition to the reasons described in the endangerment and contribution findings, there are CAA permitting programmatic and policy advantages to using the sum-of-six construct for the GHG air pollutant for PSD and title V applicability purposes. We believe now, as we did at proposal, that the benefits in using the cumulative group of GHGs outweigh any implementation advantages to using an individual-GHG-based metric. The advantages to sum-of-six definition include that it may: (1) Allow significantly more flexibility to sources for designing and implementing control strategies that maximize reductions across multiple GHGs and would also likely align better with possible future regulations that allow for such flexibility; (2) more effectively support possible future offsets or trading mechanisms that involve different source categories and different compositions of GHG emissions; and (3) could better accommodate and harmonize with future regulations because it establishes one class of pollutants that includes individual components that may, in turn, become subject to specific emission standards under future regulatory efforts.

We disagree with commenters who believe that aggregating the GHGs under one GHG metric for permitting applicability purposes would lead to an excessive amount of source permitting activity. This is because the phase-in approach addresses overwhelming permitting burdens associated with permitting of GHGs. It does so by designing our applicability thresholds to allow for a manageable amount of new permitting actions based on the emissions from sources using the sum-of-six metric. If we based applicability on individual gases, (assuming, again, that we had authority to deviate from the definition of “air pollutant” as used in the LDVR), we would still need to determine what level of permitting is manageable and appropriate based on thresholds on an individual gas basis and would expect that the final rule would result in the same levels of remaining burden. Accordingly, unless the permitting program were being implemented at the statutory thresholds, the effect of a decision to aggregate or not aggregate would not reduce workload; rather, it would simply shift work from permitting facilities that trigger based on combined GHGs to those that trigger based on individual GHGs. Although we acknowledge that this may affect applicability for a particular source, we disagree with the comment that doing so would conflict with our conclusions based on the “absurd results” or “administrative necessity” doctrines. By using a consolidated and weighted measurement, we are able to direct the limited administrative resources to those new sources and modifications with the greatest impact on GHG emissions.

We also believe that the additional flexibility resulting from the sum-of-six GHG metric will provide substantially more opportunities for sources to address emission increases of GHGs than they would have had under an individual gas based metric, and, thereby, possibly reduce their permitting burden through multi-gas mitigation strategies. We disagree with the comment that isolating BACT review on sources that emit a single GHG necessarily leads to better environmental results than it would for sources that undergo a combined review for all six gases. To the contrary, given that Congress built in considerations of energy, environmental, and economic impacts into the BACT requirement, we think that allowing consideration of those factors across six gases will likely result in decisions that more appropriately account for those impacts at the source.

2. Identifying Which GHGs Are Included in the Group

As discussed previously, we proposed to include the combination of six well-mixed GHGs as the air pollutant that triggers PSD and title V applicability: CO₂, CH₄, N₂ O, SF₆, HFCs, and PFCs. Some commenters supported including all six. They cite the proposed contribution findings that identify the pollutant through the sum-of-six construct, and they emphasize that EPA, in order to protect the public, has to control all the GHGs it has regulated and reduce the overall impact of the mix of six GHGs.

However, a substantial number of commenters, mainly from industry sectors who also disagree with grouping the GHGs together, contend that only the constituent gases that are actually subject to controls under the LDVR should be included in determining applicability under the Tailoring Rule. Some of these commenters believe that only the three compounds (CO₂, CH₄, N₂ O) for which the LDVR contains emissions standards or caps should be considered in the GHG metric for permitting, while others would also add HFCs (which are included in a credit flexibility arrangement under the LDVR) for a total of four GHGs. These commenters argued that PSD is not triggered for all six GHGs by the LDVR because under the proposed PSD interpretation in the Interpretive Memo, actual emission controls under the Act are required to trigger PSD obligations for a given pollutant. They also argue that including all six would conflict with EPA's rationale for the Tailoring Rule by leading to larger numbers of sources subject to permitting, thereby increasing the harm that EPA says it wants to avoid. They further assert that the EPA cannot exercise its discretion to widen the scope of PSD and title V applicability to six GHGs when it is relying on the judicial doctrines of “absurd results” and “administrative necessity” to narrow PSD and title V applicability. They explain that in their view, those doctrines apply only when EPA has taken all steps possible to narrow the scope of PSD and title V and thereby avoid the administrative problems that force it to rely on those doctrines.

There were a few comments on whether to include specific gases as part of the sum-of-six grouping. Several commenters representing sectors that have significant SF₆ usage specifically argue that SF₆ should not be included as a GHG, at least at this time, because there are no known SF₆ controls, it is not clear how PTE would be calculated from such facilities, and EPA has not addressed the economic burden that regulation of these facilities would create. A solid waste industry commenter asserts that the Tailoring Rule should confirm that CH₄ and N₂ O will not be regulated under PSD or title V because these pollutants are only emitted in miniscule amounts from automobiles.

We disagree with commenters who suggest that because the LDVR actually reduces only four of the six GHGs, EPA may apply PSD and title V to only those four GHGs. It is true that the LDVR standard for the single air pollutant that is comprised of the aggregate of six GHGs consists of individual standards for only four particular constituents of the single air pollutant—which are emissions limits or caps for three GHGs (CO₂, CH₄, and N₂ O) and an emission crediting option for one GHG (HFCs)—but this does not dictate that only those four compounds are subject to regulation for permitting purposes. Although the LDVR results in reductions only with respect to four specific GHGs, as a legal matter the LDVR standard covers the entire set of GHGs that as a class are the single “air pollutant” in the contribution finding. Similar to our rationale for addressing the group of six GHGs as one pollutant for PSD and title V applicability purposes, we must adhere to the definition of applicability, cited previously, in the final LDVR for GHGs and include CO₂, CH₄, N₂ O, SF₆, HFCs, and PFCs. We do not have discretion to select only a subset of these gases in defining our GHG threshold metric for the permitting applicability purposes. See LDVR, May 7, 2010, 75 FR 25398-99, section III.A.2.c. (discussing EPA's exercise of discretion under section 202(a) in setting emissions standards applicable to emission of the single air pollutant).

For the same reasons, we disagree that this approach is inconsistent with the Agency's final action in “EPA's Interpretation of Regulations that Determine Pollutants Covered by Federal Prevention of Significant Deterioration (PSD) Permit Program.” While it is the case that only four constituent gases are reduced by the LDVR, the “air pollutant” that is controlled, and thus “subject to regulation,” is the group of six, and it is this “air pollutant” to which PSD and title V apply.

We also disagree with commenters who suggested that including all six GHGs in determining permitting applicability would conflict with our “absurd results” and “administrative necessity” rationale for the phase-in periods and applicability thresholds for GHGs. Even if we did have discretion to identify the air pollutant for PSD and title V purposes as consisting of only four of the six well-mixed GHGs, we do not believe that doing so would have any meaningful impact on the administrative burdens that are at the heart of our reliance on the “absurd results” and “administrative necessity” doctrines. The number of additional permitting actions and amount of additional permitting burden resulting from including all six GHGs, rather than four, is minimal. This is because the administrative burden of GHG permitting is dominated by CO₂ and CH₄ emission sources. For example, with a major source threshold set at 100,000 tpy CO₂ e, the combined population of sources that would be major for N₂ O, HFCs, PFCs, and SF₆ accounts for fewer than two percent of the GHG sources that would remain covered.

For similar reasons, we disagree with commenters who specifically suggest SF₆ emissions should not be included in the applicability metric for GHGs. As we have stated earlier in this section, our selection of the GHG metric is driven by the definition of the “air pollutant” as defined in the LDVR, and in consideration of the final GHG endangerment finding. SF₆ is specifically included as one of the “well-mixed greenhouse gases” in the definition of air pollutant in the contribution finding, and is included in the definition of the air pollutant in the LDVR for which that rule is applicable. We do not believe we have the discretion to define the “air pollutant” differently for PSD and title V applicability purposes than the definition of the “air pollutant” that is regulated elsewhere. In any event, including SF₆ emissions based on the thresholds finalized in this rulemaking does not add an excessive administrative burden for permitting authorities. Based on our threshold evaluation study, we estimate that less than 40 sources of SF₆ nationwide would exceed the 100,000 tpy CO₂ e threshold. Furthermore, SF₆ is a high GWP gas and, as discussed elsewhere, we have included a mass-based trigger for high GWP gases that will likely have the effect of further reducing this count.

For the same reasons, we disagree with the commenters who suggest we include black carbon and other short-lived climate forcers to the list of GHGs, as well as commenters asking for an exclusion of CH₄ and N₂ O. The definition of the air pollutant, as cited in the LDVR, includes CH₄ and N₂ O and does not include black carbon or other short-lived gases.

3. Use of GWP vs. Mass-Based GHG Thresholds

For the reasons discussed previously, we are determining permit program applicability based on the sum-of-six well-mixed gases that comprise the GHG air pollutant. This section discusses our use of both the CO₂ e metric and mass emissions of the GHGs for applicability purposes.

Under our proposal, a source's emissions of all six GHGs would be combined into a single metric by multiplying the mass of each individual GHG (in tpy) by its GWP value, and summing these products to determine the total emissions of the GHG pollutant in tpy CO₂ e. We received comments on this aspect of the proposed metric. Several commenters explicitly support the use of GWP and the CO₂ e metric for GHG emissions. These commenters believe EPA has the authority to select an appropriate metric to measure GHGs in the PSD program, and policy considerations support the choice of GWP. Some of them note that GWP is a widely-used metric which employs internationally-recognized conversion factors to compare GHGs based upon their climate properties, and some add that states and local areas that have climate action plans for GHG reductions use CO₂ e. Some of these commenters believe this metric will ensure a standard measure across all permitting agencies and will lead to a more effective system for permitting authorities and create more opportunities to reduce emissions over the full class of GHGs, rather than focusing on reducing individual GHGs.

On the other hand, some commenters oppose the use of GWP and CO₂ e, believing that thresholds should be based on individual mass-based emissions for each GHG. Some of these commenters felt that EPA has no discretion to ignore the metric for regulation established by Congress for PSD in section 169 of the Act. Some commenters were also concerned that the use of CO₂ e will complicate the implementation of BACT because sources that trigger PSD will be required to install BACT for each regulated pollutant, not for CO₂ e. As a result, a source that exceeds the threshold primarily due to its CO₂ emissions would be forced to install BACT for all other individual GHGs, regardless of how minor those other emissions may be. Finally, a commenter was concerned that use of GWP would complicate implementation because GWP values can sometimes change.

In our proposal preamble discussion of GHG metric, EPA also raised the possibility of including a limitation in the metric to address the prospect (expected to occur only rarely) that high-GWP gases could be emitted in quantities less than statutory thresholds for PSD and title V but nevertheless exceed the proposed thresholds in terms of CO₂ e. Most commenters on this subject support a dual threshold under which a source would be subject to title V or PSD only if its GHG emissions exceeded both the statutory thresholds on an actual tonnage basis and the tailored thresholds on a CO₂ e basis. Commenters supporting this approach felt that it would be unlawful to apply PSD when GHGs are below the statutory thresholds, or when there is not a net emissions increase. Others added that the complexity of accounting for emissions according to both mass and GWP should be manageable and is not a reason to ignore the role of mass-based emission rates in determining the applicability of PSD requirements. Additionally, one commenter observed that a dual threshold is consistent with phasing in the Tailoring Rule and is an effective way to address the current uncertainty surrounding how to measure high-GWP gases such as SF₆. In contrast, a few commenters stated they do not support a dual threshold, primarily on the grounds that there is no benefit to the added complexity.

After considering these comments, we have decided to adopt applicability thresholds in the final rule based on a CO₂ e metric for the sum-of-six well-mixed gases, and also to adopt an additional mass-based threshold for the sum-of-six gases as discussed in the proposal. First, as discussed in the previous section, we have explained why the appropriate pollutant for PSD purposes is the single pollutant GHG, which is composed of the six well-mixed gases. Regarding the CO₂ e metric, we continue to believe there are a number of advantages, as laid out in the proposal, to a CO₂ e measure that would not be available if we used only a mass-based metric. These include: (1) A CO₂ e metric, by incorporating the GWP values, best addresses the relevant environmental endpoint, which is radiative forcing of the GHGs emitted; (2) when combined with a sum-of-six gases approach, the CO₂ e metric best allows for consideration of their combined effects when sources emit any one or combination of the six well-mixed GHGs; (3) a cumulative CO₂ e metric is consistent with the metric used in the mandatory GHG reporting rule and other related rules and guidelines; and (4) a CO₂ e metric allows more flexibility for designing and implementing control strategies that maximize reductions across multiple GHGs. We recognize the tension between the mass-based metric in the statute and the CO₂ e-based metric we are adopting in this rule, but as discussed later, we will address this by also retaining the mass-based metric. Moreover, given our need to tailor our approach to covering sources of GHGs, we believe that the considerations driving our choice to also use a CO₂ e-based metric are appropriate for defining the phase-in and allow for permitting resources to be directed at those sources and modifications that have the greatest impact on radiative forcing of the GHGs emitted.

We recognize the concern of commenters who stated that we cannot ignore the statutory thresholds based on the mass-based emissions of an air pollutant as described under CAA section 169(1). As we mentioned in the proposal, because both the PSD and title V statutory thresholds are expressed on a mass basis (i.e., tons of a pollutant with no weighting values applied) we were concerned from a legal standpoint that the metric proposed (CO₂ e) could have the effect of subjecting to PSD or title V requirements a source whose emissions fall below the statutory threshold limits on a strictly mass basis, but whose CO₂ e-based emissions exceed the CO₂ e thresholds we establish under the Tailoring Rule. As an example, in rare instances it is possible that a source may emit only a non-CO₂ GHG in very small amounts, on a mass basis, but one that carries a very large GWP. In this case, it is possible that the source may emit the GHG in amounts that fall below the PSD and/or title V statutory applicability threshold (100 or 250 tpy, as applicable) on a mass basis, but exceed the 100,000 CO₂ e PSD and title V applicability thresholds for Step 2 finalized in this action. Under these circumstances, without a mass-based threshold, the source would trigger PSD and title V for its CO₂ e emissions even though its GHG mass emissions would not, in fact, exceed the statutory triggers.

Upon review of the comments pertaining to this issue and further analysis of the legal and programmatic implications, we are adopting a two-part applicability process, for both major source applicability determinations for GHGs under PSD and title V and for determining if a net increase has occurred in PSD applicability determinations for modifications. As explained in the RTC document, we accomplish this two-step applicability approach by continuing to rely on the existing mass-based applicability provisions in the current regulations, and by including new regulatory provisions that add a definition of “subject to regulation” that in turn includes the phase-in thresholds. Similarly, for PSD modification reviews and associated netting analyses, the same two-step process must be used. Our summary in section IV.A described how we expect this provision to be implemented in practice.

We acknowledge that the possibility of changing GWP values is a downside to the use of CO₂ e for the GHG metric, and we address this comment in the next section, where we discuss our plan to codify GWP values. By codifying GWP, any changes will be manageable, and, in our judgment, will not outweigh the benefits of a CO₂ e-based approach. We also acknowledge that a CO₂ e-based approach may appear to complicate the BACT review and implementation process. However, we disagree with the commenter's ultimate conclusion that BACT will be required for each constituent gas rather than for the regulated pollutant, which is defined as the combination of the six well-mixed GHGs. To the contrary, we believe that, in combination with the sum-of-six gases approach described above, the use of the CO₂ e metric will enable the implementation of flexible approaches to design and implement mitigation and control strategies that look across all six of the constituent gases comprising the air pollutant (e.g., flexibility to account for the benefits of certain CH₄ control options, even though those options may increase CO₂). Moreover, we believe that the CO₂ e metric is the best way to achieve this goal because it allows for tradeoffs among the constituent gases to be evaluated using a common currency.

4. Determining What GWP Values Are To Be Used

At proposal, we proposed to link the calculation of CO₂ e for GHGs to GWP values in EPA's “Inventory of U.S. Greenhouse Gas Emissions and Sinks” (GHG Inventory). See, e.g., proposed 40 CFR 51.166(b)(58). Numerous commenters expressed concerns about this proposal on various grounds, including the following:

• The EPA should follow the proper notice-and-comment procedures and the requirements of the Information Quality Act for the relevant technical underpinnings of the proposal. The EPA relies upon the GWPs of the IPCC without providing the supporting data for review, and it is inappropriate to use this as a basis for this rule without first making all the raw data available for public inspection and comment.
• The EPA cannot tie the definition of GWP to the GHG Inventory because it is a non-regulatory document that may be changed without notice-and-comment rulemaking. Before EPA uses a new GWP, that GWP must be subject to notice and comment to comply with the requirements of CAA section 307 and the Administrative Procedure Act (APA).
• An annual update of GWP would create a moving target for sources conducting applicability determinations and assessing compliance with minor NSR and PSD emission limits. The EPA needs to ensure that applicability and compliance with limits continue to be based on the GWP that existed when the determination was made or the limit was established.
• The EPA should freeze the GWP at the current values by incorporating those values into the regulation. The EPA could still revise the “NSR” GWP, but would have to revise the regulation to do so.

Commenters added that it is important to ensure that all permitting agencies are using the same calculations for the determination of CO₂ e for GHGs.

We agree with commenters who suggested we should codify, either in the Tailoring Rule or through reference to codified values in another rulemaking, the GWP values to be used in permitting analyses. We agree that this approach provides certainty as to which GWP values need to be used by permitting authorities and allows sources to plan appropriately for possible changes in the GWP values. As mentioned in the comments, recommended GWP values from IPCC can change over time. While this is infrequent—the last such changes were in 2007—when it occurs, there are generally significant lag times in universal adoption of new values because of inconsistencies that could be created in national inventories and emission reporting mechanisms. In a regulatory setting, such as in the permitting programs, this could potentially create significant implementation issues, such as when a GWP change occurs while a permit action is in progress. EPA also recognized similar potential implementation issues in developing its final mandatory GHG reporting rule, and codified in the regulatory text for that rule the GWP values to be used in reporting GHGs as part of that final rulemaking.

For these reasons, we have decided to follow the approach in the mandatory GHG reporting rule and require that for PSD and title V permitting requirements, wherever emissions calculations are performed, that permitting authorities and sources use GWP values that are codified in EPA rules. We will establish the GWP values for PSD and title V rules based on a cross-reference to the values that are codified in the EPA's mandatory GHG reporting rule. 74 FR 56395, Table A-1 to subpart A of 40 CFR part 98—Global Warming Potentials. Any changes to Table A-1 of the mandatory GHG reporting rule regulatory text must go through an appropriate regulatory process. In this manner, the values used for the permitting programs will reflect the latest values adopted for usage by EPA after a regulatory process and will be consistent with those values used in the EPA's mandatory GHG reporting rule. Furthermore, the lead time for adopting changes to that rule will provide a transition time to address implementation concerns raised by commenters.

5. Use of Short Tons vs. Metric Tons

We proposed that the GHG metric would be expressed in terms of English (or short) tons, rather than metric (or long) tons. A few commenters support using short tons for this purpose. Others prefer the use of metric tons, and most of them note that the mandatory GHG reporting rule is based on metric tons and believe that the Tailoring Rule should be consistent with that rule. These commenters believe that using different units in the two rules would be confusing and could result in sources that are not subject to the mandatory GHG reporting rule becoming subject to PSD. Some of the commenters add that various “cap and trade” legislative proposals also quantify GHGs in metric tons. A few other commenters recommend that EPA harmonize the applicability thresholds established under the Tailoring Rule and the mandatory GHG reporting rule without expressing a preference for short or metric tons.

We are finalizing our proposal to use short tons because short tons are the standard unit of measure for both the PSD and title V permitting programs and the basis for the threshold evaluation to support this rulemaking. Calculation inputs for PSD are typically prepared in English units (e.g., pounds of fuel, British thermal units (Btu), etc.) which is the common convention for all PSD analyses and the units of the statutory thresholds under the Act.

It is true that the GHG reporting rule uses metric tons, but this does not create an inconsistency between permitting programs and the reporting rule because the two rules already use different applicability approaches. Although we originally proposed 25,000 tpy as the major source level for permitting programs, which was similar to the threshold in the reporting rule, we decided to adopt substantially higher thresholds in the final rule. Furthermore, even if the numbers were similar, the thresholds used for the reporting rule are based on actual emissions, while the PSD and title V programs thresholds are based on PTE. Therefore, we are less persuaded by arguments for consistency, and believe it is more important for ease of permit program implementation to ensure that GHG emissions calculations for PSD and title V will build on the same set of input variables used to develop short-ton based estimates for non-GHG pollutants. Thus, the use of short tons should actually facilitate the development of the GHG emission estimate. It would likely be more confusing to require a multi-pollutant PSD applicability analysis to present emissions information using different units for different pollutants, as would be the case if we required metric tons for GHG but continue to use short tons for every other pollutant. Finally, we do not expect this choice to introduce additional complexity because the conversion between short tons and metric tons is a very simple calculation. Therefore, based on these considerations we are requiring that short tons be used as the basis for emission calculations used to meet PSD and title V permitting requirements.

B. Rationale for Thresholds and Timing for PSD and Title V Applicability to GHG Emissions Sources

In this subsection, we describe our legal and policy rationale for our determinations concerning PSD and title V applicability to GHG emissions sources. This subsection includes: (1) An overview of our rationale; (2) data concerning costs to sources and administrative burdens to permitting authorities; (3) a review of the Chevron legal framework and the “absurd results,” “administrative necessity,” and “one-step-at-a-time” doctrines, as well as a review of how those doctrines fit into the Chevron framework; (4) an overview of the relevant PSD and title V requirements and their legislative history; (5) our application of the “absurd results” doctrine for tailoring the PSD requirements; (6) our application of the “absurd results” doctrine for tailoring the title V requirements; (7) our plans to issue further rulemaking that will address the “absurd results” basis for both PSD and title V requirements; (8) our rationale for the phase-in schedule for applying PSD and title V to GHG sources; (9) our application of the “administrative necessity” basis for tailoring the PSD and title V requirements; and (10) our application of the “one-step-at-a-time” basis for tailoring the PSD and title V requirements.

1. Overview

Under the familiar Chevron two-step approach to construction of agency-administered statutes, the agency must first, at Chevron Step 1, determine whether Congress's intent in a particular provision on a specific question is clear; and if so, then the agency must follow that intent. If the intent of the provision is not clear, then the agency may, under Chevron Step 2, fashion a reasonable interpretation of the provision. The best indicator of congressional intent is the literal meaning of the provision and generally, according to the case law, if the literal meaning addresses the specific question, then the agency should follow the literal meaning.

However, the courts have developed three doctrines relevant here that authorize departure from a literal application of statutory provisions. The first is the “absurd results” doctrine, which authorizes such a departure if the literal application would produce a result that is inconsistent with congressional intent, and particularly if it would undermine congressional intent. The judicial doctrine of “administrative necessity” authorizes an agency to depart from statutory requirements if the agency can demonstrate that the statutory requirements, as written, are impossible to administer. The “one-step-at-a-time” doctrine authorizes an agency, under certain circumstances, to implement a statutory requirement through a phased approach. Each of the three doctrines fits into the Chevron framework for statutory construction because each of the three is designed to effectuate congressional intent.

To apply the statutory PSD and title V applicability thresholds literally to sources of GHG emissions would bring tens of thousands of small sources and modifications into the PSD program each year, and millions of small sources into the title V program. These extraordinary increases in the scope of the permitting programs would mean that the programs would become several hundred-fold larger than what Congress appeared to contemplate. Moreover, the great majority of additional sources brought into the PSD and title V programs would be small sources that Congress did not expect would need to undergo permitting and that, at the present time, in the absence of streamlined permit procedures, would face unduly high permitting costs. Further, again at the present time, in the absence of streamlined permit procedures the administrative strains would lead to multi-year backlogs in the issuance of PSD and title V permits, which would undermine the purposes of those programs. Sources of all types—whether they emit GHGs or not—would face long delays in receiving PSD permits, which Congress intended to allow construction or expansion. Similarly, sources would face long delays in receiving title V permits, which Congress intended to promote enforceability. For both programs, the addition of enormous numbers of additional sources would provide relatively little benefit compared to the costs to sources and the burdens to permitting authorities. In the case of PSD, the large number of small sources that would be subject to control constitute a relatively small part of the environmental problem. In the case of title V, a great many of the sources that would be newly subject to permit requirements would have “empty” permits, that is, permits that do not include any applicable requirements, and that therefore serve relatively little purpose. For these reasons, the “absurd results” doctrine applies to avoid a literal application of the thresholds at this time. By the same token, the impossibility of administering the permit programs brings into play the “administrative necessity” doctrine. This doctrine also justifies not applying the PSD or title V applicability threshold provisions literally to GHG sources at this time.

The situation presented here is exactly the kind that the “absurd results,” “administrative necessity,” and “one-step-at-a-time” doctrines have been developed to address. Separately and interdependently, they authorize EPA and the permitting authorities to tailor the PSD and title V applicability provisions through a phased program as set forth in this rule, and to use the initial period of phase-in to develop streamlining measures, acquire expertise, and increase resources, all of which would facilitate applying PSD and title V on a broader scale without overburdening sources and permitting authorities. In this manner, the phased approach reconciles the language of the statutory provisions with the results of their application and with congressional intent.

2. Data Concerning Costs to Sources and Administrative Burdens to Permitting Authorities

This final action concerning applicability of PSD and title V to GHG-emitting sources, including the decisions on timing for the selected permitting thresholds, is based on our assessments of both the costs to the regulated sources to comply with PSD and title V permitting requirements and the administrative burdens to the permitting authorities to process PSD and title V permit actions for GHG-emitting sources. This section provides a summary of our cost and administrative burden assessments of permitting that would be required in the absence of any tailoring as well as under various tailoring options.

Our estimates of costs to the sources and administrative burdens to the permitting authorities from PSD and title V applicability for GHG emissions are based on labor and cost information from the existing Information Collection Requests (ICRs) for PSD and title V programs. We apply the same basic methodology used for the proposal, which incorporates information on numbers and types of affected sources and estimated permitting actions. We evaluate administrative burdens in terms of staffing needs, time for processing permits, and monetary costs, and we make some judgments about how those burdens would affect the permitting authorities' ability to effectively manage and administer their programs with the addition of GHG emission sources. We present the administrative burden data for applying PSD and title V requirements at the literal statutory thresholds—that is, the 100/250 tpy levels for PSD (and 0 tpy for modifications) and the 100 tpy level for title V—as well as at other thresholds, which range from 25,000 tpy CO₂ e to 100,000 tpy CO₂ e. We have significantly revised upwards our assessments of costs to sources and administrative burdens since proposal, and we summarize later our reasons for doing so. We also present significant comments concerning administrative burdens, and our responses to those comments.

In the next section, concerning legal and policy rationale for our actions, we discuss how these data on costs to the sources and administrative burdens to the permitting authorities informed our decisions that PSD and title V requirements should not, at present, be applied to GHG-emitting sources under the literal terms of the statutory thresholds as well as our decisions concerning what thresholds to apply for Steps 1 and 2 of the applicability phase-in approach and the applicability floor of 50,000 tpy CO₂ e.

a. Costs to Sources

As we did at proposal, we have estimated costs to the sources of complying with PSD and title V starting from the data in the ICRs. We recognize that the sizes of the sources, as measured by their emissions, that would be swept into the PSD and title V programs would vary greatly, and that their permitting costs would vary as well. For example, their PSD permitting costs would depend on the amount and types of their emissions and their control requirements. Accordingly, we have determined average costs, as described later.

For PSD, at proposal, we estimated that on average, an industrial source would incur costs of $84,500 to prepare the PSD application and receive the permit, and on average, a commercial or residential source would incur costs of 20 percent that amount, or $16,900. 74 FR 55337 col. 3 to 55339 col. 3. For this action, we retain the same burden estimates for an average industrial source. This type of source would need 866 hours, which would cost $84,500, to prepare the application and the PSD permit. However, based on comments received, we have determined that a more accurate estimate for an average commercial or residential source is 70 percent of that amount of time that an industrial source would need, up from our proposal of 20 percent. Thus, an average commercial or residential source would need 606 hours, which would cost $59,000, to prepare the PSD application and receive the permit. We are increasing this time over what we proposed because we now recognize that virtually all commercial and residential sources will have no experience with the PSD permitting process, and therefore will face a significant learning curve that will entail more time to complete the application, develop control recommendations, and take the other required steps. We believe this learning period could extend from 2 to possibly 4 years or more from the date that the sources become subject to PSD requirements, depending on the type and actual number of new sources that come in for permitting. In addition, we expect that in many cases, draft PSD permits for GHGs will receive comments from various stakeholders, from citizens groups to equipment vendors, who will seek to participate in the permit process, and that all this could add to the hours that the permittee will need to invest in the process. The actual costs to sources to install BACT controls, while still uncertain at this point, would likely add additional costs across a variety of sources in a sector not traditionally subject to such permitting requirements.

For title V, at proposal, we estimated that on average, an industrial source would incur costs of approximately $46,400 to prepare the title V application and receive the permit, and on average, a commercial or residential source would incur costs of 10 percent that amount, or almost $5,000. 74 FR 55338 col. 1 to 55339 col. 3. For this action, we retain the same burden estimates for an average industrial source. This type of source would need 350 hours, which would cost $46,400, to prepare the application and the title V permit. However, we have determined that a more accurate estimate for an average commercial or residential source is 50 percent of that amount of time that an industrial source would need, up from our proposal of 10 percent. Thus, an average commercial or residential source would need about 175 hours, which would cost $23,200, to prepare the title V application and receive the permit. This increase is due to the same reasons as with the PSD program just discussed. We now recognize that virtually all commercial and residential sources will have no experience with the title V permitting process and, therefore, will face a significant learning curve that will entail more time to assess, for the first time, their GHG emissions (because such sources are not covered by EPA's mandatory GHG reporting rule), complete the application, respond to permitting authority comments, meet other title V administrative requirements, and respond to interested stakeholders.

b. Administrative Burdens to Permitting Authorities

(1) Estimated Permitting Authority Burden at Proposal

As at proposal, we estimated the administrative burdens to the permitting authorities at the various threshold levels for PSD or title V applicability as follows. First, for a particular threshold level, we estimated the number of GHG-emitting sources that would be subject to PSD requirements because they would undertake new construction or modification, and the number of existing sources that would be subject to title V requirements. Second, we estimated the average additional administrative burden and cost of each PSD permitting action and each title V permitting action for the GHG-emitting sources. Third, we multiplied those two estimates, and the product is the additional administrative burden at the particular threshold level. We employed the same methodology for this final rule, but, as discussed later, and described in more detail in our final burden analysis, we have updated several key assumptions since the proposal as a result of our consideration of comments received.

First, we present the administrative burdens at the statutory levels for PSD and title V applicability. At proposal, for the PSD program, we estimated the administrative burdens that would result from applying PSD at the 100/250 tpy major emitting facility threshold levels in two ways, as described in this section. We stated that at present, 280 sources are subject to PSD each year, both for new construction and modifications. This figure served as the baseline from which to calculate increases in administrative burdens due to permitting GHG-emitting sources.

The first method that we used to calculate the administrative burdens to the permitting authorities was in terms of workload hours, which we then converted to monetary costs. To make the workload calculation, we first estimated the number of GHG-emitting sources that would become subject to PSD through new construction and modification. Based on our GHG threshold data analysis, we estimated that almost 41,000 new and modified sources per year would become subject to PSD review. We first calculated the number of new sources that would become subject to PSD. To do this, we estimated growth rates for the various sectors, and then applied those growth rates to the numbers of sources in those sectors. We then calculated the number of modifications. To do this, we first assumed that each year, two percent of sources that meet or exceed the threshold levels for PSD applicability due to their conventional pollutants undertake modifications. We then calculated the number of sources that would meet or exceed the threshold levels for PSD applicability due to their GHG emissions, and applied the same assumption that two percent of them would undertake modifications. In this manner, we estimated the number of modifications of GHG-emitting sources that would become subject to PSD.

We noted that currently, 280 PSD permits are issued each year, but that applying PSD to GHG-emitting sources at the 100/250 tpy statutory threshold levels would cause an increase in permits of more than 140-fold. The reason for the extraordinary increase in PSD applicability lies simply in the fact that it takes a relatively large source to generate emissions of conventional pollutants in the amounts of 100/250 tpy or more, but many sources combust fossil fuels for heat or electricity, and the combustion process for even small quantities of fossil fuel produces quantities of CO₂ that are far in excess of the sources' quantities of conventional pollutants and that, for even small sources, equal or exceed the 100/250 tpy levels.

Based on the 140-fold increase in permits, we then estimated the per-permit burden on permitting authorities. As we stated in the proposal:

We estimated the number of workload hours and cost a permitting authority would expend on each new source and each modification. We based these estimates on the workload hours and cost for processing permits for new sources of non-GHG emissions, which we derived from labor and cost information from the existing ICRs for PSD programs. The ICRs show that permitting authorities expend 301 hours to permit a new or modified industrial source * * *.

We then made assumptions for number of workload hours and costs for new sources of GHG emissions. We assumed that permitting new industrial GHG sources that emit in excess of the 250-tpy threshold would be of comparable complexity to permitting non-GHG emitting industrial sources that are subject to PSD. Thus, for these sources, we assumed that permitting authorities would expend the same number of workload hours and costs, on a per-permit basis, as they do for non-GHG emitting industrial sources. On the other hand, for commercial and residential GHG sources that emit GHGs above the 250-tpy threshold (and as a result would be subject to the requirements of the PSD permitting program at this threshold level), we assumed that the workload hours and cost for permitting these sources would be significantly less than—only 20 percent of—the hours and cost necessary to prepare and issue initial PSD permits or permit modifications for industrial GHG sources. This 20-percent estimate amounts to 60 hours of permitting authority time per residential or commercial permit.

Based on these assumptions, the additional annual permitting burden for permitting authorities, on a national basis, is estimated to be 3.3 million hours at a cost of $257 million to include all GHG emitters above the 250-tpy threshold.

74 FR 55301 col. 2.

Note that at the proposal, in calculating the PSD administrative burdens that would occur each year due to GHG emissions, we did not undertake separate calculations for the administrative burdens associated with permitting obligations stemming from the GHG emissions of the 280 sources already subject to PSD permitting requirements due to their conventional pollutants. In effect, we treated these 280 sources are part of the over 40,000 sources that would become subject to PSD due to their GHG emissions.

The second way that we evaluated the burden on permitting authorities was by reviewing a study conducted by state and local air permitting agencies. As we said in the preamble:

In addition to conducting our burden analysis, we also reviewed summary information from state and local air permitting agencies regarding additional resources and burden considerations if GHG sources that emit above the 100/250-tpy thresholds were subjected to the PSD and title V programs. This information covered 43 state and local permitting agencies, representing programs from different regions of the country and various permitting program sizes (in terms of geographic and source population coverage) * * *. This information showed significant burdens projected by permitting agencies with adding sources of GHG emissions in terms of staffing, budget, and other associated resource needs. Importantly, the agencies based their analysis on the assumption that, for purposes of determining whether a source is major, its emissions would be calculated on an actual emissions (“actuals”) basis, and not on a PTE basis. On an actuals basis, the agencies estimated a 10-fold increase in the number of permits.

Specifically, the agencies estimated that:

• Assuming, again, that number of permits was to increase by 10-fold (based on actual emissions), the resulting workload would require an average of 12 more [full-time equivalents (FTEs)] per permitting authority at an estimated cost of $1 million/year;
• Without the additional FTEs, the average processing time for a permit would increase to 3 years, which is three times the current average processing time;
• Permitting authorities would need 2 years on average to add the necessary staff;
• Permitting authorities would also need, on average, eight additional enforcement and judicial FTEs;
• Ninety percent of permitting agencies would need to train their staff in all aspects of permitting for sources of GHG emissions.
• A quarter of permitting agencies were currently under a hiring freeze.

We went on to explain that this state survey significantly underestimated the administrative burdens:

It is important to reiterate that the state and local permitting information on burden was based on the number of additional facilities subject to PSD because their emissions of GHGs exceed the 100/250-tpy thresholds at actual emissions rates, not PTE-based emissions rates. However, the PSD applicability requirements are based on PTE. By adjusting the increase in number of permits to account for GHG sources that exceed the 100/250-tpy applicability thresholds based on their PTE emissions, EPA estimated a 140-fold increase in numbers of PSD permits, much more than the 10-fold increase estimated by the states based on actual emissions.

74 FR 55301 col. 2-3.

In addition to PSD, we also estimated title V burdens at the statutory threshold. At proposal, for the title V program, we estimated the administrative burdens that would result from applying title V requirements at the 100 tpy major source threshold level in the same two ways as for PSD, as follows. The first method was to calculate the administrative burdens in terms of workload hours, which we then converted to monetary costs. To make the workload calculation, we first estimated the number of existing GHG-emitting sources that would become subject to title V. Based on our GHG threshold data analysis, we estimated that approximately 6 million sources would become subject to title V. Compared to the 14,700 title V permits currently issued, this would be an increase in permits of more than 400-fold. We noted, in addition, that most of the 14,700 sources already subject to title V also emit GHGs and may be affected as well.

We then described the type of work that the permitting authorities would need to do for these GHG-emitting sources—the six million that would become newly subject to title V and most of the 14,700 that are already subject to title V—as follows. Note at the outset that the permitting authorities' workload is greater for sources newly subject to title V than for existing sources that seek a revised or renewed permit. As EPA noted in the preamble:

[T]he [ ] permits [for the 6 million new sources] would need to include any requirements for non-GHGs that may apply to the source, such as provisions of an applicable SIP. For any such requirements, permitting authorities would also need to develop terms addressing the various compliance assurance requirements of title V, including monitoring, deviation reporting, six-month monitoring reports, and annual compliance certifications.

Adding to the burden described above would be the burden to add GHG terms to the 14,700 existing title V permits. While, in general, existing title V permits would not immediately need to be revised or reopened to incorporate GHG (because as noted above, there are generally not applicable requirements for GHGs that apply to such sources), permitting authorities may face burdens to update existing title V permits for GHG under two possible scenarios: (1) EPA promulgates or approves any applicable requirements for GHGs that would apply to such a source, which would generally require a permit reopening or renewal application, or (2) the source makes a change that would result in an applicable requirement for GHGs to newly apply to the source, such as PSD review, which would generally require an application for a permit revision. Permitting authorities will also need to process permit renewal applications, generally on a five-year cycle, and such renewals would need to assure that the permit properly addresses GHG. Finally they would have to process title V applications for new sources (including all the PSD sources previously discussed).

74 FR 55302 cols. 2-3.

In light of those demands, we estimated the per-permit burden on permitting authorities as follows. Note, at the outset, that as with PSD, we based the workload hours on information in ICRs for industrial sources, and we then assumed that the workload for commercial and residential sources would be the indicated percentage of the workload for industrial sources:

As with PSD, we have quantified the extent of the administrative problem that would result in workload hours and cost on the basis of information concerning hours and costs for processing existing title V permits that is indicated on ICRs. However, we recognize that more than 97 percent of these new sources would be commercial and residential sources. We estimate that for permitting authorities, the average new commercial or residential permit would require 43 hours to process, which is 10 percent of the time needed for the average new industrial permit. For an average existing permit, which permitting authorities would need to process through procedures for significant revisions and permit renewals, adding GHG emissions to the permit would result in, we estimate, 9 additional hours of processing time, which is 10 percent of the amount of time currently necessary for processing existing permits. We estimate that the total nationwide additional burden for permitting authorities for title V permits from adding GHG emissions at the 100-tpy threshold would be 340 million hours, which would cost over $15 billion.

74 FR 55302 col. 3.

As with PSD, the second way that we evaluated the burden on permitting authorities at the statutory threshold was by reviewing a study conducted by state and local air permitting agencies of the burden of applying title V to existing GHG-emitting sources at the 100 tpy statutory threshold level. As we said in the preamble to the proposed rule:

[W]e also reviewed summary information from state and local permitting agencies, which showed significant burdens associated with adding GHGs in their title V programs in terms of staffing, budget, and other associated resource needs. Again, note that the permitting agencies based their estimates on numbers of permits that would be required from sources subject to the 100-tpy title V applicability threshold on an actuals—not PTE—basis. Based on that level, the agencies assumed a 40-fold increase in numbers of permits, and estimated that:

• The resulting workload would require an average of 57 more FTEs per permitting agency at an estimated cost of $4.6 million/year;
• Without the additional FTEs, the average processing time for a permit would increase to almost 10 years, which is 20 times the current average permit processing time;
• Permitting authorities would need 2 years on average to add the necessary staff;
• On average, permitting authorities would need 29 additional enforcement and judicial staff;
• Eighty percent of permitting authorities would need to train their staff in all aspects of permitting for sources of GHG emission.
• A quarter of permitting agencies were currently under a hiring freeze.

As with PSD, we added that this state survey significantly underestimated the administrative burdens:

It is important to reiterate that, as with PSD, the state and local information on projected permitting burden is based on the number of additional facilities subject to title V because their emissions of GHGs exceed the 100-tpy thresholds at actual emissions rates, not the PTE-based emissions rates. However, the title V applicability requirements are based on PTE. As noted elsewhere in this preamble, the state and local agencies estimated a 40-fold increase in numbers of title V permits based on the amount of GHG sources' actual emissions. By adjusting the summary estimates provided by the state and local agencies to account for GHG sources that exceed the 100-tpy threshold based on their PTE emissions, EPA estimated that the average permitting authority would need 570 more FTEs to support its title V permitting program.

74 FR 55302 col. 3—55303 col. 1.

(2) Revisions to Proposal Estimates of Permitting Authority Burden

We received numerous comments from state and local authorities stating that EPA had underestimated the administrative burden on the permitting authorities in the proposal. State and local authorities stated that in particular, EPA underestimated the number of modifications and the amount of time it would take permitting authorities to process permits, particularly for commercial and residential sources. Based on the comments and additional analysis that we have conducted in response, we are revising in several respects our estimates of the administrative burdens for applying PSD and title V at the statutory threshold levels.

First we present revisions to our analysis regarding the burdens at the statutory levels. Before we present those changes, we want to note a revision to our methodology that affected our estimate of the number of permits currently issued under existing programs. We are revising upwards the number of sources that are already subject to PSD permitting requirements anyway for their conventional pollutants, which, as discussed previously, we refer to as “anyway” sources. This revision has implications both for (1) the number of sources that would become subject to PSD due to their GHG emissions; and also (2) the baseline number of sources already subject to PSD, which we use to compare the amount of increases in administrative burden due to permitting GHG sources. At proposal, we stated that 280 sources each year are subject to PSD due to their new construction or modifications. However, upon further analysis, we have realized that this figure is too low because it includes only sources that have emissions of one or more NAAQS pollutants at the 100/250 tpy thresholds and that are located in areas of the country that are designated attainment or unclassifiable for all of those pollutants, and thus are not designated nonattainment for any of those NAAQS pollutants. We estimate that another 520 sources have emissions of one or more NAAQS pollutants at the 100/250 tpy thresholds and are located in areas of the country that are nonattainment for at least one of those NAAQS pollutants. Some of these 520 sources may also emit one or more pollutants at the 100/250 tpy level for which their area is designated attainment or unclassifiable, and therefore may be subject to PSD for those pollutants. Accordingly, the correct number of “anyway” sources subject to PSD each year is the 280 sources that are located in areas that are attainment or unclassifiable for each pollutant that the sources emit at the 100/250 tpy level, plus at least some of the 520 sources that are located in areas that are nonattainment for at least one of the NAAQS pollutants that the sources emit at or above the 100/250 tpy threshold. In the absence of data on the number of nonattainment NSR permits that do not have a PSD component, and because we expect this to be a small number, we have assumed for purposes of this action, that each of the 520 sources is subject to PSD for at least one pollutant, so that we will consider all 800 sources as subject to PSD. Of this number, we estimate that 70 percent, or 560 sources will undergo a modification, while the remaining 240 permitting actions will involve new construction. Of the modifications, we assume that 80 percent, or 448, would become subject to additional requirements due to their GHG emissions because those projects have combustion-related activities that would likely emit GHGs in the requisite quantities. Our estimate of 80 percent of modification activities significantly involving combustion activities is based on a review of a random sample of PSD permits. In total we estimate that 688 sources, either upon new construction or modification, would need to add GHG requirements to their otherwise required PSD permitting action.

We should also note that in this rulemaking we are justifying our conclusions about permitting authority administrative burdens on the basis of their PSD and title V cost as calculated on both a separate basis and a combined basis. That is, we believe that the administrative burdens of the PSD program justify our tailoring approach for the PSD requirements, and the administrative burdens of the title V program justify our tailoring approach for the title V requirements, but in addition, the administrative burdens of both programs on a combined basis justify the tailoring approaches. Viewing the administrative burdens on a combined basis provides a useful perspective because most permitting authorities have a single organizational unit that is responsible for both the PSD program and the title V program, and in many cases, the same employees work on both programs. In addition, in some jurisdictions, permitting authorities issue a single, merged permit that includes both PSD and title V requirements. For these reasons, considering administrative burdens on a combined PSD and title V basis, offers a more accurate picture of the issues these agencies will face in transitioning to GHG permitting.

Turning to the revisions to our burden estimates that we made as a result of public comment, we begin by noting that many commenters believed that we significantly underestimated the administrative burdens associated with the proposed thresholds or that the administrative burden under the proposed thresholds would still overwhelm the states and result in significant permitting delays and uncertainty for sources. Many of these commenters indicate that our estimate of the number of sources that would be subject to permitting is too low, and some add that we have underestimated the per-permit effort required. (More detail on these comments is given elsewhere on the methodology used in the analysis.) Several state and local agencies provided estimates of the increased number of permits and/or staff that would be required under the thresholds we proposed that were higher than our original estimates. Specifically, commenters recommended that we increase the estimated administrative burdens for PSD permits by anywhere from 100 percent to over 2,000 percent; and that we increase the burdens for title V permits by anywhere from 29 percent to 240 percent. Many commenters indicated that EPA has not adequately accounted for “synthetic minor” sources or modification projects, stating that many such sources and projects will not be able to keep GHGs below the proposed thresholds, and those who could do so may not be able to establish enforceable synthetic minor limits. Numerous commenters also stated that the EPA has underestimated the rate of major modifications for GHGs under PSD. Some commenters assert that we underestimated the number of permits required for specific industry sectors, including the oil and gas production industry, the natural gas transmission industry, the semiconductor industry, the wood products industry, the brick industry, and landfills. Some of the state and local commenters also believe that we have overestimated their ability to hire and train sufficient staff to administer GHG permitting.

We are persuaded by the data and arguments provided by the many commenters who believe EPA underestimated the number of permitting actions and the burdens of each action, and thus the overall administrative burdens associated with permitting GHG sources. Accordingly, we have reevaluated our assessment of these administrative burdens, for both the PSD and title V programs. In conducting this reevaluation, we considered arguments made by the commenters, as well as any actual data they provided, and then we determined whether and how to modify various aspects of our detailed assessment of the burdens. Based on this consideration we have substantially revised upwards our estimate of administrative burdens, based on the analysis included in the final docket for this rulemaking. The revisions affect two elements of our analysis by showing: (1) A substantial increase in the number of PSD and title V permits that will occur at a given threshold, and (2) an increase in the average burden estimate for each such permit.

Regarding the increase in our estimate of the number of projects that will occur, we estimated an increase in both PSD and title V permit actions, though the greatest changes were for PSD. At proposal, we estimated that, if PSD requirements were to apply to GHG sources at the 100/250 tpy statutory levels, 40,496 projects—consisting of 3,299 projects at industrial sources and 37,197 projects at commercial or residential sources—would need PSD permits each year. Some of these projects involve the construction of an entirely new source, but the majority of these are modifications. We now estimate that at the 100/250 tpy levels, 81,598 projects would become subject to PSD each year. These projects include 26,089 actions at industrial sources and 55,509 at commercial and residential sources. We describe our calculation of this 81,598 amount in a TSD. The great majority of these 81,598 projects that would become subject to PSD are modifications. We base these estimates on the assumption that the significance levels would be 100 tpy regardless of category.

Our estimate of the number of PSD modifications is where we made our most significant upward revisions from our proposal, based on comments. Our doubling of the estimated PSD permitting actions—from 40,496 at proposal to 81,598—results from three separate adjustments we made to our estimates at proposal of the number of permit actions that would result from applying PSD to GHG sources. Two of these increased the number of major modifications, and one of these increased the number of major sources and modifications. The most significant adjustment, and one that was raised by multiple commenters, was that we undercounted the number of major modification projects at existing major sources because we did not include the existing projects that avoid major PSD review by either taking “synthetic minor” limits or by netting out for conventional pollutants, but that would not be able to avoid PSD through those mechanisms for GHGs.

We agree that the ability and procedures for sources to achieve reductions, or minimize increases, due to GHGs through adoption of enforceable limits or through netting out are not well established at this point. We believe that there will be numerous instances, particularly for combustion-related projects, where it will not be possible for sources to achieve the same level of reductions for CO₂ emissions as they do for emissions of nitrogen oxides (NO_X), for example, simply because there are not as many proven control techniques that can reduce CO₂ emissions to the same degree as NO_X. Also, more research will be necessary in the type of emission units and processes resulting in GHG emissions, and how they operate over a wide range of utilization patterns at a variety of source categories, before permitting authorities will be able to establish procedures and rules for developing minor source permit limitations. Therefore, we adjusted our count of major modification permits under PSD upward to account for this.

The second change to the number of permits concerns the general modification rate of 2 percent that we applied at proposal, based on historical experience across all pollutant types. Commenters provided information that suggest that this 2 percent figure is an underestimate for GHG sources because their emissions of CO₂ are high and accumulate quickly from various changes involving combustion units. Therefore, a greater percentage of their physical or operational changes will result in GHG emissions in excess of the significance levels that we identified at proposal. In light of these comments, we reviewed the source populations and pollutant mix within the various populations, and determined that we should revise our general modification rate to 4 percent for GHG sources. This 4 percent rate was obtained by dividing the current annual major NSR permit actions involving modifications by the 14,700 existing sources. We have revised our burden analysis accordingly. Again, the burden analysis in the docket describes our basis for these calculations in more detail.

The third adjustment to the number of permits involves our estimate of the number of sources with PTE that is greater than the various thresholds considered. This affects the number of major sources at the statutory thresholds, which we used to estimate the number of PSD and title V major sources, but also has an effect on the number of major modifications because the number of modifications depends on the size of the population of major sources. Commenters provided evidence that our estimates of capacity utilization (which, as described previously, we use for estimating potential-to-emit based on data for actual emissions) for the general manufacturing source category (referred to as “unspecified stationary combustion” in our analysis) and for the oil and gas industry were not accurate. In our proposal, our estimated range for capacity utilization for “unspecified stationary combustion” varied from 70 to 90 percent depending on manufacturing category. For the oil and gas industry, our estimate was 90 percent. We received comments indicating that these utilization rates are higher than what is normally achieved in real-world conditions, particularly for smaller manufacturing type facilities. Accordingly, in this action, we are using a 50 percent capacity utilization rate for both of these source categories, which better reflects what can be deemed reasonable operation under normal conditions for facilities in these source categories. This adjustment increased the overall number of affected facilities at various threshold levels and we have revised our burden analysis accordingly.

A few commenters asserted that we underestimated the number of residential homes, commercial buildings, and retail stores that would be subject to permitting requirements because these commenters believed the estimate in EPA's TSD was based on actual emissions from space heating equipment rather than PTE. We wish to clarify that our threshold analysis estimates for the number of residential and commercial sources (as well as all other sources) did use a PTE basis. To calculate the PTE amount for these sources, we extrapolated from the actual emissions data for the residential and commercial sources. Specifically, we assumed that a typical residential facility operates its fuel combustion sources at only 10 percent of its capacity and a typical commercial facility operates at only 15 percent of its capacity. Based on these assumptions, we multiplied residential actual emissions by a factor of 10, and commercial actual emissions by a factor of 6.6 to obtain PTE-based estimates. There is very little information available on the capacity utilization rates of fuel combustion equipment at different types of residential and commercial facilities, but we believe our methodology was reasonable for these types of sources and we did not adjust it in response to this comment. Information on the development of these estimates is provided in our Technical Support Document for Greenhouse Gas Emissions Thresholds Evaluation.

The second source of upward revisions to our administrative burden estimate is that we are increasing the estimated average cost to permitting authorities of issuing each PSD and title V permit at the statutory thresholds. At proposal, we estimated that for PSD permits, permitting authorities would expend, on average, 301 hours to permit an industrial source of GHG emissions, and 20 percent of that time, or 60 hours to permit a commercial or residential source. After estimating that amount of workload, we went on to estimate the monetary cost to permitting authorities of that workload. Similarly, for title V permits, we estimated at proposal that permitting authorities would expend 10 percent of the number of hours needed to process an industrial permit in order to process a commercial or residential permit for GHG sources.

We received comments from both permitting authorities and sources asserting that our methodology underestimated the administrative burden on grounds that (1) Our methodology fails to recognize that when a source triggers PSD for conventional pollutants, additional labor hours would be required to issue BACT for GHGs; (2) our estimate of 60 hours (versus 301 hours) to issue PSD permits to commercial and residential sources of GHGs is unrealistically low; (3) our estimate failed to account for the increase in the complexity of permits for criteria pollutants due to (i) increases in criteria pollutant emissions becoming newly subject to BACT at sources that are major only for GHGs, which will result in increased permitting and (ii) BACT controls for criteria pollutants (e.g., an oxidizer for VOCs) may result in significant GHG emissions, triggering an additional BACT determination; and (4) our methodology failed to account for the significant additional PSD and title V burdens due to sources that obtain federally enforceable permit limits on GHGs in order to become “synthetic minors” and thereby avoid PSD (and possibly also title V).

Based on these comments and our own reassessment of permitting actions created by the addition of GHGs, we have revised upwards in several ways our estimate of the additional per-permit costs of applying PSD and title V to GHG sources, including the following: First we have added an estimate of the additional permitting cost for adding a GHG component to “anyway” PSD and title V permitting actions for conventional pollutants. We estimated this burden based on information in the comments together with our own judgment about how to adjust the burden numbers contained in the current supporting statements for our approved permitting ICRs. These adjustments are found in our revised burden estimate document.

Second, we have raised the per-permit burden hours for commercial and residential sources for PSD and title V. At proposal, our estimates were based on the fact that many of these permits will be technically simpler due to such factors as a lower number of emissions points, simpler processes, and less required modeling. However, commenters pointed out that, until EPA streamlines its permitting procedures, there are many permitting activities that represent a fixed cost, such as public notice, hearing, and response to comment activities. In addition, we agree, as commenters pointed out, that many of these sources will need significantly more permitting authority staff time to assist them in the permit application and preparation process because of their lack of experience with these requirements. In addition, permitting authorities will have little, if any, experience in permitting commercial and residential sources, and therefore will face a learning curve that will entail more time to take permitting action. In addition, we expect that in many cases PSD and title V permit applications for GHGs will receive comments from various stakeholders, from citizens groups to equipment vendors, who will seek to participate in the permit process, and responding and revising permits accordingly will add to the hours that the permitting authority will spend.

As a result, we raised the PSD per-permit hours for various steps in the permitting process, as described in the burden estimate document. While we continue to estimate that permitting authorities will expend, on average, 301 hours to issue a PSD permit to an industrial source, and that this would cost $23,243, we now recognize that a permitting authority would expend 70 percent of that time or 210 hours, to permit a commercial or residential source, which would cost $16,216. Similarly, for title V, while we continue to estimate that permitting authorities will expend, on average, 428 hours to issue a title V permit to an industrial source, and that this would cost $19,688, we now recognize that a permitting authority would expend 50 percent of the time, or 214 hours, to permit a commercial or residential source, which would cost $9,844.

We disagree with commenters who suggested that by basing our estimates on the numbers of newly constructing and modifying sources with high enough emissions to qualify as major emitting facilities, we failed to account for the costs of sources that seek “synthetic minor” permits to avoid PSD, and possibly title V, requirements. In fact, our methodology includes sources that might take such limits as newly-major sources for their GHG emissions; and therefore we count the full administrative burden associated with a PSD permit and a title V permit for those sources. In effect, we assume that such sources would go through PSD or title V permitting, rather than take “synthetic minor” limits. We take this approach because although we suspect that there may, in fact, be significant synthetic minor activity, we do not have data that would allow us to determine whether, and how many of, these sources will be able to adopt “synthetic minor” limits or restrict their operations to obtain minor source permitting status. Nor do we have data on the amount of the administrative burden that would fall on any particular permitting authority to establish a “synthetic minor” limit, except that we understand that the amount varies widely across states. As a result, we opted to include these sources in our analysis as sources receiving a PSD or title V permit. Therefore, to the extent that synthetic minor activity occurs, our estimate would already have included the burden for that activity. In fact, our estimate would have overestimated the burden to the extent that a permitting authority would have less administrative costs to issuing a “synthetic minor” permit, as compared to a PSD or title V permit.

(3) Revised Burden Estimates at Statutory Thresholds Based on the revisions just described, we estimate that in all, if sources that emit GHGs become subject to PSD at the 100/250 tpy levels, permitting authorities across the country would face over $1.5 billion in additional PSD permitting costs each year. This would represent an increase of 130 times the current annual burden hours under the NSR major source program for permitting authorities. The permitting authorities would need a total of almost 10,000 new FTEs to process PSD permits for GHG emissions.

In addition, we estimate that in all, if sources that emit GHGs become subject to title V at the 100 tpy level, permitting authorities across the country would incur about 1.4 billion additional work hours, which would cost $63 billion. We estimate that most of this work would be done over a 3 year period, which would amount to 458 million in additional work hours, and $21 billion in additional costs, on an annual basis over that 3-year period.

We also note that the survey of state and local permitting authorities described in the proposed rulemaking continues to shed light on the extent of the administrative burdens, including staffing, budget, and other associated resource needs, as projected by the permitting authorities. As noted previously, that survey concluded that application of the PSD requirements to GHG-emitting sources at the level of 100/250 tpy or more of actual emissions would, without additional FTEs, increase the average processing time for a PSD permit from one to 3 years. The survey further concluded that application of the title V requirements to GHG-emitting sources at the level of 100 tpy or more of actual emissions would, without additional FTEs, increase the average processing time for a title V permit from 6 months to 10 years. As we noted at proposal, this survey assumed a ten-fold increase in the number of PSD permits and a 40-fold increase in the number of title V permits due to GHG-emitting sources, but those assumptions were severely underestimated because they were based on actual emissions. At proposal, our calculations, which were based on potential emissions, indicated a 140-fold increase in PSD permits and a more than 400-fold increase in title V permits. In this rulemaking, we recognize that even our estimates at proposal were severely underestimated. We now recognize that the number of PSD permits will be about twice what we estimated at proposal, and the average processing time for both PSD and title V permits will be two or three times greater than what we estimated at proposal. The survey of state and local permitting authorities provided other useful information as well, including the fact that it would take the permitting authorities 2 years, on average, to hire the staff necessary to handle a ten-fold increase in PSD permits and a 40-fold increase in title V permits, and that 90 percent of their staff would need additional training in all aspects of permitting for GHG sources.

(4) Revised Estimates of Administrative Burdens at Various Threshold Levels

In order to determine the appropriate PSD and title V applicability level for GHG sources, we not only estimated the burden at the statutory thresholds, as described previously, but we also estimated the number of sources, number of permitting actions, and amount of administrative burden at various applicability levels for both PSD and title V, based on the revised methodology described previously, that we used to estimate the administrative burdens of applying PSD and title V at the statutory levels. This information is summarized in Table V-1. Note that Table V-1 also includes, in the last column, the administrative burdens, described previously, associated with the 100/250 tpy thresholds.

As described in the TSD, we considered several different major source/major modification threshold combinations. We chose the combinations to reflect representative, incremental steps along the possible range. Because it is time- and resource- intensive to develop estimates for a given step, we chose intervals that best reflect representative points within the range, given those time and resource constraints. Here, we discuss key observations about some of the combinations that we assessed. As the table indicates, under the current PSD and title V programs, approximately 15,000 sources qualify as major PSD sources for at least one pollutant and therefore meet the applicability thresholds. Of these, approximately 668 sources are subject to PSD requirements each year for at least one pollutant—240 because they undertake new construction, and 448 because they undertake modifications. The permitting authorities' administrative burdens for the NSR program are 153,795 work hours, and $12 million. For the title V program, the 15,000 sources are, for the most part already permitted, and therefore need revised permits as required and renewal permits on a 5-year schedule. The permitting authorities' title V administrative burdens on an annual basis are 1,349,659 work hours and $62 million.

The first threshold Table 1 describes—and which, as discussed later, we are adopting for Step 1—is the “anyway” source approach. Under this approach, (i) PSD applies to the GHG emissions from projects that are subject to PSD anyway as new sources or major modifications due to their emissions of non-GHG pollutants and that result in an increase (or, in the case of modifications, a net increase) of at least 75,000 tpy CO₂ e; and (ii) title V applies to what we will call “anyway” title V sources, that is, sources that are subject to title V anyway due to their emissions of non-GHG pollutants. Under this approach, the number of sources subject to PSD each year—including new construction and modifications—is the same as under the current program, but the permitting authorities will need to address GHG emissions as part of those permitting actions each year and, to do so, will require, each year, 34,400 additional workload hours costing an additional $3 million. For title V, we estimate that the number of title V sources that require permitting actions will, on average, be the same each year, but permitting authorities will need to address GHG requirements for some of them; as a result, permitting authorities will need, each year, 27,468 additional work hours costing $1 million in additional funding.

Another threshold described in Table V-1 is the one we are adopting under Step 2, as described later, under which (i) sources will be subject to PSD on account of their GHG emissions if they newly construct and emit at least 100,000 tpy CO₂ e, or if they are existing sources that emit at least 100,000 tpy CO₂ e of GHGs and make a modification that results in a net emissions increase of at least 75,000 tpy CO₂ e; and (ii) existing sources will be subject to title V due to their GHG emissions if they emit 100,000 tpy CO₂ e in GHG emissions. Under this approach, which we will call the 100,000/75,000 approach, we estimate that each year, compared to current levels, the permitting authorities will need to issue GHG permits to two additional sources that newly construct and to 915 additional sources that undertake modifications. Doing so will require 310,655 additional workload hours costing an additional $24 million, compared to the current program. For title V, an additional 190 sources will require new title V permits each of the first 3 years, and the permitting authorities' associated costs will be 141,322 work hours and $7 million more than the current program.

The last approach we will describe here may be called the 50,000/50,000 approach, which, as discussed later, we adopt as the floor for thresholds during the first 6 years after promulgation. Under this approach, (i) sources will be subject to PSD on account of their GHG emissions if they newly construct and emit at least 50,000 tpy CO₂ e, or if they are existing sources that emit at least 50,000 tpy CO₂ e of GHGs and make a modification that results in a net emissions increase of at least 50,000 tpy CO₂ e; and (ii) existing sources will be subject to title V on account of their GHG emissions if they emit 50,000 tpy CO₂ e in GHG emissions. Under this approach, each year, the permitting authorities will need to issue GHG permits to 3 additional sources that newly construct and 1,900 that undertake modifications above current permitting levels. Doing so will require 613,986 additional workload hours costing $47 million, compared to the current program. For title V, an additional 1,189 sources will require new title V permits each of the first 3 years and the permitting authorities' associated costs will be 568,017 work hours and $26 million more than the current program.

We present the remaining entries in the table to illustrate how the cost and burden estimates vary with increasing or decreasing thresholds relative to those selected in this rule. These variations are important in understanding how alternative thresholds would compare to the ones selected. We also include entries reflecting the baseline (current program without GHG permitting) and the burdens if we immediately implemented the full statutory thresholds on January 2, 2011, without tailoring or streamlining.

3. “Absurd Results,” “Administrative Necessity,” and “One-Step-at-a-Time” Legal Doctrines

a. Introduction and Summary

Having described the factual underpinnings of our action, which are the costs to sources and administrative burdens to permitting authorities, we now describe the legal underpinnings. They involve the framework for analyzing agency-administered statutes, as established by the U.S. Supreme Court in Chevron U.S.A. Inc. v. NRDC, 467 U.S. 837, 842-43 (1984). In this case, Chevron framework must take into account the “absurd results,” “administrative necessity,” and “one-step-at-a-time” legal doctrines. We believe that each of these doctrines provides independent support for our action, but in addition, the three doctrines are directly intertwined and can be considered in a comprehensive and interconnected manner. Moreover, although each of the three doctrines pre-date the 1984 Chevron decision, in which the U.S. Supreme Court established the framework for construing agency-administered statutes, each fits appropriately into the Chevron framework.

To reiterate, for convenience, the statutory provisions at issue: Congress, through the definition of “major emitting facility,” applied the PSD program to include “any * * * source [that] emit[s], or ha[s] the potential to emit, one hundred [or, depending on the source category two hundred fifty] tons per year or more of any air pollutant.” CAA sections 165(a), 169(1). In addition, Congress, through the definition of “modification,” applied the PSD program to include “any physical change in, or change in the method of operation of, a stationary source which increases the amount of any air pollutant emitted by such source or which results in the emission of any air pollutant not previously emitted.” CAA sections 165(a), 169(2)(C), 111(a)(4). Similarly, Congress, through the definition of “major source,” specified that the title V program includes “any stationary facility or source of air pollutants which directly emits, or has the potential to emit, one hundred tons per year or more of any air pollutant.” CAA sections 502(a), 501(2)(B), 302(j). EPA, through long-established regulatory action, in the case of PSD, and long-established guidance, in the case of title V, has interpreted these definitions narrowly so that they apply only with respect to air pollutants that are subject to regulation under the CAA.

Applying these definitions by their terms, as interpreted narrowly by EPA, to GHG sources at the present time would mean that the PSD and title V programs would apply to an extraordinarily large number of small sources, the sources would incur unduly high compliance costs, and permitting authorities would face overwhelming administrative burdens. As a result, we believe Congress did not intend for us to follow this literal reading, and instead, with this action, we chart a course for tailoring the applicability provisions of the PSD program and the title V program by phasing them in over time to the prescribed extent.

For our authority to take this action, we rely in part on the “absurd results” doctrine, because applying the PSD and title V requirements literally (as previously interpreted narrowly by EPA) would not only be inconsistent with congressional intent concerning the applicability of the PSD and title V programs, but in fact would severely undermine congressional purpose for those programs. We also rely on the “administrative necessity” doctrine, which applies because construing the PSD and title V requirements literally (as previously interpreted narrowly by EPA) would render it impossible for permitting authorities to administer the PSD provisions. The tailoring approach we promulgate in this action is consistent with both doctrines. It is also consistent with a third doctrine, the “one-step-at-a-time” doctrine, which authorizes administrative agencies under certain circumstances to address mandates through phased action.

Our discussion of the legal bases for this rule is organized as follows: In this section V.B.3, we provide an overview of the three doctrines and describe how they fit into the Chevron framework for statutory construction. In section V.B.4, we discuss the PSD and title V programs, including each program's relevant statutory provisions, legislative history, and regulatory history. In sections V.B.5 and V.B.6 we discuss the “absurd results” approach for PSD and title V, respectively, that we are finalizing in our action. In section V.B.7., we discuss additional rulemaking in which we may consider exempting certain categories of sources from PSD and title V under the “absurd results” doctrine. In section V.B.8, we discuss the legal and policy rationale for the phase-in schedule that we are adopting for applying PSD and title V to GHG sources. In section V.B.9 we discuss the “administrative necessity” approach for PSD and title V, respectively. In section V.B.10, we discuss the third legal basis for our action, the “one-step-at-a-time” doctrine.

b. The “Absurd Results” Doctrine

Turning first to the “absurd results” doctrine, we note at the outset that we discussed the doctrine at length in the notice of proposed rulemaking, and we incorporate by reference that discussion, although we make some refinements to that discussion in this preamble. The starting point for EPA's interpretation of the PSD and title V applicability provisions and reliance on the “absurd results” doctrine is the familiar Chevron two-step analysis. We discuss this analysis in greater detail later, but in brief, in interpreting a statutory provision, an agency must, under Chevron Step 1, determine whether Congress's intent on a particular question is clear; if so, then the agency must follow that intent. If the intent of the provision is not clear, then the agency may, under Step 2, fashion a reasonable interpretation of the provision. Chevron U.S.A. Inc. v. NRDC, 467 U.S. 837, 842-43 (1984).

The courts consider the best indicator of congressional intent to be the plain meaning of the statute. However, the U.S. Supreme Court has held that the literal meaning of a statutory provision is not conclusive “in the `rare cases [in which] the literal application of a statute will produce a result demonstrably at odds with the intentions of the drafters' * * * [in which case] the intention of the drafters, rather than the strict language, controls.” United States v. Ron Pair Enterprises, 489 U.S. 235, 242 (1989). This doctrine of statutory interpretation may be termed the “absurd results” doctrine.

Although, as just noted, the U.S. Supreme Court has described the “absurd results” cases as “rare,” in that case the Court seemed to be referring to the small percentage of statutory-construction cases that are decided on the basis of the doctrine. The DC Circuit, in surveying the doctrine over more than a century of jurisprudence, characterized the body of law in absolute numbers as comprising “legions of court decisions.” In re Franklyn C. Nofziger, 925 F.2d 428, 434 (DC Cir. 1991). Indeed, there are dozens of cases, dating from within the past several years to well into the 19th century, in which the U.S. Supreme Court has applied the “absurd results” doctrine to avoid the literal application of a statute, or if not so holding, has nevertheless clearly acknowledged the validity of the doctrine. Some of the more recent of these cases include: Logan v. United States, 552 U.S. 23, 36-37 (2007) (“[s]tatutory terms, we have held, may be interpreted against their literal meaning where the words `could not conceivably have been intended to apply' to the case at hand [citation omitted]”); Nixon v. Missouri Municipal League, 541 U.S. 125, 132-33 (2004) (“any entity” includes private but not public entities); Raygor v. Regents of Univ. of Minn., 534 U.S. 533, 542-45 (2002) (“implying a narrow interpretation of * * * `any claim asserted' so as to exclude certain claims dismissed on Eleventh Amendment grounds”); United States v. X-Citement Video, Inc., 513 U.S. 64, 69 (1994) (rejecting a literal interpretation of the statutory term “knowingly” on grounds that Congress could not have intended the “positively absurd” results that some applications of such an interpretation would produce, “[f]or instance, a retail druggist who returns an uninspected roll of developed film to a customer “knowingly distributes” a visual depiction and would be criminally liable if it were later discovered that the visual depiction contained images of children engaged in sexually explicit conduct”); Rowland v. Cal. Men's Colony, 506 U.S. 194, 200 (1993) (finding that an artificial entity such as an association is not a “person” under the statute, and describing the absurdity doctrine as a “common mandate of statutory construction”); United States v. Ron Pair Enterprises, 489 U.S. 235, 242 (1989) (the plain meaning of a statutory provision is not conclusive “in the `rare cases [in which] the literal application of a statute will produce a result demonstrably at odds with the intentions of the drafters' * * * [in which case] the intention of the drafters, rather than the strict language, controls”); Green v. Bock Laundry Machine Company, 490 U.S. 504 (1989) (provision in Federal Rule of Evidence that protects “the defendant” against potentially prejudicial evidence, but not the plaintiff, refers to only criminal, and not civil, defendants); Public Citizen v. United States Dep't of Justice, 491 U.S. 440, 453-54 (1989) (rejecting a broad, straightforward reading of the term “utilize,” on grounds that a literal reading would appear to require the absurd result that all of FACA's restrictions apply if a President consults with his own political party before picking his Cabinet, and such a reading “was unmistakably not Congress' intention”); Watt v. Alaska, 451 U.S. 259, 266 (1981) (rejecting reliance on plain statutory language and concluding that the term “minerals” in section 401(a) of the Wildlife Refuge Revenue Sharing Act applies only to minerals on acquired refuge lands; stating “[t]he circumstances of the enactment of particular legislation may persuade a court that Congress did not intend words of common meaning to have their literal effect”); Train v. Colorado Public Interest Research Group, Inc., 426 U.S. 1, 23-24 (1976) (prohibition in Federal Water Pollution Control Act against discharging into navigable waters “pollutants,” which are defined to include “radioactive materials,” does not apply to three specific types of radioactive materials); Jackson v. Lykes Bros. S.S. Co., 386 U.S. 731, 735 (1967) (refusing to distinguish between a longshoreman hired by “an independent stevedore company” and one hired by “the shipowner * * * to do exactly the same kind of work,” despite the clear terms of the Act, and stating: “[w]e cannot hold that Congress intended any such incongruous, absurd, and unjust result in passing this Act,” when the Act was “designed to provide equal justice to every longshoreman similarly situated”); Lynch v. Overholser, 369 U.S. 705, 710, (1962) (statutory construction is not confined to the “bare words of a statute”); United States v. Bryan, 339 U.S. 323, 338 (1950) (“Despite the fact that the literal language would encompass testimony elicited by the House Committee in its questioning of respondent relative to the production of the records of the association, the Court will not reach that result if it is contrary to the congressional intent and leads to absurd conclusions. And we are clearly of the opinion that the congressional purpose would be frustrated if the words, “in any criminal proceeding,” were read to include a prosecution for willful default under R.S. § 102.”).

The DC Circuit has also handed down numerous decisions that applied the absurd results doctrine to avoid a literal interpretation or application of statutory provisions or that have acknowledged the doctrine. Some of the most recent ones include: Arkansas Dairy Cooperative Ass'n, Inc., v. U.S. Dep't of Agriculture, 573 F.3d 815 (DC Cir. 2009) (rejecting the canon of construction that presumes that Congress is aware of existing law pertinent to the legislation that it enacts, when in this case, the presumption that Congress was aware of the Departments definition of “hearing” would lead to “the absurd result that Congress intended to impose a requirement with which the Secretary could not comply;” stating: “Courts, `in interpreting the words of a statute, [have] some scope for adopting a restricted rather than a literal or usual meaning of its words where acceptance of that meaning would lead to absurd results * * * or would thwart the obvious purpose of the statute * * *.’ ” (quoting In re Trans Alaska Pipeline Rate Cases, 436 U.S. 631, (1978)); Buffalo Crushed Stone, Inc. v. Surface Transportation Board, 194 F.3d 125, 129-30 (DC Cir. 1999) (regulation of Surface Transportation Board providing that if a notice of exemption “contains false or misleading information, the use of the exemption is void ab initio” does not apply to a notice containing false information when declaring the notice void ab initio would undermine the goals of the governing statute; a conflict between the “literal application of statutory language” and maintaining the integrity of the regulatory scheme should be resolved by construing the text in accordance with its purpose); Mova Pharm. Corp. v. Shalala, 140 F.3d 1060, 1068-69 (DC Cir. 1998) (as discussed later, describes the “absurd results” doctrine in the context of the Chevron framework for statutory construction; invalidates a Food and Drug Administration's (FDA) regulation designed to remedy what the FDA described as the absurd result of a literal application of the statutory provisions governing FDA approval of successive generic drug applications, on grounds that “[i]n effect, the FDA has embarked upon an adventurous transplant operation in response to blemishes in the statute that could have been alleviated with more modest corrective surgery;” states that “[t]he rule that statutes are to be read to avoid absurd results allows an agency to establish that seemingly clear statutory language does not reflect the “unambiguously expressed intent of Congress,” Chevron, 467 U.S. at 842, and thus to overcome the first step of the Chevron analysis”); Environmental Defense Fund v. EPA, 82 F.3d 451, 468-69 (DC Cir. 1996) (although Act requires that a federal action conform to the SIP that is currently in place, EPA may instead require conformity to a revised implementation plan that state commits to develop; “[t]his is one of those rare cases * * * [that] requires a more flexible, purpose-oriented interpretation if we are to avoid `absurd or futile results.' ”); American Water Works Ass'n v. EPA, 40 F.3d 1266, 1271 (DC Cir. 1994) (holding that EPA's interpretation of the term “feasible” so as to require a treatment technique instead of a maximum contaminant level (MCL) for lead is reasonable; the court stated: “Indeed, where a literal reading of a statutory term would lead to absurd results, the term simply `has no plain meaning * * * and is the proper subject of construction by the EPA and the courts.' If the meaning of `feasible' suggested by the NRDC is indeed its plain meaning, then this is such a case; for it could lead to a result squarely at odds with the purpose of the Safe Drinking Water Act.” (quoting Chem. Mfrs. Ass'n v. Natural Res. Def. Council, Inc., 470 U.S. 116, 126 (1985)) (citation omitted); In re Nofziger, 925 F.2d 428, 434-35 (DC Cir. 1991) (provision authorizing payment of attorney fees to the subject of an investigation conducted by an independent counsel of the Department of Justice only if “no indictment is brought” against such individual does not preclude payment of attorney fees when an indictment is brought but is determined to be invalid).

c. The “Administrative Necessity” Doctrine

In the proposed rulemaking, we also described in detail the “administrative necessity” doctrine, 74 FR 55311 col. 3 to 55318 col. 3, and we incorporate that discussion by reference into this notice. Under this doctrine, if a statutory provision, however clear on its face, is impossible for the agency to administer, then the agency is not required to follow the literal requirements, and instead, the agency may adjust the requirements in as refined a manner as possible to assure that the requirements are administrable, while still achieving Congress's overall intent. The DC Circuit set out the doctrine of “administrative necessity” in a line of cases that most prominently includes Alabama Power v. Costle, 636 F.2d 323 (DC Cir. 1980). The Court cited the doctrine most recently in New York v. EPA, 443 F.3d 880, 884, 888 (DC Cir. 2006).

As we stated in the proposed rulemaking, “We believe that the “administrative necessity” case law establishes a three-step process under which an administrative agency may, under the appropriate circumstances, in effect revise statutory requirements that the agency demonstrates are impossible to administer so that they are administrable.” 74 FR 55315 col. 1. Specifically:

[T]he three steps are as follows: When an agency has identified what it believes may be insurmountable burdens in administering a statutory requirement, the first step the agency must take is to evaluate how it could streamline administration as much as possible, while remaining within the confines of the statutory requirements. The second step is that the agency must determine whether it can justifiably conclude that even after whatever streamlining of administration of statutory requirements (consistent with those statutory requirements) it conducts, the remaining administrative tasks are impossible for the agency because they are beyond its resources, e.g., beyond the capacities of its personnel and funding. If the agency concludes with justification that it would be impossible to administer the statutory requirements, as streamlined, then the agency may take the third step, which is to phase in or otherwise adjust the requirements so that they are administrable. However, the agency must do so in a manner that is as refined as possible so that the agency may continue to implement as fully as possible Congressional intent.

74 FR 55315 cols. 1-2.

It should also be noted that we believe the administrative burdens encountered by the state and local permitting authorities are fully relevant under the “administrative necessity” doctrine. Although the case law that discusses the doctrine focuses on federal agencies (see 74 FR 55312-14), under the CAA, state and local agencies are EPA's partners in implementing provisions of the CAA, and have primary responsibility for implementing the PSD program. They generally adopt EPA's PSD requirements in their SIPs, as required under CAA section 110(a)(2)(C); and they generally adopt EPA's title V requirements in their title V programs, as required under CAA section 502(d). They issue the PSD and title V permits and are responsible in the first instance for enforcing the terms of the permits. In all these respects, the law that the state and local permitting authorities administer is both federal and state law. Under certain circumstances, EPA may become responsible for permit issuance and enforcement in the first instance, but even then, EPA may, and frequently has, delegated those duties to a state, in which case, the state implements federal law directly. Thus, although the PSD and title V programs are federal requirements, for the most part, it is the states that implement those programs. For this reason, the administrative burdens that the states face in implementing the programs are relevant in determining the applicability of the “administrative necessity” doctrine.

d. “One-Step-at-a-Time” Doctrine

In addition to the “absurd results” and “administrative necessity” doctrines, another judicial doctrine supports at least part of EPA's Tailoring Rule, and that is the doctrine that agencies may implement statutory mandates one step at a time, which we will call the “one-step-at-a-time” doctrine. In the notice of proposed rulemaking, we also described this doctrine and recent case law applying it. 74 FR 55319 col. 1-3. As we noted, that the U.S. Supreme Court recently described the doctrine in Massachusetts v. EPA, 549 U.S. 497, 524 (2007), as follows: “Agencies, like legislatures, do not generally resolve massive problems in one fell regulatory swoop;” and instead they may permissibly implement such regulatory programs over time, “refining their preferred approach as circumstances change and as they develop a more nuanced understanding of how best to proceed.” We assume familiarity with our discussion in the proposal, but we expand upon it here to review the case law in greater detail and to highlight certain components of the doctrine that are particularly relevant to the Tailoring Rule. The roots of the doctrine go back at least to the DC Circuit's 1979 decision in United States Brewers Association, Inc. v. EPA, 600 F.2d 974 (DC Cir. 1979). There, the Court considered a challenge to EPA's guidelines for managing beverage containers, which EPA was required to promulgate under the Resource Conservation and Recovery Act of 1976 (RCRA). RCRA gave EPA one year to promulgate the guidelines. EPA promulgated a partial set of guidelines, started two others, and was challenged before the year was out by petitioners who objected to the initial guideline, saying it fell short of the statutory mandate. The Court upheld the initial guideline, stating: “Under these circumstances we think the question of whether the Agency has fully satisfied the mandate of the statute is not fit for judicial review at this time, when the Agency, still well within the one-year period granted by statute, is deeply involved in the process of formulating rules designed to carry out the congressional mandate. The Agency might properly take one step at a time.” States Brewers Association, Inc. v. EPA, 600 F.2d at 982.

The Court addressed the doctrine at greater length in National Association of Broadcasters v. FCC, 740 F.2d 1190, 1209-14 (DC Cir. 1984). There, the Court noted that under certain statutory schemes, step-by-step agency action might not be authorized; but the Court emphasized that when it is authorized, it may offer significant benefits; and the Court went on to delineate some of the circumstances under which its use is justified. In that case, the Court held that Federal Communications Commission (FCC) acted reasonably in making a spectrum allocation decision that granted direct broadcast satellite service priority use of a gigahertz (GHz) band in 5-years time, and—although acknowledging that fixed service users that were, at that time, using that band, would have to relocate to other bands—in postponing the details of the fixed service relocation to future proceedings. The Court described in some detail “[t]he circumstances under which * * * [an] agency may defer resolution of problems raised in a rulemaking,” as follows:

The requisite judgment is in essence a pragmatic one. In an ideal world, of course, agencies would act only after comprehensive consideration of how all available alternatives comported with a well-defined policymaking objective, and in some circumstances, statutes indeed mandate that agencies proceed by only such a course * * *. But administrative action generally occurs against a shifting background in which facts, predictions, and policies are in flux and in which an agency would be paralyzed if all the necessary answers had to be in before any action at all could be taken * * *. We have therefore recognized the reasonableness of [an agency's] decision to engage in incremental rulemaking and to defer resolution of issues raised in a rulemaking even when those issues are “related” to the main ones being considered * * *. At the same time, [an agency] cannot `restructure [an] entire industry on a piecemeal basis' through a rule that utterly fails to consider how the likely future resolution of crucial issues will affect the rule's rationale * * *.

Drawing a line between the permissible and the impermissible in this area will generally raise two questions. First the agency will likely have made some estimation, based upon evolving economic and technological conditions, as to the nature and magnitude of the problem it will have to confront when it comes to resolve the postponed issue. With regard to this aspect of the agency's decision, as long as the agency's predictions about the course of future events are plausible and flow from the factual record compiled, a reviewing court should accept the agency's estimation * * *. Second, once the nature and magnitude of the unresolved issue is determined, the relevant question is whether it was reasonable, in the context of the decisions made in the proceeding under review, for the agency to have deferred the issue to the future. With respect to that question, postponement will be most easily justified when an agency acts against a background of rapid technical and social change and when the agency's initial decision as a practical matter is reversible should the future proceedings yield drastically unexpected results. In contrast, an incremental approach to agency decision making is least justified when small errors in predictive judgments can have catastrophic effects on the public welfare or when future proceedings are likely to be systematically defective in taking into account certain relevant interests * * *.

740 F.2d at 1210-11 (citations omitted).

In City of Las Vegas v. Lujan, 891 F.2d 927 (DC Cir. 1989), the Court suggested that one component of upholding partial agency compliance with a statutory directive is evidence that the agency was on track for full compliance. There, the Court upheld the Department of Interior's decision to list the population of desert tortoises living north and west of the Colorado River (the “Mojave” population) as endangered species, but not the nearby population living south and east of the river (the “Sonoran” population). The agency explained that the Mojave population faced certain threats that the Sonoran population did not, and the Court found nothing to fault in that reasoning. The Court added: “Since agencies have great discretion to treat a problem partially, we would not strike down the listing if it were a first step toward a complete solution, even if we thought it `should' have covered both the Mojave and Sonoran populations.” City of Las Vegas v. Lujan, 891 F.2d 927, 935 (DC Cir. 1989) (footnote omitted).

In Grand Canyon Air Tour Coalition v. F.A.A., 154 F.3d 455 (DC Cir. 1998), the DC Circuit added another component to the “one-step-at-a-time” doctrine: While reiterating that “ordinarily, agencies have wide latitude to attack a regulatory problem in phases and that a phased attack often has substantial benefits,” id. at 471, the Court went on to uphold partial agency action even when that action was long-delayed. There, the relevant statute was the Overflights Act, which required the Federal Aviation Administration (FAA) to reduce aircraft noise from sightseeing tours in Grand Canyon National Park, and established the goal of “substantial restoration of natural quiet and experience of the park.” The statute required the agency to develop a plan to implement the statutory requirements within 120 days after enactment, and report to Congress within 2 years after the date of the plan as to the plan's success. In fact, the FAA did not develop, through rulemaking, a plan until ten years after enactment, and when it did, it acknowledged that the plan was only a partial one, and that it would need two more rules and another ten years to meet the statutory goal of substantial restoration. Although recognizing that the Overflights Act did not establish an explicit timetable for meeting the statutory goal, the Court stated that “[t]he language of the Overflights Act does manifest a congressional concern with expeditious agency action,” and described the agency's action variously as “tardy,” “undeniably slow,” and “slow and faltering.” Id. at 476-77. Even so, the Court upheld the FAA's action against different challenges from appellants and intervenors that (i) the agency acted unreasonably in not promulgating a complete plan to meet the statutory goal, instead of promulgating just the first step; and (ii) the agency acted unreasonably in not waiting until it had a complete plan before promulgating the first step. The Court stated: “We agree that it would be arbitrary and capricious for an agency simply to thumb its nose at Congress and say—without any explanation—that it simply does not intend to achieve a congressional goal on any timetable at all * * *,” but went on to emphasize that the FAA's rule was the first of three that the agency assured would achieve the statutory goal. The Court cited City of Las Vegas v. Lujan, discussed previously, for the proposition that “a court will not strike down agency action `if it were a first step toward a complete solution.' ” Grand Canyon Air Tour Coalition v. F.A.A., 154 F.3d 455, 477-78 (DC Cir. 1998).

e. Consistency of Doctrines With Chevron Framework

Although the formation of the “absurd results,” “administrative necessity,” and “one-step-at-a-time” doctrines pre-date the Chevron two-step analysis for construing statutes that Congress has authorized an agency to administer, we believe that the doctrines can be considered very much a part of that analysis, and courts have continued to apply them post-Chevron. Under Chevron Step 1, an agency must determine whether “Congress has directly spoken to the precise question at issue.” If so, “the court, as well as the agency, must give effect to the unambiguously expressed intent of Congress.” However, if “the statute is silent or ambiguous with respect to the specific issue, the question for the court is whether the agency's answer is based on a permissible construction of the statute.”

Chevron U.S.A. Inc. v. NRDC, 467 U.S. 837, 842-42 (1984).

Thus, Step 1 under Chevron calls for determining congressional intent for the relevant statutory directive on the specific issue presented. To determine Congress's intent, the agency must look first to the statutory terms in question, and generally interpret them according to their literal meaning, within the overall statutory context, and perhaps with reference to the legislative history. If the literal meaning of the statutory requirements is clear then, absent indications to the contrary, the agency must take it to indicate congressional intent and must implement it. Even if the literal meaning of the statutory requirements is not clear, if the agency can otherwise find indications of clear congressional intent, such as in the legislative history, then the agency must implement that congressional intent.

The DC Circuit has indicated that the “absurd results” doctrine fits into the Chevron Step 1 analysis in the following way: Recall that in the cases in which the courts have invoked this doctrine, the literal meaning of the statutory requirements has been clear, but has led to absurd results. This can occur when the literal meaning, when applied to the specific question, conflicts with other statutory provisions, contradicts congressional purpose as found in the legislative history—and, in particular, undermines congressional purpose—or otherwise produces results so illogical or otherwise contrary to sensible public policy as to be beyond anything Congress would reasonably have intended. See United States v. Ron Pair Enterprises, 489 U.S. 235, 242-43 (1989); Griffin v. Oceanic Contractors, Inc., 458 U.S. 564, 571 (1982).

Under these circumstances, the agency must not take the literal meaning to indicate congressional intent. As the DC Circuit has explained, “where a literal reading of a statutory term would lead to absurd results, the term `simply has no plain meaning * * * and is the proper subject of construction by the EPA and the court.' ” American Water Works Assn v. EPA, 40 F.3d 1266, 1271 (DC Cir. 1994) (quoting Chemical Manufacturers' Association v. NRDC, 470 U.S. 116, 126 (1985)). Under these circumstances, if the agency can find other indications of clear congressional intent, then the agency must implement that intent. See United States v. Ron Pair Enterprises, 489 U.S. 235, 242-43 (1989). This may mean implementing the statutory terms, albeit not in accordance with their literal meaning, but in a way that achieves a result that is as close as possible to congressional intent. As the DC Circuit said in Mova Pharm. Corp. v. Shalala, 140 F.3d 1060 (DC Cir. 1998):

The rule that statutes are to be read to avoid absurd results allows an agency to establish that seemingly clear statutory language does not reflect the “unambiguously expressed intent of Congress,” * * * and thus to overcome the first step of the Chevron analysis. But the agency does not thereby obtain a license to rewrite the statute. When the agency concludes that a literal reading of a statute would thwart the purposes of Congress, it may deviate no further from the statute than is needed to protect congressional intent * * *. [T]he agency might be able to show that there are multiple ways of avoiding a statutory anomaly, all equally consistent with the intentions of the statute's drafters * * *. In such a case, we would move to the second stage of the Chevron analysis, and ask whether the agency's choice between these options was “based on a permissible construction of the statute.” Otherwise, however, our review of the agency's deviation from the statutory text will occur under the first step of the Chevron analysis, in which we do not defer to the agency's interpretation of the statute.

Id. at 1068 (quoting Chevron U.S.A. Inc. v. NRDC, 467 U.S. 837, 842, 843 (1984) (citations omitted)).

The “administrative necessity” doctrine is not as well developed as the “absurd results” doctrine, so that the courts have not had occasion to explicitly describe how the doctrine fits into the Chevron analytical framework. However, we think that a reasonable approach, in line with the DC Circuit's approach to the “absurd results” doctrine as just described, is as follows: Recall that under the “administrative necessity” doctrine, an agency is not required to implement a statutory provision in accordance with the literal requirements when doing so would be impossible, but the agency must nevertheless implement the provision as fully as possible. Placed in the context of the Chevron framework, we think that that the “administrative necessity” doctrine is based on the premise that inherent in the statutory design is the presumption that Congress does not intend to impose an impossible burden on an administrative agency. See Alabama Power v. Costle, 636 F.2d 323, 357 (DC Cir. 1980) (describing the “administrative necessity” approach as one of the “limited grounds for the creation of exemptions [that] are inherent in the administrative process, and their unavailability under a statutory scheme should not be presumed, save in the face of the most unambiguous demonstration of congressional intent to foreclose them”).

Therefore, if the literal meaning of a statutory directive would impose on an agency an impossible administrative burden, then that literal meaning should not be considered to be indicative of congressional intent. Rather, congressional intent should be considered to achieve as much of the statutory directive as possible. As a result, the agency must adopt an approach that implements the statutory directive as fully as possible. This is consistent with the DC Circuit's holding in Mova Pharm. Corp that if congressional intent is clear, but the plain meaning of a statute does not express that intent, then the agency must, under Chevron Step 1, select an interpretation that most closely approximates congressional intent. Mova Pharm. Corp, 140 F.3d at 1068.

The “one-step-at-a-time” doctrine fits into the Chevron framework in much the same manner that the “administrative necessity” doctrine does. That is, inherent in the statutory design is the presumption that Congress intended an agency, under certain circumstances, to implement the statutory requirements in a one-step-at-a-time fashion, as long as the agency stays on a path towards full implementation.

Under all of the circumstances described previously, congressional intent is clear—whether it is indicated by the plain language or otherwise—and as a result, the agency must follow that intent under Chevron Step 1. On the other hand, the agency may determine that congressional intent on the specific issue is not clear. In these cases, the agencies should proceed to Chevron Step 2 and select an interpretation or an application that is a permissible construction of the statute. This situation generally occurs when the statutory provisions are ambiguous or silent as to the specific issue, and there are no other indications of clear congressional intent. In addition, in some cases in which the literal meaning of the statutory provision, when applied to the specific question, leads to an absurd result—and, therefore, the statutory provision should be considered not to have a plain meaning—there may be no other indications of clear congressional intent. Under all these circumstances, the agency is authorized, under Chevron Step 2, to develop and implement a construction of the statute that the courts will uphold as long as it is reasonable.

As noted previously, the DC Circuit, has pointed out that this situation may also occur when the literal language leads to an absurd result, and, in attempting to implement congressional intent, the agency is “able to show that there are multiple ways of avoiding a statutory anomaly, all equally consistent with the intentions of the statute's drafters * * *. In such a case, we would move to the second stage of the Chevron analysis, and ask whether the agency's choice between these options was `based on a permissible construction of the statute.' ” Mova Pharm. Corp, 140 F.3d at 1068. As the U.S. Supreme Court has recently said, although in a context different than “absurd results,” “In the end, the interpretation applied by EPA “governs if it is a reasonable interpretation of the statute—not necessarily the only possible * * * interpretation, nor even the interpretation deemed most reasonable by the courts.” Entergy Corp. v. Riverkeeper, Inc., 129 S.Ct. 1498. 1505 (2009).

As a related matter, although the courts have described Chevron Step 2 as requiring that the agency's policy be “a permissible construction of the statute,” see Mova Pharm. Corp, 140 F.3d at 1068 (quoting Chevron, 467 U.S. at 842-43), if the statutory requirements cannot be read literally because doing so would produce “absurd results,” then the agency's policy need not be completely consistent with those particular requirements. The policy must still, in order to be upheld, be consistent with Congress's actions, but those actions should be considered to afford the agency broad discretion considering that both the statutory terms cannot be considered dispositive and underlying congressional intent is not clear. As the U.S. Supreme Court has recently said, although in a context different than “absurd results,” In the end, the interpretation applied by the agency governs if it is a reasonable interpretation of the statute—not necessarily the only possible * * * interpretation, nor even the interpretation deemed most reasonable by the courts.” Entergy Corp. v. Riverkeeper, Inc., 129 S.Ct. 1498. 1505 (2009).

There is another aspect of the “administrative necessity” doctrine worth noting in this context: The doctrine applies when (i) a literal application of the statutory directive to the case at hand is impossible for the agency to administer; and (ii) even so, either Congress clearly intended the statutory directive to apply to the case at hand or, if Congress did not clearly intend that, then the agency reasonably construes the statute to apply the statutory directive to the case at hand. In contrast, if Congress did not intend the statutory directive to apply to the case at hand, or if congressional intent is uncertain and the agency considers another approach to be reasonable, then the “administrative necessity” doctrine would not apply. As a result, the agency would not be required to implement the statutory directive to the case at hand at all, much less in a more administrable fashion.

f. Interconnectedness of the Legal Doctrines

Although we believe that each of the “absurd results,” “administrative necessity,” and “one-step-at-a-time” doctrines provide independent support for our action, we also believe that in this case, the three doctrines are intertwined and form a comprehensive basis for EPA's tailoring approach. As just discussed, each of the three doctrines is tied into the Chevron analytical framework because each is designed to give effect to underlying intent. As discussed previously, each of the three doctrines comes into play in this case because a literal reading of the PSD and title V applicability provisions results in insurmountable administrative burdens. Those insurmountable administrative burdens—along with the undue costs to sources—must be considered “absurd results” that would undermine congressional purpose for the PSD and title V programs. Under the “absurd results” doctrine, EPA is authorized not to implement the applicability provisions literally—that is, not to implement them as applying on the January 2, 2011 date that PSD and title V are triggered to all GHG sources at or above the statutory thresholds—but instead to tailor them in a manner consistent with congressional intent. That means applying the PSD and title V requirements through a phase-in approach to as many sources as possible and as quickly as possible, starting with the largest sources, as EPA does with this Tailoring Rule, at least to a certain point. By the same token, the insurmountable administrative burdens bring into play the “administrative necessity” doctrine, under which EPA is, again, authorized not to implement the applicability provisions literally, but instead to apply them in a manner consistent with administrative resources. This also means phasing them in through the approach in the Tailoring Rule. Finally, the “one-step-at-a-time” doctrine, which authorizes incremental action by agencies to implement statutory requirements under certain circumstances, provides further support for the phased tailoring approach in the Tailoring Rule.

g. Application of Chevron Approach

The Chevron analytical approach, and the three legal doctrines at issue here, apply to this action in the following manner: To reiterate, for convenience, the statutory provisions at issue: Congress, through the definition of “major emitting facility,” applied the PSD program to include (i) “any * * * stationary sources of air pollutants which emit or have the potential to emit, one hundred [or, depending on the source category, two hundred fifty] tons per year or more of any air pollutant,” CAA sections 165(a), 169(1); and (ii) and such sources that undertake a physical or operational change that “increases the amount of any air pollutant emitted” by such sources, CAA sections 165(a), 169(2)(C), 111(a)(4). Similarly, Congress, through the definition of “major source,” specified that the title V program includes “any stationary facility or source of air pollutants which directly emits, or has the potential to emit, one hundred tons per year or more of any air pollutant.” CAA sections 502(a), 501(2)(B), 302(j). EPA, through long-established regulatory action, in the case of PSD, and long-established interpretation, in the case of title V, has interpreted these definitions so that they apply only with respect to air pollutants that are subject to regulation under the CAA.

For each of these applicability provisions, the approach under Chevron is as follows: Under Chevron Step 1, we must determine whether Congress expressed an intention on the specific question, which is whether the PSD or title V applicability provisions apply to GHG sources. Said differently, the specific question is whether, in the case of PSD, Congress intended that the definitions of “major emitting facility” and “modification” apply, respectively, to all GHG sources that emit at least 100 or 250 tpy or GHGs and to all physical or operational changes by major emitting facilities that “increase[ ] the amount” of GHGs; and, in the case of title V, whether the definition of “major source” applies to all GHG sources that emit at least 100 tpy GHGs.

To determine intent, we must first examine the terms of the statute in light of their literal meaning. Here, the literal reading of each provision covers GHG sources. For PSD, a GHG source that emits at least 100 or 250 tpy GHGs literally qualifies as “stationary source [ ] of air pollutants which emit[s] or ha[s] the potential to emit, one hundred [or two hundred fifty] tons per year or more of any air pollutant [subject to regulation under the CAA].” CAA section 169(1). For modifications, a physical or operational change that increases the amount of GHG emissions qualifies as a “modification” because it “increases the amount of any air pollutant emitted” by the source. Similarly, for title V, a GHG source that emits at least 100 tpy GHGs literally qualifies as “any stationary facility or source of air pollutants which directly emits, or has the potential to emit, one hundred tons per year or more of any air pollutant [subject to regulation under the CAA].” CAA sections 502(a), 501(2)(B), 302(j).

Although each definition is clear that it applies to GHG sources as a general matter, applying each definition in accordance with its literal meaning to all GHG sources at the specified levels of emissions and at the present time—in advance of the development of streamlining methods and greater permitting authority expertise and resources—would create undue costs for sources and impossible administrative burdens for permitting authorities. These results are not consistent with other provisions of the PSD and title V requirements, and are inconsistent with—and, indeed, undermine—congressional purposes for the PSD and title V provisions. Accordingly, under the “absurd results” doctrine, neither the PSD definition of “major emitting facility” or “modification” nor the title V definition of “major source,” should be applied literally to all GHG sources, and therefore none should be considered to have a literal meaning with respect to its application to all GHG sources.

In analyzing the provisions of each definition more closely, we believe that each has four terms, any one of which could be considered not to have its literal meaning, in this respect. Specifically, each provision includes (i) The term “any * * * source,” or “a stationary source,” and that term could be considered not to refer literally to all of the GHG sources; (ii) either the term “two hundred fifty tons per year” or “100 tons per year,” or the term “increases the amount,” and those terms could be considered not to refer literally to the tonnage amount of emissions from all of the GHG sources; (iii) the term “any air pollutant,” and that term could be considered not to refer literally to the emissions from all of the GHG sources; and (iv) the term “subject to regulation under the CAA” (which we have interpreted “any air pollutant” to include), and that term could be considered not to refer literally to the emissions from all of the GHG sources. As long as any one of those four terms may be considered not to have its literal meaning as applied to GHG sources, then the definition as a whole—again, for PSD, the terms “major emitting facility” or “modifications,” and for title V, the term “major source”—cannot be considered to apply literally to GHG sources. Because we read the terms together, as integral parts of each definition as a whole, we do not think that the choice of which of those four terms within each definition cannot be considered to apply literally to GHG sources has substantive legal effect. In other words, we believe that any one of these terms, or all of them together as part of each definition as a whole, should be considered not to apply literally in the case of GHG sources.

Having determined that each definition does not have a literal meaning with respect to the applicability of PSD or title V applies to all GHG sources, we must next inquire as to whether Congress has nevertheless expressed an intent on that question through other means. We discuss the statutory terms and legislative history of the PSD and title V provisions in more detail later, but for now it suffices to say that on the issue of whether PSD and title V apply to GHG sources, we believe that congressional intent is clear, and that is to apply PSD and title V to GHG sources generally. We believe that this intent is clear from the broad phrasing of the applicability provisions—as noted earlier, the definitions apply by their terms to GHG source generally, even though the definitions should not be applied literally to all GHG sources—the fact that the various components of the PSD and title V programs can be readily applied to GHG sources, and the fact that the two programs can readily accommodate at least some GHG sources. As a result, we believe that as a matter of Chevron Step 1, PSD and title V generally apply to GHG sources. Our previous regulatory action defining the applicability provisions made this clear, and we do not reopen this issue in this rulemaking. Moreover, even if this long-established regulatory position were not justifiable based on Chevron step 1—on the grounds that in fact, congressional intent on this point is not clear—then we believe that this position, that the statutory provisions to apply PSD and title V generally to GHG sources, was justified under Chevron step 2.

On the issue of how to apply PSD to GHG sources, including the specific threshold levels and the timing, we believe that Congress could be considered to have expressed a clear intent that GHG sources be included in the PSD program at as close to the statutory thresholds as possible, and as quickly as possible, and at least to a certain point, all as consistent with the need to assure that the PSD program does not impose undue costs on sources or undue administrative burdens on the permitting authorities. Under this view, EPA would be required at Chevron Step 1 to adopt the Tailoring Rule because, by phasing in PSD applicability, it most closely gives effect to Congress's intent. Under these circumstances, EPA is authorized to exercise its expert judgment as to the best approach for phasing in the application of PSD to GHG sources.

Even so, we recognize that it could be concluded that on the issue of how to apply PSD to GHG sources, congressional intent is unclear. Under these circumstances, EPA has the discretion at Chevron Step 2 to adopt the Tailoring Rule because it is a reasonable interpretation of the statutory requirements (remaining mindful that the applicability requirements cannot be applied literally). Under the Tailoring Rule, EPA seeks to include as many GHG sources in the permitting programs at as close to the statutory thresholds as possible, and as quickly as possible, although we recognize that we ultimately may stop the phase-in process short of the statutory threshold levels.

As for title V, we believe that taken together, the various statutory requirements and statements in the legislative history do not evidence a clear congressional intent for how title V is to be applied to GHG sources. As discussed later, the relevant title V requirements and statements in legislative history differ from PSD, not least because they include provisions that concern empty permits that point in different directions. As a result, here, too, EPA has the discretion at Chevron Step 2 to adopt the Tailoring Rule as a reasonable interpretation of the statutory requirements. Alternatively, even if the statute does express a clear intent as to title V that, similar to PSD, title V requirements must be phased in as closely to the statutory threshold as possible and as quickly as possible, this Tailoring Rule is consistent with that intent.

It should also be noted that although EPA has concluded that applying the PSD and title V applicability provisions literally in the case of GHG sources would produce “absurd results” and therefore is not required, this conclusion has no relevance for applying other CAA requirements—such as the requirements concerning endangerment and contribution findings under CAA section 202(a)(1) or emission standards for new motor vehicles or new motor vehicle engines under CAA section 202—to GHGs or GHG sources. EPA's conclusions with respect to the PSD and title V applicability requirements are based on the specific terms of those requirements, other relevant PSD and title V provisions, and the legislative history of the PSD and title V programs.

Within the context of the Chevron framework, the “administrative necessity” doctrine applies as follows: Under the doctrine, Congress is presumed to intend that the PSD and title V applicability requirements be administrable. Here, those applicability requirements, if applied to GHG sources in accordance with their literal meaning, would be impossible to administer. Accordingly, under Chevron Step 1, it is consistent with congressional intent that EPA and the permitting authorities be authorized to implement the applicability requirements in a manner that is administrable, that is, through the tailoring approach.

As for the “one-step-at-a-time” doctrine, we believe it applies within the Chevron framework in conjunction with the “absurd results” and “administrative necessity” doctrines. As we discuss elsewhere, the PSD and title V applicability provisions by their terms require that sources at or above the 100/250 tpy thresholds comply with PSD and title V requirements at the time those requirements are triggered, which is when GHGs become subject to regulation. Therefore, if the literal meaning of the applicability provisions as applied to GHG sources were controlling—that is, if it reflected congressional intent—it would foreclose use of the one-step-at-a-time doctrine to implement a phase-in approach. However, the literal meaning is not controlling because—in light of the absurd results, including the insurmountable administrative burdens, that would result from the literal meaning—congressional intent is not to require the application of the PSD and title V requirements to all GHG sources at or above the statutory thresholds at the time that GHGs become subject to regulation. Instead, as described previously, we consider congressional intent for the applicability provisions, as applied to GHG sources, either (i) to be clear that PSD and title V should be phased in for GHG sources as quickly as possible, or (ii) to be unclear, so that EPA may reasonably choose to phase PSD and title V in for those sources in that manner. Under either view, congressional intent for PSD and title V applicability to GHG sources accommodates the “one-step-at-a-time” approach.

4. The PSD and Title V Programs

Having discussed both the factual underpinnings and, immediately above, the legal underpinnings for our tailoring approach, we now discuss the PSD and title V programs themselves, including, for each program, the key statutory provisions, their legislative history, and the relevant regulations and guidance documents through which EPA has implemented the provisions. We start with the PSD program.

a. The PSD program

(1) PSD Provisions

Several PSD provisions are relevant for present purposes because of the specific requirements that they establish and the window that they provide into congressional intent. These provisions start with the applicability provisions, found in CAA sections 165(a) and 169(1), which identify the new sources subject to PSD, and CAA section 111(a)(4), which describes the modifications of existing sources that are subject to PSD. CAA section 165(a) provides:

No major emitting facility on which construction is commenced after August 7, 1977, may be constructed in any area to which this part applies unless—

(1) A permit has been issued for such proposed facility in accordance with this part setting forth emission limitations for such facility which conform to the requirements of this part;

(2) The proposed permit has been subject to a review in accordance with this section * * *, and a public hearing has been held with opportunity for interested persons including representatives of the Administrator to appear and submit written or oral presentations on the air quality impact of such source, alternatives thereto, control technology requirements, and other appropriate considerations;

(4) The proposed facility is subject to the best available control technology for each pollutant subject to regulation under this chapter emitted from, or which results from, such facility * * *.

The term “major emitting facility” is defined, under CAA section 169(1) to include:

* * * stationary sources of air pollutants which emit, or have the potential to emit, one hundred tons per year or more of any air pollutant from [28 listed] types of stationary sources. * * * Such term also includes any other source with the potential to emit two hundred and fifty tons per year or more of any air pollutant. This term shall not include new or modified facilities which are nonprofit health or education institutions which have been exempted by the State.

As for modification of existing sources, CAA section 169(1)(C) provides that the term “construction,” as used in CAA section 165(a) (the PSD applicability section) “includes the modification (as defined in section 111(a)(4)) of any source or facility.” Section 111(a)(4), in turn, provides:

The term “modification” means any physical change in, or change in the method of operation of, a stationary source which increases the amount of any air pollutant emitted by such source or which results in the emission of any air pollutant not previously emitted.

As interpreted by EPA regulations, these provisions, taken together, provide that new stationary sources are subject to PSD if they emit at the 100/250-tpy thresholds air pollutants that are subject to EPA regulation, and that existing stationary sources that emit such air pollutants at the 100/250-tpy thresholds are subject to PSD if they undertake a physical or operational change that increases their emissions of such air pollutants by any amount.

Other provisions of particular relevance are the requirements for timely issuance of permits. The permitting authority must “grant[ ] or den[y] [any completed permit application] not later than one year after the date of filing of such completed application.” CAA section 165(c).

In addition, the PSD provisions articulate “the purposes of [the PSD program],” which are to balance environmental protection and growth. CAA section 160. One of the purposes, in subsection (1), is specifically “to protect public health and welfare,” and another, in subsection (3), is “to insure that economic growth will occur in a manner consistent with the preservations of existing clean air resources.”

The PSD provisions also include detailed procedures for implementation. Most relevant for sources of GHG are the provisions that the proposed permit for each source must be the subject of a public hearing with opportunity for interested persons to comment, CAA section 165(a)(2), and each source must be subject to BACT, as determined by the permitting authority on a source-by-source basis, CAA section 165(a)(4), 169(3).

(2) PSD Legislative History

The legislative history of the PSD provisions, enacted in the 1977 CAA Amendments, makes clear that Congress was largely focused on sources of criteria pollutants: primarily sulfur dioxide, PM, NO_X, and carbon monoxide (CO). This focus is evident in the basic purpose of the PSD program, which is to safeguard maintenance of the NAAQS. See S 95-127 (95th Cong., 1st Sess.), at 27.

Congress designed the PSD provisions to impose significant regulatory requirements, on a source-by-source basis, to identify and implement BACT and, for criteria pollutant, to also undertake certain studies. Congress was well aware that because these requirements are individualized to the source, they are expensive. Accordingly, Congress designed the applicability provisions (i) to apply these requirements to industrial sources of a certain type and a certain size—sources within 28 specified source categories and that emit at least 100 tpy—as well as all other sources that emit at least 250 tpy, and, by the same token, (ii) to exempt other sources from these requirements.

Although Congress required that CAA requirements generally apply to “major emitting facilities,” defined as any source that emits or has the potential to emit 100 tpy of any pollutant, Congress applied PSD to only sources at 100 tpy or higher in 28 specified industrial source categories, and at 250 tpy or more in all other source categories. This distinction was deliberate: According to Sen. McClure, Congress selected the 28 source categories after reviewing an EPA study describing 190 industrial source categories. 122 Cong. Rec. 24521 (July 29, 1976) (statement by Sen. McClure).

Congress also relied on an EPA memorandum that identified the range of industrial categories that EPA regulated under its regulations that constituted the precursor to the statutory PSD program, and listed both the estimated number of new sources constructing each year and the amount of pollution emitted by the “typical plant” in the category. The memorandum was prepared by B.J. Steigerwald, Director of the Office of Air Quality Planning and Standards and Roger Strelow, EPA's Assistant Administrator for Air and Waste Management (“Steigerwald-Strelow memorandum”). The Steigerwald-Strelow memorandum makes clear that the 100 tpy cut-off for the 28 listed sources categories, and the 250 tpy cut-off for all other sources, was meaningful; that is, there were a large number of sources below those cut-offs that Congress explicitly contemplated would not be included in the PSD program. Id. at 24548-50.

Consistent with this, the legislative history on the Senate side also specifically identified certain source categories that Senators believed should not be covered by PSD. The Senate bill language limited PSD to sources of 100 tpy or more in 28 listed source categories, and to any other categories that the Administrator might add. Sen. Muskie stated that the Senate bill excluded “houses, dairies, farms, highways, hospitals, schools, grocery stores, and other such sources.” 123 Cong. Rec. 18021 (June 8, 1977) (statement of Sen. Muskie). Sen. McLure's list of excluded source categories were “[a] small gasoline jobber, or a heating plant at a community college, [which] could have the potential to emit 100 tons of pollution annually.” 122 Cong. Rec. 24548-49 (July 29, 1976) (statement of Sen. McClure). The Senate Committee Report included a comparable list, and in describing it, concisely articulated the cost-conscious basis for the line-drawing: “[the PSD] procedure * * * must include an effective review-and-permit process. Such a process is reasonable and necessary for very large sources, such as new electrical generating plants or new steel mills. But the procedure would prove costly and potentially unreasonable if imposed on construction of storage facilities for a small gasoline jobber or on the construction of a new heating plant at a junior college, each of which may have the potential to emit 100 tons of pollution annually.” S. Rpt. 95-127 at 96-97.

The enacted legislation differs from the Senate bill by replacing the authorization to EPA to include by regulation source categories in addition to the listed 28 source categories with an inclusion of all other sources if they exceed 250 tpy, and with an authorization for the states to exempt hospitals and educational institutions. But Congress's overall intention remains clear, as the DC Circuit described in Alabama Power: “Congress's intention was to identify facilities which, due to their size, are financially able to bear the substantial regulatory costs imposed by the PSD provisions and which, as a group, are primarily responsible for emissions of the deleterious pollutants that befoul our nation's air * * *. [With respect to] the heating plant operating in a large high school or in a small community college * * * [w]e have no reason to believe that Congress intended to define such obviously minor sources as `major' for the purposes of the PSD provision.” 636 F.2d at 353-54.

A particularly important indication of congressional intent to limit the PSD program it was designing to larger sources comes in considering the emissions profile of the small-sized boilers. Congress focused closely on identifying which sources with emissions in excess of 100 tpy should not be subject to PSD even though they are subject to CAA requirements generally. But Congress viewed a large set of sources as emitting below 100 tpy and therefore not included in the PSD program. Chief among these sources, in terms of absolute numbers of sources, were small boilers. The Steigerwald-Strelow memorandum identified two categories of these boilers, differentiated by size. The first ranges in size from 10 to 250 x 10 ⁶ Btu per hour (Btu/hr), and has a “typical plant” size of 10 ⁷ Btu/hr, with “BACT emissions from typical plant” of 53 tpy, and a total of 1,446 sources in the category. The second category ranges in size from 0.3 to 10 x 10 ⁶ Btu/hr, and has a “typical plant” size of 1.3 x 10 ⁶ Btu/hr, with “BACT emissions from typical plant” of 2 tpy, and a total of 11,215 sources in the category. The memorandum discusses these two categories in the context of explaining which source categories exceed a size of 100 tpy—and therefore would be subject to PSD if a 100 tpy threshold were set—by stating, “Fortunately, most truly small boilers and typical space heating operations would not be covered.” 122 Cong. Rec. 24549 (July 29, 1976).

The legislative history also provides a window into the scope of the program that Congress anticipated and related administrability concerns. According to the Steigerwald-Strelow memorandum, the number of new sources each year whose “BACT emissions from typical plant” exceed 100 for the 28 listed source categories and 250 for all other source categories is less than 100 per year. Although the Steigerwald-Strelow memorandum does not attempt to estimate the number of modifications, it appears that based on this information, Congress had reason to expect the total size of the PSD program to be measured in the hundreds or perhaps thousands of permits each year. A program of this size would be manageable by EPA and the permitting authorities.

(3) PSD Regulatory History: Regulations Concerning the Definition of “Major Stationary Source”

For present purposes, the regulatory history of the PSD program is most noteworthy because it shows that since the inception of the program following the 1977 CAA Amendments, EPA has interpreted the statutory PSD applicability provisions to apply more narrowly—to any air pollutant subject to regulation—than their literal meaning (“any air pollutant”). EPA's initial rulemaking implementing the PSD program, which was proposed and finalized in 1977-1978, made explicit that the entire PSD program applied to only pollutants regulated under the Act. 43 FR 26380, 26403, 26406 (June 19, 1978) (promulgating 40 CFR 51.21(b)(1)(i)). In 1979-1980, EPA revised the PSD program to conform to Alabama Power v. Costle, 636 F.2d 323 (DC Cir. 1980). 44 FR 51924 (September 5, 1979) (proposed rule); 45 FR 52676 (August 7, 1980) (final rule). In this rulemaking, EPA did not disturb the pre-existing provisions that limited the applicability of the PSD program to regulated air pollutants. In 1996 EPA proposed, and in 2002 finalized, a set of amendments to the PSD provisions that included revisions to conform with the 1990 CAA Amendments, which, in relevant part, exempted hazardous air pollutants (HAPs) from PSD, under CAA section 112(b)(6). See 61 FR 38250 (July 23, 1996), 67 FR 80186 (December 31, 2002). In the preamble to the final rule, EPA noted that based on a request from a commenter, EPA was amending the regulations to “clarify which pollutants are covered under the PSD program.” EPA accomplished this by promulgating a definition for “regulated NSR pollutant,” which listed categories of pollutants regulated under the Act, and by substituting that defined term for the phrase “pollutants regulated under the Act” that was previously used in various parts of the PSD regulations. 67 FR 80240. The definition of “regulated NSR pollutant” includes several categories of pollutants (including, in general, NAAQS pollutants and precursors, pollutants regulated under CAA section 111 NSPS, Class I or II substances regulated under CAA title VI) and a catch-all category, “[a]ny pollutant that otherwise is subject to regulation under the Act.” E.g., 40 CFR 52.21(b)(50). As in the previous rulemakings, EPA did not address the difference between the definition of “major emitting facility” and its regulatory approach or indicate that it had received comments on this issue. While the definition of “major modification” in the PSD regulations has changed over time with respect to how emission increases are calculated, the regulatory history with respect to pollutant coverage parallels that of major emitting facility.

We recount this regulatory history as background information. We are not reconsidering or reopening these regulations to the extent they interpret the definition of “major emitting facility” and “modification” narrowly to be limited to pollutants subject to regulation under the Act.

b. Title V Program

Having reviewed the key statutory provisions, their legislative history, and the relevant administrative interpretations for the PSD program, we now do the same for the title V program.

(1) Title V Provisions

The key title V provisions for present purposes start with the applicability provisions, which are found in CAA sections 502(a), 501(2)(B), and 302(j). These provisions provide that it is unlawful for any person to operate a “major source” without a title V permit, section 502(a), and define a “major source” to include “any major stationary facility or source of air pollutants which directly emits, or has the potential to emit, one hundred tons per year or more of any air pollutant.” CAA sections 501(2)(B) and 302(j). As noted elsewhere, these provisions, taken together and as interpreted by EPA, provide that stationary sources are subject to title V if they emit at the 100-tpy threshold air pollutants that are subject to EPA regulation.

In addition, although title V does not have a set of provisions describing its purpose, it is clear from its provisions and its legislative history, discussed later, that its key goal is to gather into one permitting mechanism the CAA requirements applicable to a source and impose conditions necessary to assure compliance with such requirements, and thereby promote the enforceability of CAA requirements applicable to the covered sources. Section 503(b)(1) requires that the source's permit application contain a compliance plan describing how the source will ”comply with all applicable requirements” of the CAA, and section 504(a) requires that “[e]ach permit issued under [title V] shall include * * * such * * * conditions as are necessary to assure compliance with applicable requirements of [the Act].” See H.R. Rep. No. 101-490, at 351 (1990) (“It should be emphasized that the operating permit to be issued under this title is intended by the Administration to be the single document or source of all of the requirements under the Act applicable to the source.”).

Importantly, title V is replete with provisions designed to make the permitting process as efficient and smooth-running as possible, including the expeditious processing of permit applications and the timely issuance of permits. Section 503(c) requires that “the permitting authority shall approve or disapprove a completed application * * * and shall issue or deny the permit, within 18 months after the date of receipt thereof * * *.” Section 502(b)(6) requires the permitting authority to develop “adequate, streamlined, and reasonable procedures for expeditiously determining when applications are complete, for processing such applications, for public notice * * * and for expeditious review of permit actions, including * * * judicial review in State court of the final permit action by [specified persons].” Section 502(b)(7) includes a “hammer” provision designed to reinforce timely permit issuance, which is that the permitting authority's program must include:

To ensure against unreasonable delay by the permitting authority, adequate authority and procedures to provide that a failure of such permitting authority to act on a permit application or permit renewal application (in accordance with the time periods specified in [CAA] section 503 * * *) shall be treated as a final permit action solely for purposes of obtaining judicial review in State court of an action brought by any person referred to in paragraph (6) to require that action be taken by the permitting authority on such application without additional delay.

Section 502(b)(8) requires the permit program to include “[a]uthority and reasonable procedures consistent with the need for expeditious action by the permitting authority on permit applications and related matters, to make available to the public [certain permit-related documents]”. Section 502(b)(9) requires a permit revision to incorporate requirements promulgated after issuance of the permit, but only if the permit is for a major source and has a term of 3 or more years remaining. In addition, the revision must occur “as expeditiously as practicable.” Section 502(b)(10) requires the permit program to include operational flexibility provisions that “allow changes within a permitted facility * * * without requiring a permit revision, if the changes are not modifications * * * and * * * do not exceed the emissions allowable under the permit * * *.”

In addition, title V includes a comprehensive and finely detailed implementation schedule that mandates timely issuance of permits while building in EPA and affected state review, public participation, and timely compliance by the source with reporting requirements. Following the date that sources become subject to title V, they have 1 year to submit their permit applications. CAA section 503(c). As noted previously, the permitting authority then has 18 months to issue or deny the permit. CAA section 503(c). Permitting authorities must provide an opportunity for public comment and a hearing. CAA section 502(b)(6). If the permitting authority proposes to issue the permit, the permitting authority must submit the permit to EPA, and notify affected states, for review. CAA section 505(a)(1). EPA then has 45 days to review the permit and, if EPA deems it appropriate, to object to the permit. CAA section 505(b)(1). If EPA does object, then the permitting authority must, within 90 days, revise it to meet the objections, or else EPA becomes required to issue or deny the permit. CAA section 505(c). If EPA does not object, then, within 60 days of the close of the 45-day review period, any person may petition EPA to object, and EPA must grant or deny the petition within 60 days. CAA section 505(b)(2). If a permit is issued, it must include a permit compliance plan, under which the permittee must “submit progress reports to the permitting authority no less frequently than every 6 months,” and must “periodically (but no less frequently than annually) certify that the facility is in compliance with any applicable requirements of the permit, and [ ] promptly report any deviations from permit requirements to the permitting authority.” CAA section 503(b).

(2) Title V Legislative History

The legislative history of title V, enacted by Congress in the 1990 CAA Amendments, indicates the scope of the program that Congress expected: Congress expected the program to cover some tens of thousands of sources, which would approximate the scope of the permit program under the Clean Water Act. The Senate Committee on Environment and Public Works stated:

EPA estimates that the new permit requirements will cover about 8,200 major sources that emit 100 tons per year or more of criteria pollutants (which are regulated under SIPs). In addition, many smaller sources are (or, as EPA promulgates additional regulations, will be) covered by new source performance standards under section 111 of the Act, hazardous air pollutant standards under section 112 of the Act, and nonattainment provisions of this legislation. By comparison, under the Clean Water Act, some 70,000 sources receive permits, including more than 16,000 major sources. Although many air pollution sources have more emission points than water pollution sources, the additional workload in managing the air pollution permit system is estimated to be roughly comparable to the burden that States and EPA have successfully managed under the Clean Water Act.

S. Rep. 101-228, at 353 (1990). Sen. Mitchell, the Senate Majority Leader, stated that he expected “over 10,000 permits [to] * * * be issued under this program.” 136 Cong. Rec. S3239-03 (March 27, 1990). Others in Congress had similar estimates. See, e.g., 136 Cong. Rec. S3166 (“thousands and thousands of permit applications * * * will be required to be submitted”) (statement of Sen. Nickles).

Furthermore, the legislative history indicates that Congress did not contemplate that large numbers of very small sources would be subject to title V's requirements. This becomes clear by reviewing the legislative history of a companion piece of legislation to the operating permits provisions that Congress enacted into CAA section 507, which is the “Small business stationary source technical and environmental compliance assistance program.” CAA section 507. Under this provision, sources that, among other things, “are not major stationary source[s]” and that emit less than 50 tpy of any regulated pollutant, as well as less than 75 tpy or all regulated pollutants, are eligible for assistance under CAA section 507. CAA section 507(c)(1). The House Committee Report described this provision—including what types of sources it expected this provision to benefit—as follows:

New section [507] is a small source/small business provision added by the Committee. It seeks to help small businesses to comply with the problems that are likely to occur under the Act as amended by this bill. For purposes of this section, small businesses or small emitters are defined as sources that are emitting 100 tons or less per year and that have a number of employees that would qualify them for assistance from the Small Business Administration (SBA). As we look to the future of environmental protection under the Act, we take special steps here to ensure that it is possible for these small businesses to comply with minimum hassle and in recognition of the problems that are unique to them. Such small businesses include printers, furniture makers, dry cleaners, and millions of other small businesses in this country.

House Committee Report, H.R. 101-590, at 354. In this manner, the House Committee Report made clear that it expected “millions of * * * small businesses”—including “printers, furniture makers, dry cleaners” and many others—to benefit from the CAA section 507 small source/small business program, but Congress did not expect them to become subject to the operating permit requirements of title V because their emissions fell below 100 tpy, which is, in general, the threshold for title V applicability as a “major source.”

The legislative history of title V confirms that Congress viewed a principal purpose of title V as providing a vehicle to compile the requirements applicable to the source. As the report of the House Committee on Energy and Commerce (“House Committee Report”) stated, “It should be emphasized that the operating permit to be issued under this title is intended by the Administration to be the single document or source of all of the requirements under the Act applicable to the source.” H.R. Rep. No. 101-490, at 351 (1990). Combined with the source's reporting requirements, this compilation of applicable requirements would facilitate public awareness of a source's obligations and compliance and would facilitate compliance and enforcement.

On the Senate side, Sen. Chafee, one of the floor managers of the bill, made a similar point:

The permits will serve the very useful function of gathering and reciting in one place—the permit document itself—all of the duties imposed by the Clean Air Act upon the source that holds the permit. This would clearly be an improvement over the present system, where both the source and EPA must search through numerous provisions of state implementation plans and regulations to assemble a complete list of requirements that apply to any particular plant * * *.

Once these permits are in place, plant managers will be better able to understand and to follow the requirements of the Clean Air Act. At the same time, EPA will be better able to monitor how well each plant is complying with those requirements. This is a highly sensible approach for all concerned.

136 Cong. Rec. S213 (January 24, 1990) (statement of Sen. Chafee). Sen. Lieberman made a similar statement. 136 Cong. Rec. 3172-73 (March 26, 1990) (statement of Sen. Lieberman). Thus, a central purpose of the title V permit program is to compile all the requirements applicable to the source into a single place, the permit. Implicit in this purpose is that the sources subject to title V will have applicable requirements to be compiled. As Sen. Chafee directly stated, “[T]he vast majority of these permit applications will * * *, in all likelihood, only codify the existing requirements of the applicable State implementation plan.” 136 Cong. Rec. S2720 (March 20, 1990) (statement of Sen. Chafee).

More broadly, the legislative history also indicates congressional concern about the costs of permitting for small businesses, and a determination to minimize those costs to the extent possible. This concern is reflected in several provisions of title V. For example, section 502(a) authorizes EPA to exempt all or part of a source category—except for any major source from the title V permit program if EPA “finds that compliance with [title V] requirements is impracticable, infeasible, or unnecessarily burdensome on such categories.” Similarly, the permit fee provisions include a presumptive minimum fee amount, but authorize an exemption from that presumptive amount upon a showing that a lesser amount will meet overall fee requirements, CAA section 502(b)(3)(B)(iv). One of the drafters of this provision, Rep. Wyden, explained that its purpose was to preserve the flexibility of states to impose lower fees of small businesses:

I note that the provision on fees allows reductions for small sources where appropriate. The state has some flexibility, under the general permit fee provisions, to adjust fee levels for any source so long as the average fee charged meets the statutory minimum.

136 Cong. Rec. H12884 (Oct. 26, 1990) (statement of Rep. Wyden). See, e.g., 136 Cong. Rec. H2559 (May 21, 1990) (statement of Rep. Wyden) (discussing need to “help small businesses through the air permit labyrinth”).

The legislative history also indicates that Congress was deeply concerned both about the need not to burden sources generally with undue costs and to assure the administrability of the title V program, and as a result, was determined to make the program as smooth-running as possible. These goals are reflected in many of the title V requirements, as discussed previously. See, e.g., CAA section 502(b)(6) (requiring “adequate, streamlined, and reasonable procedures for expeditiously determining when applications are complete, for processing such applications, for public notice * * * and for expeditious review of permit actions); CAA section 502(b)(7) (includes a “hammer” provision designed to reinforce timely permit issuance); CAA section 502(b)(9)-(10) (limiting circumstances under which permit revision is required; requiring revision to occur “as expeditiously as practicable;” including operational flexibility provisions).

The legislative history confirms that these provisions were designed to reduce costs to sources and promote administrability. The “Chafee-Baucus Statement of Senate Managers” for the bill explained the purpose of the CAA section 502(b)(6) requirement for “[a]dequate, streamlined, and reasonable procedures for expeditious[ ]” permit actions as follows:

[M]uch concern has been expressed that this new permitting process will unduly delay the proper functioning of many sources, and we intend to mitigate any delay by directing that the process be expeditious.

In addition to this general directive for expeditious processing, we mandate in new section 503 that permitting authorities approve or reject permit applications within certain specified time periods following filing. In this fashion, we have taken explicit steps to protect against undue delays.

136 Cong. Rec. S16941 (statement of Sen. Chafee). The same statement explained that the permit revision procedures of CAA section 502(b)(9) reflect a—

careful effort to ensure that the permit program works effectively and efficiently. Succinctly, this provision accommodates two competing concerns. On the one hand, it is important to ensure that permit requirements remain up-to-date as the provisions of the Clean Air Act are developed and new requirements are imposed. On the other hand, it also is important to be sure that we do not reduce the permit program to a shambles by requiring sources to engage in a continuous process of revising their permits as these new requirements are imposed.

136 Cong. Rec. 16942 (Oct. 27, 1990) (Chafee-Baucus statement of Senate Managers) (statement of Sen. Chafee).

In addition, these concerns were at the bottom of the following statement by Sen. Chafee, in which he described how the bill's drafters had revised it in response to a concern by industry that an earlier version of the bill would have put undue costs on industry:

We have also heard concerns from industry that S. 1630 would burden sources unduly by requiring them to submit—along with their permit applications—plans explaining how they intend to comply with all requirements of the Clean Air Act that apply to them.

But, Mr. President, we emphatically do not intend to burden industry with preparation and submission of unnecessary compliance plans. The substitute clarifies that any compliance plans would address only those matters by which the sources would comply with new requirements imposed by this act as it is finally signed into law. These plans would not need to address compliance with any existing Clean Air Act requirements, unless the source is in violation of those requirements.

136 Cong. Rec. S2107 (March 5, 1990) (statement of Sen. Chafee).

As another indication of congressional concern over administrability, Congress recognized that at the beginning of the program, large numbers of permit applications might overwhelm the permitting authorities. To protect against this, Congress included in CAA section 503(c) a phase-in schedule for permitting authorities to act on the initial set of permit applications. Under 503(c), permitting authorities were not required to act on the initial set of permit applications within 18 months after it received the application, but rather could act on one-third of them on an annual basis over a 3-year period. Sen. Chafee, in describing an early version of this provision—which would have allowed permitting authorities to phase in the submission of permit applications—explained that its purpose was “to avoid a logjam of permit applications[,] * * * ensure that [regulatory] gridlock can be avoided, and [ensure] that the permitting process will work with a minimum of disruption and delay.” 136 Cong. Rec., S2106 (March 5, 1990) (statement of Sen. Chafee).

(3) Title V Regulatory History

As with PSD, for present purposes, the regulatory history of the title V program is most noteworthy because it shows that beginning shortly after the inception of the program following the 1990 CAA Amendments, EPA has interpreted the statutory title V applicability provisions to apply more narrowly—to any air pollutant subject to regulation—than their literal meaning (“any air pollutant”). As discussed previously, title V applies to any “major source,” defined, as relevant here, under CAA sections 501(2)(B) and 302(j), as “any stationary facility or source of air pollutants which directly emits, or has the potential to emit, one hundred tons per year or more of any air pollutant * * *.” EPA's regulations mirror the CAA definitional provisions. 40 CFR 70.2.

However, since 1993, EPA has interpreted the applicability provisions more narrowly. At that time, which was shortly after title V was enacted, EPA issued a guidance document making clear that it interprets this requirement to apply to sources of pollutants “subject to regulation” under the Act. Memorandum from Lydia N. Wegman, Deputy Director, Office of Air Quality Planning and Standards, U.S. EPA, “Definition of Regulated Air Pollutant for Purposes of Title V” (Apr. 26, 1993) (Wegman Memorandum). The interpretation in this memorandum was based on: (1) EPA's reading of the definitional chain for “major source” under title V, including the definition of “air pollutant” under section 302(g) and the definition of “major source” under 302(j); (2) the view that Congress did not intend to require a variety of sources to obtain title V permits if they are not otherwise regulated under the Act (see also CAA section 504(a), providing that title V permits are to include and assure compliance with applicable requirements under the Act); and (3) consistency with the approach under the PSD program.

While the specific narrow interpretation in the Wegman Memorandum of the definition of “air pollutant” in CAA section 302(g) is in question in light of the holding in Massachusetts v. EPA, 549 U.S. 497, 533 (2007) (finding this definition to be “capacious”), we believe that the overall rationale for our interpretation of the applicability of title V remains sound. EPA continues to maintain its interpretation, consistent with CAA sections 302(j), 501, 502 and 504(a), that the provisions governing title V applicability for “a major stationary source” can only be triggered by emissions of pollutants subject to regulation. This interpretation is based primarily on the purpose of title V to collect all regulatory requirements applicable to a source and to assure compliance with such requirements, see, e.g., CAA section 504(a), and on the desire to promote consistency with the approach under the PSD program.

In the Tailoring Rule notice of proposed rulemaking, EPA acknowledged the Wegman Memorandum and affirmed the memorandum's continued viability, stating that “EPA continues to maintain this interpretation.” 74 FR 55300, col. 3, fn. 8; see also 75 FR 17022-23 (Interpretive Memo reconsideration).

As with PSD, we recount this regulatory history as background information, and we are not reconsidering or re-opening this interpretation of the definition of “major source” narrowly to be limited to pollutants subject to regulation under the Act.

5. Application of the “Absurd Results” Doctrine for the PSD Program

Having reviewed the factual background, legal doctrines, and the key components of the PSD and title V programs, we now turn towards interpreting the PSD and title V requirements in accordance with the Chevron framework, accounting for the applicable legal doctrines. We begin with the “absurd results” doctrine, and apply it first to the PSD requirements.

In this action, we finalize, with some refinements, the “absurd results” basis we proposed. Specifically, we are revising our regulations to limit PSD applicability to GHG emitting sources by revising the regulatory term, “regulated NSR pollutant,” and although our revised regulations do not accord with a literal reading of the statutory provisions for PSD applicability, which are incorporated into the definition of “major emitting facility” and “major modification,” we have concluded that based on the “absurd results” doctrine, a literal adherence to the terms of these definitions is not required. Even so, we believe Congress did intend that PSD apply to GHG sources as a general matter. Further, we may apply PSD to GHG sources in a phased-in manner, as we do through the tailoring approach, because either congressional intent is clear on that issue and the tailoring approach best reflects it, or congressional intent is unclear and the tailoring approach is a reasonable interpretation of the statute.

a. Congressional Purpose for the PSD Program

To reiterate, for convenience, CAA section 169(1) defines a “major emitting facility” to include “any * * * source[] [that] emit[s], or ha[s] the potential to emit, [depending on the source category], one hundred [or two hundred fifty] tons per year or more or more of any air pollutant.” CAA section 169(1); and a “modification” as any physical or operational change in “a stationary source which increases the amount of any air pollutant emitted by such source,” CAA section 169(2)(C), 111(a)(4). We also reiterate that, as discussed above, beginning with our initial rulemaking in 1977-1978 to implement the PSD program, we have interpreted these definitions more narrowly by reading into them the limitation that a source is subject to PSD only if the air pollutants in question are “subject to regulation under the Act.” 40 CFR 51.166(b)(49)(iv). EPA is not re-opening this interpretation in this regulation in this action.

Under the current interpretation of the PSD applicability provision, EPA's recent promulgation of the LDVR will trigger the applicability of PSD for GHG sources at the 100/250 tpy threshold levels as of January 2, 2011. This is because PSD applicability hinges on the definition of “major emitting facility,” which, under EPA's long-standing narrowing interpretation, but absent further tailoring, applies PSD to sources of any air pollutant subject that is subject to regulation under another provision of the CAA. EPA's promulgation of the LDVR means that GHGs will become subject to regulation on the date that the rule takes effect, which will be January 2, 2011.

But absent tailoring, the January 2, 2011 trigger date for GHG PSD applicability will subject an extraordinarily large number of sources, more than 81,000, to PSD each year, an increase of almost 300-fold. And the great majority of these new sources will be small commercial or residential sources. We believe that for many reasons, this result is contrary to congressional intent for the PSD program, and in fact would severely undermine what Congress sought to accomplish with the program. As a result, under our Chevron analysis, accounting for the “absurd results” doctrine, the statutory definition for “major emitting facility” (as interpreted narrowly to include “subject to regulation”) should not be read to apply to all GHG sources at or above the 100/250 tpy threshold as of the January 2, 2011 date. Rather, the definitions of “major emitting facility” and “modification” should be tailored so that they apply to GHG sources on a phased-in basis, with the largest sources first, as we describe in this rule.

As explained previously, Chevron Step 1 calls for a determination of congressional intent, and the courts consider the best indicator of congressional intent to be the plain meaning of the statute. However, the U.S. Supreme Court has held that the literal meaning of a statutory provision is not conclusive “in the `rare cases [in which] the literal application of a statute will produce a result demonstrably at odds with the intentions of the drafters' * * * [in which case] the intention of the drafters, rather than the strict language, controls.” United States v. Ron Pair Enterprises, 489 U.S. 235, 242 (1989). To determine whether “the intentions of the * * * drafters” differ from the result produced from “literal application” of the statutory provisions in question, the courts may examine the overall context of the statutory provisions, including whether there are related statutory provisions that either conflict or are consistent with that interpretation; and the legislative history to see if it exposes what the legislature meant by the terms in question. In addition, the courts may examine whether a literal application of the provisions produces a result that the courts characterize variously as absurd, futile, strange, or indeterminate, and therefore so illogical or otherwise contrary to sensible public policy as to be beyond anything Congress would reasonably have intended. In such cases, the literal language cannot be said to reflect the intention of the drafters, and therefore does not control. See United States v. Ron Pair Enterprises, 489 U.S. 235, 242-43 (1989); Griffin v. Oceanic Contractors, Inc., 458 U.S. 564, 571 (1982).

Here, applying the definitions of “major emitting facility” and “modification” literally (as EPA has interpreted them more narrowly) at the present time—in the absence of streamlining measures or additional permitting authority resources, and without tailoring—would be contrary to congressional purpose for the PSD provisions, as found in the statutory provisions and legislative history, especially in light of the impact from applying those definitions literally. Congress established the PSD program in large measure because it was concerned that around the country, industrial development, which was confronting barriers to locating in nonattainment areas (that is, areas that do not meet the NAAQS), would attempt to locate in clean air areas (that is, attainment areas or unclassifiable areas), but that as a consequence, the clean air areas would see their air quality deteriorate to the point where they, too, would no longer meet the NAAQS. The end result would be the spread of environmental and health problems to those formerly clean air areas, as well as more barriers to further industrial development. With these concerns in mind, Congress designed the PSD program to require newly constructing or modifying sources in areas with air quality that meets the NAAQS (or that is unclassifiable) to analyze their emissions of NAAQS pollutants and to implement controls as needed to assure that those emissions do not significantly deteriorate air quality. Many of the PSD requirements, and much of the discussion in the legislative history, reflect these aspects of the PSD program. E.g., CAA sections 162, 163, 164, 165(a)(3), 165(d)(2), 165(e), 166; see generally H. Rep. 95-294, 95th Cong., 1st Sess. (1977) 103-78.

Congress also designed the PSD program to impose controls on non-NAAQS pollutants, through the requirement under CAA section 165(a)(4) that the source be “subject to the best available control technology for each pollutant subject to regulation under this chapter emitted from, or which results from, such facility.” For example, when Congress enacted the PSD provisions in 1977, sources emitting HAPs were required to implement BACT for those pollutants, although in the 1990 CAA Amendments, Congress redesigned CAA section 112, which includes the requirements for HAPs, and excluded HAPs from PSD. CAA section 112(b)(6).

Congress was keenly aware that the PSD program needed to serve two purposes: Protect the environment and promote economic growth. Congress explicitly identified these two goals in the “purposes” section of the PSD provision, CAA section 160, and various PSD requirements clearly reflect them. For example, to protect economic growth, the PSD program expedites the permit process to include a 1-year limitation on the time that the permitting authority has act on permit applications. To protect the environment, in addition to including many provisions that focus on NAAQS pollutants, the PSD program requires that the preconstruction permit impose emission limits that reflect BACT for each pollutant subject to regulation under another CAA provision. CAA section 165(a)(4). This BACT provision also makes clear, by its terms, that although Congress designed the PSD program largely with NAAQS pollutants in mind, Congress also intended that sources subject to PSD control the emissions of their other pollutants as well. The DC Circuit has recognized the twin goals of environmental protection and economic development that underlie PSD, and has upheld EPA interpretations of the PSD program that reflect a balancing of those goals. See, e.g., New York v. EPA, 413 F.3d 3, 27 (DC Cir.), rehearing en banc den. 431 F.3d 801 (2005).

Congress was also keenly aware that the PSD analyses and controls that it was mandating had to be implemented on a source-by-source basis, and that this process would be expensive for sources. As a result, Congress intended to limit the PSD program to large industrial sources because it was those sources that were the primary cause of the pollution problems in question and because those sources would have the resources to comply with the PSD requirements. Congress's mechanism for limiting PSD was the 100/250 tpy threshold limitations. Focused as it was primarily on NAAQS pollutants, Congress considered sources that emit NAAQS pollutants in those quantities generally to be the large industrial sources to which it intended PSD to be limited.

That Congress paid careful attention to the types and sizes of sources that would be subject to the PSD program and designed the thresholds deliberately to limit the program's scope is evident from the legislative history. Several Senate floor statements and the Committee Report made clear that PSD should not apply to small sources. As discussed later, Congress scrutinized information that EPA provided as to types and sizes of sources, found largely in the Steigerwald-Strelow memorandum. Sen. Muskie stated that the Senate bill excluded “houses, dairies, farms, highways, hospitals, schools, grocery stores, and other such sources.” 123 Cong. Rec. 18021 (June 8, 1977) (statement of Sen. Muskie). Sen. McClure stated that PSD should be limited to “industrial plants of significant impact,” and should exclude ”[a] small gasoline jobber, or a heating plant at a community college, [which] could have the potential to emit 100 tons of pollution annually.” 122 Cong. Rec. 24548-49 (July 29, 1976) (statement of Sen. McClure). The Senate Committee Report mirrored Sen. McClure's statement, and concisely articulated the cost-related basis for the line-drawing: “[The PSD] procedure * * * must include an effective review-and-permit process. Such a process is reasonable and necessary for very large sources, such as new electrical generating plants or new steel mills. But the procedure would prove costly and potentially unreasonable if imposed on construction of storage facilities for a small gasoline jobber or on the construction of a new heating plant at a junior college, each of which may have the potential to emit 100 tons of pollution annually.” S. Rpt. 95-127 at 96-97.

The DC Circuit had occasion, in Alabama Power, to acknowledge this legislative history. “Congress's intention was to identify facilities which, due to their size, are financially able to bear the substantial regulatory costs imposed by the PSD provisions and which, as a group, are primarily responsible for emissions of the deleterious pollutants that befoul our nation's air.” Alabama Power, 636 F.2d at 353. The Court added, “Though the costs of compliance with [the PSD] requirements are substantial, they can reasonably be borne by facilities that actually emit, or would actually emit when operating at full capacity, the large tonnage thresholds specified in section 169(1).”. Id. at 354.

It is not too much to say that applying PSD requirements literally to GHG sources at the present time—in the absence of streamlining or increasing permitting authority resources and without tailoring the definition of “major emitting facility” or “modification”—would result in a program that would have been unrecognizable to the Congress that designed PSD. Congress intended that PSD be limited to a relatively small number of large industrial sources. Without phasing in PSD and title V applicability to GHG sources so as to allow the development of streamlining methods and increases in permitting authority resources, the PSD program would expand by January 2, 2011, from the current 280 sources per year to almost 82,000 sources, virtually all of which would be smaller than the sources currently in the PSD program and most of which would be small commercial and residential sources. Until EPA could develop streamlining methods, all of the sources that would become newly subject to PSD—whether they be larger or smaller sources, whether industrial or commercial/residential sources—would have to undergo source-specific BACT determinations for their GHG emissions, as well as their emissions of conventional pollutants in amounts in excess of the significance levels. We estimate that the commercial and residential sources—the great majority of which are small business—would each incur, on average, almost $60,000 in PSD permitting expenses. This result would be contrary to Congress's careful efforts to confine PSD to large industrial sources that could afford these costs.

A closer look at the legislative history confirms the view that Congress did not expect PSD to apply to large numbers of small sources, including commercial and residential sources, and instead expected the 100/250 tpy thresholds to limit PSD's applicability to larger sources. As noted previously, Congress relied on an EPA memorandum—the Steigerwald-Strelow memorandum—that identified the range of industrial categories that EPA regulated under its program that constituted the precursor to the statutory PSD program, and listed both the estimated number of new sources constructing each year and the amount of pollution emitted by the “typical plant” in the category. The Steigerwald-Strelow memorandum makes clear that the 100 tpy cut-off for the 28 listed sources categories, and the 250 tpy cut-off for all other sources, would exclude from PSD a large number of sources. 122 Cong. Rec. 24548-50 (July 29, 1976). However, virtually all, if not all, of the sources in half the 28 source categories emit CO₂ in quantities that equal or exceed the 100 tpy threshold, and almost all of the sources in the remaining categories emit CO₂ in quantities that equal or exceed the 100 tpy threshold. Therefore, applying the “major emitting facility” definition to GHG sources, in the absence of streamlining methods and without tailoring, would, as a practical matter, vitiate much of the purpose of the 100 tpy cut-off for industrial sources.

Most telling, in this regard, is the small-sized boilers, which the Steigerwald-Strelow memorandum describes, in terms of size, pollutants emitted, and numbers of sources, as follows: The memorandum identified two categories of these boilers, differentiated by size. The first ranges in size from 10 to 250 x 10⁶ Btu/hr, and has a “typical plant” size of 10⁷ Btu/hr, with “BACT emissions from typical plant” of 53 tpy, and a total of 1,446 sources in that category. The second category ranges in size from 0.3 to 10 x 10⁶ Btu/hr, and has a “typical plant” size of 1.3 x 10⁶ Btu/hr, with “BACT emissions from typical plant” of 2 tpy, and a total of 11,215 sources in the category. That memorandum makes clear that EPA did not believe that sources in these two categories—and especially the smallest one—would be subject to PSD under a 100 tpy threshold, by stating, “Fortunately, most truly small boilers and typical space heating operations would not be covered.” 122 Cong. Rec. 24549 (July 29, 1976). However, these data and conclusions were all based on emissions of NAAQS pollutants, the amounts of which placed these boilers well below the PSD threshold limitations. In general, most boilers of these small sizes are fired with natural gas, and a natural gas boiler greater than 0.5 x 10⁶ Btu/hr emits at least 250 tpy CO_2. Therefore, if the CO₂ emissions of these small boilers are considered—as would occur by applying the definition of “major emitting facility” to GHG sources without tailoring—then most of them would in fact be subject to PSD. Again, this result would directly contravene Congress's intention to limit PSD to “industrial plants of significant impact.” 122 Cong. Rec. 24548-49 (statement of Sen. McClure).

Perhaps the most compelling reason why applying the PSD program to GHG sources without tailoring, and before the development of streamlining methods, would be inconsistent with congressional intent, is that the resulting program would prove unadministrable. Although the legislative history of the PSD program does not reveal much explicit congressional focus on administrability issues, the Steigerwald-Strelow Memorandum, which identifies the source categories and numbers of sources that were before Congress as it considered PSD, suggests that the program that Congress fashioned could be expected to cover at most a few thousand sources each year. This appears to be approximately the size of the program that EPA administered before the 1977 CAA Amendments, so that it seems reasonable to assume that Congress expected the PSD program it enacted to be within EPA's and the states' administrative capacities.

Moreover, the Alabama Power court stressed the importance of administrability concerns: Most importantly, the Court held that EPA, in interpreting the “modification” provisions that apply PSD to physical or operational changes by major emitting facilities that “increase the amount of any air pollutant emitted,” CAA section 111(a)(4), may “exempt from PSD review some emission increases on grounds of de minimis or administrative necessity,” and went on to state that in establishing the exemption thresholds, “[t]he Agency should look at the degree of administrative burden posed by enforcement at various de minimis threshold levels.” 636 F.2d at 400,405. In addition, the Court based its holding that potential-to-emit for purposes of the applicability thresholds should be defined as emissions at full capacity with implementation of control equipment, in part on its view that with this definition, the number of sources subject to PSD would be manageable:

Though the costs of compliance with section 165 requirements are substantial, they can reasonably be borne by facilities that actually emit, or would actually emit when operating at full capacity, the large tonnage thresholds specified in section 169(1). The numbers of sources that meet these criteria, as we delineate them, are reasonably in line with EPA's administrative capability.

Alabama Power, 636 F.2d at 354. However, applying PSD to GHG sources before streamlining and without tailoring would increase the size of the PSD program at least an order of magnitude beyond what Congress seems to have expected, which would have been far beyond the “administrative capability” that Alabama Power described EPA as having.

Beyond this disconnect with congressional expectations, what is most important is that the extraordinarily large number of permit applications would overwhelm permitting authorities and slow their ability to process permit applications to a crawl. Our best estimate at present is that permitting authorities would need to process almost 82,000 permit applications per year, compared to, at most, 800 in the current PSD program. The total additional workload, in work hours, for PSD permits would be more than 19.5 million more work hours, compared to 150,795 work hours for the current PSD program, and the total additional costs would be over $1.5 billion, compared with $12 million for the current PSD program.

At proposal, we noted that the states had estimated that the influx of permit applications that would result from applying the 250 tpy threshold at actual emissions would, without additional resources, result in permitting delays of 3 years. In fact, as we noted at proposal, a literal reading of the PSD requirements would require their application at the 250 tpy PTE level, which would result in ten times more permit applications than were assumed when the states made the 3-year estimate. Further, our current estimates of the numbers of sources that would be subject to PSD requirements are about twice what we estimated at proposal, as described elsewhere. Moreover, our estimate of the number of hours that permitting authorities would need to process a permit application from a source in the commercial or residential sector—which is, by far, the largest single sector—is three and one-half times as long as we estimated at proposal. And under a literal reading of the PSD applicability provisions as applied to GHG sources, the permitting authorities would be required to implement a program of this size beginning on January 2, 2011, less than 9 months from now. We received many comments from states and industry raising concerns about the cost to sources and administrative burdens of PSD permitting if the statutory threshold were to apply for GHG emissions. One commenter estimated a cost of over $5 billion and the dedication of over 17,000 FTEs to this effort.

We consider it difficult to overstate the impact that applying PSD requirements literally to GHG sources as of January 2, 2011—before streamlining or increasing permitting resources and without tailoring—would have on permitting authorities and on the PSD program, and we are concerned that this impact could adversely affect national economic development. The number of PSD permits that would be required from such an approach is far beyond what the PSD program has seen to date. It is clear throughout the country, PSD permit issuance would be unable to keep up with the flood of incoming applications, resulting in delays, at the outset, that would be at least a decade or longer, and that would only grow worse over time as each year, the number of new permit applications would exceed permitting authority resources for that year. Because PSD is a preconstruction program, during this time, tens of thousands of sources each year would be prevented from constructing or modifying. In fact, it is reasonable to assume that many of those sources will be forced to abandon altogether plans to construct or modify. As a result, a literal application of the PSD applicability provisions to GHG sources would slow construction nationwide for years, with all of the adverse effects that this would have on economic development.

The remedies for this scenario would be for permitting authorities to increase their PSD funding by over 100-fold, from $12 million to over $1.5 billion, or the development by EPA and the permitting authorities of streamlining techniques. But it is not possible for permitting authorities to increase their funding to those levels in the foreseeable future, partly because of the sheer magnitude of those levels and partly because of the financial challenges that states currently face. And, for the reasons discussed later, although streamlining offers genuine promise to improve the manageability of the PSD workload, streamlining cannot do so in the very near term and, in any event, the extent to which it can do so has not yet come into focus.

So clear are at least the broad outlines of this picture that EPA did not receive any substantive comments arguing that permitting authorities could in fact administer the PSD program with the applicability requirements applied literally to GHG sources beginning in the very near future. Every permitting authority that addressed this issue in their comments on the proposed Tailoring Rule stated unequivocally that it could not administer the PSD program at the statutory levels. To cite a few examples (each of which considered both the PSD and title V programs together): NACAA, which represents air pollution control agencies in 53 states and territories, stated it “* * * agrees with the EPA that immediately attempting to implement the PSD and title V programs using the statutory thresholds meets the test for invoking the administrative necessity and absurd results doctrines.” Similarly, the California Air Resources Board stated that it “* * * concurs with the United States, EPA that if more appropriate applicability thresholds [as opposed to the statutory thresholds] are not set for GHG it will not be administratively possible to implement these [the PSD and tile V] permitting programs.” All other state and local permitting agencies that commented on the proposed tailoring provided similar comments that they would not have the adequate staff capacity or resources to be able to successfully administer their permitting programs with the addition of GHG emission sources at the statutory thresholds for PSD and title V.

It is the many-year delays in permit issuance and the consequent chilling of economic development that provide perhaps the clearest indication that applying the PSD applicability provisions to GHG sources without tailoring produces absurd results. These effects would undermine one of Congress's central purposes in establishing the PSD program, which was to promote development in clean air areas by large industrial sources (as long as they included environmental safeguards). As discussed previously, this goal is manifest in the structure of the PSD provisions, and Congress even went so far as to make this goal explicit in the purposes section of the PSD provisions.

Moreover, at the present time, there is relatively little environmental benefit in subjecting large numbers of small GHG sources to the expensive, source-by-source PSD permitting requirements. They represent a relatively small share of the GHG inventory and the control options available to them, at present, are limited. As a result, approaches other than source-by-source permitting presently offer more promise for generating emissions reductions in an efficient manner. These approaches, which may be developed through both federal and state efforts, include requirements, incentives, and educational outreach to promote efficiency improvements to boilers and furnaces and energy efficient operations, including, for example, weatherization programs.

For all these reasons, interpreting the definition of “major emitting facility” and “modification” literally—that is, as EPA has interpreted them more narrowly, but without tailoring and before the program requirements can be streamlined or permitting authority resources can be increased—would produce results that are not consonant with, and, in fact, would severely undermine, Congress's purpose for the PSD program. These results may fairly be characterized as the type of absurd results that support our view that the literal terms of the PSD applicability provisions do not indicate congressional intent for how those provisions should applied to GHG sources.

b. Congressional Intent for the Applicability Provisions

(1) Congressional Intent for Whether and How PSD Applies to GHG Sources

Several of the PSD provisions and statements in the legislative history are particularly important in determining whether and how the PSD program should apply to GHG sources, as discussed elsewhere:

(1) The applicability provisions, under CAA section 165(a) and 169(1). These provisions are written broadly, and although, as we explain above, they cannot be read literally to apply to GHG sources at or above the 100/250 tpy, they nevertheless can be read to indicate that directionally, Congress intended that PSD be applied inclusively.

(2) The various PSD provisions that identify the pollutants subject to PSD. Compare, e.g., CAA sections 162, 163, 164, 165(a)(3), 165(d)(2), 165(e), and 166 (NAAQS pollutants) with CAA sections 165(a)(3)(C), 165(a)(4) (other pollutants). These provisions indicate that a major purpose of the PSD program is to control NAAQS pollutants, but that the program also covers non-NAAQS pollutants.

(3) The requirement that permitting authorities act on PSD applications within 1 year. CAA section 165(c). This provision indicates that Congress anticipated the PSD program would be of a size that would allow permitting authorities to meet this deadline.

(4) The purpose provision. CAA section 160. This provision makes clear that PSD is designed both to protect public health and welfare and to promote economic growth.

(5) In addition, we consider important the legislative history indicating the Congress intended PSD to apply to large industrial sources because they were the primary source of the air pollution problems and they have the resources to manage the demands of the PSD permitting process; and that, by the same token, Congress expected that small sources would not be subject to PSD. The legislative history does not specifically mention GHG sources.

Looking at these provisions and the legislative history together, we think Congress can be said to have intended that the PSD program apply to GHG sources as a general matter. The most important indication of congressional intent in this regard is the applicability provisions, which provide, in part, that PSD applies to (i) “any * * * source[ that] emit[s], or ha[s] the potential to emit [the specified quantity] of any air pollutant,” CAA section 169(1); and (ii) to any such source that undertakes a physical or operational change that “increases the amount of any air pollutant emitted.” CAA section 169(2)(C), 111(a)(4). These terms are quite broad, and should be read to include GHG sources and GHGs. See Massachusetts v. EPA, 549 U.S. 497, 533 (2007) (“Because greenhouse gases fit well within the Clean Air Act's capacious definition of `air pollutant,' we hold that EPA has the statutory authority to regulate the emission of such gases from new motor vehicles.”). Moreover, including GHG sources—under certain circumstances—is consistent with the PSD provisions that refer to other pollutants, establish the time-frame for acting on PSD applications, and establish the overall purpose of the program. In addition, including GHG sources—again, under certain circumstances—is consistent with the legislative history that PSD be limited to sources that cause a meaningful part of the air pollution problem and have the resources to manage the PSD requirements. No PSD provision explicitly imposes any limitation of PSD to large industrial sources, and Congress's reasoning for focusing on large industrial sources—which was that these sources are best suited to handle the resource -intensive analyses required by the PSD program—could extend to GHG sources under certain circumstances (that is, large sources first, and smaller sources after streamlining methods are developed). Similarly, as discussed previously, it is reasonable to read into Congress's intent that the PSD program be limited to a size that permitting authorities would be able to administer, but it is consistent with that reading to recognize that the permitting authorities could take certain steps—including adoption of streamlining measures and ramping up resources—that would allow them to handle a higher volume of permitting. Finally, we find nothing in the PSD provisions or legislative history that would indicate congressional intent to exclude GHG sources. Accordingly, we believe that Congress must be said to have intended an affirmative response for whether PSD applies to sources of GHGs as a general matter. Our previous regulatory action defining the PSD applicability provisions made this clear, and we do not reopen this issue in this rulemaking. Moreover, even if this long-established regulatory position were not justifiable based on Chevron Step 1—on the grounds that in fact, congressional intent on this point is not clear—then we believe that this position, that the statutory provisions to apply PSD to GHG sources in general, was justified under Chevron Step 2.

As to how PSD applies to GHG sources, although, for reasons discussed previously, the 100/250 tpy threshold provision, which establishes the scope of PSD applicability, should not be read as applying literally to GHG sources—and as a result, the applicability provision as a whole cannot be said to have a plain meaning as to the scope of coverage of GHG sources—we believe that the applicability provisions and legislative history nevertheless indicate a congressional intent for how PSD should apply to GHG sources. That is to apply PSD to as many sources as possible as quickly as possible, at least to a certain point. We believe that this intent can be inferred from the inclusiveness of the applicability provision, combined with the legislative history that focuses on Congress's desire to include in the PSD program sources that have the resources to comply with the requirements and, as the Court in Alabama Power recognized, Congress's concern about administrability. That is, at first, PSD may apply to the largest GHG sources because they may be expected to have the resources to comply with PSD's requirements and permitting authorities may be expected to accommodate those sources; and over time, with streamlining and increases in permitting authority resources, PSD may apply to more GHG sources. As discussed later, the tailoring approach is consistent with congressional intent in this regard.

We recognize the tension between the applicability provisions, which are inclusive, and the statements in the legislative history that express Congress's expectation that PSD be limited to large industrial sources. At least to a point, the applicability provisions and these statements can be reconciled by recognizing that the reason why Congress expected that PSD would be limited to large industrial sources was that Congress recognized that PSD applied on a source-by-source basis, that this would be costly to sources, and that only the large industrial sources could afford those costs. Taking certain actions—including streamlining PSD requirements—can render PSD more affordable and thereby allow its application to smaller sources in a more cost-effective manner. In this way, PSD's inclusive applicability provisions can be reconciled with the narrower scope Congress expected, and this is part of the reason why we characterize congressional intent as being consistent with phasing in the applicability of PSD to GHG sources through the tailoring approach.

On the other hand, if Congress cannot be said to have expressed an intent as to the manner and scope of PSD applicability to GHG sources, then, under Chevron Step 2, EPA may apply a reasonable interpretation of the applicability provisions to determine the scope of coverage of GHG sources that is consistent with the statutory requirements. The Tailoring Rule is a reasonable interpretation under Chevron Step 2. It is consistent with (1) The applicability provisions, recognizing that as we have seen, those provisions cannot be applied literally under these circumstances, (2) the provisions described above concerning which pollutants the PSD provisions cover and the timetable for permitting authority action on PSD applications; (3) the purpose provisions of PSD, and the accompanying legislative history, because it protects public health and welfare without inhibiting economic development; and (4) the legislative history indicating Congress intended that PSD be limited to sources that cause a meaningful part of the problem and can manage its requirements, because it will expand PSD's applicability only after streamlining methods and greater permitting authority resources will allow for such an expansion in an orderly manner.

(2) Criteria for Establishing Phase-in Schedule

The specific phase-in schedule under the tailoring approach will depend on several things. The first is our progress in developing streamlining methods that will render the permitting authority workload more manageable by taking some sources off the table (through regulations or guidance interpreting PTE), and by allowing for more efficient permit processing (through general permits and presumptive BACT). At the same time, streamlining techniques will lower permitting costs to sources or even eliminate some sources' obligations to obtain permits altogether. The second is the time that permitting authorities need to ramp up their resources in an orderly and efficient manner to manage the additional workload. The third is information we have as to the sources' abilities to meet the requirements of the PSD program and the permitting authorities' ability to process permits in a timely fashion. That information will be based on the real-world experience the permitting authorities will accumulate as they proceed to process permit application for the larger GHG sources.

Thus, under our present approach, we will develop streamlining techniques, we expect the permitting authorities to ramp up resources in response to the additional demands placed upon them in the first two steps, and we will gather real-world information about the GHG permitting process; and based on all that, we will address expanding the PSD program in a step-by-step fashion to include more sources over time. We intend to follow this process to establish both the PSD applicability thresholds and, as we describe next, the significance levels.

(3) Criteria for Establishing Significance Levels

The criteria for establishing the significance levels are the same as for establishing the “major emitting facility” thresholds. As noted previously, under the applicable CAA sections, any physical or operational change at a stationary source that “increases the amount of any air pollutant emitted by such source” or that results in the emission of a new pollutant is treated as a “modification” that is subject to PSD requirements. Although the CAA, by its terms, treats as an “increase” any amount of emissions that is greater than zero, the DC Circuit held in Alabama Power v. Costle that EPA may establish a threshold—called the significance level—on de minimis grounds for the amount of any particular pollutant that may be increased. 636 F.2d at 400.

Of particular importance, the Court in Alabama Power indicated that EPA may rely on administrative considerations to establish significance levels. Id. To reiterate, the Court held that “EPA does have discretion, in administering the statute's `modification' provision, to exempt from PSD review some emission increases on grounds of de minimis or administrative necessity.” 636 F.2d at 400. The Court added a more detailed exposition of its views in a subsequent part of its opinion, where it discussed the BACT provision, under CAA section 165(a)(4), and the Court made clear that those views applied as well to the “modification” provision. There, the Court invalidated an EPA regulation that established a 100- and 250-tpy exemption from the BACT requirement. Both the BACT provision and the modification provision apply by their terms to all emissions from a source, but the Court stated that each provision must be read to incorporate an exemption based on de minimis or administrative considerations, and explained:

We understand that the application of BACT requirements to the emission of all pollutants from a new facility, no matter how miniscule some may be, could impose severe administrative burdens on EPA, as well as severe economic burdens on the construction of new facilities. But the proper way to resolve this difficulty is to define a de minimis standard rationally designed to alleviate severe administrative burdens, not to extend the statutory 100 or 250-ton threshold to a context where Congress clearly did not apply it. Just as for the applicability of PSD to modifications, the de minimis exemption must be designed with the specific administrative burdens and specific regulatory context in mind. This the Agency has failed to do. We do not hold that 100 tons per year necessarily exceeds a permissible de minimis level; only that the Agency must follow a rational approach to determine what level of emission is a de minimis amount.

A rational approach would consider the administrative burden with respect to each statutory context: what level of emission is de minimis for modification, what level de minimis for application of BACT. Concerning the application of BACT, a rational approach would consider whether the de minimis threshold should vary depending on the specific pollutant and the danger posed by increases in its emission. The Agency should look at the degree of administrative burden posed by enforcement at various de minimis threshold levels. It is relevant that our decision requires the Agency, in its evaluation of emissions of facilities, to take into account the facility's air pollution controls. It may also be relevant, though it is certainly not controlling, that Congress made a judgment in the Act that new facilities emitting less than 100 or 250 tons per year are not sizeable enough to warrant PSD review.

Id. at 405. As just quoted, the Court acknowledged the 100 and 250 tpy thresholds for a major emitting facility, and did not indicate whether the modification exemption level could exceed those statutory levels, but nevertheless, the Court made clear that EPA may “consider the administrative burden” associated with modifications to establish an exemption level for modifications.

EPA has established significance levels for various pollutants, generally relying on a de minimis basis. See, e.g., 45 FR 52676, 52705-52710 (August 7, 1980). In these actions, EPA generally established the level based on the triviality of the amount of emissions excluded. To this point, we have not attempted to determine de minimis—that is, trivial—levels for GHGs. Instead, in this rulemaking, EPA is establishing a phase-in schedule for significance levels based on the Chevron framework, accounting for the “absurd results,” “administrative necessity,” and “one-step-at-a-time” doctrines. It is not necessary to establish a permanent de minimis level in this rulemaking. For one thing, the Court in Alabama Power explicitly authorized an administrative basis for significance levels. Moreover, were EPA to establish a de minimis level, that amount could be below—perhaps even well below—the “major emitting facility” thresholds established in this rulemaking on grounds of “administrative necessity” and the other doctrines. Accordingly, at present, if we were to establish a permanent significance level on a de minimis basis, that level could result in too many small sources being required to submit permit applications while the phase-in is occurring. This would give rise to the same problems concerning undue costs to the sources and administrative burdens for the permitting authorities for which we are fashioning a remedy. Accordingly, the significance levels we establish with this action are the lowest levels that sources and permitting authorities can reasonably be expected to implement at the present time in light of the costs to the sources and the administrative burdens to the permitting authorities.

c. Other Possible Approaches to Reconciling a Literal Reading of PSD Applicability Provisions and Congressional Intent

Commenters have suggested another approach to reconciling the inconsistency between the definition of “major emitting facility” and congressional intent. They urge that the “major emitting facility” definition should be applied so that only sources that emit NAAQS pollutants, for which the area is designated attainment or unclassifiable, in the requisite quantities would be subject to PSD, and sources would not be subject to PSD based solely on their emissions of non-NAAQS pollutants or a NAAQS pollutant for which an area has been designated nonattainment. Some commenters argue that this approach is mandated by several of the PSD provisions, read together or at least that the relevant statutory provisions are ambiguous and that this approach is a reasonable reading of them. Under this approach, we would not need to phase in the application of PSD by lowering the applicability threshold for GHG emitters.

Specifically, many commenters have questioned whether EPA has the authority to regulate GHGs under the PSD provisions. Although the specific lines of reasoning vary somewhat from one commenter to another, in general, they based their arguments largely on CAA sections 161 and 165(a). Under CAA section 161:

In accordance with the policy of section 101(b)(1), each applicable implementation plan shall contain emission limitations and such other measures as may be necessary, as determined under regulations promulgated under this part, to prevent significant deterioration of air quality in each region (or portion thereof) designated pursuant to section 107 as attainment or unclassifiable.

Commenters point out that section 107 applies only to NAAQS pollutants and directs that areas be designated as attainment, nonattainment, or unclassifiable on a pollutant-by-pollutant basis. Under CAA section 165(a), a “major emitting facility” cannot be constructed “in any area to which this part applies” unless it meets certain requirements. According to some commenters, these provisions, read together, limit PSD's applications to only NAAQS pollutants that are emitted from sources in areas that are designated attainment or unclassifiable for those pollutants. Other comments make a similar point, except to state that PSD applies more broadly to pollutants with a local, ambient impact.

Some commenters go on to take the position that NAAQS pollutants for which the area is designated attainment or unclassifiable are the only pollutants that can be regulated under any provision of the PSD requirements; while others take the position that once PSD is triggered for a source on the basis of its NAAQS pollutants, then other, non-NAAQS, pollutants may be regulated under certain PSD provisions, in particular, the BACT provision under CAA section 165(a)(4). These commenters agree, however, that emissions of GHGs, by themselves, cannot trigger PSD applicability. Finally, some commenters state that even if the PSD provisions cannot be read by their terms to preclude GHGs from triggering PSD, then they can be read to authorize EPA to determine that GHG emissions do not trigger PSD.

We recognize, as we have said elsewhere, that a major purpose of the PSD provisions is to regulate emissions of NAAQS pollutants in an area that is designated attainment or unclassifiable for those pollutants. However, we do not read CAA sections 161 and the “in any area to which this part applies” clause in 165(a), in the context of the PSD applicability provisions, as limiting PSD applicability to those pollutants. The key PSD applicability provisions are found in sections 165(a) and 169(1). Section 165(a) states, “No major emitting facility on which construction is commenced after August 7, 1977, may be constructed in any area to which this part applies unless [certain requirements are met].” A “major emitting facility” is defined, under CAA section 169(1), as “any * * * stationary source[s] which emit[s], or ha[s] the potential to emit, one hundred [or, depending on the source category, two hundred fifty] tons per year or more of any air pollutant.” As discussed elsewhere, EPA has long interpreted the term “any air pollutant” to refer to “any air pollutant subject to regulation under the CAA,” and for present purposes, will continue to read the “subject to regulation” phrase into that term.

Although section 165(a) makes clear that the PSD requirements apply only to sources located in areas designated attainment or unclassifiable, it does not, by its terms, state that the PSD requirements apply only to pollutants for which the area is designated attainment or unclassifiable. Rather, section 165(a) explicitly states that the PSD requirements apply more broadly to any pollutant that is subject to regulation. Moreover, another requirement in CAA section 165(a) also applies to air pollutants broadly. Under CAA section 165(a)(3), one of the requirements for securing a preconstruction permit is to demonstrate that the source's emissions “will not cause, or contribute to, air pollution in excess of any (A) maximum allowable increase or maximum allowable concentration for any pollutant in any area [to which the PSD requirements apply], (B) [NAAQS] in any air quality control region, or (C) any other applicable emission standard or standard of performance under this chapter.” As just quoted, subparagraph (C), by its terms clearly applies to non-NAAQS pollutants. This is because it refers to (1) “any other applicable emission standard,” which distinguishes it from subparagraph (B) and therefore from NAAQS pollutants; and (2) “any * * * standard of performance under this chapter,” which refers to standards of performance under section 111, several of which are for non-NAAQS pollutants. See, e.g., 40 CFR 60.33c(a) “municipal solid waste landfill emissions.” By the same token, CAA section 110(j) specifically contemplates that a source required to hold a permit under title I of the Act, which includes a PSD permit, demonstrate that the source complies with “standards of performance,” which may include requirements for pollutants other than NAAQS.

In addition, CAA section 163(a)(4) includes as a PSD requirement that “the proposed facility is subject to the best available control technology for each pollutant subject to regulation under this chapter emitted from, or which results from, such facility.” Section 163(a)(4)'s broad reference to “each pollutant subject to regulation under this chapter” clearly indicates that it applies to non-NAAQS pollutants, as long as they are regulated under other provisions of the Act. The DC Circuit, in Alabama Power v. Costle, 636 F.2d 323, 361 n.90 (DC Cir. 1980) indicated that, under the law applicable at the time the Court handed down the decision in 1980, PSD applies to HAPs.

In addition, PSD requirements are part of SIPs, and although SIPs generally are limited to provisions that implement the NAAQS, and therefore generally are limited to controlling NAAQS pollutants (or non-NAAQS pollutants that affect ambient air quality), see generally CAA section 110, Congress explicitly required SIPs to include requirements to protect visibility, under CAA section 169A-B. See CAA sections 110(a)(2)(D)(i)(II), 169A(b)(2)(A). Congress took much the same approach with the PSD program, which was to require that PSD requirements be included in the SIPs, but to explicitly require that PSD apply to non-NAAQS pollutants.

These provisions—sections 165(a)(3), 165(a)(4), and 110(j)—all indicate by their terms that PSD requirements apply to non-NAAQS pollutants. As such, they lend credence to our view that Congress intended the PSD applicability provisions to include GHG sources. At the very least, they demonstrate that Congress certainly knew how to specifically describe certain air pollutants—e.g., “air pollution in excess of * * * any other applicable emission standard or standard of performance under this chapter,” CAA section 165(a)(3)(C)—which indicates that its decision not to specifically describe air pollutants in the applicability provisions suggests an intent to cover air pollutants broadly.

To return to sections 161 and the “in any area to which this part applies” phrase in 165(a), which commenters rely on as the cornerstone of their argument, commenters in effect take the position that Congress intended the geographic references in these provisions—that is, the references to areas designated as attainment or unclassifiable—to limit the scope of the permitting provisions. We think it unpersuasive that Congress would have taken such an indirect, and silently implied, route to limit the scope of the permitting provisions. As noted previously, the permitting provisions apply broadly by their terms. Had Congress intended to limit PSD permitting in the manner urged by commenters, it certainly could have done so directly, such as by limiting PSD permitting to “any pollutant for which an area is designated attainment or unclassifiable.” Indeed, Congress did so in other PSD provisions, discussed previously. Similarly, in other sections of the CAA, Congress also directly limited the scope of pollutant applicability by specifying which pollutants are or are not subject to the provision. See, e.g., section 111(d) (performance standards for existing sources apply only to pollutants other than NAAQS or HAPs), section 112(a)(1) (applying air toxics requirements in section 112 to sources that emit above the specified tonnage thresholds of “hazardous air pollutants”).

In addition, although section 161 requires that SIPs contain emission limitations and other measures as necessary to prevent significant deterioration in areas designated as attainment or unclassifiable, it does not by its terms limit SIPs to only those measures.

Most broadly, we read the PSD provisions and their legislative history to evidence Congress's intent that PSD apply throughout the country to large sources that undertake new construction or modifications, and that Congress's overall purpose was to assure that, as the industrial stock of the nation turned over, it would become cleaner for all air pollutants emitted. Greenhouse gas sources, as a general matter, fit readily into this overall vision. At the time that Congress enacted the PSD provisions in 1977, every area of the nation was designated attainment or unclassifiable for at least one air pollutant, and that has remained the case to the present time. Accordingly, at all times, PSD has applied in every area of the country. The PSD requirements clearly cover all air pollutants emitted by the source, and provide a process for reviewing those emissions and determining BACT for them under CAA section 165(a)(4). It is true that at the time Congress adopted the PSD provisions, it was primarily concerned about the NAAQS pollutants—or, as some commenters assert, pollutants with local, ambient impact—because those pollutants represented a major component of the air pollution problems it was aware of and was addressing. But its overall purpose was broad enough to cover additional pollutants; the process it enacted for establishing BACT was broad enough to encompass additional pollutants; and the applicability provisions it established were phrased broadly enough to encompass additional pollutants, see section 169(1). As a result, we believe that the PSD applicability provisions, which, again, refer to, as we have interpreted them, “any air pollutant [subject to regulation under the CAA],” should be seen as “capacious” and therefore encompass GHG sources, in much the same manner as the U.S. Supreme Court viewed the definition of “air pollutant” to be “capacious” and therefore encompass GHGs. Massachusetts v. EPA, 549 U.S. 497, 533 (2007).

In addition, it should not be overlooked that we have applied PSD to non-NAAQS pollutants since the inception of the program over 30 years ago. For example, prior to the 1990 CAA Amendments, PSD applied to HAPs regulated under CAA section 112; and over the years, EPA has established significance levels for fluorides, sulfuric acid mist, hydrogen sulfide, TRS, reduced sulfur compounds, municipal waste combustor organics, municipal waste combustor metals, municipal waste combustor acid gases, and municipal solid waste landfill emissions, see 40 CFR 51.166(b)(23)(i); and EPA has proposed a significance level for ozone depleting substances. See 61 FR 38307 (July 23, 1996). Of course, the basis for all these actions is PSD's applicability to these non-NAAQS air pollutants. We are not aware that EPA's actions in establishing significance levels for these pollutants gave rise to challenges on grounds that the PSD provisions do not apply to them. As the U.S. Supreme Court recently stated in upholding an EPA approach in another context: “While not conclusive, it surely tends to show that the EPA's current practice is a reasonable and hence legitimate exercise of its discretion * * * that the agency has been proceeding in essentially this fashion for over 30 years.” Entergy Corp. v. Riverkeeper, Inc., 129 S.Ct. 1498, 1509 (2009) (citations omitted).

Finally, we note that excluding GHG sources from PSD applicability would create inequitable results. Consider the hypothetical case of two sources that construct in the same area, each of which emits the same amount of GHGs, and that amount is large enough to trigger PSD applicability. Assume that the first one, but not the second, also emits NAAQS pollutants amounts large enough to trigger PSD applicability. If GHG sources are excluded from PSD applicability, then the first of those sources, but not the second, would be subject to PSD requirements for its GHG emissions. Similarly, consider the hypothetical case of two sources that emit identical amounts of the same NAAQS pollutant and identical amounts of GHGs, all amounts of which are large enough to trigger PSD applicability requirements. Assume that the first source constructs in an area that is an attainment or unclassifiable area for the NAAQS pollutant that it emits, and that the second source constructs in an area that is not an attainment or unclassifiable area for that NAAQS pollutant. Here again, if GHG sources are excluded from PSD applicability, then the first of those sources, but not the second, would be subject to PSD requirements for its GHG emissions. These results are inequitable and would create an uneven playing field and for this reason, too, support our view that the PSD applicability provisions apply to GHG sources.

Accordingly, we reject the argument that section 165 must be, or may reasonably be, limited in scope to pollutants for which an area has been designated as attainment or unclassifiable. Rather, the PSD applicability provision—the definition of “major emitting facility” in CAA section 169(1)—applies by its terms (as we have interpreted them narrowly through regulation) to sources emitting any air pollutant subject to regulation, and is not limited to any NAAQS air pollutant. Our research has not disclosed any explicit statements in the legislative history that Congress intended to limit PSD applicability to sources of NAAQS pollutants.

6. Application of the “Absurd Results” Doctrine for the Title V Program

Having discussed the application of the Chevron framework, taking account of the “absurd results” doctrine, for the PSD applicability requirements, we now turn towards applying the same approach to the title V applicability requirements. Because of the parallels between the PSD and title V applicability provisions, much of the discussion later parallels the previous discussion of PSD. As with PSD, we finalize, with some refinements, the “absurd results” basis we proposed. Specifically, we are revising our regulations to limit title V applicability to GHG emitting sources by revising the regulatory term, “major source,” and although our revised regulations do not accord with a literal reading of the statutory provisions for title V applicability, which are incorporated into the statutory definition of “major source,” we have concluded that based on the “absurd results” doctrine, a literal adherence to the terms of this definition is not required. Rather, we may apply title V to GHG sources in a phased-in manner, as we do through the tailoring approach, because although congressional intent is clear that title V applies to GHG sources in general, congressional intent is unclear on the question of how title V applies, and the tailoring approach is a reasonable interpretation of the statute.

To reiterate, for convenience, the title V applicability provisions provide that after the effective date of a title V program, it is unlawful for any person to operate a “major source” without a title V permit (CAA section 502(a), and define a “major source” to include “any major stationary facility or source of air pollutants which directly emits, or has the potential to emit, one hundred tons per year or more of any air pollutant.” CAA sections 501(2)(B) and 302(j).

Under the current interpretation of the title V applicability provisions, EPA's recent promulgation of the LDVR will trigger the applicability of title V for GHG sources at the 100 tpy threshold levels as of January 2, 2011. This is because title V applicability hinges on the definition of “major source,” which, under EPA's long-standing narrowing interpretation, but absent further tailoring, applies title V to sources of any air pollutant that is subject to regulation under another provision of the CAA. EPA's promulgation of the LDVR means that GHGs will become subject to regulation on the date that the rule takes effect, which will be January 2, 2011.

But absent tailoring, the January 2, 2011 trigger date for GHG PSD applicability will see an extraordinarily large number of sources—some 6.1 million—become subject to title V, an increase of over 400-fold over the 14,700 sources that currently are subject to title V. The great majority of these will be small commercial or residential sources.

We believe that for many reasons, this result is contrary to congressional intent for the title V program, and in fact would severely undermine what Congress sought to accomplish with the program. As a result, under Chevron, accounting for the “absurd results” doctrine, the statutory definition for “major source” (as EPA has already narrowed it to refer to any air pollutant “subject to regulation”) should not be read to apply to all GHG sources at or above the 100 tpy threshold as of the January 2, 2011 date. Rather, the definition of “major source” should be tailored so that it applies to GHG sources on a phased-in basis, with the largest sources first, as we describe in this rule.

a. Congressional Intent for the Title V Program

As we said, previously, in a similar circumstance involving the PSD program, applying title V requirements to GHG sources without tailoring the definition of “major source”—and, as discussed later, without streamlining the title V requirements or allowing for time for permitting authorities to ramp up resources—would result in a program unrecognizable to the Congress that enacted title V, and one that would be flatly unadministrable. Without tailoring, the PSD program would expand from the current 14,700 sources to some 6.1 million, with the great majority of the sources being small commercial and residential sources that not only have never been permitted before, but that in many cases have no applicable requirements under the CAA to include in the permit. In the next several sections, we will describe some of the specific ways that this literal application of title V would not only differ from, but would undermine, congressional intent. But the big picture is readily drawn: The influx of millions of permit applications would do nothing less than overwhelm the program Congress finely crafted for thousands of sources, with its multi-step deadlines measured in days and months, its multiple mandates for expeditious permit processing, its nuanced limitations on the need for permit revisions, its efforts to save smaller sources permit fees. Regulatory gridlock, precisely what Congress strove to avoid, would result.

Most visibly, interpreting the applicability provisions literally to include GHG sources at the 100 tpy level immediately would revise the program from what Congress envisioned in three major ways, the legislative history of each of which was discussed previously:

• It would immediately expand the program to cover several-hundred-fold more sources than Congress anticipated.
• It would immediately expand the program to cover very small sources that Congress expected would not be included in the program.
• It would immediately expand the program so that a large number of sources have empty permits, that is, permits without applicable requirement, and undermine the implementation of the program for sources with applicable requirements.

Revising the program in this way through a literal interpretation of the applicability provisions—without tailoring the applicability requirements and without streamlining the program requirements—is clearly inconsistent with Congress's conception of the program's scope, and these inconsistencies are foundational. Most importantly, the program that would result would be unduly costly to sources and impossible for permitting authorities to implement, and therefore would frustrate the purposes that Congress intended to achieve with the program that it did design.

As discussed previously, Congress was fully aware that with the title V program, it was subjecting sources and permitting authorities to additional costs and administrative burdens, and it was fully aware of concerns that absent careful design, the program could become a formula for regulatory gridlock. Determined to make the program workable, Congress crafted the provisions to be efficient and workable.

However, if title V were to apply to GHG sources at the 100 tpy level, until EPA could develop streamlining methods, all of these sources newly subject to title V would need to apply for permits. We estimate that the commercial and residential sources would incur, on average, expenses of $23,175, while an industrial source would incur expenses of $46,350, to prepare a permit application and receive a permit. The great majority of these sources would be small commercial and residential sources of the type that Congress did not expect would be included in title V. For example, as discussed above, the legislative history of title V, including both the permit program under CAA sections 501-506 and the “small business stationary source technical and environmental compliance assistance program” under CAA section 507, indicated that Congress did not expect that “printers, furniture makers, dry cleaners, and millions of other small businesses” would become subject to title V. House Committee Report, H.R. 101-590, at 354. These sources generally do not have the potential to emit conventional pollutants at or above the 100 tpy threshold. However, many do have the potential to emit GHGs above that threshold. Many printers and furniture makers use a variety of combustion equipment that has the potential to emit at least 100 tpy CO₂, and many commercial dry cleaners have gas-fired driers that have the potential to emit at least 100 tpy of CO₂. All told, there are in fact “millions of * * * small businesses” that would become subject to title V—of the 6.1 million sources that would become subject to title V, the great majority are small businesses—if the title V applicability provisions are applied literally to GHG sources.

Moreover, the overall cost to all 6.1 million sources—before the development of streamlining methods—would be a staggering $49 billion per year over a 3 year period. Imposing burdens of this magnitude on these sources—individually and in total—would of course be contrary to Congress's efforts to minimize the expenses of title V, especially to small sources. The magnitude of the costs is, in a sense, heightened because a great many of these sources will not have applicable requirements to include in their permits; therefore, much of the costs will produce relatively little benefit.

Yet, the most important reason why applying the title V program to GHG sources without tailoring, and before the development of streamlining methods, would be inconsistent with congressional intent, is that the resulting program would prove unadministrable. Adding some 6.1 million permit applications to the 14,700 that permitting authorities now handle would completely overwhelm permitting authorities, and for all practical purposes, bring the title V permitting process to a standstill.

The costs to permitting authorities of this multi-million-source program would again be staggering. On average, and without streamlining, a permitting authority would expend 214 hours, which would cost $9,844, to issue a permit to a commercial or residential source; and 428 hours, which would cost $19,688, to issue a permit to an industrial source. In all, permitting authorities would face over $21 billion in additional permitting costs each year due to GHGs, compared to the current program cost of $62 million each year.

Beyond this disconnect with congressional expectations as to scope of the program, the extraordinarily large number of permit applications would overwhelm permitting authorities and slow their ability to process permit applications to a crawl. As described at proposal, the survey of permitting authorities conducted by NACAA found that a literal application of the title V applicability provisions to all GHG sources would result in permitting delays of some 10 years. However, as we further noted at proposal, this estimate was based on the assumption that the applicability threshold would be 100 tpy based on actual emissions; in fact, the applicability threshold would be 100 tpy based on PTE, which would sweep in many more sources. Moreover, as stated elsewhere, we currently estimate the amount of per-permit work hours for permitting authorities in processing title V permit applications to be several times higher than what we estimated at proposal. As with PSD, such a program would be beyond anything within our experience, and it is difficult to give a meaningful estimate for how long the permitting process would take for each permit on average. But it is clear that the period would be many years longer than even the 10 years estimated by NACAA.

In addition, applying title V to all GHG sources without tailoring would be in tension with a specific CAA requirement, that of CAA section 503(c), which imposes a time limit of 18 months from the date of receipt of the completed permit application for the permitting authority to issue or deny the permit. It would be impossible for permitting authorities to meet this statutory requirement if their workload increases from some 14,700 permits to 6.1 million, and without streamlining. Instead, as just noted, permit applications would face multi-year delays in obtaining their permits.

Moreover, these delays would undermine the overall statutory design that promotes the smooth-running of the permitting process, and the underlying purpose of the title V program itself. As noted elsewhere, Congress intended through title V to facilitate sources' compliance with their CAA obligations by establishing an operating permit program that requires the source to combine all of its CAA requirements, and explain how it will assure compliance with such requirements. Congress established a comprehensive process to implement the operating permit program. Through this process, following the date that sources become subject to title V, they have 1 year to submit their permit applications. CAA section 503(c). As noted, the permitting authority then has 18 months to issue or deny the permit. CAA section 503(c). Permitting authorities must provide an opportunity for public comment and a hearing. CAA section 502(b)(6). If the permitting authority proposes to issue the permit, the permitting authority must submit the permit to EPA for review, and notify affected states. CAA section 505(a)(1). EPA then has 45 days to review the permit and, if EPA deems it appropriate, to object to the permit. CAA section 503(b)(1). If EPA does object, then the permitting authority must, within 90 days, revise it to meet the objections, or else EPA becomes required to issue or deny the permit. CAA section 503(c). If EPA does not object, then, within 60 days of the close of the 45-day review period, any person may petition EPA to object, and EPA must grant or deny the petition within 60 days. CAA section 505(b)(2). This set of applicant, permitting authority, and EPA actions and deadlines establishes the process for the prompt and efficient issuance of operating permits for the appropriate universe of sources.

But at least for an initial period, until resources could be ramped up and streamlining methods could be developed, the extraordinary numbers of these permit applicants would sweep aside this carefully constructed program, and instead, backlog the permit authorities. This initial period would last for many years. As discussed elsewhere, it would take several years to develop and apply streamlining measures—in particular, general permits—and during that time, the permit backlog would grow so large that it would take many more years for permitting authorities to catch up by raising the requisite funds and hiring and training the necessary employees.

What's more, only a fraction of these millions of sources newly covered by title V will be subject to any CAA requirements due to their GHG emissions, and we suspect that a larger number will not be subject to any CAA requirements at all. As a result, for most of these sources, although they would need to apply for and receive a permit, there would be no applicable requirements to include in the permit and thus the exercise would not improve compliance.

The picture that emerges from a literal application of title V's requirements to all GHG sources—at the 100 tpy level, beginning on January 2, 2011—shows multi-year delays in issuance of all permits, for both the sources that have applicable requirements and that Congress clearly intended the program to cover, and for the millions of sources that may not be subject to any applicable requirements. In short, this literal interpretation would apply title V to millions of sources that Congress did not expect be covered, and the ensuing administrative burdens—at least initially—would impede the issuance of permits to the thousands or perhaps tens of thousands of sources that Congress did expect be covered. This is the type of “absurd results” from a literal application of statutory provisions that the courts have held should be avoided. And even beyond all that, the sheer magnitude of the numbers involved—millions of permits requiring thousands of FTEs at a cost to the permitting authorities of billions of dollars, all this beginning immediately at the time that GHGs become subject to regulation—makes clear that this result of a literal application of the title V provisions to GHG sources cannot be what Congress intended.

b. EPA's Reconciliation of Applicability Provisions With Congressional Intent

For the reasons just described, we should not consider the literal meaning of the applicability provisions to be determinative of congressional intent as to the applicability of title V to all GHG sources; rather, we should examine other provisions of the statute and the legislative history to determine congressional intent on that question. If congressional intent is clear, we must adopt and implement an applicability approach that is as close as possible to congressional intent; and if congressional intent is not clear, then we must select an interpretation that is reasonable and consistent with the statutory requirements. This section explains EPA's view of congressional intent for the applicability of the title V program to GHG sources and the principles and approach EPA is using for tailoring. In addition, we also respond to other approaches that were suggested by commenters.

To determine congressional intent, we consider the statutory provisions and legislative history, and this analysis is similar to that for PSD. The most important title V provisions and legislative history for this purpose are the following:

(1) The applicability provisions themselves, which, as we have interpreted them, apply title V to all sources that emit at least 100 tpy of any air pollutant subject to regulation. CAA sections 502(a), 501(2)(B), 302(j). Although we do not believe these provisions should be applied literally to GHG sources, their broad phrasing indicates, directionally, a congressional intent towards inclusiveness of sources in title V, including GHG sources.

(2) The provisions for general permits, CAA section 504(d); and title V fees, CAA section 502(b)(3)(A). These provisions give title V an important measure of flexibility as to its scope. The explicit authorization of general permits means that title V may be applied to more sources and more efficiently, thereby saving costs to both source and permitting authority. The requirements for permit fees provide a mechanism for permitting authorities to, over time, develop their programs to cover more sources. In this sense, these provisions could be construed to indicate congressional intent to apply title V inclusively, to the extent that permitting authorities can accommodate additional sources through general permits and permit fees.

(3) The detailed procedural requirements—including time periods, such as the 18-month time period for action on permit applications—for title V permit processing. CAA sections 503, 505. Although these requirements are consistent with applying title V to GHG sources—in the sense that at least in theory, there is nothing intrinsic to GHG sources that would mean that permitting authorities could not comply with these requirements—these requirements cast doubt on whether Congress can be said to have intended that title V cover the many small GHG sources (at least immediately) in light of the risk that including all those sources in title V would strain the process.

(4) The provisions and legislative history concerning applicable requirements, which indicate that a purpose of title V is to include sources' applicable requirements in their permits. CAA sections 503(b)(2), 504(a). These provisions, and the accompanying legislative history, discussed previously, suggest an intent to include within title V GHG sources that have applicable requirements, but may also suggest that Congress would not have intended to include in title V the large numbers of GHG sources that have “empty permits,” at least where their inclusion would undermine implementation of the program for sources with applicable requirements.

(5) The small-business-assistance provisions of section 507 and the legislative history of title V—both the permitting program and the small-business-assistance program—concerning the scope of the permitting program and small businesses. These indicate that Congress intended title V to cover some tens of thousands of sources, and did not intend that title V apply to small businesses. These provisions and legislative history suggest that Congress did not intend for title V to apply to include large numbers of small GHG sources.

Finally, the legislative history of title V does not explicitly mention GHG sources, which could suggest that Congress did not have occasion to focus on whether and how title V would apply to GHG sources.

With all this, we believe that Congress had a clear intent on the question of whether title V generally applies to GHG sources, and that was that it does. As with PSD, the most important indication of congressional intent in this regard is the applicability provisions, which provide, in part, that title V applies to “any stationary facility or source of air pollutants which directly emits, or has the potential to emit, [the requisite quantity] of any air pollutant.” CAA sections 502(a), 501(2)(B), 302(j). This term is quite broad, and should be read to include GHG sources. See Massachusetts v. EPA, 549 U.S. 497, 533 (2007) (“Because greenhouse gases fit well within the Clean Air Act's capacious definition of `air pollutant,' we hold that EPA has the statutory authority to regulate the emission of such gases from new motor vehicles.”). Moreover, including GHG sources—under certain circumstances—is consistent with the various statutory provisions and statements in the legislative history described previously.

In the alternative, if it is concluded that Congress did not express a clear intent on that question, then, under Chevron S tep 2, EPA exercises its discretion to conclude that title V applies to GHG sources as a general matter. This is a reasonable policy because applying the title V program to at least the larger GHG sources will assure promote accountability and enforceability for those sources, which is a key goal of the title V program, and will not impose obligations that are beyond the resources of those sources or insurmountable burdens on the permitting authorities. This policy is a reasonable interpretation of the statutory provisions for the same reasons just discussed.

As to the question of how title V applies to GHG sources, we believe that Congress cannot be said to have expressed a clear intent. A central aspect of how title V is to apply to GHG sources concerns “empty permits,” and on this aspect, some of the above-described provisions and statements in the legislative history point in different directions. This is particularly true of, on the one hand the title V applicability provisions, which apply by their terms inclusively and, on the other hand, the requirement that sources include applicable CAA requirements in their permits, and the statements in the legislative history indicating that Congress intended title V to cover sources subject to other CAA requirements.

Because Congress cannot be said to have expressed an intent as to the manner and scope of title V applicability to GHG sources, then, under Chevron Step 2, EPA may apply a reasonable interpretation of the applicability provision to determine the scope of coverage of GHG sources that is consistent with the statutory requirements. The Tailoring Rule qualifies as such an interpretation. The Tailoring Rule in effect reads the applicability provisions not to apply title V to GHG sources at or above the 100 tpy level, but instead to apply title V to as many of the GHG sources at or above that level as possible and as quickly as possible, starting with the largest sources first, that is consistent with both the permitting authorities' ability to administer the program and with a sensible imposition of costs to sources. This tailoring approach is consistent with the inclusive direction of the applicability provision, the flexibility in title V's scope that is inherent in the provisions authorizing general permits and requiring permit fees, the detailed process requirements, and the legislative history that focuses on Congress's concern about costs to sources and administrability. With the tailoring approach, over time, more sources may be included in title V, consistent with those provisions and legislative history. This reconciles the inclusiveness of the applicability provisions with Congress's expectations of a more limited scope for the title V program. However, as part of the tailoring approach, we recognize that we may at some point determine that it is appropriate to exclude certain sources, such as the smallest of the GHG sources. In addition, we intend to address the issue of sources with “empty permits” in a later rulemaking, as discussed previously.

The specific phase-in schedule will depend on the following: We will gather information about the permitting authorities' ability to process permits, and we will develop streamlining techniques. Based on that information, we will address expanding the title V program in a step-by-step fashion to include more sources over time. Each step will be based on our assessment of the permitting authorities' and sources' ability to comply with their respective obligations under the title V program.

We recognize that the availability of permit fees to support title V permit actions creates a potentially important source of resources, and that this has implications for the permitting authorities' ability to implement the title V program for sources of GHGs. At least in theory, permitting authorities could assess and collect sufficient fees to support hiring and training sufficient personnel so that they could expand their programs to match the expansion in the number of sources covered by the program.

Even so, title V fees cannot be considered a panacea that will resolve all resource problems that permitting authorities will have, for several reasons. Permitting authorities will likely be constrained as to the rate in which they can increase fees in light of the costs to sources. As indicated elsewhere, at least at the outset of the program before streamlining techniques have been developed, a literal application of the title V applicability provisions to GHG sources would, on average, cost each industrial source $46,400 and each commercial or residential source $23,200 to complete the permit application and take other associated actions; and it would cost each permitting authority, on average, $19,688 to process the industrial source permit and $9,844 to process the commercial or residential source permit. Particularly in light of the high costs to sources of applying for a permit, it is not likely that permitting authorities would be able to pass on to the sources in the form of fees, the entirety of the permitting authorities' own high costs for processing those permits, at least not right away. Even to the extent it would be possible to raise permit fees, permitting authorities would have to undergo a process to assess, impose, and collect those fees, and then hire and train personnel. The survey from the state and local agencies described previously forecast a 2-year period for hiring and training, without counting time for the fee process. For these reasons, we do not believe that the authorization for fees will allow the permitting authorities either to accelerate Steps 1 or 2 of the tailoring schedule or to permit a larger number of sources at those steps. Step 1 will take effect on January 2, 2011, Step 2 will take effect on July 1, 2011, and the process for determining and collecting fees, and then hiring and training personnel will take at least several years after July 1, 2011.

Moreover, we do not believe that the authorization for fees means that permitting authorities can reasonably be expected to permit title V sources at levels below 50,000 tpy CO₂ e before 2016. The next level below 50,000 tpy CO₂ e for which we have data is 25,000 tpy CO₂ e, and the costs to permitting authorities to run their programs at that level ($126 million) is more than double their current costs ($62 million). We do not consider it reasonable to expect permitting authorities to more than double their program within the first 6 years of title V applicability to GHG sources. That it is not reasonable to expect that is made even clearer when the permitting authorities' burdens in implementing their PSD programs are considered. The ability of permitting authorities to impose fees may have more important implications for subsequent steps, and as we address those subsequent steps in future rulemakings, we will consider the fees. EPA's approach to fees in this rulemaking is discussed elsewhere.

c. Other Possible Approaches to Reconciling Literal Reading of Title V Applicability Provisions and Congressional Intent

Having described how the Chevron framework, accounting for the “absurd results” doctrine, applies to title V requirements in this case and why it supports this Tailoring Rule—under which we expect to apply title V to more sources, in a step-by-step fashion, over time—we turn to the last part of our discussion of this doctrine. Here, we address another possible approach suggested by comments, which is that EPA should apply the title V program only to sources that are subject to applicable requirements, so that sources should not be required to hold “empty permits” (e.g., permits issued to a source that is not subject to any applicable requirement for any pollutant). To the extent that commenters argue that the statute requires EPA to adopt a “no-empty-permits” theory, we disagree. We believe that although various provisions of title V indicate that one of title V's purposes is to gather a source's applicable requirements into a single permitting mechanism, see CAA sections 503(b)(1), 504(a), we do not read those provisions as expressly limiting, as a matter of Chevron Step 1, title V to sources with applicable requirements. The applicability provisions, by their terms, include sources based on amount of emissions, and do not include any explicit limits to applicability based on whether the sources has applicable requirements. As described previously, we believe that Congress, although clearly expressing an intent that title V apply to GHG sources generally, did not express a clear intent as to how title V applies to GHG sources. The tension between these two sets of provisions, which we identified in the proposal and commenters further discussed, provides further support for that conclusion. Accordingly, we have discretion under Chevron Step 2 to determine a reasonable approach, consistent with the statutory requirements, concerning the application of title V to GHG sources with empty permits.

We note that to date, we have issued permits to sources without applicable requirements, albeit on rare occasions. We have little reason to believe that the “empty-permits” issue will arise in Steps 1 and 2 of our tailoring approach because we believe there will be no “empty permits” in Step 1 or Step 2 or, if there are, that they will be very few in number. As stated elsewhere, we believe that the tailoring approach we adopt in this rulemaking for Steps 1 and 2 is a reasonable approach that is consistent with statutory requirements.

We need to gather more information concerning the potential number and utility of “empty permits” for GHG sources, in light of the fact that the need for requirements in title V permits will vary based on the requirements of each SIP, and the fact that some SIPs contain broadly applicable requirements. As stated elsewhere, we intend to consider the issue of the applicability of title V to GHG sources with “empty permits” in Step 3 of our tailoring approach. When we do so, we will further assess the potential for the approach of excluding empty permits from title V to relieve burden consistent with statutory requirements.

7. Additional Rulemaking for the PSD and Title V Programs

The previous sections 5 and 6 discussed our application of the Chevron framework, accounting for the “absurd results” doctrine, to the PSD and title V applicability requirements, respectively. As another point in this regard, which is relevant for both PSD and title V purposes, we also commit to subsequent rulemakings in which we may further address the “absurd results” doctrine.

Specifically, we will propose or solicit comment on establishing a further phase-in, that is, a Step 3, that would apply PSD and title V to additional sources, effective July 1, 2013, and on which we commit to take final action, as supported by the record, by no later than July 1, 2012. We further commit to completing another round of rulemaking addressing smaller sources by April 30, 2016. Our action in that rulemaking would take into account the severity of the remaining problems associated with permitting authority burden and source costs.

While committing to future action, we do not decide in this rule when the phase-in process will ultimately end, or at what threshold level, because all that depends on uncertain variables such as our progress in developing streamlining approaches and on permitting authorities' progress in developing permitting expertise and acquiring more resources. We may continue the phase-in process with further rulemaking(s) after 2016. Alternatively, we may make a final determination through future rulemaking that, under the “absurd results” doctrine, PSD and/or title V do not apply to GHG sources that, while small and relatively inconsequential in terms of GHG contribution, are above the statutory tonnage thresholds for these programs, and thereby end the phase-in process. In addition, we may consider whether to limit title V applicability to GHG sources in order to minimize the number of GHG sources with “empty” permits.

8. Rationale for the Phase-In Schedule for Applying PSD and Title V to GHG Sources

Having discussed in sections V.B.5, V.B.6, and V.B.7 the reasons for tailoring the PSD and title V programs, we now describe our rationale for selecting the specific phase-in schedule in this rule for applying PSD and title V to GHG-emitting sources. To reiterate for convenience, under Step 1 of this schedule, which begins on January 2, 2011, (1) PSD applies to the GHG emissions of “anyway” PSD sources, that is, sources that are subject to PSD anyway due to their emissions of conventional pollutants and that undertake a modification that results in an increase of at least 75,000 tpy CO₂ e; and (2) title V applies to “anyway” title V sources, that is, sources that are subject to title V anyway due to their emissions of conventional pollutants. Under Step 2, which begins on July 1, 2011, (1) sources will be subject to PSD on account of their GHG emissions if they newly construct and emit at least 100,000 tpy CO₂ e, or if they are existing sources that emit at least 100,000 tpy CO₂ e of GHGs and make a modification that results in the emission of at least 75,000 tpy CO₂ e; and (2) existing and new sources will be subject to title V on account of their GHG emissions if they emit 100,000 tpy CO₂ e in GHG emissions. In addition, EPA intends to begin another round of rulemaking—Step 3—in 2011 and commits to complete it by July 1, 2012. In that rulemaking, we will propose or solicit comment on a further phase-in of GHG sources for PSD and title V applicability, and we may propose or solicit comment on another application of the “absurd results” doctrine that excludes categories of sources from PSD or title V. However, under this rule, in no event will EPA apply PSD or title V to sources below the 50,000 tpy CO₂ e levels in Step 3, or any other step we might promulgate prior to April 2016. In addition, EPA commits to conduct a study, to be concluded by April 30, 2015, evaluating the status of PSD and title V applicability to GHG sources, and, based on the study, complete a rulemaking by April 30, 2016 that addresses another round of a phase-in.

a. Rationale for Step 1

In Step 1 of our tailoring approach, which begins on January 2, 2011, PSD and title V requirements will apply to only those sources that are subject to PSD or title V requirements anyway due to their conventional pollutants (“anyway” sources) and that, in the case of PSD, make modifications that result in an increase in GHG emissions of at least 75,000 tpy CO₂ e. No sources would become major for PSD or title V under this step based on their GHG emissions alone. This section describes our proposal, comments on the proposal and our response to those comments, and our rationale for Step 1.

(1) Proposal

In our proposal, we proposed (1) the application of PSD and title V requirements to sources that emit at least 25,000 tpy CO₂ e, (2) a PSD significance level of between 10,000 and 25,000 tpy CO₂ e, and (3) a commitment to undertake a study to be followed by further rulemaking after 6 years. In addition, we solicited comment on the alternative of limiting PSD and title V applicability to “anyway” sources for at least the first 6 years. Under this approach, PSD and title V applicability would be determined based on non-GHG pollutants, and without regard to GHGs, but those sources subject to PSD would also be subject to BACT requirements for GHGs if their GHG emissions exceeded the significance level established in the final rule, and those sources subject to title V would be required to include any applicable requirements for GHGs in their permits.

(2) Comments

Many commenters supported this “anyway”-source approach, and offered a variety of reasons: According to the commenters, (1) This approach is a better reading of Congress's intent in the Act and is consistent with Alabama Power v. Costle, 636 F.2d 323 (DC Cir. 1980); (2) this approach would reduce the permitting workload on sources currently considered minor and focus PSD and title V requirements on large sources of non-GHG pollutants, as intended by Congress; (3) it is appropriate to base PSD and title V applicability on non-GHG emissions until data on GHG emissions are available from the mandatory GHG reporting rule; (4) in the initial phase, this approach would be more straightforward to administer, would provide a more predictable permitting workload, and would prevent a flood of newly regulated sources from overburdening state agencies; (5) this approach would provide permitting agencies time to develop experience handling GHG sources and requirements under the PSD and title V programs; (6) this approach would provide EPA and the permitting agencies the time needed to develop streamlining techniques; (7) this approach is consistent with the “absurd results” and “administrative necessity” doctrines because the scope of the permitting programs would remain consistent with both congressional intent and current administrative practice, but EPA and state agencies would still be allowed to begin regulating GHG emissions from existing PSD and title V sources; and (8) sources already required to obtain PSD permits are best equipped to work through BACT issues with permitting authorities.

Commenters added that if BACT is applied for GHGs due to permit actions involving non-GHG pollutants, EPA would need to set a significance threshold for the application of BACT, without which BACT could apply to very small (e.g., 1 ton) GHG increases associated with projects that otherwise triggered PSD for increases of non-GHG.

(3) Determination as to Step 1, PSD and Title V Applicability and PSD Significance Level

After considering the administrative burdens from increased permitting actions and the need for permitting authorities to have sufficient time to develop necessary expertise and staffing resources to address that burden, we have decided in this final action to establish the “anyway” source approach as Step 1. Beginning on January 2, 2011, sources subject to PSD requirements for their conventional pollutants anyway will be required to apply BACT to their GHG emissions if they construct or modify and in so doing, emit at least 75,000 tpy CO₂ e in GHGs. Similarly, sources subject to title V requirements anyway due to their conventional pollutants will be required to meet certain requirements for their GHGs, as described elsewhere. These requirements at Step 1 for PSD and title V will not expire. On July 1, 2011, a further phase-in of PSD and title V applicability—Step 2—will kick in.

At Step 1, by definition, all of the covered sources are already subject to PSD and title V permitting requirements, and will simply be adding a GHG component to what would be an otherwise occurring permitting action for conventional pollutants. These sources include fossil fuel-fired power plants, petroleum refineries, cement plants, iron and steel plants, pulp and paper plants, petroleum refineries, large landfills, and other large industrial sources. These sources will need to perform some additional analysis that is unique to GHG emission units, particularly related to the BACT review and selection process, but they will likely be able to utilize information developed as part of other permitting requirements for conventional pollutants, such as equipment fuel usage and operational parameters. Also, because these facilities are familiar with the case-by-case permitting processes, including all the steps from the application to the final review process, they will not confront a high PSD or title V learning curve.

The “anyway” source approach has particular appeal during the first step of the phase-in approach because it begins to apply key PSD and title V program requirements as soon as January 2, 2011 to large sources of emissions, but because it applies only to sources that are already subject to PSD for other pollutants, it can be implemented efficiently and with an administrative burden that is manageable in the next 8 months. We expect that under this approach, the sources and permitting authorities will still face substantial additional work to address the GHG emissions. In addition to the activities discussed elsewhere, there will be significant and complex policy questions about how BACT will be implemented for GHGs that must be resolved. These issues will include how to determine BACT for GHGs, how to do netting, and other similar issues. Even with EPA guidance, many case-specific policy issues will arise and will have to be resolved by the permitting authority in the context of a specific permit application. Nevertheless, with the “anyway” source approach, this work will be manageable because the associated permitting burden will be limited to adding a GHG component to each existing permit action for which it will be required, and will avoid the significantly greater burdens associated with large numbers of new permit actions that would be required for sources and modifications that would be subject to PSD for the first time. Instead, this “anyway” source approach allows permitting authorities sufficient time to develop necessary expertise and staffing resources to address GHG BACT.

We agree with commenters that the establishment of a significance level—which, in effect, is a BACT threshold—is appropriate, and we have decided to establish this level at 75,000 tpy CO₂ e because, for reasons discussed later, that is the level that will apply during Step 2. At this level, the administrative burdens, described later, will be manageable. Importantly, we believe a consistent significance level between Steps 1 and 2, as opposed to a lower significance level in Step 1, will provide for a smoother transition and avoid the problems that would arise if PSD applied to modifications during Step 1 that PSD would not apply to in Step 2. Otherwise, we would create a perverse incentive for companies to delay such projects until Step 2 to avoid BACT.

We estimate that Step 1 will result in a 23 percent increase in permitting authority work hours and a $3 million increase—which amounts to a 25 percent increase from the current program cost of $12 million—in their annual costs for running PSD programs. This is primarily due to the GHG BACT review requirements. For title V programs, we estimate a 2 percent increase in permitting authority work hours and a $1 million increase in the title V annual program costs for permitting authorities under Step 1 as compared to the current program cost of $62 million. These work hours and costs will be needed primarily to review GHG emissions information, add any GHG-related requirements to title V revisions and renewal actions that would otherwise be occurring, respond to comments and petitions from the public, as well as develop fee requirements and make fee determinations associated with issuing new or revised title V permits that add GHG-related information. For both the PSD and title V programs on a combined basis, the additional costs for Step 1 will be $4 million, which amounts to a 5 percent increase in the current combined program cost of $74 million.

In addition to these workload and monetary costs, permitting authorities will confront additional burdens before and during Step 1, which we have not attempted to quantify. One of the most significant of these is training staff in the PSD-related areas of GHG emissions calculations and BACT evaluations. In addition, permitting staff will need to build staff expertise and capacity for addressing GHG requirements in preparation for Step 2, which will begin only 6 months after Step 1; and in communicating and providing outreach to sources addressing GHG emissions for the first time. Based on comments we received on the proposal from permitting authorities, we believe these additional training and outreach requirements—for both the PSD and title V programs—will add significantly to the permitting authorities' burden during the initial 6-month period under Step 1.

We believe that these administrative burdens are substantial but manageable. Following this action, permitting authorities will have only 8 months to prepare for Step 1, when they will need to increase their resources by 5 percent for both the PSD and title V programs combined, and be able to implement BACT requirements for GHG sources. During Step 1, they will need to prepare for Step 2, when, as discussed later, they will need to process over 900 additional PSD permits each year and begin to process over 1,100 additional title V permit actions.

We have decided to limit Step 1 to the “anyway” source approach, and not apply PSD or title V to sources based on their GHG emissions, for several reasons. First, we believe that the administrative burdens described previously are the most that the permitting authorities can reasonably be expected to manage before and during Step 1. Tighter PSD and title V applicability requirements would mean greater administrative burdens.

Second, we believe that the costs of GHG permitting to the sources, as described previously, are substantial and as a result, necessitate that we wait for the permitting authorities to develop the PSD and title V programs for GHG sources during the first 6 months of 2011 before subjecting sources to PSD and title V requirements on account of their GHG emissions. By July 1, 2011, when Step 2 takes effect, the PSD and title V programs will be better developed. For example, the permitting authorities will have more experience making BACT determinations. In addition, by that time, sources will have had more time to prepare for the permitting processes. In addition, as suggested by one commenter, the additional time will allow sources and permitting authorities to address the current uncertainty surrounding how to measure high-GWP gases.

Third, we estimate that “anyway” sources account for approximately 65 percent of total national stationary source GHG emissions. As a result, limiting Step 1 to these sources will still capture a large portion of the GHG inventory.

A large number of commenters urged us to leave this “anyway” source approach in place until such time as we complete an assessment and conduct further rulemaking, which we proposed would be 6 years from now. We are not taking this action; rather, for the reasons discussed next, we believe it is reasonable to use GHG thresholds to begin to phase in PSD and title V applicability to additional sources in Step 2.

b. Rationale for Step 2

(1) Proposal

We proposed to establish the applicability level for PSD and title V to GHG sources at 25,000 tpy CO₂ e, and we proposed a PSD significance level in the range of 10,000 to 25,000 tpy CO₂ e. Our burden estimates at proposal led us to conclude that at those threshold levels, for the PSD program, “approximately 400 additional new or modified facilities would be subject to PSD review in a given year. These include approximately 130 new facilities and approximately 270 modifications * * *.” 74 FR 55331, col. 1. We estimated that processing these numbers of additional permits, along with doing the additional work associated with GHG emissions from sources subject to PSD anyway due to their conventional emissions, would increase permitting authority burdens by “approximately 112,000 staff hours at an additional cost of approximately $8 million. This workload amount represents an increase of about 1.3 times, or 32 percent, in the current burden for permitting authorities on a nationwide basis.” Id. col. 3. We concluded that “this additional burden is manageable,” but that “any threshold lower than 25,000 tpy CO₂ e, would create undue administrative burdens.” Id.

For the title V program, we estimated that at a 25,000-tpy CO₂ e permitting threshold, “about 13,600 existing facilities” would become subject to title V, and that to manage the additional workload associated with permitting those sources and with the other permit revisions and modifications that would result from the 25,000 tpy CO₂ e threshold, permitting authorities would require an additional 492 FTEs, which would be an estimated 50 percent increase over current title V staffing levels. 74 FR 55335, cols. 1-2.

(2) Comments

We received a significant number of comments from both permitting authorities and industry representatives that our proposed GHG threshold of 25,000 tpy CO₂ e for major source applicability was too low and would result in an unmanageable amount of permitting actions in the near term. Many offered evidence that we severely underestimated both the number of permitting actions and the per-permit administrative burden, for both PSD and title V programs.

Commenters also asserted that the proposed 25,000 tpy threshold is too low because it will subject small sources (including many small businesses) to PSD and title V, which is not in keeping with Congress's intent to limit PSD and title V to large sources when Congress set the 100/250 tpy thresholds for the permitting programs. EPA, in collaboration with the SBA, conducted an outreach meeting designed to exchange information with small entities that may be interested in these regulations. The EPA took this small business outreach effort into account when finalizing this rule. Many commenters from this outreach effort said that there were many more small businesses that would become subject to PSD and title V due to the proposed permitting thresholds than EPA estimated at proposal.

Many commenters recommended specific major source thresholds for PSD and title V, including levels of 25,000 (as proposed), 40,000, 50,000, 100,000, 150,000, 250,000, and 1,000,000 tpy CO₂ e. A majority of the commenters—including both industry and state agency commenters—recommended major source thresholds of 100,000 tpy CO₂ e. However, several state agency commenters recommended thresholds of 50,000 tpy CO₂ e. Other commenters recommended sector-specific thresholds. For example, solid waste industry commenters suggested thresholds of 820,000 tpy CO₂ e for PSD [which they calculate to be equivalent to the existing PSD threshold for “municipal solid waste landfill emissions,” i.e., 250 tpy nonmethane organic compounds (NMOC)] and 320,000 tpy CO₂ e for title V (calculated to be equivalent to the existing major source applicability threshold of 100 tpy NMOC). Other commenters urged EPA to set the GHG thresholds at levels that correspond to emissions of conventional pollutants at the 100/250 tpy level.

Many of the commenters that recommended increasing the thresholds cited EPA's estimates that a particular threshold would significantly reduce the number of sources subject to the rule while causing only a slight reduction in the percentage of GHGs captured. Several of these commenters noted that Table VIII-2 in the proposal preamble indicates that shifting the major source threshold for PSD from 25,000 to 100,000 tpy CO₂ e would reduce the number of major sources from 13,661 to 4,850 while reducing the coverage of U.S. stationary source GHG emissions by only about 4 percent. Other commenters referred to the regulatory impact analysis (RIA) for the mandatory GHG reporting rule to conclude that raising the threshold from 25,000 to 100,000 tpy CO₂ e would exclude thousands of entities that, on a combined basis, emit only one percent of the nation's GHG emissions. See the RTC document for this final rulemaking for more detailed description of comments received on our proposed burden assessment.

Many commenters also recommended specific PSD GHG significance thresholds, including levels of 10,000 (as proposed), 15,000 (within the proposed range), 25,000 (also as proposed), 40,000, 50,000, and 100,000, and 150,000 tpy CO₂ e, as well as suggesting sector-specific thresholds. These recommendations were based on the view that we had underestimated the number of modifications and that the burden of permitting at the proposed levels would therefore be much worse than we projected. A number of the commenters argued that the significance threshold should be no less than the major source threshold, at whatever level that is set. The largest number of commenters recommended a PSD significance threshold of 100,000 tpy CO₂ e, although significant numbers also support 25,000 and 50,000 tpy CO₂ e.

(3) Rationale for Step 2

Based on these comments, we reassessed our original burden estimates from our proposal. This reassessment is discussed at the beginning of this section. We decided that, once this adjustment is taken into account, the burdens at the proposed 25,000 threshold and the proposed 10,000-25,000 significance levels would be unmanageable. We therefore evaluated higher thresholds ranging from a 25,000 tpy CO₂ e major source applicability level for PSD and title V to a 50,000, 75,000, or 100,000 tpy CO₂ e level, with associated PSD GHG significance levels of equal or lesser magnitude; and we selected the 100,000/75,000 tpy CO₂ e level. Central to our decision to promulgate higher thresholds than what we proposed is our recognition, based on comments and further analysis, that applying PSD to GHG sources at the statutory or any other threshold level or significance level that we have considered would result in (1) a greater number of sources, and significantly greater number of modifications than we first estimated becoming subject to those programs; and (2) a greater per-permit cost than we first estimated to the permitting authority of processing those permit actions. We discussed our revised estimates and reasoning at the beginning of this section.

We now estimate that the 25,000/25,000 tpy level would result in 250 additional PSD permit actions for new construction (either for GHG-only sources or additions to otherwise occurring permits) and an additional 9,200 PSD permits for modifications each year (compared to our estimate at proposal of 130 for new construction and 270 for modifications). This level of permitting would require an additional 2,815,927 work hours, or 1,400 FTEs (compared to our estimate at proposal of 112,000 additional work hours, or 57 FTEs); and would cost an additional $217 million each year (compared to our estimate at proposal of an additional $8 million). See 74 FR 55331 (proposal). This $217 million amount represents approximately a 1,800 percent increase over current permitting authority annual cost of $12 million for the major NSR programs.

For title V, under our final burden analysis at a 25,000 tpy CO₂ e threshold, we estimate a $64 million annual increase in program costs to permitting authorities to add GHG emission sources, which reflects a greater than 100 percent increase over current program costs of $62 million. We estimate that this increased burden would result in the need for almost 700 new FTEs nationwide at permitting authorities (compared to our estimate at proposal of 492 additional FTEs, or about a 50 percent increase in existing program size). This increase in burden is due to an estimated annual increase of 2,500 new title V permits, over 9,500 permit revisions, and over 2,600 permit renewal actions due to GHG emission sources. These additional title V actions compare to current annual program actions of approximately 50 new title V permits, 1,394 significant revisions, and 3,267 permit renewals.

Based on this information, we have decided not to finalize our proposal to apply a 25,000 tpy CO₂ e applicability threshold to GHG sources at the time that PSD and title V take effect. At that level, too many sources—many more than we thought at proposal—would be subject to high permitting costs. In addition, permitting at that level and at that time would not be administratively feasible. The resulting increase in the number of PSD and title V permitting actions and workload would create insurmountable resource demands for permitting agencies in the near term, which would jeopardize the functioning of these permitting programs. We are mindful that not only would the permitting programs have to bear the costs that our estimates are able to monetize, but they would also incur burdens associated with hiring and training staff to make and implement GHG BACT determinations, GHG emissions evaluations, and other evaluations required under the PSD program for a wide variety of formerly unpermitted sources, including significant numbers and types of small manufacturing and commercial or residential establishments. They would also incur burdens associated with reviewing applications, citizen comment and petitions, and the need to communicate and provide outreach to new categories of sources, including, again, significant numbers and types of small manufacturing and commercial or residential sources. Thus, the increased administrative burdens at the 25,000/25,000 tpy CO₂ e levels are so great that we have concluded that they would not be consistent with the goals of avoiding absurd results that contravene congressional intent, including avoiding a permitting burden that would overwhelm the capacity of permitting authorities to effectively implement their programs.

Based on our revised burden analysis, in this final action, we have decided to establish a multi-step, phase-in approach that contains a significantly higher initial threshold level. We have determined that a 100,000 tpy CO₂ e major source threshold level for PSD and title V purposes, and a 75,000 tpy CO₂ e significance level, produce a level of permitting activity that would certainly be an increase over current workload, but that would be administratively feasible by July 1, 2011. As a result, we have decided to finalize these thresholds as Step 2.

In reaching this conclusion, we needed to consider both the sources' abilities to manage the permitting process and the permitting authorities' capacity to address newly-major sources as expeditiously as possible. As to the former, sources subject to Step 2 will, for the most part, continue to include the “anyway” sources subject to Step 1. In addition, we estimate that Step 2 will include about 500 additional sources that are not already subject to permitting. Most of them will become subject to PSD and title V because of fuel burning. In order to meet the 100,000/75,000 threshold, they will have to burn a significant quantity of fuel, and that means they will be a significant size. In general, these sources include municipal or commercial landfills that are large, but not large enough to be covered by the NSPS, pulp and paper facilities, electronics manufacturing plants, chemical production plants, and beverage producers. Although these sources have not been subject to PSD permitting before, some of them have already been subject to minor source permitting, and so will have some familiarity with the permitting process. In addition, in general, these sources are in source categories that have larger sources that are already subject to PSD and title V. As a result, they are in industries that have experience in the permitting process. Because of their relatively large size and access to knowledge about the permitting processes, we believe these sources will be able to manage the permitting requirements.

As to the permitting authorities' capacity to handle the Step 2 workload, we note first that our Step 1 approach does not cover newly-major sources. As a result, the Step 2 threshold and timing has to be established in a way that takes into account permitting authority challenges in addressing many sources and categories that would be subject to major source permitting for the first time.

We considered the various PSD and title V threshold applicability and significance level options in our final burden analysis, summarized in Table VI-1, including levels at 50,000 CO₂ e and 100,000 CO₂ e. As Table VI-1 indicates, we estimate that a 100,000 tpy CO₂ e major source applicability threshold would result in approximately 550 sources becoming newly classified as major sources for PSD based on their GHG emissions, while a 50,000 tpy CO₂ e threshold would result in 3,500 newly classified major sources.

We then considered the impact on both PSD and title V programs of different PSD significance level options for GHGs. The choice of a PSD significance level has a direct impact on title V burdens because PSD permit requirements resulting from modification activities will result in required title V permit revisions. We developed PSD and title V burden estimates based on significance levels of 50,000 tpy, 75,000 tpy and 100,000 tpy CO₂ e, combined with a major source applicability level of 100,000 tpy CO₂ e.

At a 50,000 tpy CO₂ e significance level, we estimated an annual increase of approximately 1,800 PSD permitting actions and almost 2,000 additional title V permitting actions, as compared to Step 1. At a 75,000 tpy CO₂ e significance level, we estimated an annual increase of approximately 900 PSD permitting actions and just over 1,000 additional title V permitting actions as compared to Step 1. At a 100,000 tpy CO₂ e significance level we estimated an annual increase of approximately 25 PSD permitting actions and 210 additional title V permitting actions as compared to Step 1. For title V, under these different scenarios, the major source applicability level of 100,000 tpy CO₂ e results in approximately 200 new permits annually, but, as noted, the choice of significance levels affects the number of required permit revisions.

Based on this information, we have decided to set our final Step 2 thresholds at 100,000 tpy CO₂ e for major source applicability under PSD and title V and at a 75,000 tpy CO₂ e significance level for PSD. Overall, we estimate that the almost 900 additional PSD permitting actions (virtually all of which would be modifications) per year at these levels will result in an approximately $21 million increase (from Step 1) in states' annual costs for running PSD programs. In addition, we estimate that the 1,000 additional title V permit actions will cause the total title V burden for permitting authorities to increase by $6 million annually from Step 1. This total increase in permit program burdens of $27 million represents a 34 percent increase over the $78 million in total cost of PSD and title V programs at Step 1. We consider this a substantial increase particularly because Step 2's start date of July 1, 2011, is only 6 months after Step 1's start date of January 2, 2011. What's more, Step 1 will entail a substantial increase in permitting authority obligations, so that adding the costs of Step 1 and Step 2 together—$31 million—means that permitting authorities will be required to increase their permitting resources by approximately 42 percent between now and Step 2. In addition to the administrative burdens we have been able to monetize, we must be mindful that permitting authorities will incur other burdens, including the significant support and outreach activities by permitting staff for the many newly permitted sources. We believe that any lower thresholds in this timeframe, whether in the PSD and title V applicability levels or in the significance level, would give rise to administrative burdens that are not manageable by the permitting authorities.

Although the burdens at the 100,000 tpy CO₂ e/75,000 tpy CO₂ e levels are steep, we consider them manageable. Step 2 permitting for GHGs will mostly involve source categories in which some sources have traditionally been subject to permitting, which should render applying even the new GHG requirements more manageable. These source categories include fossil fuel-fired power plants, petroleum refineries, cement plants, iron and steel plants, and petroleum refineries, in addition to other large industrial type source categories. A full description of the type of sources that we expect will have GHG emissions that exceed the 100,000 tpy CO₂ e threshold is provided in the “Technical Support Document for Greenhouse Gas Emissions Thresholds Evaluation” located in the public docket for this rulemaking. In addition, because Step 2 does not begin until July 1, 2011, permitting authorities have about 14 months to prepare for it.

In addition, we believe that the sources that will become subject to PSD and title V requirements at the 100,000/75,000 tpy CO₂ e levels will be able to accommodate the additional costs of permitting. For the most part, these sources will be of a comparable size and activity level as those sources that are already subject to those requirements.

Because the administrative burdens at the 100,000/75,000 tpy CO₂ e level are as heavy as the permitting authorities can reasonably be expected to carry, adopting these threshold levels is consistent with our legal basis under the “absurd results” doctrine. Under this basis, we are reconciling the statutory levels with congressional intent by requiring that the PSD and title V requirements be applied to GHG sources at levels as close as possible to the statutory thresholds, and as quickly as possible, in light of costs to sources and administrative burdens.

Because the administrative burdens at the 100,000/75,000 tpy CO₂ e level are manageable, we do not believe that higher threshold levels are justifiable for Step 2. Specifically, at the 100,000/100,000 level—which would entail a 100,000 tpy CO₂ e significance level, rather than a 75,000 tpy CO₂ e level—permitting sources would need to handle only 20 additional modifications beyond current levels, and thus would not incur substantial additional costs. By the same token, we disagree with commenters who suggested that we needed to set permanent GHG permitting thresholds for major sources at a rate equivalent to the amount of GHGs that would be emitted by conventional pollutants at the 100 and 250 tpy level in order to meet the legal bases of the “absurd results” and “administrative necessity” doctrines. These levels would likely be well above 300,000 tpy CO₂ e, depending on fuel types and assumptions regarding the relative emissions of GHGs compared to the conventional pollutants. Our data show that none of the levels above 100,000/75,000 tpy CO₂ e would result in significant increases in administrative burdens. As a result, establishing these levels would not apply PSD or title V requirements to GHG sources as quickly as possible, and thus would not be consistent with our approach in the Tailoring Rule.

We estimate that facilities meeting the Step 2 major source applicability thresholds account for approximately 67 percent of total national stationary source GHG emissions. Many commenters felt that this should be an important basis for our selection of a threshold, stating that there is no significant loss in GHG emissions coverage of source categories at the 100,000 tpy CO₂ e threshold, and in some cases arguing that as a result, we should set the level even higher. We agree that it is important that the coverage in Step 2 represents 86 percent of the coverage at full implementation of the statutory 100/250 thresholds.

c. Rationale for EPA's Plan Beyond Step 2

EPA commits that after Step 2, EPA will begin another rulemaking in 2011 and complete it by July 1, 2012, and in that rulemaking take comment on a further phase-in of GHG sources for PSD and title V applicability (Step 3). However, under this rule, in no event will EPA apply PSD or title V to sources below the 50,000 tpy CO₂ e levels prior to 2016. In addition, EPA commits to conduct a study, to be concluded by April 30, 2015, evaluating the status of PSD and title V applicability to GHG sources, and, based on the study, complete a rulemaking by April 30, 2016, that addresses another round of a phase-in.

(1) Proposal

In our proposal, we noted that following implementation of the first phase of PSD and title V applicability to GHG sources, generally at the 25,000 tpy CO₂ e threshold, additional action would be required over time to assure full compliance with the statute. We did not establish more steps in the schedule, but we did commit to conduct a study, to be completed by 5 years after promulgation, evaluating the status of PSD and title V applicability to GHG sources, and, based on the study, complete a rulemaking by 6 years after promulgation that addressed an additional step of the phase-in.

(2) Comments

A number of commenters supported the proposal's overall approach to phase in the permitting of GHGs, mainly because this approach will allow permitting of the largest sources of GHGs immediately while collecting more information about smaller sources and more fully considering streamlining options for subsequent phases. Many of these commenters made clear that they do not support implementation of the statutory 100/250 tpy thresholds, even through a phase-in approach. On the other hand, one commenter asserted that EPA has failed to demonstrate that it needs 6 years to study and implement NSR and title V for sources emitting less than 25,000 tpy. The commenter contends that EPA has not analyzed, among other things, what combined effect the full implementation of its streamlining proposals in the 15 months before the due-date for title V permit applications would be to reduce the cost, complexity, and number of title V permit applications that would have to be submitted.

(3) Rationale for Further Steps

We agree with commenters who support a phased-in approach to the Tailoring Rule. Our final action reflects a multi-step process that we believe will facilitate a manageable expansion of PSD and title V applicability, as appropriate, to GHG-emitting sources. In our final action, we have established the initial two steps of a multi-step phase-in of lower threshold applicability with a commitment to take further regulatory activity to consider adopting lower thresholds. We believe this process will provide substantial opportunity for permitting authorities and sources to establish enough experience and information, and to provide significant real-world feedback to EPA, so as to better inform decisions on future phase-in steps.

With this overall phase-in approach in mind, in this final rule, EPA includes an enforceable commitment to undertake a notice-and-comment rulemaking that would begin with an SNPR that we expect to be issued in 2011 and that we commit will be finalized in 2012. The notice will propose or solicit comment on further reductions in the applicability levels. This rulemaking will take effect by July 1, 2013, and therefore, in effect, constitute Step 3. In this action, we are committing to a rulemaking for Step 3, but are not promulgating Step 3, because it is important to allow EPA and the permitting authorities to gain experience permitting sources under Steps 1 and 2, and to allow time to develop streamlining methods, before attempting to determine what would be the next phase-in levels for PSD and title V applicability. While committing to future action, we do not decide in this rule when the phase-in process will ultimately end, or at what threshold level, because all that depends on uncertain variables such as our progress in developing streamlining approaches and on permitting authorities' progress in developing permitting expertise and acquiring more resources. We may continue the phase-in process with further rulemaking(s) after 2016. Alternatively, we may make a final determination through future rulemaking that, under a Chevron analysis, accounting for the “absurd results” doctrine, PSD and/or title V do not apply to GHG sources that, while small and relatively inconsequential in terms of GHG contribution, are above the statutory tonnage thresholds for these programs, and thereby end the phase-in process.

In addition, in this action, we are determining that in no event—whether through Step 3 or a subsequent step—will we apply PSD or title V to sources at the 50,000/50,000 tpy CO₂ e level or lower prior to May 1, 2016. We have several reasons for making this determination at this time. Most importantly, our examination of the expected burdens to the permitting authorities of applying PSD and title V to GHG sources convinces us that extending the permitting programs to sources at or below the 50,000/50,000 tpy CO₂ e level within 6 years of promulgation would result in prohibitively heavy burdens. This threshold option would result in close to 2,000 additional annual PSD permitting actions per year over the current program and more than 1,000 over Step 2, including both new construction and modifications. For title V, we estimated an increase of over 1,000 new title V permits (all newly permitted sources because of GHG emissions) over 2,000 permit revisions per year over the current program, and about 980 new title V permits and 900 permit revisions more than the Step 2 amounts.

These increases, which could occur between 2013 and 2016 under our approach depending on the outcome of the Step 3 rulemaking, represent very substantial additions to the permitting program. In terms of cost, we estimate that these additional actions would result in a $73 million per year increase in joint PSD and title V program costs over the current programs—which is almost a doubling of costs—and $42 million annual cost increase over Step 2 for the current programs. We believe that it would take permitting authorities some time to adjust to this workload. This is particularly true because at the 50,000/50,000 tpy CO₂ e level, smaller sources—including ones not previously subject to permitting requirements—will become subject to PSD and title V. It will take some time for both the permitting authorities and the sources to absorb these new obligations.

Importantly, the next lower cut-off—below 50,000 tpy CO₂ e for the major threshold level—is the 25,000/25,000 tpy CO₂ e level. For the reasons discussed previously, this level is clearly not manageable within the first 6 years after this action. This applicability level would bring in over 7,000 sources that would be newly subject to title V permitting and result in close to 10,000 new PSD permitting actions. This would result in a 380 percent increase over current program costs for PSD and title V to run these programs. Based on comments we received from state and local permitting agencies on our proposed Tailoring Rule, these levels of permitting activities would far exceed the administrative capabilities of the permitting agencies for at least the near future. Thus, the 6-year exclusion is necessary to provide these agencies and their permittees certainty that this will not occur.

We recognize that at present, we do not have data that would allow us to compile administrative burden estimates for specific levels between the 50,000/50,000 and 25,000/25,000 tpy CO₂ e levels we assessed. However, it is clear that the burdens begin to rise sharply below the 50,000/50,000 tpy CO₂ e level. To reiterate, the combined PSD and title V administrative burdens at the 50,000/50,000 tpy CO₂ e level cost almost twice as much as the current programs, but the burdens at the 25,000/25,000 tpy CO₂ e level cost almost four times as much as the current programs. As a result, we conclude that dropping the level below 50,000/50,000 tpy CO₂ e too soon would quickly expose the permitting authorities to unacceptably high burdens.

As a further reason for concluding that we will not reduce thresholds beyond 50,000/50,000 tpy CO₂ e during the first 6 years, we recognize that the PSD permitting process in particular carries important ramifications for the permitting authorities and the affected sources. If we have underestimated the permitting burden or the ability of states to respond to their additional workload, then permitting backlogs will result, and PSD permit issuance will be delayed, and sources seeking a PSD permit will not be able to construct or modify. If this were to happen on a large enough scale, it could have potentially serious consequences for the national economy.

Moreover, we need to be mindful that the best information we currently have as to permitting authority burdens represents a national average, as described previously. Our information at the individual state and local level, where permitting occurs, is not as robust. Accordingly, we recognize that a particular state may encounter permitting costs that are higher than average, and this may result in permitting backlogs in that state, with the consequence that sources in that state will face long delays in constructing or modifying. Similarly, even if a particular state's costs are in line with the national average, that state may not be able to find the additional resources to cover those costs as readily as other states. For this reason, too, sources in that state could face long delays in constructing or modifying.

Beyond the administrative burdens to permitting authorities, we recognize that the costs of PSD and title V permitting to sources may be high, and we are not inclined to allow their imposition at this time on sources smaller than the 50,000/50,000 tpy CO₂ e threshold. At that level, the permitting programs will apply to a significant number of newly permitted sources, including a variety of small manufacturing, commercial and residential categories. The next level that we have analyzed is the 25,000/25,000 tpy CO₂ e threshold. At that level, more than 7,000 more sources would become subject to PSD each year—almost all due to modifications—and another 4,000 sources would become subject to title V each year. These sources would be even smaller than those that already will have become subject to PSD and title V due to their GHG emissions. We do not think it reasonable to subject more of those types of sources, and smaller ones, to permitting costs within the next 6 years.

Finally, we note that moving from a 50,000 tpy CO₂ e threshold to 25,000 tpy CO₂ e will increase the emissions coverage of GHG stationary sources from 70 percent to 75 percent nationwide, which we consider to be a relatively small amount.

We recognize that our progress in developing streamlining methods will be a key determinant to the ability of permitting authorities to administer, and sources to comply with, PSD and title V at GHG emission levels below 50,000/50,000 tpy CO₂ e. Although we commit to pursue streamlining, we cannot predict our progress. This uncertainty may be problematic for stakeholders, primarily permitting authorities and industry. That is, permitting authorities will face uncertainty in planning the scope of their programs over the next few years, and industry will face uncertainty as to what new construction projects and modifications will be subject to PSD for GHGs. By determining now that for the next 6 years we will not impose PSD requirements below a floor at the 50,000/50,000 tpy CO₂ e level, we add a measure of needed certainty.

We also recognize that selecting a level that is too high or keeping a level for too long means that some sources may construct or modify without implementing BACT level controls, and this could result in additional emissions of GHGs. We need to be vigilant and to protect against this outcome. Even so, all things considered, we believe that our determination not to apply the PSD or title V permitting requirements to sources below the 50,000/50,000 tpy CO₂ e level for the first 6 years also represents a reasonable balancing of protection of the environment with promotion of economic development. This type of balancing is consistent with our authority under the PSD provisions.

We also raised the issue of “hollow” or “empty” permits in discussing our rationale for why it may make sense to delay title V permitting under our proposal. We were concerned that many title V permits for GHG sources would contain no applicable requirements, and their issuance would therefore be of little value and would not be the best use of scarce resources. Several commenters agreed that implementing title V for GHGs will, at least initially, require “empty permits” to be issued to GHG sources because such sources will not be subject to “substantive” requirements, and that this would not be the best use of scarce resources.

We believe that the amount of resources that would be spent on, and the limited value that would result from, “empty permits” does warrant consideration under the Chevron analysis, taking account of the “absurd results” doctrine. Therefore, we intend to consider the role of “empty permits” when we undertake future rulemaking. However, we believe the issue of “empty permits” has limited or no relevance to the first two steps of the phase-in that we are promulgating in this rule. During Step 1, permitting for GHGs is only required if the source is otherwise subject to permitting for its emissions of non-GHGs. Those sources very likely will be subject to existing substantive applicable requirements for non-GHGs (e.g., NSPS, Maximum Achievable Control Technology (MACT), and SIP requirements, including PSD). Thus, there should be no, or at least no additional, “empty permits” during Step 1. For Step 2, it is possible that sources that become subject to title V requirements for GHG emissions may not be subject to other requirements, but our assessment suggests that this is very unlikely. We estimate that virtually all of the 550 newly-major sources in Step 2 will be subject to applicable requirements under the CAA because they are from categories that have been traditionally subject to regulations, such as smaller industrial sources from already regulated categories, large landfills, and oil/gas/coal production. Even the approximately 50 newly-subject commercial sources in Step 2, which we estimate to be comprised of very large hospitals, are likely to be covered by standards for medical waste incinerators. In addition, we expect these sources may well be subject to SIP requirements. Thus, we do not expect any, or at most very few “empty permits” during Step 2.

In later stages of implementation (e.g., prospective Step 3) or in the event that we permit smaller, non-traditional sources of GHGs that have never otherwise been subject to major source permitting, there would be a greater potential for “empty permits” to be issued under title V. Cognizant of this, we intend to further explore in the rulemaking for Step 3 “empty permit” theories under the “absurd results” rationale that may serve to permanently narrow the scope of title V to exclude sources that would potentially be required to obtain an “empty permit” due to GHG emissions.

In this action, EPA is also finalizing its proposal to commit to conduct an assessment of the threshold levels—to be completed in 2015, 5 years after this action—that will examine the permitting authorities' progress in implementing the PSD and title V programs for GHG sources as well as EPA's and the permitting authorities' progress in developing streamlining methods. We further commit to undertake another round of rulemaking—beginning after the assessment is done, and to be completed by April 30, 2016—to address smaller sources.

We disagree with the commenter who asserted that we do not need 6 years to study and implement PSD and title V for smaller sources. As we discussed in the proposal, and reiterate in this final action, we do not have sufficient information at this time to determine the applicability and effectiveness of the various permitting streamlining techniques. For reasons discussed in more detail in section V.E.1 regarding streamlining, we are not now able to determine how such techniques will be implemented or whether they will prove viable or effective. We agree with the commenter that these measures may reduce the scope, cost, and complexity of these programs, but there is considerable uncertainty as to the extent of this effect. We do commit in this action to fully investigate, propose, and evaluate permit streamlining techniques to determine where they may have applications, how they would be applied, and whether they can withstand legal challenge. Even for those techniques that may ultimately be deemed viable, there is a significant time period necessary for rulemaking and state adoption, all of which could take up to 3 years or more. We also note that we will be required to complete our study of the effectiveness of these techniques within 5 years, meaning that, in order to complete it in time, we will essentially need to begin the study as soon as relevant data are starting to become available. Finally, the sixth year, in which EPA must complete rulemaking, requires proposal and promulgation of a rule within 1 year, which is an ambitious schedule. Therefore we believe that 6 years is appropriate for this type of effort. We also have received a substantial number of comments from permitting authorities that agreed with our 5-year timeframe, or a greater timeframe, to get more prepared for permitting smaller sources.

d. Other Comments on “Absurd Results” Doctrine

We received other comments on our application of the “absurd results” doctrine, which we respond to in the RTC document. One comment was overarching, and so we respond to it here: Commenters have asserted that under the “absurd results” doctrine, EPA does not have authority to, or at least should not, promulgate the endangerment/cause or contribute findings (which we will sometimes refer to as the “findings”) or the LDVR because doing so would trigger the PSD and title V requirements, which in turn would give rise to “absurd results”. According to commenters, under the “absurd results” case law, EPA is obliged to avoid taking any action that would trigger absurd results and in this case that means foregoing the endangerment/cause or contribute findings and/or the LDVR, or at least deferring finalizing them until EPA has time to streamline PSD and title V requirements so as to avoid “absurd results”. Commenters made the related comment that if we promulgate the LDVR, and thereby trigger PSD, we cannot rely on the “absurd results” doctrine because it is our own actions—the promulgation of the LDVR—that will have given rise to the “absurd results,” and under those circumstances, the doctrine is not available.

The comments that EPA had no authority to promulgate, or should not have promulgated, the endangerment/cause or contribute findings or the LDVR at the times that EPA did are not relevant to this rule, the Tailoring Rule. EPA has already promulgated the findings and the LDVR, and the LDVR triggers PSD and title V applicability, as we have seen. These comments would have been relevant only to the proposed findings and LDVR, and we are not, in this rulemaking, revisiting or reopening the findings or the LDVR.

Commenters claim that if EPA promulgates the LDVR, the “absurd results” doctrine will no longer apply to the Tailoring Rule because it will have been EPA's own action—promulgation of the LDVR—that gives rise to the “absurd results”. We disagree for several reasons. For one thing, commenters have not cited case law, and our research has disclosed none, in which a court specifically addressed a similar situation and issued a holding along the lines of what commenters urge. Moreover, commenters' approach would be punitive because the absurd results would occur absent this rule going final. Such an outcome would be counter to the purpose of the doctrine. That is, it would mean that PSD and title V would apply to GHG sources by their terms—at the statutory levels, as of January 2, 2011—with all the adverse consequences described elsewhere.

In any event, and although we are not obligated to respond to these comments on the merits, they are incorrect on the merits, for the reasons that follow. This discussion should not be viewed as reopening the endangerment/cause or contribute findings or the LDVR because, as stated previously, we are not reconsidering or reopening those two actions in this rule.

In determining and implementing congressional intent, it is important that the statutory provisions at issue be considered together—(1) The obligation to make a determination on endangerment and contribution under CAA section 202(a); (2) if affirmative endangerment/cause or contribute findings are made, the obligation to promulgate standards applicable to the emission of any air pollutant from new motor vehicles or new motor vehicle engines under CAA section 202(a); and (3) the PSD and title V applicability provisions. The most appropriate reading, and certainly a reasonable reading, is that we are required to take the action we have taken, and are taking with this rule, and that is to issue the findings, promulgate the LDVR, and promulgate the Tailoring Rule. Our approach gives effect to as much of Congress's intent for each of these provisions, and the CAA as a whole, as possible.

With respect to the endangerment/cause or contribute findings under CAA section 202(a), congressional intent is clear that, as we stated in making the findings and the Supreme Court held in Massachusetts v. EPA, we are precluded from considering factors other than the science based factors relevant to determining the health and welfare effects of the air pollution in question. Accordingly, EPA determined that under Massachusetts v. EPA, 549 U.S. 497 (2007) we were precluded from deferring or foregoing the findings due to concern over impacts on stationary sources affected by PSD or title V requirements. See 74 FR at 66496, 66500-01 (“Taken as a whole, the Supreme Court's decision clearly indicates that policy reasons do not justify the Administrator avoiding taking further action on the questions here.”); see also Massachusetts v. EPA, 549 U.S. at 533; see also 74 FR at 66515-16 (December 9, 2009). (The Administrator “must base her decision about endangerment on the science, and not on the policy considerations about the repercussions or impact of such a finding). Moreover, as EPA also noted, “EPA has the ability to fashion a reasonable and common-sense approach to address greenhouse gas emissions and climate change.” 74 FR at 66516.

Regarding the timing of the LDVR, Congress's intent was that endangerment/cause or contribute findings under section 202(a) would in fact lead to control of the air pollutants from new motor vehicles and new motor vehicle engines contributing to the harm. The primary goal of section 202(a) is to achieve such reductions by requiring that EPA adopt emissions standards, and as a result, proceeding with the LDVR is consistent with that goal. In contrast, deferring the LDVR and thereby delaying achievement of the public health and welfare benefits Congress expected and required under section 202(a) would run directly counter to what Congress intended under section 202(a)—EPA issuing emissions standards to address the public health and welfare problems that were identified, not EPA refusing to do so.

Moreover, we have compelling reasons to proceed with the LDVR, in the manner that we did. As we stated in the LDVR, in response to similar comments that we were not obligated to conduct that rulemaking, or to conduct it at the time that we did:

Some of the comments relating to the stationary source permitting issues suggested that EPA should defer setting GHG standards for new motor vehicles to avoid * * * [adverse] stationary source permitting impacts. EPA is issuing these final GHG standards for light-duty vehicles as part of its efforts to expeditiously respond to the Supreme Court's nearly three year old ruling in Massachusetts v. EPA, 549 U.S. 497 (2007). In that case, the Court held that greenhouse gases fit within the definition of air pollutant in the Clean Air Act, and that EPA is therefore compelled to respond to the rulemaking petition under section 202(a) by determining whether or not emissions from new motor vehicles cause or contribute to air pollution which may reasonably be anticipated to endanger public health or welfare, or whether the science is too uncertain to make a reasoned decision. The Court further ruled that, in making these decisions, the EPA Administrator is required to follow the language of section 202(a) of the CAA. The Court stated that under section 202(a), “[i]f EPA makes [the endangerment and cause or contribute findings], the Clean Air Act requires the agency to regulate emissions of the deleterious pollutant.” 549 U.S. at 534. As discussed above, EPA has made the two findings on contribution and endangerment. 74 FR 66496 (December 15, 2009). Thus, EPA is required to issue standards applicable to emissions of this air pollutant from new motor vehicles.

The Court properly noted that EPA retained “significant latitude” as to the “timing * * * and coordination of its regulations with those of other agencies” (id.). However it has now been nearly three years since the Court issued its opinion, and the time for delay has passed. In the absence of these final standards, there would be three separate federal and state regimes independently regulating light-duty vehicles to increase fuel economy and reduce GHG emissions: NHTSA's CAFE standards, EPA's GHG standards, and the GHG standards applicable in California and other states adopting the California standards. This joint EPA-NHTSA program will allow automakers to meet all of these requirements with a single national fleet because California has indicated that it will accept compliance with EPA's GHG standards as compliance with California's GHG standards. 74 FR at 49460. California has not indicated that it would accept NHTSA's CAFE standards by themselves. Without EPA's vehicle GHG standards, the states will not offer the federal program as an alternative compliance option to automakers and the benefits of a harmonized national program will be lost. California and several other states have expressed strong concern that, without comparable federal vehicle GHG standards, the states will not offer the federal program as an alternative compliance option to automakers. Letter dated February 23, 2010 from Commissioners of California, Maine, New Mexico, Oregon and Washington to Senators Harry Reid and Mitch McConnell (Docket EPA-HQ-OAR-2009-0472-11400). The automobile industry also strongly supports issuance of these rules to allow implementation of the national program and avoid “a myriad of problems for the auto industry in terms of product planning, vehicle distribution, adverse economic impacts and, most importantly, adverse consequences for their dealers and customers.” Letter dated March 17, 2010 from Alliance of Automobile Manufacturers to Senators Harry Reid and Mitch McConnell, and Representatives Nancy Pelosi and John Boehner (Docket EPA-HQ-OAR-2009-0472-11368). Thus, without EPA's GHG standards as part of a federal harmonized program, important GHG reductions as well as benefits to the automakers and to consumers would be lost.165 In addition, delaying the rule would impose significant burdens and uncertainty on automakers, who are already well into planning for production of MY 2012 vehicles, relying on the ability to produce a single national fleet. Delaying the issuance of this final rule would very seriously disrupt the industry's plans.

Instead of delaying the LDV rule and losing the benefits of this rule and the harmonized national program, EPA is directly addressing concerns about stationary source permitting in other actions that EPA is taking with regard to such permitting. That is the proper approach to address the issue of stationary source permitting, as compared to delaying the issuance of this rule for some undefined, indefinite time period.

75 FR 25,402 cols. 1-3 (May 7, 2010) (footnote omitted).

With respect to both the endangerment/cause or contribute findings and the LDVR, it would require speculation and conjecture to defer—or, certainly, to forego altogether—the findings or LDVR until EPA completed streamlining the PSD and title V requirements on grounds that doing so would allow full compliance in the future with all PSD and title V statutory provisions. That is the gist of commenters' argument—that EPA should defer or forego issuance of the findings and the LDVR to avoid causing an absurd result from implementation of the separate PSD and title V programs. Underlying this claim is the assumption that this would allow EPA to avoid the “absurd results”. As we discuss elsewhere in this rulemaking, there is no basis at this point to determine that streamlining will ultimately allow full compliance with the PSD and title V requirements. Rather, it is possible that EPA may conclude that none of the available streamlining techniques will allow all GHG sources at the statutory thresholds to comply with PSD and title V requirements in a manner that does not impose undue costs on the sources or undue administrative burdens on the permitting authorities. Under these circumstances, EPA may then permanently exclude GHG source categories from PSD or title V applicability under the absurd results doctrine. Moreover, it may well take many years before EPA is in a position to come to a conclusion about the extent to which streamlining will be effective and therefore be able to come to a conclusion as to whether any source categories should be permanently excluded from PSD or title V applicability. In our rulemaking today, we describe what actions we expect to take in the first 6 years after PSD and title V are triggered for GHG sources, and we may well be in a situation in which we continue to evaluate streamlining measures and PSD and title V applicability to GHG sources after this 6-year period.

Accordingly, deferring the endangerment/cause or contribute findings and LDVR until such time that PSD and title V streamlining would allow full implementation of these programs at the statutory limits would serve only to delay the benefits of the LDVR, as well as the benefits that come from phasing in implementation of the PSD program to cover larger sources first. It would rely on an assumption that is unfounded at this point, that is, that such full compliance will be required at some point in the future. Delaying the emissions benefits of the LDVR and the related emissions benefits from partial implementation of the PSD program fails to implement Congress' intent that the endangerment/cause or contribute findings “shall” lead to emissions standards for new motor vehicles contributing to the endangerment, and related emissions controls for the same air pollutant under the PSD program. EPA need not determine at this time what approach would be appropriate if there was a determination that full compliance with PSD and title V would in fact occur at some point in the future. In this case, absent such a determination, it would be improper to rely on speculation of such a future possibility as a basis under section 202(a) to defer or forego issuance of the LDVR on the grounds that EPA should defer or forego the LDVR to avoid causing an absurd result. Likewise there is no basis to defer proceeding at this time with the streamlining of the PSD and title V programs.

With respect to the PSD and title V applicability requirements, as we discuss elsewhere, we believe that Congress expressed a clear intent to apply PSD and title V to GHG sources and that the phase-in approach incorporated in the Tailoring Rule is fully appropriate. Proceeding now with the endangerment/contribution findings and LDVR, even if phasing-in of the PSD and title V programs is required, is consistent with our interpretation of the PSD and title V applicability requirements. Delaying the endangerment/contribution findings or LDVR, and thereby delaying the triggering of PSD and title V requirements for GHG sources, would lead to the loss of a practicable opportunity to implement the PSD and title V requirements in important part, and thereby lead to the loss of important benefits. As discussed elsewhere, promulgating the LDVR and applying the PSD and title V requirements to the largest GHG sources, as we do in this Tailoring Rule, is practicable because the sources that would be affected by the initial implementation steps we promulgate in this rule are able to bear the costs and the permitting authorities are able to bear the associated administrative burdens. Promulgating the LDVR now provides important advantages because the sources that would be affected by the initial steps are responsible for most of the GHG emissions from stationary sources.

It should also be noted that as discussed elsewhere in this rulemaking, our ability to develop appropriate streamlining techniques for PSD and title V requirements is best done within the context of actual implementation of the permitting programs, and not in isolation of them. That is, because the great majority of GHG sources have not been subject to PSD and title V requirements, we will need to rely on the early experience in implementing the permitting requirements for the very large sources that initially will be subject to those requirements in order to develop streamlining techniques for smaller sources. It is the real world experience gained from this initial phase that will allow EPA to develop any further modifications that might be necessary. This would not and could not occur if the LDVR were delayed indefinitely or permanently, so that PSD and title V requirements were not triggered. It is unrealistic to expect that delaying action until a future tailoring rule could resolve all of the problems identified in this rulemaking, absent any real world implementation experience.

At its core, commenters' argument is that EPA should delay (if not forego altogether) doing anything to address GHG emissions and the problems they cause until it can do so in a way that does not cause any implementation challenges, even if that delay results in continued endangerment to public health and welfare. EPA does not take such a myopic view of its duties and responsibilities under the CAA. Congress wrote the CAA to, among other things, promote the public health and welfare and the productive capacity of the population. CAA § 101(b)(1). EPA's path forward does just this. Thus, proceeding with the endangerment/cause or contribute findings, the LDVR, and with PSD and title V through the phase-in approach of the Tailoring Rule maximizes the ability of EPA to achieve the Congressional goals underlying sections 202(a) and the PSD and title V provisions, and the overarching CAA goal of protecting public health and welfare. Congress called for EPA (1) To determine whether emissions from new motor vehicles contribute to air pollution that endangers, (2) if that the determination is affirmative, to issue emissions standards for new motor vehicles to address the endangerment, and (3) to implement the PSD and Title V program to address similar emissions in their permitting program as another tool to address the air pollutant at issue. Delaying both the LDVR and PSD/title V implementation, as commenters have called for, would run directly counter to these Congressional expectations. Commenters' calls for deferral or foregoing of the findings or LDVR are generally phrased in a conclusory fashion, and do not demonstrate how EPA could take the required CAA actions concerning GHGs while remaining within the requirements of each of the various CAA provisions, and achieving the overall goals of the CAA. As such the comments do not provide a valid basis for the deferral of agency action they suggest.

9. “Administrative Necessity” Basis for PSD and Title V Requirements in Tailoring Rule

EPA believes that the “administrative necessity” doctrine, within the Chevron framework, also justifies this rulemaking. Applying the applicability requirements of the PSD and title V programs according to a literal reading of their terms (as EPA has narrowed them in the past through interpretation) to GHG sources beginning on the January 2, 2011 date that regulation of GHGs takes effect would sweep so many sources into those programs as to render the programs impossible for the permitting authorities to administer. Although streamlining the PSD and title V programs offers some promise to improve the administrability of the programs, given the time needed to implement such streamlining, the step-by-step expansion of PSD and title V requirements to GHG sources that we are promulgating is the most that the permitting authorities can reasonably be expected to administer.

This section discusses the application of the “administrative necessity” doctrine. Our views concerning this doctrine remain similar to what we said at proposal, except that in this rulemaking we place the doctrine more clearly in the Chevron analytical framework, we revise our assessment of the administrative burdens due to new analysis we have conducted and information we have received since proposal, and we make certain revisions to the tailoring approach. This analysis and information, as well as the revisions to the tailoring approach, have already been presented previously, in the discussion of the “absurd results” basis. In addition, it is not necessary to reiterate the lengthy discussion of the “administrative necessity” doctrine that we included in the proposal or the factual data presented previously; as a result, this section briefly highlights the conclusions we have reached about the application of this doctrine.

As noted previously, under the PSD and title V applicability provision—read literally, as we have long interpreted them—EPA's recent promulgation of the LDVR will trigger the applicability of PSD and title V for GHG sources at the 100/250 tpy and 100 tpy threshold levels, respectively, as of January 2, 2011. This is because PSD applicability hinges on the definition of “major emitting facility” and title V applicability hinges on the definition of “major sources,” and those terms, read literally, and under EPA's long-standing narrowing interpretation, apply PSD and title V, respectively, to sources of any air pollutant that is subject to regulation under another provision of the CAA. EPA's promulgation of the LDVR means that GHGs will become subject to regulation on the date that the rule takes effect, which will be January 2, 2011.

Absent tailoring, the January 2, 2011 trigger date for GHG PSD applicability will give rise to an extraordinarily large number of PSD permitting actions—we estimate more than 81,000 per year—representing an increase of almost 300-fold over the current 280 PSD permitting actions each year. In addition, over 6 million sources will become subject to title V, an increase of more than 400-fold over the 14,700 sources that currently are subject to title V. The permitting authorities will find it impossible to administer programs of these sizes as of that date.

All this results from a literal application of the PSD and title V applicability provisions to GHG sources. However, under Chevron, we must interpret and apply statutory requirements on the basis of congressional intent. Although the literal meaning of the statutory provisions is the first and generally the best indicator of congressional intent, there are cases in which that is not so. As discussed previously, we believe that as a general matter, statutory directives should be considered to incorporate Congress's intent that they be administrable, and we believe that this proposition is implicit in the “administrative necessity” doctrine that the DC Circuit has established and that we believe applies here. See Alabama Power v. Costle, 636 F.2d 323, 356-57 (DC Cir. 1980). This doctrine authorizes EPA to undertake a process for rendering the PSD and title V requirements administrable. Indeed, the Court in Alabama Power established this doctrine specifically in the context of the PSD provisions, including, in particular, the modification provision. As noted elsewhere, the Court held that EPA may “consider the administrative burden” associated with applying PSD for emissions increases, and establish significance levels designed to avoid “severe administrative burdens on EPA, as well as severe economic burdens” on sources. Id. at 405.

As we said in the proposal, we read the case law to establish a three-step approach for implementing the “administrative necessity” doctrine: An agency is not required to adhere to literal statutory requirements if the agency, as the first step, makes every effort to adjust the requirements within the statutory constraints, but concludes with justification—at the second step—that it would be impossible to comply with the literal reading of the statute. Under those circumstances, the agency may—at the third step—develop what is in effect a compliance schedule with the statutory requirements, under which the agency will implement the statute as much as administratively possible and as quickly as administratively possible. See 74 FR 55315—55316.

a. First Step of the “Administrative Necessity” Analysis: Streamlining

In the proposed rulemaking, EPA discussed at length the prospect of streamlining both PSD and title V. EPA described “several potentially useful tools available in the streamlining toolbox for the PSD permitting threshold level, the PSD significance level, and the title V permitting threshold,” specifically:

For the PSD permitting threshold level and significance level, there are at least three such tools: The first is interpreting the definition of “potential to emit” so that the amount of a source's emissions that counts in determining whether it qualifies as a major source and therefore is above the permitting threshold requirements is closer to the amount of its emissions when it is in actual operation, rather than the amount of emissions that the source would emit if it were operating continuously. Narrowing the definition of PTE is a potentially extremely important tool in this context because identifying the amount of a source's emissions as closer to its actual emissions in this manner would mean that very large numbers of residential and commercial sources would have significantly lower emissions and would fall below the statutory threshold requirements for triggering PSD. Second, EPA believes it may be able to develop programs involving general permits, under which large numbers of similarly situated sources would each be covered by essentially the same permit established through a regulatory action by the permitting authority. This approach could achieve economies of scale and thereby reduce administrative burden. Third, EPA believes it may be able to streamline the single most time-consuming element of the PSD permit program, which is the determination of BACT as required under CAA § 165(a)(4), by establishing presumptive BACT levels for certain source categories that comprise large numbers of sources. As for title V, as discussed below in detail, EPA believes that defining “potential to emit” to reflect more closely a source's actual operation and developing a program of general permits could streamline the administration of title V permits.

74 FR 55315 col. 2-3.

At proposal we stated that we would, and we still commit to, vigorously pursue development of these streamlining measures, and, as indicated in our discussion of streamlining methods in section V.E.1 and in response to comments, we have already begun developing those measures. For example, as described elsewhere, we have done much work—both with stakeholders and in-house—to begin to develop recommendations for what controls would qualify as BACT for various industries. This work is important as a foundation for developing presumptive BACT, which is a potentially efficient streamlining measure.

However, it is not possible for us or the state and local permitting authorities to develop and implement streamlining techniques by the time that PSD and title V are triggered for sources emitting GHGs—January 2, 2011—or shortly thereafter. Developing streamlining methods would entail acquiring more information about the affected industry, may entail rulemaking, and would likely entail some type of public review of proposals for streamlining even if not done through rulemaking. As discussed in section V.E, we do not expect that we could complete all those steps for meaningful streamlining measures within 2 years.

b. Second Step of the “Administrative Necessity” Analysis: Demonstration of Administrative Impossibility

With no streamlining measures available at the time that PSD and title V would apply to sources of GHGs or shortly thereafter, under the second step of the “administrative necessity” analysis, we must determine whether implementation of the statutory requirements at that time would be administratively impossible for the permitting authorities. We are mindful that the DC Circuit has cautioned that this showing is a high hurdle. See 74 FR 55317.

Even so, we believe there is no question that a literal application of the PSD and title V programs to GHG sources as of January 2, 2011 would be flatly impossible for the state and local permitting authorities to administer for at least an initial period of time. The key facts have been recounted previously, and no more than a brief recitation is necessary here. On the PSD side, annual permit applications would increase by over 300-fold, from 280 to almost 82,000; costs to the permitting authorities would increase more than 100-fold, from $12 million to $1.5 billion; and the permitting authorities would need to hire, train, and manage 9,772 FTEs. For title V, total permit applications would increase by over 400-fold, from 14,700 to 6.1 million; costs to the permitting authorities would increase from $62 million to $21 billion; and the permitting authorities would need to hire, train, and manage 229,118 FTEs.

We have elaborated upon these burdens elsewhere in this notice. They bespeak an impossible administrative task. It is not hyperbole to say that if these administrative responsibilities are not considered impossible within the meaning of the “administrative necessity” doctrine, then it is difficult to imagine what would be considered impossible.

c. Third Step of the “Administrative Necessity” Analysis: Tailoring

Under the third step of the “administrative necessity” analysis, we must demonstrate that the steps we intend to take towards implementation of the statutory requirements are the most that can be done during the indicated time frames, in light of administrative resources. In this manner, we adhere most closely to the statutory requirements. See 74 FR 55318. This amounts to establishing a schedule for phasing in PSD and title V applicability to GHG sources. Because this step is based on the administrative resources of the permitting authorities, our analysis is similar, and leads to the same conclusions, as we described previously concerning the “absurd results” basis. That is, we believe that our tailoring approach—including Step 1, to be implemented as of January 2, 2011; Step 2, to be implemented as of July 1, 2011; the additional rulemaking that we commit to finalize by July 1, 2012, and that will address further threshold reductions as a Step 3; the study and subsequent rulemaking to address smaller sources by April 30, 2016; and the determination not to lower the threshold below 50,000/50,000 tpy CO₂ e before April 30, 2016 at the earliest—is the most that we can do to expand the PSD and title V programs, based on administrative resources and the information we currently have about the prospects for streamlining and increasing permitting resources.

As noted previously, at some point in the process of additional rulemaking, we may conclude under the “absurd results” doctrine that we will not apply PSD or title V to GHG sources below a certain size level. The same conclusion may be supportable under the “administrative necessity” doctrine if we decide, based on the information available to us, that even with all of the streamlining that we are able to accomplish and even with a significant expansion of permitting resources, it may not be administratively feasible to implement PSD or title V to sources below that level. See Alabama Power v. Costle, 636 F.2d at 358 (acknowledging, in discussing the “administrative necessity” doctrine, that “[c]ategorical exemptions from the clear commands of a regulatory statute [are] sometimes permitted,” although emphasizing that such exemptions “are not favored”).

In addition, as noted above, in a subsequent rulemaking, we may conclude that title V should not apply to GHG sources with “empty permits,” under the “absurd results” doctrine. The basis for this conclusion could be a determination that (1) although the applicability provisions apply by their terms to sources on the basis of their emissions, and without regard to whether the sources would hold “empty permits,” those provisions cannot be read literally under the “absurd results” doctrine; and (2) it is not clear whether Congress intended that title V apply to such sources, and EPA has reasonably determined, under Chevron Step 2, that title V does not. If we come to that conclusion, then, at that point in time, the “administrative necessity” doctrine would remain relevant for title V purposes only if it is necessary, for administrative reasons, to phase in the application of title V to GHG sources that have applicable requirements, and that therefore do not have “empty permits.” This is because the “administrative necessity” doctrine is relevant only when a statutory directive, read literally, imposes impossible administrative obligations, and Congress may be presumed to have intended that the directive be administrable. The “administrative necessity” doctrine would not come into play if it is concluded either that under the “absurd results” doctrine Congress did not intend the statutory directive or that, under that doctrine, Congress's intent was not clear and EPA reasonably decided that the directive does not apply.

10. “One-Step-at-a-Time” Basis for Tailoring Rule

In addition to the “absurd results” and “administrative necessity” doctrines, the “one-step-at-a-time” judicial doctrine, within the Chevron framework, supports EPA's Tailoring Rule. The case law under this doctrine, described previously, indicates that the doctrine justifies an agency's step-by-step approach under the following circumstances or conditions: (1) The agency's ability to comply with a statutory directive depends on facts, policies, or future events that are uncertain; (2) the agency has estimated the extent of its remaining obligation; (3) the agency's incremental actions are structured in a manner that is reasonable in light of the uncertainties; and (4) the agency is on track to full compliance with the statutory requirements. EPA's Tailoring Rule fulfills each of those four.

First, as the DC Circuit stated in National Association of Broadcaster v. FCC, 740 F.2d 1190, 1210 (DC Cir. 1984) (“National Association of Broadcasters”), incremental agency action is most readily justifiable “against a shifting background in which facts, predictions, and policies are in flux and in which an agency would be paralyzed if all the necessary answers had to be in before any action at all could be taken.” Those circumstances are present here, and so is that fact that the task at hand is extraordinarily demanding. As discussed previously, EPA and the permitting authorities' progress in implementing the PSD and title V programs for GHG sources will depend in large measure on the development of streamlining measures and increases in permitting authorities' resources, and those things carry some uncertainty and in any event, under the best of circumstances, cannot have much impact for at least several years. It will take EPA that long to develop streamlining measures, and it will take permitting authorities that long to begin to raise money and hire and train FTEs.

Second, as the Court stated in National Association of Broadcasters, “the agency [should] ma[k]e some estimation, based upon evolving economic and technological conditions, as to the nature and magnitude of the problem it will have to confront when it comes to [undertake the remaining steps]” and that estimation must be “plausible and flow from the factual record compiled.” Id. at 1210. Here, EPA has done this by estimating the number of PSD and title V permits and the costs of issuing them, and has provided as much information as possible about the development of streamlining methods and permitting authority resources.

Third, again as the Court stated in National Association of Broadcasters, it must be “reasonable, in the context of the decisions made in the proceeding under review, for the agency to have deferred the issue to the future. With respect to that question, postponement will be most easily justified when an agency acts against a background of rapid technical and social change and when the agency's initial decision as a practical matter is reversible should the future proceedings yield drastically unexpected results.” Id. at 1211. Here, our tailoring approach is reasonable in light of changes in permitting authority capacity that may occur with the development of streamlining methods and increased resources. In addition, the first two steps that EPA promulgates today are reasonable initial steps that we expect to build on by lowering thresholds, as appropriate, in the future. We have no reason to suspect that we may need to reverse either of the first two steps. Having received and analyzed extensive comment on the number of permitting actions to expect and on permitting authority resources, we consider it unlikely that we would need to establish a higher threshold level than what we have established in Steps 1 and 2. In addition, if we were to adopt an “empty permits” approach for title V, we would not need to reverse either of Steps 1 and 2, as explained above.

Finally, as the DC Circuit stated in Grand Canyon Air Tour Coalition v. F.A.A., 154 F.3d 455, 477-78 (DC Cir. 1998), the Courts will accept an initial step towards full compliance with a statutory mandate, as long as the agency is headed towards full compliance, and we believe that the doctrine is applicable here. EPA intends to require full compliance with the CAA applicability provisions of the PSD and title V programs, but we believe that in the case of GHG-emitting sources, by application of the “absurd results” doctrine or the “administrative necessity” doctrine, full compliance with the applicability provisions does not necessarily mean full compliance with the literal terms of those provisions. Rather, as we have explained elsewhere, in the case of GHG sources, full compliance may mean compliance with higher levels that are consistent with congressional intent, under the “absurd results” doctrine, or that are within the reach of permitting authorities in light of their administrative constraints, under the “administrative necessity” doctrine. This rulemaking constitutes a package of initial steps towards that full compliance, and, seen in that light, is supported by the “one-step-at-a-time” doctrine.

Even if the doctrine were found to apply only when an agency is committed to fully implementing statutory requirements according to their literal terms, we believe that the steps we promulgate in this notice would be considered valid under the one-step-at-a-time doctrine. This is because even if we are incorrect about the applicability of the “absurd results” and “administrative necessity” doctrines, so that GHG sources are required to comply with the literal terms of the PSD and title V applicability provisions, the “one-step-at-a-time” doctrine would allow PSD and title V applicability to be phased in, and the first two steps we promulgate in this notice would be upheld as reasonable initial steps toward full compliance with the literal terms of the CAA. As we have described elsewhere, there is little question but that sources and permitting authorities cannot reasonably be expected to comply with or implement PSD and title V applicability requirements in the near term—by January 2, 2011 and July 1, 2011—except to the limited extent described under Steps 1 and 2. Nor is applicability of the PSD and title V requirements at levels below 50,000 tpy CO₂ e reasonable before 6 years from promulgation of this rule, as discussed elsewhere. If further steps resulting in full compliance with the literal terms of the applicability provisions of PSD and title V were required, it would be reasonable for those steps to occur in the future, as part of the rulemaking to be completed by the sixth year after promulgation, to which EPA commits itself as part of this action, or as part of subsequent actions. See Grand Canyon Air Tour, 891 F.2d at 476-77 (upholding agency action as a step towards full compliance with statutory mandate when the agency expected full compliance to occur some 20 years after the deadline in the statute).

C. Mechanisms for Implementing and Adopting the Tailoring Approach

In this section, we discuss three issues related to adoption of the tailoring approach within our regulations and by permitting agencies. The first is the regulatory mechanism for implementing the tailoring approach—that is, the specific way we are revising the PSD and title V applicability provisions to incorporate the tailoring approach—and our rationale. The second is the process by which state or local permitting authorities may incorporate the tailoring approach into their PSD SIP and title V permit programs. Finally, we discuss our reasons for delaying action on our proposal to limit approval of both SIP-approved PSD programs and title V programs, and we request certain information from states on both of their programs and their actions in response to this rule.

In brief, we proposed to exempt sources emitting GHGs below certain threshold levels from the definition of the regulatory terms “major stationary source” and “major modification” in PSD programs and the definition of the regulatory term “major source” in title V programs. We further proposed to effectuate this change in SIP-approved PSD programs (as included in SIPs) and EPA-approved part 70 title V programs by limiting our prior approval of those programs to the revised applicability thresholds for GHGs. These changes would have the effect of putting the higher thresholds adopted under the Tailoring Rule in place in states PSD and title V programs as a matter of federal law. However, state commenters expressed concern that they would not be able to adopt the Tailoring Rule under state laws on an expeditious basis. To address this, our final action differs from our proposed rule in the way we incorporate the limitations promulgated in this Tailoring Rule into the “major stationary source,” “major modification” and “major source” definitions. This approach relies on further defining the term “subject to regulation” and although this approach is not substantively different in effect from the proposed rule, it will facilitate more rapid adoption and implementation of the Tailoring Rule by states through interpretation of language in existing state regulations. We believe these differences are a logical outgrowth of our proposed rule. We are also delaying action on our proposed limited approval of EPA-approved PSD programs and part 70 title programs to determine how each state will implement the final rules.

1. PSD Approach: Background and Proposal

Under CAA section 165(a), no “major emitting facility” may construct or modify unless it receives a preconstruction permit that meets the requirements of the PSD program. CAA section 169(1) defines a major emitting facility as “any * * * source[]” in one of 28 specified source categories that “emit[s], or ha[s] the potential to emit, one hundred tons per year or more of any air pollutant;” or “any other source with the potential to emit two hundred and fifty tons per year or more of any air pollutant.” EPA's regulations replace the term “major emitting facility” with the term “major stationary source” and define the term as “[a]ny of * * * [28 types of] stationary sources of air pollutants which emits, or has the potential to emit, 100 tons per year or more of any regulated NSR pollutant” or “any stationary source which emits, or has the potential to emit 250 tons per year or more of a regulated NSR pollutant.” 40 CFR 51.166(b)(1)(i)(a)-(b). The term “regulated NSR pollutant” is defined to include, among other things, “any * * * air pollutant that otherwise is subject to regulation under the Clean Air Act.” 40 CFR 51.166(b)(50). Note that the regulatory definition in effect interprets the statutory definition more narrowly to read “one hundred [or two hundred and fifty] tons per year or more of any air pollutant subject to regulation under the Clean Air Act” (emphasis added).

Similarly, under the statute, a modification occurs if there is a physical change or change in the method of operation “which increases the amount of any air pollutant emitted * * * .” CAA section 165(a), 169(2)(c), and 111(a)(4). As with the major stationary source definition, we have limited coverage of the modification provision to physical changes or changes in the method of operation that result a significant net emissions increase in emissions of a “regulated NSR pollutant.” 40 CFR 51.166(b)(2)(i).

Our proposed rule revised the definition of “major stationary source to (1) exempt GHG from the regulated NSR pollutants that, if emitted by a source in the 100 or 250 tpy quantities, would cause the source to qualify as a “major stationary source,” and (2) add a specific threshold at which a source that emits a specified quantity of GHGs (at proposal, that quantity was 25,000 tpy CO₂ e) would qualify as a “major stationary source.” 74 FR 55351, proposed 40 CFR 51.166(b)(1)(i)(a), (b), and (d). We also proposed a significance threshold, which is the amount of an increase needed to trigger PSD for a modification or to require BACT for a new source, at a level between 10,000 and 25,000 tpy CO₂ e. 74 FR 55351; 40 CFR 51.166(b)(23)(i).

Additionally, we recognized that it may take some time before states could change their SIP-approved PSD programs and that as a result, absent additional action on our part, GHG-emitting sources would remain subject to the 100 or 250 tpy thresholds, and subject to a zero significance threshold for major modifications as a matter of federal law. To address this issue, we proposed to narrow our previous approval of those SIPs. The effect of our proposal would be that EPA would have approved the SIP PSD programs only to the extent they apply PSD and requirements to GHG sources at or above the thresholds established in the Tailoring Rule (which, generally, were 25,000 tpy CO₂ e), and EPA would have taken no action on the SIP PSD programs to the extent they apply PSD requirements to GHG sources below that threshold. We relied on the authority of the APA and the general authority of CAA section 301 and, in the alternative, on the error correction mechanism under CAA section 110(k)(6). Our limited approval would revise existing EPA-approved SIP PSD programs to authorize permitting under the CAA only for GHG sources at the appropriate levels.

In response to our proposed approach, we received numerous comments from state and local permitting agencies expressing significant concern. They observed that our proposed approach could meet its objectives to avoid applying PSD requirements to small sources under federal law, but would not succeed in avoiding the application of PSD requirements to those small sources under state law. The commenters explained that, although EPA was changing federal PSD applicability thresholds; for GHG-emitting sources to incorporate the tailoring approach, and limiting the scope EPA approval of SIPs consistent with these thresholds, the state rules containing the originally-approved SIP thresholds would continue to apply as a matter of state law. As commenters explained, for the most part, the laws and regulations states adopt to implement federal PSD programs mirror EPA's regulations, so that the state laws, apply PSD to sources that emit air pollutants subject to regulation at the 100/250 tpy threshold. Commenters reasoned that, until the states can change their state laws, the 100/250 tpy thresholds will continue to apply as a matter of state law, even though the higher thresholds apply as a matter of federal law.

Importantly, these commenters emphasized, their state process requires that they promulgate a rulemaking, or in some cases, a legislative change, to incorporate the higher thresholds for GHG sources in their SIPs. These processes would require many months and in some cases as long as 2 years. As a result, sources that emit GHGs below the federally established levels in the final rule, but at above the 100/250 tpy levels in state laws and rules, would still be required to obtain PSD permits under state law. As a result, states, in attempting to implement state permitting requirements, would be faced with the same administrative difficulties that EPA recognized in the proposed rule as impossible. Commenters emphasized that this situation was untenable.

In addition to the state comments just described, we received comments that took issue with our view that we were in effect revising the numerical thresholds for PSD applicability as the legal mechanism for the tailoring approach. They asserted that in fact, our mechanism consisted of interpreting the term “any source” to exclude small GHG-emitting sources. Other commenters objected to our proposed mechanism of narrowing our previous SIP approval, arguing that this mechanism was without legal basis.

2. Rationale for Our Final Approach To Implementing PSD

In response to these concerns, we are adding another mechanism to implement the tailoring approach for PSD, and that is to adopt a definition, within our PSD regulations, the phrase “subject to regulation,” as found within the phrase “any regulated NSR pollutant,” which, in turn, is part of the definitions of “major stationary source” and “major modification.” To implement this mechanism, we are defining the phrase “subject to regulation” so that the GHGs emitted by sources that fall below the thresholds or scope established in Steps 1 and 2 are not treated as “subject to regulation,” and therefore do not trigger PSD for the sources that emit them. As discussed in section V.B.3., the term “subject to regulation” is one of four terms that should be considered not to apply literally in the case of GHG sources.

To understand this approach, it is useful to return to the definition of “major stationary source,” which, again, is central to PSD applicability. The definition, quoted previously, employs the term “regulated NSR pollutant,” which is a defined term. The definition incorporates many other elements as well (e.g., the 100/250 threshold requirements), but for convenience, we quote it as follows: A “major stationary source” is “[a]ny * * * source[-] of air pollutants, which emits, or has the potential to emit, [depending on the source category, either] 100 [or 250] tons per year or more of any air pollutant that is subject to regulation under the Clean Air Act.” 40 CFR 51.166(b)(1)(i)(a)-(b). Applying our definition of “subject to regulation” to exclude GHG sources that emit below specified thresholds, the definition may now be paraphrased as follows: A “major stationary source” is any source of air pollutants, which emits, or has the potential to emit, depending on the source category, either 100 or 250 tpy or more of any air pollutant subject to regulation under the CAA, except that the source's GHGs are considered to be subject to regulation under the CAA only the extent indicated under Steps 1 and 2 of the Tailoring Rule, e.g., for Step 2, only if the source's GHG emissions exceed the threshold established in Step 2. We adopt the same approach for the definition of the regulatory term “major modification.”

Although EPA is revising its regulations to apply the phrase subject to regulation in this manner, we have been advised that states may be able to adopt our approach without having to undertake a rulemaking action to revise their state regulations or without requiring an act of the state legislature. Instead, it is our understanding that states may adopt our approach by interpreting the term “subject to regulation” reflected in their regulations to have the same meaning that we are assigning to that term in our regulations in this rulemaking. This is particularly—although not exclusively—the case in a state that has taken the position, or determines now, that the state's definition of “subject to regulation,” or, more broadly, “regulated NSR pollutant” or “major stationary source” or “major modification,” is intended to be interpreted in a way that tracks the meanings that EPA has assigned to these phrases. Such states can adopt the meaning of “subject to regulation” that we establish in this rule by January 2, 2011, and thereby avoid the situation in which, as a matter of state law, GHG-emitting sources above the 100 or 250 tpy thresholds become subject to PSD by that date. The following explains our basis for concluding that states may apply EPA's approach under existing regulations that use the term “subject to regulation.” On December 18, 2008, EPA issued the Interpretive Memo, establishing EPA's interpretation of the definition “regulated NSR pollutant” found at 40 CFR 52.21(b)(50). EPA intended this memorandum to resolve ambiguity in subparagraph (iv) of this definition, which includes “any pollutant that otherwise is subject to regulation under the Act.” Specifically, the memorandum stated that EPA will interpret the definition of “regulated NSR pollutant” to exclude pollutants for which EPA regulations only require monitoring or reporting but to include pollutants subject to either a provision in the CAA or regulation adopted by EPA under the CAA that requires actual control of emissions of that pollutant.

After reconsidering this interpretation through a formal notice-and-comment process, EPA refined its interpretation to establish that the PSD permitting requirements will not apply to a newly regulated pollutant until a regulatory requirement to control emissions of that pollutant “takes effect.” 75 FR 17704. Importantly, as stated previously, because the term “regulated NSR pollutant” is embedded within the definition of “major stationary source,” this interpretation effectively defines which major stationary sources are subject to PSD permitting. As a result, for example, EPA explained that PSD and title V permitting requirements for GHGs will not apply to GHGs until at least January 2, 2011, following the anticipated promulgation of EPA regulations requiring control of GHG emissions under title II of the CAA. Id.

In the RTC document for EPA's reconsideration of the PSD interpretative memorandum, we stated that,

Absent a unique requirement of state law, EPA believes that state laws that use the same language that is contained in EPA's PSD program regulations at 52.21(b)(50) and 51.166(b)(50) are sufficiently open-ended to incorporate greenhouse gases as a regulated NSR pollutant at the appropriate time consistent with EPA's interpretation of these regulations (emphasis added). (Docket ID No. EPA-HQ-OAR-2009-0597-0128).

Because the state regulations that include EPA's definition of the term “subject to regulation” in the reconsideration of the Interpretive Memo are “sufficiently open-ended to incorporate greenhouse gases as a regulated pollutant,” those state regulations are also sufficiently open-ended to incorporate the further refinement to the meaning of the phrase “subject to regulation” that we make in this rulemaking.

By the same token, EPA has historically interpreted certain state SIP-approved programs as sufficiently open-ended such that the rules provide for the “automatic assumption for the responsibility for review” of new pollutants before the general deadline for states to revise their PSD programs. See, e.g., 52 FR 24682. Conversely, we have also read federal rules and state rules approved in SIPs to provide for the automatic removal of a pollutant when such pollutant is no longer “subject to regulation.” For example, the 1990 CAA Amendments exempted HAPs listed in section 112(b)(1) from the PSD requirements. See CAA section 112(b)(6). Following passage of the amendments, EPA issued “New Source Review (NSR) Program Transitional Guidance,” a memorandum from John S. Seitz, Director, Office of Air Quality Planning and Standards to Regional Air Division Directors on March 11, 1991. In that guidance, EPA interpreted its PSD regulations to automatically cease to apply to listed HAPs (with some noted exceptions), and implicitly stated that a state with an open-ended SIP-approved PSD rule could also take the position that its SIP-approved rule automatically ceased to regulate HAPs.

After reviewing these past practices in the PSD permitting program, and EPA's prior statements regarding pollutants subject to the PSD program, we conclude that states with SIP-approved rules that contain the same language as used in 40 CFR 52.21(b)(50) or 40 CFR 51.166(b)(49), or that otherwise have sufficiently open-ended PSD regulations, would be able to implement our Tailoring Rule approach to permitting by interpreting their regulations, and without needing to promulgate a regulation or seek state legislative action. This is particularly—although not exclusively—the case for states that take the position that they intend their rules to apply in the same manner as EPA's counterpart rules. If states adopt this reading of their regulations, GHG sources falling below the specified cutoffs would not be emitting pollutants “subject to regulation” within the definition of “regulated NSR pollutant” and therefore would not be subject to PSD permitting as a major stationary source or for making a major modification.

During our consideration of this action, we participated in teleconferences with one local and six state agency permitting authorities to discuss this issue of whether they could implement the proposed rule without the need for state law or regulation changes or a revision of the provisions of state law that are a part of the SIP. We specifically discussed whether defining the phrase “subject to regulation” would better facilitate state incorporation of the limitations in this final rule. The state and local agencies participating in the calls generally agreed that defining the phrase “subject to regulation” would, compared to our proposed approach, better facilitate state incorporation of the limitations in the final rule in states with regulations that mirror the existing federal rules, or in states whose rules are otherwise sufficiently open-ended to incorporate the limitations in the final rule by interpretation. Participants from each agency also indicated that their rules contain the term “subject to regulation” and that term has not been previously interpreted in ways that would preclude application of the meaning assigned to the term by EPA. We therefore concluded it is likely the state rules are sufficiently open-ended to apply EPA's approach by interpretation (although some states indicated they may elect to pursue rulemaking in addition to or instead of interpretation). Accordingly, we selected the “subject to regulation” regulatory approach as the mechanism for implementing the final rule.

3. Other Mechanisms

As just described, we selected the “subject to regulation” mechanism because it most readily accommodated the needs of states to expeditiously revise—through interpretation or otherwise—their state rules. Even so, it is important to recognize that this mechanism has the same substantive effect as the mechanism we considered in the proposed rule, which was revising numerical thresholds in the definitions of major stationary source and major modification. Most importantly, although we are codifying the “subject to regulation” mechanism, that approach is driven by the needs of the states, and our action in this rulemaking should be interpreted to rely on any of several legal mechanisms to accomplish this result. Thus, our action in this rule should be understood as revising the meaning of several terms in these definitions, including: (1) The numerical thresholds, as we proposed; (2) the term, “any source,” which some commenters identified as the most relevant term for purposes of our proposal; (3) the term, “any air pollutant; or (4) the term, “subject to regulation.” The specific choice of which of these constitutes the nominal mechanism does not have a substantive legal effect because each mechanism involves one or another of the components of the terms “major stationary source”—which embodies the statutory term, “major emitting facility”—and “major modification,” which embodies the statutory term, “modification,” and it is those statutory and regulatory terms that we are defining to exclude the indicated GHG-emitting sources.

4. Codification of Interpretive Memo

As noted previously, we recently affirmed and refined our interpretation of the term “subject to regulation” as it applies broadly to the PSD program through a formal notice and comment process. “Interpretation of Regulations that Determine Pollutants Covered by Clean Air Act Permitting Programs,” 75 FR 17004. In the proposal associated with that action, we requested comment on whether we should codify our interpretation in the regulatory text. 74 FR 51535, 51547 (October 7, 2009). We elected not to codify our interpretation in the final action on reconsideration of the Interpretive Memo because we concluded such an action was not necessary and that it was important to apply the refined interpretation immediately. 75 FR 17015. However, in the RTC document for that action, we indicated that we had not ruled out the option of codifying our interpretation at a later time. Since we are otherwise adopting a definition of “subject to regulation” in this rule as the mechanism for implementing the phase-in, it makes sense in this final rule to codify the interpretation reflected in the Interpretive Memo and the final action on reconsideration at the same time to bring clarity to our rules. Specifically, the definitions of the term “subject to regulation” contain a paragraph that reflects our existing interpretation of that term (i.e., prior to adopting the provisions that implement the phase-in). Codification of the Interpretive Memo in this action makes sense to ensure the regulations reflect a complete picture of the meaning of “subject to regulation” applied by EPA. We also are moving existing exceptions (e.g., section 112 HAPs) to a new paragraph within the definition of “subject to regulation.” This minor reorganization of these regulations is not intended to effect any change in how they are to be implemented, but merely simplifies and clarifies the regulations by clearly delineating different terms and concepts.

This codification of this interpretation of “subject to regulation” from the reconsideration for the Interpretive Memo is not necessary to assure the effectiveness of the interpretation, and it does not disturb states' existing authority to adopt the definition through interpretation of their existing rules. Codifying our existing interpretation in this action will ensure that parties reading the regulations have a full understanding of how EPA applies the PSD program requirements. Since the interpretation described in the Interpretive Memo and the April 2, 2010 final action are otherwise applicable at this time, the particular time sensitivity discussed in the latter action is not the same for this final action tailoring the PSD requirements.

5. Delaying Limited Approvals and Request for Submission of Information From States Implementing a SIP-Approved PSD Program

Because we now anticipate that many states will be able to implement our tailoring approach through interpretation of the term “subject to regulation,” and without the need to revise their SIPs, we are delaying further action on our proposal to limit our approval of SIPs until we better understand how permitting authorities will, in fact, implement our tailoring approach. For this purpose, we ask each state to submit a letter to the appropriate EPA Regional Administrator no later than August 2, 2010. In that letter, the state should explain whether it will apply EPA's meaning of the term “subject to regulation” and if so, whether the state intends to incorporate that meaning of the term through interpretation, and without undertaking a regulatory or legislative process. If a state must undertake a regulatory or legislative process, then the letter should provide an estimate of the time needed to adopt the final rules. If a state chooses not to adopt EPA's meaning by interpretation, the letter should address whether the state has alternative authority to implement either our tailoring approach or some other approach that is at least as stringent, whether the state intends to use that authority. If the state does not intend to interpret or revise its SIP to adopt the tailoring approach or such other approach, then the letter should address the expected shortfalls in personnel and funding that will arise if the state attempts to carry out PSD permitting for GHG sources under the existing SIP and interpretation.

For any state that is unable or unwilling to adopt the tailoring approach by January 2, 2011, and that otherwise is unable to demonstrate adequate personnel and funding, we will move forward with finalizing our proposal to limit our approval of the existing SIP. Although we received comments questioning our authority to limit approval as proposed, using our general rulemaking and CAA section 110(k)(6) authorities, we are not responding to those comments at this time. We will address these comments in any final action we take to implement a limited approval.

In our proposed rule, we also noted that a handful of EPA-approved SIPs fail to include provisions that would apply PSD to GHG sources at the appropriate time. This is generally because these SIPs specifically list the pollutants subject to the SIP PSD program requirements, and do not include GHGs in that list, rather than include a definition of NSR regulated pollutant that mirrors the federal rule, or because the state otherwise interprets its regulations to limit which pollutants the state may regulate. At proposal, we indicated that we intended to take separate action to identify these SIPs, and to take regulatory action to correct this SIP deficiency.

We ask any state or local permitting agency that does not believe its existing SIP provides authority to issue PSD permits to GHG sources to notify the EPA Regional Administrator by letter, and to do so no later than August 2, 2010. This letter should indicate whether the state intends to undertake rulemaking to revise its rules to apply PSD to the GHG sources that will be covered under the applicability thresholds in this rulemaking, or alternatively, whether the state believes it has adequate authority through other means to issue federally-enforceable PSD permits to GHG sources consistent with this final rule. For any state that lacks the ability to issue PSD permits for GHG sources consistent with this final rule, we intend to undertake a separate action to issue a SIP call, under CAA section 110(k)(5). As appropriate, we may also impose a FIP through 40 CFR 52.21 to ensure that GHG sources will be permitted consistent with this final rule.

6. Title V Programs

Our final action also differs from the proposal in the specific regulatory mechanism by which we tailor the definition of “major source” for title V permit programs, but is a logical outgrowth of our proposed rule. EPA proposed to implement tailoring for GHGs under title V by excluding sources of GHGs from the general definition of “major source” under 40 CFR 70.2 and 71.2, and adding a separate definition of “major source” with tailored thresholds for sources of GHGs. In response to comments, particularly from states concerned with implementation of the proposed approach under state law, EPA is adopting an approach in the final rule that (1) amends the definition of “major source” by codifying EPA's longstanding interpretation that applicability for a “major stationary source” under CAA sections 501(2)(B) and 302(j) and 40 CFR 70.2 and 71.2 is triggered by sources of pollutants “subject to regulation,” and (2) adds a definition of “subject to regulation.” Further, we are delaying our action to move forward with limiting our previous approval of existing state part 70 programs.

We are finalizing this alternative approach to address concerns similar to those we received with respect to state implementation of SIP-approved PSD programs. Specifically, we received comments that the mechanism we proposed would not address the significant administrative and programmatic considerations associated with permitting GHGs under title V, because the 100 tpy threshold would continue to apply as a matter of state law. Commenters stated that states would need to undertake a regulatory and/or legislative process to change the threshold in their state laws which they could not complete before the laws would otherwise require issuance of operating permits to GHG sources.

After considering the commenters' concerns, we are finalizing an approach designed to address the state law concerns for states. As a result, it is unnecessary to move forward at this time with our proposed approach to limit approval of existing part 70 programs in many states.

EPA's approach involves the interrelationship of terms within the part 70 definition of “major source” in title V and EPA's implementing regulations, and EPA's historical practice of interpreting the term “any air pollutant” in the “major stationary source” component of that definition. EPA believes the approach in the final rule will allow many states to adopt the final rule through interpretation of existing state laws. Specifically, paragraph (3) within the definition of “major source” found in 40 CFR 70.2 and 71.2 defines a major source as “a major stationary source of air pollutants, as defined in section 302 of the Act, that directly emits or has the potential to emit, 100 tpy or more of any air pollutant * * *.” The EPA previously articulated the Agency's interpretation that the regulatory and statutory definitions of “major source” under title V, including the term “any air pollutant,” applies to pollutants “subject to regulation.” Memorandum. EPA recently re-affirmed this position in EPA's Reconsideration of Interpretation of Regulations that Determine Pollutants Covered by Clean Air Act Permitting Programs. 75 FR 17704, 17022-23 (April 2, 2010) (Interpretive Memo reconsideration final action).

Accordingly, under our long-established policy, states historically have interpreted the term “any air pollutant” under the title V definition of “major source” to mean any pollutant “subject to regulation” under the Act. Thus, as a matter of established interpretation, EPA and states effectively read the definition of “major source” under title V to include a source “* * * that directly emits or has the potential to emit, 100 tpy or more of any air pollutant subject to regulation under the Act” (emphasis added). By amending our regulations to expressly include and define “subject to regulation” to implement our tailoring for GHGs under title V, we are seeking to enable states to adopt and implement this approach through a continued interpretation of the phrase “any air pollutant” within the “major source” definition, without the need for changes to state regulations or statutes. States may be able to track EPA's approach to tailoring for GHG permitting without regulatory or statutory changes, for example, where a state has taken the position, or determines now, that the state's interpretation of “major source,” “subject to regulation” and/or “any air pollutant” is intended to track EPA's interpretation.

Thus, EPA is adding the phrase “subject to regulation” to the definition of “major source” under 40 CFR 70.2 and 71.2. EPA is also adding to these regulations a definition of “subject to regulation.” Under the part 70 and part 71 regulatory changes adopted, the term “subject to regulation,” for purposes of the definition of “major source,” has two components. The first component codifies the general approach EPA recently articulated in the “Reconsideration of Interpretation of Regulations That Determine Pollutants Covered by Clean Air Act Permitting.” 75 FR 17704. Under this first component, a pollutant “subject to regulation” is defined to mean a pollutant subject to either a provision in the CAA or regulation adopted by EPA under the CAA that requires actual control of emissions of that pollutant and that has taken effect under the CAA. See id. at 17022-23; Wegman Memorandum at 4-5. To address tailoring for GHGs, EPA includes a second component of the definition of “subject to regulation,” specifying that GHGs are not subject to regulation for purposes of defining a major source, unless as of July 1, 2011, the emissions of GHGs are from a source emitting or having the potential to emit 100,000 tpy of GHGs on a CO₂ e basis.

As explained previously, we find no substantive difference between the alternative mechanisms for implementing GHG tailoring in the final rule. Whether we add GHG thresholds directly to the definition of “major source” (as we proposed), or alternatively, expressly add and define the term “subject to regulation,” both approaches revise the definition of “major source” to implement the Tailoring Rule. Accordingly, we adopt the later approach to facilitate state implementation of the final rule through an interpretation of existing state part 70 programs. Similar to our explanation previously for PSD, while we adopt the “subject to regulation” mechanism for implementing GHG tailoring in the final rule, the thrust of our rulemaking is to apply the title V definition of “major source”—which includes the statutory term, “major stationary source”—to GHG sources by treating only GHG sources that emit at levels above the Steps 1 and 2 thresholds as meeting that definition. Further, we believe that our action may reasonably be construed to revise any of several terms in that definition, including (1) The numerical thresholds, as we proposed; (2) the term “any air pollutant,” (3) the term “a major stationary source,” (4) the term “subject to regulation,” which, as discussed previously, our regulations graft into the definition of “major source.” We believe that the specific choice of which term constitutes the legal mechanism does not have a substantive legal effect because each mechanism involves one of the components of the regulatory term “major source”—which embodies the meaning of the statutory term, “major source”—and it is that term that we are interpreting to tailor title V applicability for GHG-emitting sources. Thus, while the “subject to regulation” mechanism facilitates expeditious implementation by states, and we are therefore revising our regulations to adopt this approach, we otherwise find no substantive difference between the alternative mechanisms we may use to finalize the proposed rule.

Further, similar to our revised approach for addressing state SIP-approved PSD programs, we are delaying our action to limit our previous approvals for state part 70 operating permit programs. In our proposed rule, we explained our concern that states lack adequate personnel and resources to carry out part 70 operating permit programs for GHG sources that emit or have the potential to emit 100 tpy of GHGs. Accordingly, we proposed to use our general rulemaking authority under section 301(a) of the CAA and APA section 553 to limit our prior approval of state operating permit programs. This limited approval action would have had the effect of applying CAA permitting requirements only to sources that exceed the permitting thresholds established in this rule for the phase-in, because only those sources would be covered by the federally approved part 70 programs. 74 FR 55345. As discussed previously, we are proceeding with a slightly revised approach to address concerns similar to those raised with our proposed approach for addressing SIP-approved PSD permit programs. Because we now recognize that, like the PSD program, many states will be able to implement the final rules without the need to revise their existing part 70 operating permit programs, we are delaying further action on our proposal to limit approval of existing part 70 programs until we better understand how permitting authorities will implement our final rule.

In addition to the information requested previously on SIP-approved PSD permit programs, we ask each state to submit a letter to the appropriate EPA Regional Administrator no later than August 2, 2010 detailing the state's plan for permitting of GHG sources under the state's part 70 program. In that letter, states should explain whether they will adopt an interpretation of the terms “major source” or any of its component terms—“a major stationary source,” “any air pollutant,” or “subject to regulation,” or the numerical thresholds—that is consistent with EPA's regulatory interpretation of these terms as codified at 40 CFR 70.2, and whether the state intends to adopt the interpretation without undertaking a regulatory or legislative process. This approach may be available, for example, where a state has taken the position, or determines now, that the state's interpretation of these terms is intended to track EPA's interpretation, resulting in title V permitting for sources of GHGs as described in EPA's regulations adopted in this rule. If a state must revise its title V regulations or statutes to implement the interpretation, we ask that it provides an estimate of the time to adopt final rules or statutes in its letter to the Regional Administrator. If a state chooses not to (or cannot) adopt our interpretation, the letter should address whether the state has alternative authority to implement the GHG tailoring approach or some other approach that is at least as stringent, but which also addresses the expected shortfalls in personnel and funding and delays in permitting that would exist if the state carried out permitting under part 70 program thresholds lower than those adopted by EPA in this final rule. For any state that is unable or unwilling to adopt the permitting thresholds in the final rules, and otherwise is unable to demonstrate adequate personnel and funding, EPA will move forward with finalizing a narrowed limited approval of the state's existing part 70 program. If we do so, then we will respond in that action to comments on our proposal.

In our proposed rule, we also noted that a handful of part 70 operating permit programs may include provisions that would not require operating permits for any source of GHG emissions because, for example, the programs may apply only to pollutants specifically identified in the program provisions, and the provisions do not specifically identify GHGs. In these cases, states may be unable to interpret their regulatory provisions to interpret the term “any pollutant” to include pollutants “subject to regulation.” We indicated that we intended to take separate action to identify these programs, and to take regulatory action to correct this deficiency. Accordingly, we ask any state or local permitting agency that does not believe its existing part 70 regulations convey authority to issue title V permits to GHG sources consistent with the final rule to notify the EPA Regional Administrator by letter as to whether the state intends to undertake rulemaking to revise its rules consistent with these applicability thresholds. This notification should be done no later than the previously described letter regarding adoption of the Tailoring Rule, and could be combined with similar notifications we request regarding the PSD program. We intend to undertake a separate regulatory action to address part 70 programs that lack the ability to issue operating permits for GHG sources consistent with the final rule. We also intend to use our federal title V authority to ensure that GHG sources will be permitted consistent with the final rule.

D. Rationale for Treatment of GHGs for Title V Permit Fees

The title V program requires permitting authorities to collect fees “sufficient to cover all reasonable (direct and indirect) costs required to develop and administer [title V] programs.” To meet this requirement, permitting authorities either collect an amount not less than a minimum amount specified in our rules (known as the “presumptive minimum”), or may collect a different amount (usually less than the presumptive minimum). We did not propose to change the title V fee regulations in our notice of proposed rulemaking for this action, nor did we propose to require new fee demonstrations when title V programs begin to address GHGs. However, we did recommend that each state, local or tribal program review its resource needs for GHGs and determine if the existing fee approaches will be adequate. If those approaches will not be adequate, we suggested that states should be proactive in raising fees to cover the direct and indirect costs of the program or develop other alternative approaches to meet the shortfall. We are retaining this proposed approach, and are not changing our fee regulations as part of this final action establishing Steps 1 and 2 of the phase-in. However, we are offering some additional clarification of our fee approach during these steps in response to comments we received on this issue. Additional discussion of fees will be included as part of subsequent actions establishing Step 3 and beyond.

A few state commenters suggested that EPA should modify part 70 to adopt a presumptive minimum fee (or range for such fee) for GHGs, some of whom suggested that current fees may be insufficient to cover the costs of their program. It is important to clarify that altering the presumptive minimum would only affect those states that chose to charge the presumptive minimum fee to sources. Most states—including some of the commenters asking EPA to raise the presumptive minimum—collect a lower amount that is not based on the presumptive minimum, but rather, relies on another fee schedule that it developed and EPA approved as adequate to cover costs. Therefore, it is useful to first discuss our approach to programs that have fee schedules resulting in a different amount before discussing our approach to the presumptive minimum.

Because of the added GHG title V permitting workload described elsewhere in this notice, any state that will not, under its current fee structure, collect fees adequate to fund the permitting of GHG sources must alter its fee structure in order to meet the requirement that fees be adequate to cover costs. Changes may not be required in every instance; circumstances will vary from state to state. For example, a state may see increases in revenue from newly-covered sources (based on emissions of pollutants already subject to fees) that fully cover the state's increased costs, or a state may be over-collecting fees now and could use the surplus to offset the increased costs. Nonetheless, in many cases, we think states will need to adjust their fee structures to cover the costs of GHG permitting in order to meet the requirements of the Act and our regulations.

For this reason, although we are not calling for new fee demonstrations at this time, we plan to closely monitor state title V programs during the first two steps of the Tailoring Rule to ensure that the added workload from incorporating GHGs into the permit program does not result in fee shortfalls that imperil operating permit program implementation and enforcement, whatever the basis of the states' fee schedule. As described in the proposal, such fee oversight by EPA may involve fee audits under the authority of 40 CFR 70.9(b)(5) to ensure that adequate fees are collected in the aggregate to cover program costs, with emphasis on whether the additional GHG workload is being appropriately funded. Also, EPA retains the ability to initiate a program revision under 40 CFR 70.4(i)(3) or issue a notice of deficiency under the process described in 40 CFR 70.10(b) to address fee adequacy issues, which may be uncovered during a fee audit. By relying on existing oversight measures, we are ensuring that the fee requirements are met with a minimum of disruption to existing programs at a time when they will already be facing significant challenges related to GHG permitting.

Turning to the minority of states that do use the presumptive minimum, we did not propose to change the presumptive minimum calculation method to account for GHGs. Currently under the statute and our rules, the presumptive minimum is based on a subset of air pollutants (i.e., VOCs, NAAQS pollutants except for CO, and pollutants regulated under the NSPS and MACT standards promulgated under sections 111 and 112 of the Act, respectively) that does not include GHGs. The amount is specified on a per-ton basis and changes with inflation (it is currently set at $43.75/ton), but does not apply to emissions over 4,000 tpy of a given pollutant from a given source. We noted several difficulties in applying the presumptive minimum to GHG, including the large amounts of GHG emissions relative to other pollutants and the need for better data to establish a GHG-specific amount. Noting that GHGs are not currently included in the Act's list of pollutants to which the presumptive fee applies, we also invited comment on whether we should raise the fee for listed pollutants to cover the added cost of GHG permitting.

A few state commenters asked us to set a presumptive fee for GHGs, which we take to mean we should add GHGs to the list of pollutants to which a presumptive fee would apply. However, many commenters noted that the current presumptive minimum fee is unreasonable for GHGs because GHGs are emitted in greater quantities than the pollutants currently subject to presumptive fees, which would result in excessive fees. These commenters believe that EPA needs to limit the fees that states can charge for GHGs. Moreover, one commenter read the statute to prohibit us from listing GHGs in the presumptive fee calculation in the first place. Several commenters disagreed with the idea of increasing the presumptive fee for other pollutants to cover the cost of regulating GHGs, some of whom believed that this would unfairly punish existing sources or would bring in no new revenue from sources triggering title V for the first time.

After considering these comments, we remain disinclined, as we were at proposal, to change the presumptive fee calculation regulations. While there is some support for changing the regulations, the comments confirm the challenges in doing so. While we expressly rejected charging the full presumptive cost per ton amount for GHG, we also did not propose language to establish a different amount just for GHG, to establish whether a different tpy cap would apply, or to assess whether GHGs could even be added to the list. Thus, many commenters were very concerned about whether the full $43.75 or the 4000 tpy cap would apply to GHG if we listed it as a regulated pollutant for fee purposes. Furthermore, we noted at proposal, and commenters did not disagree, that more data would be needed to establish the appropriate basis for the GHG presumptive minimum. We are not taking a final position in this notice on whether the statute is amenable to including GHG in the presumptive fee calculation currently, but these comments illustrate some of the difficulties of such an approach.

At the same time, we are not increasing the presumptive minimum for other pollutants already included in the fee calculation. We disagree with the commenter who said such an approach would bring in no new revenue from newly-subject sources. Many of the newly-subject sources would emit already-included pollutants. If new revenue from these pollutants were insufficient, and because the Act does not specify how the shortfall must be addressed, the amount of any projected shortfall could be made up by increasing fees on these pollutants. In fact, the projected shortfall could be addressed without having to inventory GHG emissions from title V sources, since the emissions of already-included pollutants are well-known. We also note that, although some commenters are concerned that failing to assess fees for GHGs directly would be unfair, the statute does not provide that the presumptive fee be proportional to each type of pollutant or be proportionally allocated to all sources. Rather, the presumptive fee approach provides a backstop for states that do not wish to adopt a more tailored approach. Nonetheless, we have decided not to increase the presumptive fee amounts for other pollutants because we lack information about the extent to which shortfalls exist due to GHG permitting, and which mix of sources and fees is appropriate for addressing any such shortfall in a state. This decision also provides greater flexibility to states and minimizes disruption to existing programs.

We note that, contrary to the statements of some commenters, the CAA provisions allowing for a presumptive fee calculation do not override the basic requirement that fees be adequate to cover costs. As noted previously, we expect states to see a revenue increase from emissions of listed pollutants at newly-major sources for GHGs, and it is also possible that the presumptive minimum may currently be resulting in over-collection of fees in a state. Thus, a state continuing to use the presumptive minimum may not have a shortfall. However, if states using the presumptive minimum approach do have a revenue shortfall due to GHG permitting, the statute requires the shortfall to be addressed. The EPA has had, and will continue to have, the ability to require states that use the presumptive minimum to increase their fees if the presumptive minimum results in a revenue shortfall that imperils operating permit program implementation and enforcement. Thus, although we are not changing the presumptive minimum in our regulations, we plan to follow the same oversight approach for states using the presumptive minimum as for those collecting less based on a resource demonstration. As described previously, this approach may involve fee audits with emphasis on whether the additional GHG workload is being appropriately funded, and other appropriate follow-up.

Consistent with our proposal, EPA is not modifying its own part 71 fee structure (which closely mirrors the presumptive minimum) in order to charge an additional fee for GHGs. EPA must revise its fee schedule if the schedule does not reflect the costs of program administration. We have not determined that the existing fee structure will be inadequate to fund the part 71 programs costs during the first two phases of permitting GHGs as set forth in this action. However, we are required to review the fee schedule every 2 years, and make changes to the fee schedule as necessary to reflect permit program costs. 40 CFR 71.9(n)(2). Thus we will continue to examine the increases in part 71 burden due to GHG permitting, the current revenue collection, and the increases in revenue from newly-subject part 71 sources, and will adjust the part 71 fee approach accordingly.

Finally, several state and industry commenters asked EPA to provide guidance and recommendations for an appropriate GHG fee structure. We note that title V grants permitting authorities considerable discretion in charging fees to sources for title V purposes and does not require or prohibit fees specifically for GHGs, provided the states collect fees in the aggregate that are sufficient to cover all the direct and indirect program costs. In responding to requests for guidance, we do not wish to limit state discretion. For example, some commenters suggest that EPA prohibit emissions-based fees for GHGs or cap the amount that can be collected, while others suggest we provide a range of acceptable fees. We are concerned that, given the wide variety of fee approaches that states now take, providing specific guidance may be disruptive, rather than helpful, to states.

On the other hand, we recognize that it will initially be difficult for states to establish an appropriate emissions fee for GHGs. As noted previously, there are currently limited data available for establishing such a fee, and, due to the large quantities of GHG emissions, such a fee may only amount to a few cents per ton. At the same time, as noted in the proposal, a number similar to that used for other pollutants (e.g., the presumptive minimum of approximately $45/ton of GHG) would be inappropriate because it would likely result in huge over-collection. Because of this challenge, we note that 40 CFR 70.9(b)(3) allows the state to charge fees to individual sources on any basis (e.g., emission fee, application fee, service-based fees, or others, in any combination). While most states use emissions-based fees, there is merit to considering all the available fee bases to address increased GHG workload, including approaches that do not require a GHG emissions inventory for fee purposes. For example, where it is possible to estimate a revenue shortfall as a percentage of fee revenue, it may be appropriate to simply attach a percentage-based surcharge to each source's fee to match that shortfall. Similarly, where the shortfall could be estimated as a total dollar amount, a flat surcharge could be added to each source's fee to address the shortfall.

These suggestions should not be read to indicate that EPA prefers any particular approach, or that EPA rejects a cost per ton approach. Rather, they illustrate that it is possible to address a revenue shortfall without establishing a GHG per-ton fee. While the EPA is declining to recommend specific approaches in this preamble, we are committed to assisting states in implementing the fee requirements for GHG. Therefore, we will work with any state that requests assistance from EPA in developing a workable fee approach.

E. Other Actions and Issues

1. Permit Streamlining Techniques

In our proposal, we stated that while we were phasing-in permitting requirements, we would make a concerted effort to assess and implement streamlining options, tools, and guidance to reduce the costs to sources and permitting authorities of GHG permitting. We recognized that the development and implementation of these techniques should be an integral part of our strategy during the phase-in period, and we stated that we would undertake as many streamlining actions as possible, as quickly as possible. We discussed several streamlining techniques in particular, including: (1) Defining PTE for various source categories, (2) establishing emission limits for various source categories that constitute presumptive BACT, (3) establishing procedures for use of general permits and permits-by-rule, (4) establishing procedures for electronic permitting, and (5) establishing “lean” techniques for permit process improvements. The first three of these approaches have the potential to have the greatest impact in reducing the numbers of sources subject to PSD or title V (the definition of PTE) or of reducing permitting costs (presumptive BACT and general permits or permits-by-rule).

In our proposal, we also described the timing for development and implementation of these streamlining techniques. We explained that each of the first three techniques would generally take 3-4 years to develop and implement, and therefore would be of limited use in the near-term. This time frame is necessary because EPA will first need to collect and analyze small source data that we do not currently have—because these are sources that EPA has not traditionally regulated—in order to assess which of these techniques are viable or effective for such sources. In general, EPA will then need to conduct notice-and-comment rulemaking to establish the approaches, and that rulemaking will need to address various legal and policy aspects of these approaches. After that, the permitting authorities will need some time to adopt the streamlining techniques as part of their permitting programs.

We received several comments on streamlining techniques. In general, the comments indicate widespread support for our pursuit of streamlining approaches, but some commenters were concerned that one or more of EPA's identified streamlining options were complex, vague, ineffective, and questionable legally. Noting our proposal to phase in permitting, in part to allow more time to develop streamlining options for smaller sources, some commenters suggested that we should delay permitting for larger sources for the same reasons. We disagree. Such a delay is not justified under our legal basis for this rule. While implementation of Steps 1 and 2—which will cover larger sources—will pose implementation challenges, and some of the streamlining tools could assist with meeting these challenges, we have assessed the burdens associated with GHG permitting and have established a phase-in schedule that represents a manageable workload, even in the absence of streamlining techniques. On the other hand, we do agree with these commenters that, absent streamlining, applying PSD and title V requirements to the much larger number of small sources would lead to absurd results and administrative impossibility. The sources for whom the phase-in delays applicability are precisely the sources that have the greatest need for streamlining measures, and thus the greatest need for a deferral while we develop and implement streamlining options.

In addition, commenters generally echoed many of our concerns about why it will take time to put these measures in place, and no commenter presented any information to suggest that our 3-4 year estimate for the PTE, presumptive BACT, and general permit measures was invalid.

For these three techniques, we continue to believe that as we noted at proposal, we will require collection of significant category-specific data for source and emission unit types that have heretofore generally not been regulated by the CAA (e.g., furnaces, water heaters, etc.), which could take up to 1 year. Moreover, commenters had differences of opinion as to whether and how we should move forward on these approaches, and some raised policy and legal issues that we would likely want to explore through a notice and comment process in order to assess which of these measures are viable to pursue further. Even if a rulemaking were done expeditiously, it would likely require 1 year. Finally, unlike lean and electronic permitting, these approaches, once finalized by EPA, will likely require additional time of up to 2 years for states to adopt. Thus, it is clear that these approaches will not be in place in time to ease any burden prior to the planned rulemaking for Step 3.

Some commenters did observe that the fourth and fifth techniques, lean and electronic permitting, could, at least theoretically, become available sooner. However, these commenters also noted that successful design and implementation of these approaches will require implementation experience with GHG permitting that is not now available. We expect that for the lean and electronic permitting techniques, at least 1 year of implementation experience (of the type that we will gain starting in 2011) would be required, plus at least an additional year to extrapolate that experience to small sources and put these approaches into effect for small source permitting. Thus, we do not think the lean and electronic permitting would be in place before the beginning of 2013. Moreover, a handful of commenters questioned whether lean and/or electronic permitting would alleviate significant burden. Thus we are not able, at this time, to presume that these approaches will ease any burden prior to the planned rulemaking for Step 3.

It is also important to note that, as a practical matter, while these efforts to streamline the program for small sources are underway, EPA and states will also be devoting a significant amount of their permitting resources and expertise to implementing the PSD and title V programs for the GHG-emitting sources covered under Steps 1 and 2. We have established these steps in a manner that they will be feasible for EPA and state/local/tribal authorities, but even so, they will not only consume current permitting authority resources, but they will also require substantial additional resources. As a result, the efforts to develop and implement streamlining techniques will have to compete with the work necessary to administer existing programs. For example, during the remainder of 2010, as described elsewhere in this notice, EPA permitting program resources will, in addition to continuing to administer programs for non-GHG pollutants, be used to conduct at least the following GHG-related activities in addition to streamlining: (1) Develop BACT and other information and guidance for implementing programs for sources covered by Step 1, followed by additional guidance and information for sources covered by Step 2; (2) review and act on information we receive regarding state adoption of GHG permitting requirements, which may entail narrowing of previous SIP approvals or processing of other programmatic revisions; and (3) propose and finalize measures to address programs with deficiencies in GHG coverage. As the beginning of Step 2 nears, we will also begin to receive and process the first applications for permits that will incorporate GHG requirements (i.e., those that will be issued after January 2, 2011). States seeking to implement streamlining approaches will face similar competition for permitting resources.

These time frames and resource considerations for streamlining confirm the approach to phase-in that we are taking in this rule. First and foremost, they make clear that it will not be possible to have streamlining measures in place in time for either Step 1 or Step 2. Therefore our selection of threshold for those steps is not built on assumptions that streamlining will remove some or all of the burden during those steps.

Second, they make clear that, while no significant streamlining can be in place by the time we must begin to develop the Step 3 rule (i.e., latter half of 2011, to promulgate by July 2012, effective July 2013), it is likely that by that time EPA and states will have had an opportunity to gain implementation experience that could serve as the basis for beginning to implement streamlining techniques that do not require rulemaking or state adoption (e.g., lean and electronic permitting). It is also likely that we will have had an opportunity to gather technical information—which we have already begun to gather—for certain source and emissions unit categories that would be necessary to support proposal of PTE or presumptive BACT approaches for those categories. We expect that the Step 3 rulemaking will provide an opportunity for us to use that experience and data to begin to propose streamlining approaches that need notice and comment rulemaking. We can also begin to take into account any burden reductions from possible early streamlining efforts—that is, through lean and electronic permitting—in the establishment of Step 3.

Third, it is clear that the potential availability of streamlining measures does not call into question our decision that in no event will we broaden PSD and title V applicability to cover GHG-emitting sources below the 50,000 tpy CO₂ e level prior to July 2016, as discussed elsewhere. EPA cannot now predict the resources that will be required to implement PSD and title V programs for GHG-emitting sources once various streamlining techniques are ultimately completed. This is uncertain not only because we need data and implementation experience with GHG permitting during Steps 1, 2, and 3 that we can apply to estimates for small sources, but also because, as comments indicate, there is a broad range of legal and policy issues to consider in crafting the streamlining approaches we ultimately adopt. We have presented an initial assessment of options and obtained views of commenters both supporting and opposing them, and it is the result of these future actions, whose outcomes are uncertain at this time, that will ultimately determine the extent to which streamlining approaches will allow for the administration of PSD and title V programs for numerous small sources. Thus, while we are optimistic that we can craft workable, common-sense solutions, we nonetheless, believe it is important to preserve our small source exclusion until we have not only had time to put the streamlining approaches in place, but also have had time to assess the burdens that remain, before we bring in additional sources below the 50,000 tpy CO₂ e levels. We believe that the 6-year timeframe will require a sustained intensive effort by EPA and states to develop, adopt, and implement streamlining techniques, and will require EPA to then evaluate those techniques and complete a rulemaking concerning PSD and title V applicability to small-sources based on that evaluation. In this manner, the 6-year period will give us the necessary time to make the best decisions about the actions we should take beyond Step 3.

While comments make clear that there are issues to be addressed, nothing in the comments has persuaded us that we should abandon our streamlining efforts. To the contrary, the strong support for these efforts shown by many commenters reinforces our intention, as stated at proposal, to move forward with these approaches as an integral part of our phase-in approach. Moreover, notwithstanding the competition for GHG permitting resources and expertise, we believe it is critical that we move forward expeditiously. As noted previously, we are already taking a first step by initiating permitting for larger sources, beginning January 2011, that will begin to provide valuable implementation experience. This experience can be useful in allowing states to begin implementing early streamlining measures, like lean and electronic permitting, which do not require EPA action. We have also already begun, and will continue, developing data necessary to support possible rulemakings addressing approaches such as PTE, presumptive BACT, and/or general permits. We expect to be able to use these data to support possible rulemakings on these topics, as appropriate, at about the same time as our Step 3 rulemaking. There may also be available streamlining options that were not described in our proposal that warrant further consideration. Because of the uncertainty surrounding such approaches, we are not committing to finalize rules on any particular approach, but we do plan to explore all streamlining options as expeditiously as possible, beginning immediately and proceeding throughout the phase-in period, and we encourage permitting authorities to do the same. We commit to consider a wide array of possible streamlining measures, and we commit to propose and take comment on, in the Step 3 rulemaking, a set of those measures that we determine are viable to pursue further.

2. Guidance for BACT Determinations

The CAA requires that a PSD permit contain, among other things, emissions limits based on the BACT for each pollutant subject to regulation under the Act emitted from the source that triggers PSD. 42 U.S.C. 7475(a)(4); 42 U.S.C. 7479(3). BACT is defined as follows:

(3) The term “best available control technology” means an emission limitation based on the maximum degree of reduction of each pollutant subject to regulation under this Act emitted from or which results from any major emitting facility, which the permitting authority, on a case-by-case basis, taking into account energy, environmental, and economic impacts and other costs, determines is achievable for such facility through application of production processes and available methods, systems, and techniques, including fuel cleaning, clean fuels, or treatment or innovative fuel combustion techniques for control of each such pollutant. In no event shall application of “best available control technology” result in emissions of any pollutants which will exceed the emissions allowed by any applicable standard established pursuant to section 111 or 112 of this Act. Emissions from any source utilizing clean fuels, or any other means, to comply with this paragraph shall not be allowed to increase above levels that would have been required under this paragraph as it existed prior to enactment of the Clean Air Act Amendments of 1990.

42 U.S.C. 7479(3).

Thus, the BACT process is designed to determine the most effective control strategies achievable in each instance, considering energy, environmental, and economic impacts. However, the case-by-case nature of BACT, together with the range of factors and technologies that must be considered, presents a challenge in determining BACT for newly regulated pollutants. When a new pollutant is regulated, the first permit applicants and permitting authorities that are faced with determining BACT for a new pollutant will likely need to invest more time and resources in gathering and analyzing information necessary to make an assessment of BACT under the statutory criteria. Once the PSD permitting program matures with respect to the new pollutant, successive BACT analyses will establish precedents that can inform subsequent BACT determinations. While the BACT provisions clearly contemplate that the permitting authority evaluate control strategies on a case-by-case basis, EPA recognizes the need to develop and share policy guidance and technical information for sources and permitting authorities as they begin to permit sources of newly regulated pollutants, such as GHGs. When applied in a practical manner, this additional EPA guidance and technical information should reduce time and resource needs when evaluating BACT for newly regulated pollutants.

As described in the proposed Tailoring Rule, EPA intends to compile and make available technical and background information on GHG emission factors, control technologies and measures, and measurement and monitoring methodologies for key GHG source categories. We expressed our intent to work closely with stakeholders in developing this supporting information and to ensure this information is available in sufficient time to assist permitting agencies in their BACT determinations. The proposal took comment on what other types of support or assistance EPA can provide to initially help air pollution control agencies with the permitting of GHGs.

Commenters on the proposed Tailoring Rule generally supported EPA providing technical information and policy guidance for sources of GHGs. Several commenters specifically requested guidance to clarify GHG-related issues, such as how to compute CO₂ e emissions, how to evaluate emissions of CO₂ from biomass fuel, and whether an air quality analysis will be required for GHGs. Additionally, commenters requested that EPA issue “white papers” and other tools that would provide information on a range of control technologies and measures for major stationary source categories, such as power plants, cement kilns, glass furnaces, and other sources. Many of these commenters further requested that EPA provide an opportunity for stakeholder input on the guidance, and a few commenters insisted that permitting for sources of GHGs should not begin prior to issuing final guidance.

Consistent with our commitment at proposal to involve all stakeholders in our guidance development, EPA called upon the CAAAC in September 2009, to provide assistance and recommendations for what types of guidance and technical information would be helpful. Specifically, our charge to the CAAAC was “* * * to discuss and identify the major issues and potential barriers to implementing the PSD Program under the CAA for greenhouse gases * * * [and] focus initially on the BACT requirement, including information and guidance that would be useful for EPA to provide concerning the technical, economic, and environmental performance characteristics of potential BACT options.” This charge also requested the CAAAC to “identify and discuss approaches to enable state and local permitting authorities to apply the BACT criteria in a consistent, practical and efficient manner.”

At its October 6, 2009 meeting, the CAAAC established a Climate Change Work Group, made up of 35 representatives from a variety of industries, state and local governments, and environmental and public health non-profit organizations, organized under CAAAC's Permits, New Source Review and Toxics Subcommittee. The Work Group initially focused its attention on the procedure for evaluating BACT and decided that the process and criteria for determining BACT for criteria pollutants represented a workable and acceptable framework for GHGs. The Work Group also recommended a second phase, in which the Work Group would consider member proposals regarding possible alternative or supplementary approaches to applying the PSD program to GHG sources.

In February 2010, the CAAAC completed work on the first phase of its effort and sent EPA a list of recommendations that highlighted areas of the BACT determination process that are in need of technical and policy guidance. For more information, see the Interim Phase I Report on Issues related to BACT for GHGs, February 3, 2010 that is located in the public docket for this rulemaking and at http://www.epa.gov/air/caaac/climate/2010_02_InterimPhaseIReport.pdf. In response, we are working on a number of fronts to develop technical information, guidance, and training to assist states in permitting large stationary sources of GHGs, including identifying GHG control measures for different industries. EPA is currently working with states on technical information and data needs related to BACT determinations for GHGs. This includes developing the EPA Office of Research and Development GHG Mitigation Strategies Database, enhancing the RACT/BACT/LAER Clearinghouse to include GHG-specific fields, and preparing technical information on sector-based GHG control measures. Also, EPA is actively developing BACT policy guidance for GHGs that will undergo notice and comment and will culminate in training courses for state, local, and tribal permitting authorities. The results of all of these efforts will roll out over the remainder of 2010. EPA currently awaits the Work Group's recommendations from its second phase of deliberations, which is underway as of the date of this notice.

EPA does not agree with some commenters' suggestion that EPA should delay permitting of any sources until final BACT guidance is issued. As discussed in the final action on reconsideration of the Interpretive Memo, delaying the application of BACT to enable the development of guidance or control strategies is not consistent with the BACT requirements. 63 FR 17008. Furthermore, as just described, EPA expects such a delay to be unnecessary because EPA will soon begin providing technical information to inform BACT decisions, and will continue to provide additional guidance prior to the date that GHG permitting begins. However, even in the absence of such guidance, a delay would not be justified under the legal doctrines of “absurd results” and “administrative necessity.” While implementation of the BACT requirement during Steps 1 and 2 will pose implementation challenges, EPA has assessed the burden associated with GHG permitting with consideration given to these challenges, and has established a phase-in schedule that represents a manageable workload.

Thus, while BACT will remain a case-by-case assessment, as it always has been under the PSD program, EPA is confident that this guidance development effort will help support a smooth transition to permitting emissions of GHGs. Furthermore, EPA will continue to work to provide the most updated information and support tools to allow permitting authorities to share and access the most updated information on GHG BACT determinations as they are made once permitting of GHGs begins. EPA remains committed to involving stakeholders in the upcoming efforts to develop guidance to help permitting authorities in making BACT determinations for sources of GHGs.

3. Requests for Higher Category-Specific Thresholds or Exemptions From Applicability

Although we did not propose any categorical exemptions, many commenters requested exemptions from major source and major modification applicability determinations under title V and PSD for certain types of GHG-emitting sources or certain types of GHG emissions as follows:

Source Categories. Many commenters requested various exemptions or exclusions from source applicability for GHGs under both PSD and title V permitting, either during the phase-in period or permanently, citing anticipated burdens, societal costs, and differences in emission characteristics. Commenters representing non-traditional sources or source categories (sources that have not historically been required to get permits) requested exemptions from permitting based on GHG emissions, including agricultural sources, residential sources, and small businesses. In general, these commenters sometimes, but not always, cited “absurd results” and “administrative necessity” arguments in their exemption requests.

Several commenters from sectors that consume a great deal of energy in their industrial processes and that are subject to international competitiveness, such as aluminum, steel, cement, glass, pulp and paper, and other manufacturers, requested that they be exempt from permitting under this final rule. These commenters state that we have not carefully considered the environmental and economic consequences of this action because if we had, we would have exempted them for several reasons, including (1) other countries typically exempt similar sources from GHG cap and trade programs because the industries are making significant energy efficiency improvements even in the absence of GHG regulation, and (2) permitting such sources may cause many facilities to move to countries that have less regulation or no regulation for GHGs.

Other industry groups cited unique characteristics of their emissions, or the quantities in which they are emitted, that they argued should justify exclusion or unique thresholds. Semiconductor production facilities asked for exemptions, arguing that combustion-related GHG emissions are different from their GHG emissions, which result from the use of high-GWP industrial gases, such as PFCs, with higher GWP values that are more likely to trigger permitting requirements at relatively low tpy values. One lime production commenter stated that EPA could encourage energy efficiency projects at its plants by excluding calcination and other process emissions, arguing that these emissions are a relatively small portion of the national inventory that will have no material effect on air quality and global warming. Another commenter requested that EPA exclude emissions from poultry production (natural bird respiration) from permitting consideration because the IPCC excludes them from its GHG emission estimates. Representatives of the landfill industry pointed to the relationship between current statutory thresholds that apply to their regulated emissions, primarily NMOC, and the equivalent amount of GHG emissions this corresponds to. They argued for a source-category specific threshold that is at least equivalent to their current NMOC threshold, or roughly 750,000 tpy CO₂ e according to their estimate.

Although the proposal for the Tailoring Rule generally addressed how the statutory requirements for major source applicability (100/250 tpy thresholds) could be phased in in ways that would offer relief to traditional and non-traditional sources, such as residences, farms, small business, and semiconductor manufacturers, it did so by establishing relatively high CO₂ e thresholds during the early implementation period and lowering the thresholds over time as streamlining mechanisms become available to reduce administrative burdens. We did not propose any permanent exemptions of any kind or temporary exemptions based on source category. Also, note that the proposal discussed energy efficiency, process efficiency improvements, recovery and beneficial use of process gases, and certain raw material and product changes in the context of short-term, low-cost means of achieving GHG emission reductions for small-scale stationary sources, but not in the context of exemptions.

As discussed previously, we are still considering whether permanent exemptions from the statute are justified for GHG permitting based on the “absurd results” legal doctrine. We do not have a sufficient basis to take final action at this time to promulgate any of the suggested exclusions on the grounds, described previously, suggested by the commenters. We note, however, that nothing in this rule forecloses the opportunities we may have to explore such options in the future. Therefore, we are taking no action in this rule on these various commenters' requests for exclusions.

Some commenters also recommended that we create exclusions for their particular source categories for the specific purpose of avoiding overwhelming permitting burdens. We did solicit comment on alternative approaches to burden relief in the proposal. Some commenters suggested that the “administrative necessity” or “absurd results” rationale, each of which would be based on extraordinary administrative burdens, could be used to create at least temporary exclusions that would allow more sources to escape permitting than what we proposed. However, commenters have not, to date, provided specific information about the costs and administrative burdens associated with permitting their source categories.

Regarding the specific concerns about the need for a small business exclusion, we note that the Office of Advocacy of the SBA made several recommendations on the proposal to address concerns about large numbers of small businesses becoming subject to the permit programs. For example they recommended that EPA adopt major source thresholds of 100,000 tpy and major modification thresholds of 50,000 tpy CO₂ e. They also recommended that we adopt an interpretation of the effective date of the LDVR to provide additional time to prepare. We took action consistent with the latter recommendation in the Interpretive Memo, and we are taking action consistent with the former recommendation in this rule (although the threshold for modifications we are adopting is higher, for reasons explained previously). We are finalizing Steps 1 and 2 using the threshold-based approach, which applies the various legal doctrines, in the context of the Chevron framework, in a way that effectively exempts all small sources during this part of the phase-in, while assuring the administrability of the permitting programs for the sources that remain subject to them. We anticipate that virtually all small businesses not already subject to PSD and title V would be excluded under this approach. Similarly, with respect to high GWP gases as discussed previously, we are maintaining the statutory mass-based threshold, and this should address commenters' concerns regarding the inclusion of those gases. Therefore, we reiterate that we are not finalizing any such exclusions in this rule and, as noted above, we are not taking final action in the commenters' requests for exclusions.

Concerning the comment that we did not take appropriate economic and environmental considerations into account for this rulemaking action, we disagree. The approach we finalize in this notice for Steps 1 and 2 minimizes economic burdens by limiting permitting to the largest GHG emission sources. We further note that the PSD program as applied to the sources that are covered in Steps 1 and 2 contains an express requirement to take energy, environmental, and economic considerations into account when making control technology (i.e., BACT) decisions and accordingly many of the concerns about control costs will be able to be accounted for in that analysis.

Biomass Combustion/Biogenic Emissions. Several commenters request that EPA exempt emissions from biogenic activities or biomass combustion or oxidation activities, including solid waste landfills, waste-to-energy projects, fermentation processes, combustion of renewable fuels, ethanol manufacturing, biodiesel production, and other alternative energy production that uses biomass feedstocks (e.g., crops or trees). For example, commenters urged that EPA exclude emissions from biomass combustion in determining the applicability of PSD to GHGs based on the notion that such combustion is “carbon neutral” (i.e., that combustion or oxidation of such materials would cause no net increase in GHG emissions on a lifecycle basis). Some commenters oppose the exemption of biogenic/biomass activities, claiming the lack of a valid scientific basis for treating these GHG emissions differently than other GHG emissions and expressing concern that we should not assume all biomass combustion is carbon neutral.

The proposed Tailoring Rule did not address this issue of exemptions for biomass combustion or biogenic emissions. We are mindful of the role that biomass or biogenic fuels and feedstocks could play in reducing anthropogenic GHG emissions, and we do not dispute the commenters' observations that many state, federal, and international rules and policies treat biogenic and fossil sources of CO₂ emissions differently. We note that EPA's technical support document for the endangerment finding final rule (Docket ID No. EPA-HQ-OAR-2009-0472-11292) states that “carbon dioxide has a very different life cycle compared to the other GHGs, which have well-defined lifetimes. Instead, unlike the other gases, CO₂ is not destroyed by chemical, photolytic, or other reaction mechanisms, but rather the carbon in CO₂ cycles between different reservoirs in the atmosphere, ocean, land vegetation, soils, and sediments. There are large exchanges between these reservoirs, which are approximately balanced such that the net source or sink is near zero.”

Nevertheless, we have determined that our application of the “absurd results,” “administrative necessity,” and one-step-at-a-time legal rationales that support this rule, which are based on the overwhelming permitting burdens described previously, does not provide sufficient basis to exclude emissions of CO₂ from biogenic sources in determining permitting applicability provisions at this time. This is because such an exclusion alone, while reducing burdens for some sources, would not address the overwhelming permitting burdens described above, and a threshold-based approach would still be needed. As noted above, we have not examined burdens with respect to specific categories and thus we have not analyzed the administrative burden of permitting projects that specifically involve biogenic CO₂ emissions taking account of the threshold-based approach, nor did the commenters provide information to demonstrate that an overwhelming permitting burden would still exist, justifying a temporary exclusion for biomass sources.

At the same time, the decision not to provide this type of an exclusion at this time does not foreclose EPA's ability to either (1) provide this type of an exclusion at a later time when we have additional information about overwhelming permitting burdens due to biomass sources, or (2) provide another type of exclusion or other treatment based on some other rationale. Although we do not take a final position here, we believe that some commenters' observations about a different treatment of biomass combustion warrant further exploration as a possible rationale. Therefore, although we did not propose any sort of permanent exclusion from PSD or title V applicability based on lifecycle considerations of biogenic CO₂, we plan to seek further comment on how we might address emissions of biogenic carbon dioxide under the PSD and title V programs through a future action, such as a separate Advance Notice of Proposed Rulemaking (ANPR). This action would seek comment on how to address biogenic carbon under PSD and title V, the legal and policy issues raised by options regarding implementation. We will provide an opportunity for public comment before adopting any final approach.

We further note that, while we are not promulgating an applicability exclusion for biogenic emissions and biomass fuels or feedstocks, there is flexibility to apply the existing regulations and policies regarding BACT in ways that take into account their lifecycle effects on GHG concentrations. This topic has already been explored by the CAAAC workgroup on BACT issues related to GHGs that recently provided recommendations to EPA. These recommendations are located in the public docket for this rulemaking and at http://www.epa.gov/air/caaac/climate/2010_02_InterimPhaseIReport.pdf. While that group was unable to come to a consensus on how biomass-based emissions should be treated, it provided us with information that we will consider as we issue guidance on BACT. As previously discussed, we plan to issue BACT guidance later this year, but are not doing so as part of this rulemaking. Without prejudging the outcome of our process to seek comment whether and how we might address emissions of biogenic carbon under the PSD and title V programs through a future action, this issue warrants further exploration in the BACT context as well, and we plan to fully explore it and take action if appropriate.

Fugitive Emissions. Numerous commenters believe that fugitive GHG emissions should be excluded from major source determinations, citing difficulties in measuring or estimating such emissions. Others believe EPA did not address fugitive emissions in the proposal and they ask for clarification of the treatment of fugitive GHGs in applicability determinations under PSD and title V. Some of these commenters state that EPA has not undertaken a rulemaking under CAA section 302(j) for any source category of fugitive GHGs, so they should not be included. Several commenters representing the solid waste disposal industry requested exemptions for fugitive emissions for landfills and waste-to-energy projects, pointing out that current practice under PSD is for fugitive emissions from certain landfills to not be counted toward major source determinations.

In the proposal, EPA did not offer any specific guidance or discuss exemptions for fugitive emissions of GHGs. Commenters did not suggest that a fugitive exemption would address the overwhelming permitting burdens described previously, or that it was necessary to specifically tailor GHG applicability through the use of a fugitive emissions exclusion for categories that would otherwise be required to include them.

We do agree with commenters who stated that we should clarify how to count fugitives in determining applicability under this rule. In response, we note that we are not taking final action with respect to commenters' request, and we are not finalizing any special rules for fugitive emissions related to GHG. Thus, EPA's rules related to the treatment of fugitives would apply. Regarding the comment that a CAA section 302(j) rulemaking is required before fugitive emissions may be counted, we disagree. As we read section 302(j), once EPA has established by rule that fugitive emissions are to be counted for a specific source category, nothing in section 302(j) requires EPA to conduct new rulemaking to allow for the counting of additional pollutants from that category. We read section 302(j) as imposing an obligation to determine if fugitive emission generally should be counted from a source or source category and not requiring that EPA list both source categories and relevant pollutants. Indeed, our practice in listing categories has not been to limit the pollutants to which the listing applies. Therefore, we are applying our existing rules and policies for fugitive emissions for GHG as we would any other pollutant.

Pollution Control Projects. Other commenters request exemptions for pollution control projects from PSD major modification requirements, particularly projects that increase the efficiency or thermal performance of a unit or facility, resulting in emission reductions on a pounds/megawatt-hour or production basis. The current PSD rules do not exclude pollution control projects from being considered a physical change or change in the method of operation that would—if it resulted in a significant net emissions increase—constitute a major modification, and the case law makes clear that we could adopt a permanent exclusion in the future. To the extent that the commenters seek an exclusion for pollution control projects that relies solely on “absurd results” or “administrative necessity” for reasons similar to those described previously for other requested exclusions, we take no action on this request in this rulemaking.

4. Transitional Issues Including Requests for Grandfathering

In the Tailoring Rule proposal, we did not discuss or specifically ask for comment on any provisions to address the transition from a permitting regime that does not incorporate GHGs to one that does, such as “grandfathering” provisions or similar approaches that would exempt previously issued permits or pending applications from having to incorporate requirements for GHGs. We nonetheless received several public comments that addressed a variety of transitional issues. One group of comments addresses situations prior to permit issuance where a PSD or title V application is either administratively complete or more generally being processed prior to the trigger date for GHG permitting (“in process” applications). Another group of comments addresses situations where a PSD or title V permit is issued prior to the GHG trigger date and the commenters request that the application and/or permit be exempt from any requirements for updates related to GHGs after permit issuance.

With respect to PSD, many commenters requested that we adopt a “grandfathering” approach to applicability to exempt projects that have administratively complete PSD or minor NSR permit applications pending when the GHG permitting requirements go into effect. Several commenters urged us to promulgate transition provisions (without specifically using the term “grandfathering”), pointing out that we have provided transition periods for revising pending PSD permits, in the past, when new PSD rules were issued (e.g., in late 1970s and 1980). These commenters assert that GHG requirements will cause more disruption than those previous rule changes. Several commenters asked that PSD applications be evaluated on the basis of the PSD requirements effective when the application is submitted and if submitted prior to the trigger date, then the application and permit would not need to address GHGs. Several commenters also asked that PSD sources with a valid permit that commences construction within 18 months of the trigger date not be required to seek a revised PSD permit for GHGs. Similarly, several commenters asked that PSD permits issued prior to the GHG trigger date not be required to be reopened only for the purpose of addressing GHG emissions. Additional commenters asked that we clarify that sources or projects not be required to obtain PSD permits if they obtained a determination that PSD did not apply (a “non-applicability” determination) prior to the GHG trigger date. Finally, many commenters also requested “grandfathering” for title V so that existing title V applications and permits do not need to be amended, revised, or resubmitted to address GHGs after they become “subject to regulation.” Other commenters asked that transition provisions for title V be provided in the final action that would be similar to those requested for PSD.

We partially addressed transitional issues for PSD permitting in our April 2, 2010 final action on reconsideration of the Interpretive Memo. 75 FR 17021. This action addressed the applicability of PSD permitting requirements for GHGs to pending PSD permit applications that were (or will be) submitted prior to January 2, 2011 based on emissions of pollutants other than GHGs. However, we have not yet addressed the questions raised by public comments concerning sources that obtain PSD permits, minor NSR permits, or determinations that no such permits are needed prior to the Step 1 period set forth in this rule. We have also not yet addressed questions about the applicability of PSD permitting requirements for sources that are not currently required to submit an application for a PSD permit but that could be required to do so in Step 2 of the phase-in established in this action. In addition, our April 2, 2010 action did not address transitional issues concerning the application of the title V provisions to GHGs.

a. Transition for PSD Permit Applications Pending When Step 1 Begins

In our action on April 2, 2010, EPA explained that the Agency did not see grounds to establish a transition provision for pending PSD permit applications because we had determined that PSD permitting requirements would not apply for GHGs for another 9 months. We explained that permit applications submitted prior to April 2, 2010 should in most cases be issued prior to January 2, 2011 and, thus, effectively have a transition period of 9 months to complete processing before PSD requirements become applicable to GHGs. We also observed that, in the case of any PSD permit application review that cannot otherwise be completed within the next 9 months based on the requirements for pollutants other than GHGs, it should be feasible for permitting authorities to begin incorporating GHG considerations into permit reviews in parallel with the completion of work on other pollutants without adding delay to permit processing. Additional discussion of EPA's reasons for not developing transition provisions for PSD permit applications that are pending on January 2, 2011 are provided in the April 2, 2010 notice. 75 FR 17021-22.

For these same reasons, we continue to feel that a transition period is not warranted to incorporate GHG requirements into any PSD permit applications that are pending when Step 1 of the permitting phase-in begins for those sources that would otherwise need to obtain a PSD permit based on emissions of pollutants other than GHGs. Thus, this action makes no change to the position we expressed on this particular issue in the April 2, 2010 notice. In this final rule on tailoring the PSD program to address GHGs, we have determined that the additional burden of incorporating GHG requirements into PSD permits for the sources already required to obtain such permits is manageable in the Step 1 period. Thus, this rule has added no additional requirements or limitations that would justify deferring the establishment of pollution controls for this category of GHG sources once PSD permitting requirements are initially triggered for GHGs.

While we do not provide for grandfathering of PSD applications, we do note that there are more than 7 months left before GHG BACT requirements will be triggered at anyway sources for projects that increase GHG emissions by more than 75,000 tpy CO₂ e and more than a year before the requirements would be triggered at sources solely because of emissions of GHGs (more than 100,000 tpy of CO₂ e). We intend to work constructively and affirmatively with permitting authorities to use this time to ensure expeditious processing of pending permits and to further assure that the triggering of BACT requirements at such sources will not result in adverse impacts on pending projects. We have separately described our plans to expeditiously issue GHG BACT guidance, but we understand that for pending projects that will be permitted soon after January 2, 2011, an opportunity for earlier engagement with EPA on BACT issues would be beneficial for permitting authorities to issue these permits without delay.

Therefore, following the issuance of this rule, we will contact permitting authorities that have pending PSD permit applications to identify those applications with a reasonable likelihood that final issuance will occur after January 2, 2011, and therefore will be required to contain GHG BACT limits. We will then work closely with those permit agencies to provide technical, legal, or policy assistance to help prepare BACT analysis and provide additional support as necessary to expedite permitting for those pending applications. Similarly, when EPA is the permitting authority, we will provide assistance to applicants with pending permits to ensure that GHG permitting decisions are made promptly, and that administrative processes move forward expeditiously.

b. PSD Permits Issued Prior to Step 1

EPA has not historically required PSD permits to be updated or reopened after they are issued in the absence of an action by the applicant to change the physical or operational characteristics of the source described in the permit application. EPA's PSD permitting regulations contain no provisions that address the modification or amendment of a PSD permit or require a PSD permit to be reopened or modified on the basis of new PSD permitting requirements that take effect after the final permit is issued. Since PSD permits are construction permits, EPA has not required updates to PSD permits in the same manner as is typically required for operating permits that incorporate a variety of applicable requirements (such as title V permits and National Pollutant Discharge Elimination System (NPDES) permits under the Clean Water Act). In addition, unlike operating permits, PSD permits are not required to be renewed. However, if construction under a PSD permit is not commenced in a timely manner or is discontinued for an extended period, a PSD permit may expire if an extension is not requested or justified. See 40 CFR 52.21(r)(2); 40 CFR 124.5(g).

With respect to the application of PSD permitting requirements for GHGs beginning on January 2, 2011, we do not see any cause to deviate from our historical practice of not requiring PSD permits to be reopened or amended to incorporate requirements that take effect after the permit is issued. Thus, we are not promulgating any new rules or requirements pertaining to PSD permits issued prior to Step 1 of the phase-in described in this rule. There is no mandatory requirement to reopen a previously issued PSD permit to incorporate GHG requirements that were not applicable at the time the permit was issued.

A major source that obtains a PSD permit prior to January 2, 2011 will not be required under EPA regulations to reopen or revise the PSD permit to address GHGs in order for such a source to begin or continue construction authorized under the permit. Our current PSD permitting regulations provide that “[n]o new major stationary source or major modification to which the requirements of paragraphs (j) through (r)(5) of this section apply shall begin actual construction without a permit that states the major stationary source or major modification will meet those requirements.” 40 CFR 51.166(a)(7)(iii); 40 CFR 52.21(a)(2)(iii). The term “begin actual construction” generally means “initiation of physical onsite construction activities on an emissions unit which are of a permanent nature” and includes activities such as “installation of building supports and foundations, laying underground pipework and construction of permanent storage structures.” 40 CFR 51.166(b)(11); 40 CFR 52.21(b)(11). A source that begins actual construction authorized under a PSD permit prior to January 2, 2011 will not be in violation of the prohibition described previously if it continues construction after that date. This portion of the regulation precludes only beginning construction without the appropriate preconstruction permit and does not require a permit to be updated to continue actual construction that has already begun.

Furthermore, a source that is authorized to construct under a PSD permit but has not yet begun actual construction on January 2, 2011 may still begin actual construction after that date without having to amend the previously-issued PSD permit to incorporate GHG requirements. Sections 51.166(a)(7)(iii) and 52.21(a)(2)(iii) require “a permit that states that the major stationary source or major modification will meet those requirement,” which refers to the “requirements in paragraphs (j) through (r)(5)” referenced earlier in those provisions. EPA construes this language to describe a permit that meets the requirements of paragraph (j) through (r)(5) that are in effect at the time the permit is issued. Permitting and licensing decisions of regulatory agencies must generally reflect the law in effect at the time the agency makes a final determination on a pending application. See Ziffrin v. United States, 318 U.S. 73, 78 (1943); State of Alabama v. EPA, 557 F.2d 1101, 1110 (5th Cir. 1977); In re: Dominion Energy Brayton Point, LLC, 12 E.A.D. 490, 614-616 (EAB 2006); In re Phelps Dodge Corp., 10 E.A.D. 460, 478 n. 10 (EAB 2002).

Thus, a source may begin actual construction on or after January 2, 2011 under a PSD permit that authorized construction to begin prior to January 2, 2011 because such a permit states that the source will meet the requirements of paragraphs (j) through (r)(5) of these regulations (or state equivalents) that were in effect at the time the permit was issued. However, this would not be the case if the permit has expired because the applicant has discontinued construction or failed to commence construction by the necessary date. See 40 CFR 52.21(r)(2).

This approach is consistent with EPA's practice when the preconstruction permitting requirements change by virtue of the designation of an area as a nonattainment area after a PSD permit is issued. In transitional guidance issued by EPA in 1991, EPA explained that “the area designation in effect on the date of permit issuance by the reviewing agency determines which regulations (part C or D) apply to that permit.” Memorandum from John S. Seitz, Director OAQPS, New Source Review (NSR) Program Transitional Guidance, page 6 (March 11, 1991). This memorandum explained further that “where a source receives a PSD or other permit prior to the date the area is designated as nonattainment, the permit remains in effect” as long as the source meets the conditions necessary to prevent the permit from expiring. Id. at 6.

This approach does not apply if the source engages in a major modification after January 2, 2011 that is not authorized by the previously issued permit. Once Step 1 of the phase-in begins, if the PSD requirements for GHGs are applicable to a previously-permitted source that engages in a major modification not covered by the permit, such a source will need to obtain a new PSD permit to authorize the modification and that permit may need to include GHG requirements depending on the level of increase in GHGs that results from the modification.

c. Additional Sources for Which PSD Applies in Step 2

In light of the terms of existing PSD regulations and the lead time provided in this action for sources that will first become subject to PSD permitting in Step 2, we do not believe there is presently a need to establish transition provisions for sources that will be required to obtain PSD permits for the first time in Step 2 of the phase-in. As described previously, under our current PSD permitting regulations, a new major stationary source or major modification may not begin actual construction without a PSD permit that meets the applicable preconstruction permitting requirements. 40 CFR 51.166(a)(7)(iii); 40 CFR 52.21(a)(2)(iii).

Since a permit must be obtained before a major source may begin actual construction, the major source preconstruction permitting requirements in 40 CFR 51.166 and 52.21 of the regulation do not generally apply to a source that begins actual construction at a time when it was not a major source required to obtain a PSD permit. One exception, however, is the unique circumstance when a source becomes a major source solely by virtue of the relaxation of an enforceable limitation on the source's PTE. 40 CFR 51.166(r)(2); 40 CFR 52.21(r)(4). But absent these circumstances, PSD preconstruction permitting requirements do not generally preclude a source from continuing actual construction that began before the source was a source required to obtain a PSD permit. Thus, a source that began actual construction under the authorization of any previously required minor source or state construction permit is not required to meet any PSD preconstruction permitting requirement that becomes applicable after actual construction begins unless the source engages in a major modification after PSD permitting requirements are applicable. Likewise, a PSD permit is not required after a source begins actual construction based on a valid determination (by the source or the permitting authority) that the source need not obtain either a major PSD permitting requirements or and minor NSR permit. Based on these provisions in existing regulations, EPA will not require any sources to which PSD permitting requirements begin to apply in Step 2 to obtain a PSD permit to continue construction that actually begins before Step 2 begins.

However, we will expect Step 2 sources that begin actual construction in Step 2 (i.e., beginning July 1, 2011) to do so only after obtaining a PSD permit in accordance with 40 CFR 52.21 or 51.166, or any applicable state regulation that meets the requirements of 40 CFR 51.166. We recognize the potential for the triggering of Step 2 to result in a change in status where a project may legally have begun actual construction before Step 2 but did not do so and would then need a PSD permit. However we also note that we are providing over a year of lead time before PSD permitting requirements become applicable to Step 2 sources. If projects would be adversely affected by this change in status, this lead time affords an opportunity for sources planning such projects to secure appropriate minor NSR permits (which generally take less than a year to issue), non-applicability determinations, etc. in time to avoid such a change in status. If a new or modified source that would become newly subject to PSD in Step 2 plans to begin actual construction before Step 2, it has more than a year to obtain the applicable preconstruction approvals and begin actual construction. Likewise, a Step 2 source that does not anticipate the ability to begin actual construction before Step 2 begins should have enough lead time to submit a PSD permit application and obtain the necessary permit without significantly delaying the project further. Therefore, we do not think it is necessary or appropriate to promulgate a transition provision that would exempt Step 2 sources from PSD permitting requirements that will apply based on construction that begins after Step 2 takes effect.

This approach for Step 2 sources that have obtained a minor source construction permit or non-applicability determination differs from the approach described previously for source that obtained a PSD permit prior to Step 1. As described previously, a Step 1 source that is authorized to begin actual construction before January 2, 2011 under a previously-issued PSD permit may begin actual construction under that permit after January 2, 2011 without modifying the PSD permit to address GHGs. However, a Step 2 source that was not required to obtain a PSD permit before Step 2 begins would need to obtain a PSD permit addressing GHGs if it has not yet begun actual construction prior to Step 2, even if the source had obtained any preconstruction approvals that were necessary to authorize construction prior to Step 2. This is because such a Step 2 source that begins actual construction after Step 2 would likely be doing so without having any permit meeting the requirements of paragraphs (j) through (r)(5) of 40 CFR 52.21 or 51.166, or a state equivalent. A source that has obtained only a minor source permit prior to Step 2 but that begins actual construction after July 1, 2011 would violate the requirements of 40 CFR 52.21(a)(2)(iii) or 51.166(a)(7)(iii), or a state equivalent, unless the source took care to ensure that it was authorized to construct under a PSD permit or could demonstrate that the source's minor source construction permit makes clear that requirement of paragraphs (j) through (r)(5) of 40 CFR 52.21 or 51.166, or a state equivalent, would be met by the source even though such a permit was not nominally a PSD permit. This difference in approach for non-PSD sources is driven by the terms of 40 CFR 52.21(a)(2)(iii) and 51.166(a)(7)(iii). Since we have not provided any prior notice that we might be considering revisions to 40 CFR 52.21 and 51.166 to address this topic, we are unable to revise the regulations in this action to achieve the same result for non-PSD sources as for PSD sources. Furthermore, at the present time, we see no indication that this difference in approach is unreasonable since non-PSD sources will not trigger permitting for GHG until Step 2 (only anyway PSD source trigger in Step 1). Thus sources will have until July 1, 2011, an additional 6 months of lead time (for a total of more than 14 months), to prepare for the transition described here. Nevertheless, we recognize that the transition to the increased coverage of new sources and modifications that occurs in July will represent an unusual occurrence that may have unanticipated impacts. For this reason it is important to note that nothing in this rule forecloses our ability to further address such impacts, as necessary, by adopting rule changes or using other available tools.

EPA has previously promulgated exemptions that have authorized some sources that were not previously subject to the PSD regulations to commence construction on the basis of minor source permits after the date new PSD requirements have took effect in 1978 and 1980. See, e.g., 40 CFR 52.21(i)(1)(iv)-(v). There is a notable distinction between these provisions, which use the term “commence construction,” and the terms of 40 CFR 52.21(a)(2)(iii) and 51.166(a)(7)(iii), which use the term “begin actual construction.” “Commence construction” is defined more broadly than “begin actual construction” to include obtaining all necessarily preconstruction approvals and either beginning actual on-site construction or entering into binding contracts to undertake a program of actual construction. 40 CFR 52.21(b)(9); 40 CFR 51.166(b)(9). The term “commence construction” is also defined in the CAA. 42 U.S.C. 7479(2)(A). Among other purposes, the term “commence construction” is generally used in the Act and EPA regulations to distinguish construction activities that are exempt from new PSD permitting requirements from those that are not. See, e.g., 42 U.S.C. 7475(a); 40 CFR 52.21(i)(1)(i)-(v). In the absence an explicit exemption in the CAA or the PSD regulations that uses the term “commence construction,” we do not believe we can use the date a source “commences construction” under a minor source construction permit approval as a demarcation point for Step 2 sources that may continue ongoing construction activities without having to obtain a PSD permit based on emissions of GHGs. Since we did not provide prior notice of an intention to adopt transition provisions applicable to this situation, we are unable to adopt such an exemption in this action that applies the term commence construction in this context. Consequently, the approach described previously applies the term “begin actual construction” based on the language in 40 CFR 52.21(a)(2)(iii) and 51.166(a)(7)(iii).

d. Transitional Issues for Title V Permitting

Since the title V permitting regulations already include a robust set of provisions to address the incorporation of new applicable requirements and other transitional considerations, we do not see grounds to establish unique transition or grandfathering provisions for GHGs in this action. Furthermore, since the purpose of title V is to collect all regulatory requirements applicable to a source and ensure compliance, we do not believe special exemptions for GHG requirements are likely to be justified. The existing title V rules do not provide any exemptions that relieve the obligation to incorporate all applicable requirements into a title V permit. However, the title V regulations contain numerous provisions that allow a reasonable period of time for incorporating new applicable requirements or applying for a title V permit that was not previously required. Transitional issues for incorporation of GHG requirements into title V permitting generally involve questions in the following categories: (1) Permit application requirements for sources not previously subject to title V that will become subject to title V requirements in Step 2 of the phase-in; (2) the need for updates or amendments to title V permit applications that are pending when GHGs become subject to regulation in Step 1 of the phase-in; and (3) the incorporation of new applicable requirements for GHGs into existing permits for sources currently subject to title V.

With respect to the first category, a title V source applying for the first time must submit its permit application within 12 months after the source “becomes subject to the [operating] permit program” or such earlier time that the permitting authority may require (see 40 CFR 70.5(a)(1)). Sources not otherwise subject to title V can become major sources subject to title V due to emissions of GHG no sooner than July 1, 2011. If a source becomes “subject to the [operating] permit program” on July 1, 2011, then its permit application under the title V operating permit program would typically have to be submitted no later than July 1, 2012.

There are also existing regulations relevant for the second category of GHG transition issues, where sources currently subject to title V have title V permit applications pending with a permitting authority as of January 2, 2011. Where additional applicable requirements become applicable to a source after it submits its application, but prior to release of a draft permit, the source is obligated to supplement its permit application. See 40 CFR 70.5(b); 71.5(b). Furthermore, title V permits are generally required to contain provisions to assure compliance with all applicable requirements at the time of permit issuance. See CAA section 504(a); 40 CFR 70.6(a)(1) and 71.6(a)(1). If a permitting authority determines that additional information is necessary to evaluate or take final action on an application (e.g., because of uncertainty over whether a draft permit assures compliance with all applicable requirements), it may, and should, request additional information from the source in writing and set a reasonable deadline for a response. See 40 CFR 70.5(a)(2); 71.5(a)(2).

Likewise, the existing title V regulations provide sufficient transition for the third category of issues, where a source has additional GHG-related applicable requirements (such as the terms of a PSD permit) that must be incorporated into its existing title V permit. Where a source is required to obtain a PSD permit, the source must apply for a title V permit or permit revision within 12 months of commencing operation or on or before such earlier date as the permitting authority may establish (or prior to commencing operation if an existing title V permit would prohibit the construction or change in operation). See 40 CFR 70.5(a)(1)(ii); 71.5(a)(1)(ii); see also 40 CFR 70.7(d) and (e); 71.7(d) and (e) (permit modifications). In addition, where a source becomes subject to additional applicable requirements, the permitting authority is required to reopen the permit to add those applicable requirements if the permit term has three or more years remaining and the applicable requirements will be in effect prior to the date the permit is due to expire. See 40 CFR 70.7(f)(1)(i); 71.7(f)(1)(i).

Finally, EPA notes that the existing title V regulations require sources to furnish permitting authorities, within a reasonable time, any information the permitting authority may request in writing to determine whether cause exists for modifying, revoking, and reissuing, or terminating the permit, and for other reasons, and further provide that permitting authorities shall reopen and revise permits if EPA or the permitting authority determine that the permit must be revised or revoked to assure compliance with applicable requirements. See 40 CFR 70.6(a)(6)(v); 71.6(a)(6)(v) and 70.7(f)(1)(iv); 71.7(f)(1)(iv).

Thus, EPA believes that the existing title V regulations provide an adequate regulatory framework for managing the transition to incorporating GHG requirements in title V permits and additional specific exemptions or transition rules for title V are not currently warranted.

VI. What are the economic impacts of the final rule?

This section of the preamble examines the economic impacts of the final rule including the expected benefits and costs for affected sources and permitting authorities. The final rule uses a phased-in approach for requiring sources of GHG emissions to comply with title V operating permit and PSD statutory requirements, essentially lifting this burden for the phase-in period for a large number of smaller sources of GHG. Thus, this rule provides regulatory relief rather than regulatory requirements for these smaller GHG sources. For larger sources of GHGs that will be required to obtain title V permits and/or comply on PSD requirements, there are no direct economic burdens or costs as a result of this final rule, because these requirements are not imposed as a result of this rulemaking. Statutory requirements to obtain a title V operating permit or to adhere to PSD requirements are already mandated by the CAA and by existing rules, not by this rule. Similarly, this rule will impose costs to society in the form of foregone environmental benefits resulting from GHG emission reductions that, absent this rule, might otherwise have occurred at sources deferred from permitting during the phase-in period.

The RIA conducted for this final rule provides details of the benefits or regulatory relief that smaller GHG sources will experience in terms of costs avoided as a result of this final rule and the potential for social costs in terms of foregone environmental benefits during this 6-year period. Complete details of the RIA conducted for this final rule may be found in the document “Regulatory Impact Analysis for the Final Prevention of Significant Deterioration and Title V Greenhouse Gas Tailoring Rule,” in the docket for this rulemaking.

This rulemaking provides permitting thresholds for sources of GHG that exceed levels contained in the CAA, and these levels are phased-in steps based upon application of the “administrative necessity” and “absurd results” doctrines as explained in section V.B. For Step 1, which is effective from January 2, 2011, through June 30, 2011, only sources required to undergo title V or PSD permitting based upon non-GHG air pollutants are required to obtain an operating permit or PSD permit to include GHG emissions (referred to as the “anyway” threshold). Step 2, effective from July 1, 2011, until such time as EPA acts on a rule to amend it (which for reasons described previously, we assume is June 30, 2013, for the purposes of this analysis), will phase in title V permit requirements for larger sources emitting GHG above 100,000 tpy CO₂ e (if they do not already have one) and phase in for such sources, PSD requirements when they are newly constructed or modify in a way that increases emissions by more than a 75,000 tpy CO₂ e significance level. Step 2 is referred to as the 100,000 tpy CO₂ e threshold. Thereafter, EPA makes an enforceable commitment to consider a possible Step 3 to further lower thresholds below 100,000 tpy CO₂ e and/or permanently exclude some sources from the program(s), but only after a regulatory process is conducted addressing “administrative necessity” and “absurd results” considerations based upon the actual permitting experiences in the first two steps of the phase-in. In addition, EPA provides a deferral of permitting until we take required action in April 2016 for sources and modifications that emit below 50,000 tpy CO₂ e. The deferral will end when a required study is conducted of the permitting process for sources of GHG and EPA acts, based on the study, to promulgate a rule that describes the additional GHG permitting requirements beyond 2016. In the 6 years following promulgation of this rule, the EPA estimates that compared to baseline estimates that do not include the effects of this rule, over six million sources of GHG emissions in total will be allowed to continue to operate without a title V operating permit. During this period, tens of thousands of new sources or modifying sources each year will not be subject to PSD requirements for GHG. For this large number of smaller sources, this rule alleviates the regulatory burden associated with obtaining an operating or PSD permit or complying with NSR BACT requirements. Therefore, this final action may be considered beneficial to these small sources because it provides relief from regulation that would otherwise be required.

This decision does potentially have environmental consequences in the form of higher emissions during the 6-year period of time (generally because emissions increases would have been lower if BACT were applied). These consequences are limited due to the fact that sources between 100/250 and 100,000 tpy CO₂ e account for an estimated 11 percent of the six directly emitted GHG nationally from industrial, commercial, and residential source categories, while representing over 95 percent of the total number of sources potentially requiring an operating or PSD permit for GHG under current permitting thresholds in the CAA. Moreover, requiring such a large number of small sources to obtain permits for the first time would overtax the permitting authorities' abilities to process new permits and would therefore interfere with any such benefits actually being achieved. Moreover, reductions from these small sources will still be occurring, notwithstanding the fact that permitting requirements would not apply to them. These smaller sources of GHGs will be the focus of voluntary emission reduction programs and energy efficiency measures that lead to reductions in GHGs. We will also reevaluate this decision after a 6-year period and complete a study of the implications for those sources and permitting authorities of permitting smaller GHG sources beyond 2016.

In reaching the preceding decisions for this final rule, we carefully considered comments received on the Tailoring Rule proposal. We received several comments specifically on our description of the impacts of this rule. Most of these comments disagreed with our assertion that the rule is a “relief” rule. Others assert that we should have prepared a more comprehensive RIA than prepared for the rule proposal. Those commenting contend: (1) We understated the burdens of the rule while overstating its relief at proposal; (2) we erroneously omitted the impacts for “larger sources” of GHGs from the proposal RIA and should have recognized the burden to “larger sources” due to other GHG actions; (3) the economic impacts the rule will have on industry and the U.S. economy and society in general will be burdensome, especially given the current state of the economy; and (4) we need to propose a full RIA or a complete estimation of impacts to comply with CAA section 307(d) and the APA.

EPA has carefully considered the comments addressing the issue of whether the Tailoring Rule is a regulatory “relief rule,” and we are not persuaded that we erred in concluding that the effect of the Tailoring Rule is to provide regulatory relief to a large number of sources of GHG for a period of up to 6 years. This final rule will provide relief from title V permitting to over 6 million sources of GHG in this country. Likewise tens of thousands of sources potentially subject to PSD permitting requirements annually for GHG will have regulation postponed for a period of up to 6 years under this rule, followed by an additional required rule addressing the period beyond 6 years. While larger sources of GHG may be required to obtain title V permits or modify existing permits and to comply with PSD requirements, these burdens result not from the Tailoring Rule but rather from the CAA requirements to apply PSD and title V to each pollutant subject to regulation, which are triggered when the LDVR takes effect. To clearly illustrate this, consider what would occur if EPA did not complete the Tailoring Rule. Sources would not be relieved of the requirement to obtain permits addressing each pollutant subject to regulation when they construct or modify, nor would they be relieved of their obligation to obtain title V permits. Instead, these requirements would simply apply to a much larger population of sources and modifications, and would lead to the absurd results and severe impairment to program implementation that this rule is designed to address.

In response to comments asserting that the RIA completed for proposal of this rulemaking: (1) Understated the burdens of the rule and overstated the benefits, (2) did not fully recognize the rule will be burdensome, especially given the current state of the economy; and (3) does not consider a complete estimation of impacts to comply with the APA and CAA section 307(d) and needs to correct flawed or erroneous assumptions, EPA did make improvements and modifications to the RIA completed for this final rule. Based upon comments, EPA modified estimates of the number of sources affected at various threshold levels upward. EPA also improved the burden estimates associated with obtaining permits for sources and permitting authorities.

After consideration of the burden imposed by the proposed rule with these improved estimates for affected sources, the EPA modified the steps of the phase-in period to include two initial steps, described in section V, that are higher, and therefore cover fewer sources and are less burdensome than the proposal threshold of 25,000 tpy CO₂ e emissions. EPA also increased the threshold below which permitting would not apply for 6 years from 25,000 to 50,000 tpy CO₂ e. After the initial two step period, EPA has committed to consider lower thresholds but only down to 50,000 tpy CO₂ e, and only after a regulatory process that uses information gathered on actual permitting activity during the first two steps of the phase-in period. The RIA conducted for the final rule also incorporates improvements in our estimates of the number of sources affected at alternative thresholds and improved estimates of the costs of obtaining permits by sources and processing permits by permitting authorities. The EPA acknowledges that the regulatory relief associated with the control costs due to BACT requirements for PSD new and modifying sources is not included in the RIA for the final rule due to the lack of sufficient data about the nature of those requirements. However, it is the case that, as it relates to burden, those estimates would simply increase the amount of regulatory relief associated with this final rule.

Finally, with regard to comments that the RIA should have been a more comprehensive analysis to include the larger sources of GHG that will be required to obtain permits when GHG are regulated, the EPA maintains as previously explained that there are no direct economic burdens or costs as a result of this rule for these sources. Requirements for larger GHG sources to obtain title V or PSD permits are already mandated by the Act and by existing rules and are not imposed as a result of the Tailoring Rule. Thus the economic impacts for larger sources of GHG do not occur because of this Tailoring Rule. To include these larger sources in the RIA would actually be an inaccurate assessment of how this rule affects sources and would ignore the fact that this rule provides regulatory relief.

A. What entities are affected by this final rule?

As previously stated, this final rule does not itself result in the application of permitting requirements to any industrial, commercial, or residential entities. Entities affected by this rule are those who experience regulatory relief due to the higher thresholds and deferred applicability set forth in this rule. This action increases the threshold to obtain a title V and PSD permitting from statutory CAA levels using a phased-in step process as previously discussed. As Table VI-1 shows, this action lifts permitting requirements for over six million potential title V sources in total and tens of thousands of potential PSD new sources annually that would be otherwise required by the CAA to obtain permits. Under Step 1, over six million title V sources in total and approximately 20 thousand new PSD sources per year will not be required to obtain permits. Under Step 2, requiring sources over a 100,000 tpy CO₂ e to obtain a permit, over six million title V sources in total and approximately 19.9 thousand new PSD sources per year will obtain regulatory relief. While the threshold approach differs for Steps 1 and 2 of the phase-in plan, the estimated number of sources affected does not differ greatly as shown in Table VI-1. Sectors experiencing this regulatory relief include electricity, industrial, energy, waste treatment, agriculture, commercial and residential.

B. What are the estimated annual benefits to sources due to regulatory relief from the statutory requirements?

EPA estimated the annual benefits (avoided costs) to sources of GHG emissions and permitting authorities anticipated from this final rule. In addition, an accounting of the benefits from this action as measured by avoided permit processing costs for state, local, and tribal permitting authorities is provided. These benefits or avoided costs relate specifically to permit burden costs postponed for smaller sources of GHG emissions otherwise required to obtain an operating permit under title V or required to modify an existing permit to address GHG emissions. Avoided costs shown also include permit burdens for additional PSD permits postponed for new or modifying smaller sources of GHG, as well as the avoided costs to state, local and tribal permitting authorities. We are providing an illustrative monetary estimate of statutory permitting requirements to show the magnitude of the savings that hypothetically result from this rulemaking. While we believe it is impossible to implement these permit requirements by January 2, 2011, for the reasons laid out in this preamble, it is useful to understand the scale of what the burden may have been. For sake of simplicity, we refer to this illustrative monetary estimate as the monetized benefits of the regulatory relief presented by this rulemaking or regulatory relief benefits for brevity.

These benefit estimates do not consider avoided emission control costs associated with PSD requirements for potential BACT requirements. Estimates for BACT are unavailable at this time because of the difficulty predicting the results of the BACT process as it would be applied to new pollutants and classes of sources for which there is no previous BACT experience on which to rely.

1. What are annual estimated benefits or avoided burden costs for title V permits?

Table VI-2 shows that the estimated annual title V benefits to sources and to permitting authorities in terms of avoided information collection cost resulting from this final action to be approximately $70,535 million under Step 1 of the phase-in. These avoided costs become $70,520 million annually under Step 2 of the phase-in. where permitting is required for sources at or above the 100,000 tpy CO₂ e threshold. Under the anyway threshold Step 1, approximately $49,457 million in regulatory relief will accrue to sources and approximately $21,078 million to permitting authorities annually in the form of avoided permit processing costs. With the 100,000 tpy CO₂ e threshold for phase-in Step 2, these annual regulatory relief benefits are expected to be quite similar at $49,447 million for sources of GHG emissions and $21,072 million for permitting authorities. Industrial sources permitting costs are estimated to be $46.4 thousand per permit for a new permit and $1.7 thousand for a permit revision. The EPA estimates that over tens of thousands of industrial sources per year will avoid incurring these permitting costs under Steps 1 and 2 of the phase-in period. The cost for a permit for new commercial and residential sources is estimated to be $23.2 thousand per permit with approximately 2 million of these permits avoided annually.

State, local, and tribal permitting authorities will also benefit in terms of avoided permitting administrative costs of over $21 billion as a result of the decisions final in this action. For industrial sources, the cost for permitting authorities to process a new industrial title V permit is approximately $19.7 thousand per permit and $1.8 thousand for a permit revision. Similarly, permitting authority avoided permit processing costs are approximately $9.8 thousand per permit for a new commercial or residential title V permit. All estimates are stated in 2007 dollars.

2. What are annual benefits or avoided costs associated with NSR permitting regulatory relief?

Table VI-3 summarizes the estimated annual permit burden costs avoided by sources and permitting authorities for PSD permitting due to this Tailoring Rule. The benefits associated with avoided cost of compliance for BACT for these sources is not included in these estimates due to a lack of available data. The estimated avoided burden or reporting and recordkeeping cost that would occur absent this rule for new industrial sources to obtain permits is estimated to be $84.5 thousand for a modifying PSD industrial source and $59.2 thousand for a modifying commercial or multi-family residential source. New PSD sources will also be required to obtain a title V permit increasing these costs to $130.9 thousand per permit for new industrial sources and to $82.3 thousand per permit for new commercial or multi-family residential sources. (Note the title V costs for these new PSD sources have been included in title V estimates shown in Table VI-2.) New and modifying sources avoid approximately $5.5 billion annually in PSD permitting costs with this rule under the phase-in Step 1 threshold. Under the phase-in Step 2, 100,000 tpy CO₂ e threshold and 75,000 tpy CO₂ e significance level, this avoided PSD permitting cost estimate becomes $5.4 billion annually. State, local, and tribal permitting authorities are expected to avoid about $1.51 billion annually in administrative expenditures associated with postponing PSD program requirements for these GHG sources under Step 1 and $1.49 billion under Step 2. All estimates are shown in 2007 dollars.

C. What are the economic impacts of this rulemaking?

This final rulemaking does not impose economic burdens or costs on any sources or permitting authorities, but should be viewed as regulatory relief for smaller GHG emission sources and for permitting authorities. Although sources above the thresholds set in this rule will become subject to permitting on January 2, 2011, those impacts are not attributable to the present rulemaking. Rather they are mandated by the CAA and existing regulations and automatically take effect independent of this action.

In addition to considering the regulatory relief expected for affected entities as a result of this final rule, the EPA considered the impact of this rulemaking to small entities (small businesses, governments and non-profit organizations) as required by the Regulatory Flexibility Act (RFA) and the Small Business Regulatory Enforcement Fairness Act (SBREFA). For informational purposes, the RIA includes the SBA definition of small entities by industry categories for stationary sources of GHG and potential regulatory relief from title V and NSR permitting programs for small sources of GHG. Since this rule does not impose regulatory requirements but rather lessens the regulatory burden of the CAA requirements to smaller sources of GHG, no economic costs are imposed upon small sources of GHG as a result of this final rule. Rather this action provides regulatory relief for small sources. These avoided costs or benefits accrue because small sources of GHG are not required to obtain a title V permit and new or modifying small sources of GHG are not required to meet PSD requirements. Some of the small sources benefitting from this action are small entities, and these entities will benefit from the regulatory relief finalized by this rule. For discussion of comments received and EPA responses regarding small entities impacts, see section VII of this preamble.

D. What are the costs of the final rule for society?

EPA examined the social costs of this final rule. These social costs represent the foregone environmental benefits that will occur as a result of the regulatory relief offered to sources of GHG emissions. This action is one of regulatory relief since it increases the emissions thresholds for the title V and PSD programs, as they apply to sources of GHG emissions, to levels above those in the CAA. In this preamble section, the benefits or avoided regulatory costs of such relief are discussed, but there is also a social cost imposed by such relief, because this rule may forego some of the possible benefits associated with title V and PSD programs for sources of GHG emissions below the permitting thresholds established. These benefits are those attributed to title V and PSD permitting programs in general. These benefits are based upon the relevance of these programs to policymaking, transparency issues, and market efficiency, and therefore are very difficult to quantify and monetize. For title V, they include the benefits of improved compliance with CAA requirements that stem from (1) Improved clarity regarding applicability of requirements, (2) discovery and required correction of noncompliance prior to receiving a permit, (3) improving monitoring, recordkeeping, and reporting concerning compliance status, (4) self-certification of compliance with applicable requirements initially and annually, and prompt reporting of deviations from permit requirements, (5) enhanced opportunity for the public to understand and monitor sources' compliance obligations, and (6) improved ability of EPA, permitting authorities, and the public to enforce CAA requirements. However, it is important to remember that a title V permit generally does not add new requirements for pollution control itself, but rather collects all of a facility's applicable requirements under the CAA in one permitting mechanism. Therefore, the compliance benefits above are less when title V permits contains few or no CAA applicable requirements. During the initial steps of the phase-in plan established under this action, we expect that the vast majority of sources excluded from title V would be sources that have no CAA applicable requirements for GHG emissions and few or no requirements for other pollutants because their emissions of those pollutants are so small. For this reason, while it is extremely difficult to measure the degree of improved compliance, if any, that would be foregone, or to quantify the social costs that would be imposed, we expect that they would be small. We will be evaluating this issue further during subsequent phases.

For PSD, the primary social cost imposed by the Tailoring Rule stems from the foregone benefit of applying BACT to the tens of thousands of small new sources and modifications that will be below our final thresholds during the first steps of the phase-in. This social cost potentially weighs against the cost savings described previously that stem (in part) from avoiding the administrative and control costs of applying BACT to these sources. The BACT requirement assures that new and modified sources, when they increase their emissions are using state-of-the-art emission controls and affords the public an opportunity to comment on the control decision. It does not prohibit increases but it assures that such controls are applied. Delaying the BACT requirement for numerous small sources during the first steps of the phase-in for this final rule could allow increases from these smaller sources that are greater than they would be if BACT were applied. A detailed analysis of this difference is beyond the scope of this rule, because we do not have detailed information on the universe of these tens of thousands of small PSD actions, the candidate BACT technologies for each of them, how permitting authorities would make the BACT decisions, and how the BACT limit would compare to what would otherwise be installed absent BACT.

It is not possible at this time to quantify the social costs of avoided BACT. However, we note that the universe of possible emissions that would be regulated by sources excluded under the Tailoring Rule is small compared to those that would remain subject to PSD. The sources excluded in these first two steps of the phase-in plan of this action comprise only 11 percent of total stationary source GHG emissions, while 67 percent remain subject to regulation. Furthermore, we expect the emissions differences due to BACT controls for such sources to be relatively small due to the lack of available capture and control technologies for GHG at such sources that are akin to those that exist for conventional pollutants and sources, as well as the likelihood that even in the absence of BACT such sources would already be installing relatively efficient GHG technologies to save on fuel costs. Thus, while potential benefits would be foregone by excluding smaller sources from the permitting programs, these benefits are likely to be small. Under the Tailoring Rule, we will be working during the 6-year period to greatly improve our understanding of both the administrative costs of regulating and the social costs of not regulating smaller sources under PSD and title V, and we will be relying on that information to support our future threshold analyses called for under the action.

In reaching the decisions for this Tailoring Rule, the EPA recognizes that GHG emissions can remain in the atmosphere for decades to centuries, meaning that their concentrations become well-mixed throughout the global atmosphere regardless of emission origin, and their effects on climate are long lasting and significant. A detailed explanation of climate change and its impact on health, society, and the environment is included in EPA's TSD for the endangerment finding action (Docket ID No. EPA-HQ-OAR-2009-0171). The EPA recognizes the importance of reducing climate change emissions for all sources of GHG emissions including those sources afforded regulatory relief in this rule and plans to address potential emission reductions from these small sources using voluntary and energy efficiency approaches. Elsewhere, we have discussed EPA's interest in continuing to use regulatory and/or non-regulatory tools for reducing emissions from smaller GHG sources because we believe that these tools will likely result in more efficient and cost-effective regulation than would case-by-case permitting.

E. What are the net benefits of this final rule?

The net benefits of this GHG tailoring rule represent the difference between the benefits and costs of this rule to society. As discussed in this preamble, this rule is one of regulatory relief and the benefits to society are estimates the regulatory relief (avoided permit burden costs) to sources and permitting authorities for Steps 1 and 2 of the phase-in period. The social costs of the rule are the foregone environmental benefits in the form of potential GHG emission reductions that could occur during the phase-in period and are discussed qualitatively.

This rulemaking provides regulatory relief for a phase-in period to smaller sources of GHG by phasing in the statutory permitting threshold at levels above statutory requirements. This final rule establishes thresholds and PSD significance levels for Steps 1 and 2 of the phase-in period (the 2.5 year period between January 2, 2011 and July 1, 2013), commits to considering a further Step 3, and indicates floor title V and PSD threshold levels from July 1, 2013 through April 30, 2016. The net benefits of the final rule for Steps 1 and 2 are $193,598+B-C million for the 2 and one-half year period where B denotes the unquantified benefits and C the quantified costs of this final rule. These unquantified benefits of this rule include the avoided PSD BACT costs for new and modifying sources. The unquantified costs previously discussed relate to the foregone environment benefits or GHG emission reductions that might be possible during the 2.5 year Step 1 and 2 phase-in period. These estimates are subject to significant uncertainties that are discussed at length in the Regulatory Impact Analysis for the Prevention of Significant Deterioration and Title V GHG Tailoring Rule contained in the docket to this final rule. All dollar estimates shown are based upon 2007$.

VII. Comments on Statutory and Executive Order Reviews

In this section, we provide responses to comments we received for various Executive Orders.

A. Comments on Executive Order 12866—Regulatory Planning and Review

At proposal, EPA prepared an analysis of the potential costs and benefits associated with EPA's Tailoring Rule proposal in an RIA. Several commenters state that EPA's failure to estimate the full costs of the effects of its interpretation of PSD applicability in the proposed Tailoring Rule violates Executive Order 12866. Some of these commenters maintain that Executive Order 12866 directs EPA to submit to the Office of Management and Budget (OMB) new significant regulations under consideration by the EPA. These commenters assert that, in the section 202 rule, EPA failed to analyze the effect on stationary sources in the cost benefit analysis and there is no indication that EPA included these impacts in its submission to OMB. According to the commenters, in EPA's proposal for this rulemaking, EPA has similarly failed to analyze the costs and benefits of triggering PSD for stationary sources. The commenters assert that without this key information, OMB could not fully review the impacts of the proposed rule. The commenters believe that EPA's failure to account for known costs that will occur as a direct result of the promulgation of the proposed rule in conjunction with the section 202 rule violates several applicable requirements of Executive Order 12866, including sections 6(B)(ii) and 6(C)(iii), which require assessments of the potential costs and benefits of the regulatory action and “reasonably feasible alternatives to the planned regulation, identified by the Agencies or the public * * *” thereby violating both the APA and CAA section 307(d) because they deprive businesses and permitting authorities alike of a meaningful opportunity to comment on the rule.

The EPA has prepared a revised RIA assessing the benefits and costs of the final Tailoring Rule to support this rulemaking in accordance with Executive Order 12866, as was done with the proposal for this rulemaking. Similarly, the RIA completed for this action is subject to review by an Inter-agency review panel that includes OMB, as was the case with the proposal RIA. Further, the RIA completed for this final rule fully assesses the known benefits and costs associated with the Tailoring Rule. This final rule is one of regulatory relief from statutory requirements in which a large number of sources of GHGs will be relieved of the burden of title V and PSD permitting for a period of at least 6 years. This final rule will provide relief from title V permitting to over 6 million sources of GHG in this country. Likewise tens of thousands of sources potentially subject to PSD permitting requirements for GHGs will have regulation postponed for a period of at least 6 years. While larger sources of GHG may still be required to obtain title V permits or modify existing permits and to comply with PSD requirements, these burdens result from existing statutory requirements, not from this final Tailoring Rule.

B. Comments on the Paperwork Reduction Act

At proposal, we stated in the preamble that we did not believe that the proposal would impose any new information collection burden. We concluded that the proposed action would reduce costs incurred by sources and permitting authorities relative to the costs that would be incurred if EPA did not revise the rule and provided estimates of those reduced costs. Further, we stated that, despite our estimated burden reductions, it was unnecessary for us to submit a new ICR to the OMB because the ICR contained in the existing regulations for PSD (see, e.g., 40 CFR 52.21) and title V (see 40 CFR parts 70 and 71) had already been approved under the provisions of the Paperwork Reduction Act, 44 U.S.C. 3501 et seq. and was assigned OMB control number 2060-0003 and OMB control number 2060-0336, respectively.

However, several commenters disagree that it was unnecessary for us to submit a new ICR for the proposed action. These commenters believe that (1) prior approval of an ICR for the PSD and title V programs ignores the fact that there would be an increase in the paperwork burden as a result of applying PSD and title V permitting requirements; and (2) unless EPA resubmits the information collection approval request to OMB with a proper and fully-inclusive analysis, EPA will lack authority to collect information from stationary sources for PSD and title V GHG emissions permitting.

As we stated in the proposal, this is a burden relief rule and as such it does not impose any new requirements for the NSR or title V programs that are not currently required. For that reason, we concluded that for purposes of this rule it was unnecessary for us to submit a new ICR to the OMB and that the ICR contained in the existing regulations for PSD (see, e.g., 40 CFR 52.21) and title V (see 40 CFR parts 70 and 71) that had already been approved under the provisions of the Paperwork Reduction Act, 44 U.S.C. 3501 et seq. and was assigned OMB control number 2060-0003 and OMB control number 2060-0336, respectively, still applies.

Nevertheless, we understand that once GHGs are regulated under the PSD and title V programs, there might be an increase in the overall paperwork burden for these programs. EPA will have to assess this possible burden during the normal course of 3-year renewal ICR process.

C. Comments on the RFA

At proposal, EPA certified that the proposed rule would not have a significant impact on a substantial number of small entities and therefore we are not obligated to convene a formal Small Business Advocacy Review (SBAR) panel. This certification was based upon the fact that the proposed action would relieve the regulatory burden associated with the major PSD and title V operating permits programs for new or modified major sources that emit GHGs, including small businesses. Nevertheless, EPA was aware at proposal that many small entities would be interested in the various GHG rulemakings currently under development and might have concerns about the potential impacts of the statutory imposition of PSD requirements that may occur as a result of the group of EPA actions, notwithstanding the relief provided to small businesses by the Tailoring Rule. For these reasons, and in collaboration with the SBA, EPA conducted an outreach meeting designed to exchange information with small entities that may be interested in these regulations. The outreach effort was organized and led by representatives from EPA's Office of Air Quality Planning and Standards within the Office of Air and Radiation, EPA's Office of Policy Economics and Innovation, the Office of Information and Regulatory Affairs within OMB, and the Office of Advocacy of the SBA. This meeting was conducted on November 17, 2009 in Arlington, VA, and documentation of this meeting, which includes a summary of the advice and recommendations received from the small entity representatives identified for the purposes of this process, can be obtained in the docket for this rulemaking. (See Docket No. EPA-HQ-OAR-2009-0517-19130.)

During the comment period, several commenters alleged that EPA inappropriately limited its RIA and RFA/SBREFA analysis, and that had we done a comprehensive analysis, we would not have been able to certify that any of the proposed rules will not have a significant economic impact on a “substantial number of small entities.” Thus they conclude that EPA failed to prepare and publicize an initial regulatory flexibility analysis (IRFA). Additional commenters stated that EPA's failure to conduct an IRFA to assess the full costs of the effects of its interpretation of PSD applicability in the proposed Tailoring Rule violates a host of statutes and Executive Orders requiring analysis and public review of regulatory burdens. These commenters conclude that EPA should have convened one or more SBAR Panels.

We are not persuaded that we should have taken into account effects beyond those caused by the Tailoring Rule when we made our certification of no significant economic impact on a substantial number of small entities for this rule. No permitting requirements are imposed by this final Tailoring Rule. Instead, this final Tailoring Rule offers regulatory relief to over an estimated six million sources of GHG emissions that would otherwise be required to obtain a title V permit and tens of thousands of sources of GHG emissions subject to PSD permitting requirements that would otherwise be required statutorily to obtain permit. The RFA does not require that an agency complete a regulatory flexibility analysis or conduct an SBAR panel where the rule does not have any negative impact on small entities. For more discussion of RFA issues, please see the RTC document.

D. Comments on the Unfunded Mandates Reform Act

At proposal, EPA asserted that the Tailoring Rule does not impose unfunded mandates on any entities including sources and permitting authorities. Since the proposed Tailoring Rule is one of regulatory relief, it alleviates the burden of adhering to statutorily required permitting thresholds and does not impose regulatory requirements.

Some commenters on the proposed rule assert that EPA has failed to comply with the requirements of the Unfunded Mandates Reform Act (UMRA), pursuant to which EPA must assess the effects of the proposed rule on state, local, and tribal governments and the private sector. Specifically, these commenters state that section 202 of the UMRA requires EPA to prepare a written statement, including a cost-benefit analysis, for proposed rules with “federal mandates” that may result in expenditures to state, local, and tribal governments, in the aggregate, or to the private sector, of $100 million or more in any 1 year. According to the commenters, in concluding that “the revisions would ultimately reduce the PSD and title V program administrative burden that would otherwise occur in the absence of this rulemaking,” EPA did not account for the billions of dollars that permitting authorities and stationary sources will soon be required to spend once PSD is triggered for GHGs. Additionally, a few commenters contend that the EPA underestimated the impacts to public utilities which are owned/operated by local governments and also to state regulatory agencies.

The EPA has carefully considered the comments on unfunded mandates expressed by commenters to the proposed rule. The EPA did complete a RIA for the final rule assessing the benefits and costs of the Tailoring Rule, including any unfunded mandates. As previously discussed, the Tailoring Rule is one of regulatory relief because it increases the GHG emissions threshold for NSR and title V permitting substantially above otherwise statutory requirements. As such, the EPA has determined that this Tailoring Rule does not impose unfunded mandates on any entities. This RIA of the final rule incorporates the extensive changes made in this final rule, including increased threshold levels for title V and PSD above those contained in the proposed rule. While we also incorporated improved estimates of the costs for sources to obtain permits and for permitting authorities to process permits, they do not change our conclusion that this final rule does not impose unfunded mandates on any entities.

E. Comments on Executive Order 13132—Federalism

Some comments received on the proposed rule assert that federalism concerns were ignored, in violation of Executive Order 13132. According to the commenters, EPA cannot maintain that the Tailoring Rule “will not have a substantial direct effect on the states, on the relationship between the national government and the states, or on the distribution of power and responsibilities between various levels of government,” such that Executive Order 13132 does not apply. Some of these commenters assert that the proposed rulemaking would require radical changes in state laws, interjects GHGs into permit programs never once conceived for that purpose (any more than was EPA's), requires massive staff hiring at state agencies, and rewrites SIPs in place for years or even decades.

As we stated previously, this is a burden relief rule and as such it does not impose any requirements for the NSR or title V programs that are not currently required. In addition, this action does not interject GHGs into the permit programs, nor does it change state laws or SIPs to impose any new permitting requirements. Instead, this action will significantly reduce the burden and costs incurred by sources and permitting authorities relative to the burden and costs that would be incurred if EPA did not revise the permitting provisions to account for higher applicability thresholds for GHG emissions.

However, since this rule finalizes burden reducing thresholds that will not otherwise apply to the PSD and title V programs, we are aware that a few states may have to amend their SIPs to incorporate these new thresholds if they do not incorporate federal rules by reference and cannot adopt our approach through interpretation. Executive Order 13132 is still not implicated by this rule because it finalizes burden reducing thresholds that would not otherwise apply to the PSD and title V programs.

F. Comments on Executive Order 13175—Consultation and Coordination With Indian Tribal Governments

The National Tribal Air Association (NTAA) supports EPA's proposed rule but requests that tribal air grant funding be increased to reflect the air quality-related needs of tribes across the nation, and to allow these tribes the opportunity to implement the CAA's PSD and title V programs. The NTAA states that, not only are tribes eligible for section 103 grant funding to conduct air quality monitoring, emissions inventories, and other studies and assessments, but they may also obtain section 105 grant funding to implement CAA regulatory programs. According to the NTAA, tribes are facing many of the same air-related issues that neighboring state and local jurisdictions are facing, but are significantly underfunded to address such issues.

The Agency is aware and concerned about the resource needs for the tribal air program and we are working to see how grant funding might be increased in the future. Nevertheless and for the purpose of the permitting programs, we want to clarify that tribes that develop Tribal Implementation Plans (TIPs) can charge for permits and tribes with delegation or authorization would develop permit fee programs under their authority (e.g., Navajo's permit fee program for their delegated title V permit program) to fund both the NSR and title V programs. For these reasons, there are a number of ways we would like to work with tribes to address the funding concern, including encouraging delegation or authorization of permitting programs and having model codes available for tribes that want to do TIPs for NSR and title V permitting.

G. Comments on Executive Order 13211—Actions That Significantly Affect Energy Supply, Distribution, or Use

Other commenters assert that EPA's analysis under Executive Order 13211 is insufficient because it addresses only smaller sources. These commenters contend that EPA has not meaningfully examined the energy implications of its proposed actions and interpretations of the CAA. The commenters disagree with EPA's conclusion that the imposition of costly PSD obligations on power plants would have no impact on power supply, distribution, or use, when those plants will have had no time to prepare for compliance and no idea what BACT may be for GHG emissions. Other commenters opine that the adoption of BACT for some industries newly-subject to PSD permitting requirements for GHGs could involve fuel-switching, and increased energy costs (due to the need for a source to convert from coal to natural gas to meet BACT).

Again, this action is a burden relief rule and as such it does not create any new requirements for sources in the energy supply, distribution, or use sectors. For the purpose of the BACT determinations for GHGs, the long-standing top-down BACT selection process still applies. Under the CAA and EPA's implementing regulations, BACT is still an emission limitation based on the maximum degree of emission reduction achievable through application of production processes and available methods, systems, and techniques that considers energy, environmental, and economic impacts. In other words, BACT determinations for GHGs will still have to consider energy, environmental and economic feasibility for the various control technologies under consideration before selecting a particular technology as BACT for a specific source. For that reason, what BACT may be for GHG emissions will vary by source, and the technology that is ultimately selected has to be one that is feasible based on the current energy, environmental and economic impacts that the planned technology might have. Thus, we do not believe that this action is likely to have a significant adverse effect on the supply, distribution, or use of energy.

VIII. Statutory and Executive Order Reviews

A. Executive Order 12866—Regulatory Planning and Review

Under section 3(f)(1) of Executive Order 12866 (58 FR 51735, October 4, 1993), this action is an “economically significant regulatory action” because it is likely to have an annual effect on the economy of $100 million or more. Accordingly, EPA submitted this action to the OMB for review under Executive Order 12866 and any changes made in response to OMB recommendations have been documented in the docket for this action.

In addition, EPA prepared an analysis of the potential costs and benefits associated with this action. This analysis is contained in the RIA for this final rule. A copy of the analysis is available in the docket for this action and the analysis is briefly summarized in section VII of this preamble.

This rule uses a phased-in approach for requiring larger sources of GHG emissions to comply with title V operating permit and PSD statutory requirements, essentially lifting this burden for a period of at least 6 years for a large number of sources of GHG. Thus, this rule provides regulatory relief rather than regulatory requirements for these GHG sources. For sources of GHG that will be required to obtain title V permits and/or comply with PSD requirements, there are no direct economic burdens or costs as a result of this final rule, because these requirements are not imposed as a result of this rulemaking. Statutory requirements to obtain a title V operating permit or to adhere to PSD requirements are already mandated by the CAA and by existing rules, not by this rule. As a result, this Tailoring Rule annual effect on the economy will be positive because it will result in billions of dollars of regulatory relief during the phase-in period.

B. Paperwork Reduction Act

This action does not impose any new information collection burden. Instead, this action will significantly reduce costs incurred by sources and permitting authorities relative to the costs that would be incurred if EPA did not revise the rule. Based on our revised GHG threshold data analysis, we estimate that over 80,000 new and modified facilities per year would be subject to PSD review based on applying a GHG emissions threshold of 100/250 tpy using a CO₂ e metric. This is compared to 280 PSD permits currently issued per year, which is an increase of more than 280-fold. Similarly, for title V, we estimate that over six million new sources would be affected at the 100-tpy threshold for GHGs using the CO₂ e metric. By increasing the volume of permits by over 400 times, the administrative burden would be unmanageable without this rule.

However, OMB has previously approved the information collection requirements contained in the existing regulations for PSD (see, e.g., 40 CFR 52.21) and title V (see 40 CFR parts 70 and 71) under the provisions of the Paperwork Reduction Act, 44 U.S.C. 3501 et seq. and has assigned OMB control number 2060-0003 and OMB control number 2060-0336. The OMB control numbers for EPA's regulations in 40 CFR are listed in 40 CFR part 9.

C. Regulatory Flexibility Act

The RFA generally requires an agency to prepare a regulatory flexibility analysis of any rule subject to notice and comment rulemaking requirements under the APA or any other statute unless the agency certifies that the rule will not have a significant economic impact on a substantial number of small entities. Small entities include small businesses, small organizations, and small governmental jurisdictions.

For purposes of assessing the impacts of this final action on small entities, small entity is defined as: (1) A small business that is a small industrial entity as defined in the U.S. SBA size standards (see 13 CFR 121.201); (2) a small governmental jurisdiction that is a government of a city, county, town, school district, or special district with a population of less than 50,000; or (3) a small organization that is any not-for-profit enterprise that is independently owned and operated and is not dominant in its field.

After considering the economic impacts of this final rule on small entities, I certify that this final action will not have a significant economic impact on a substantial number of small entities. In determining whether a rule has a significant economic impact on a substantial number of small entities, the impact of concern is any significant adverse economic impact on small entities, since the primary purpose of the regulatory flexibility analyses is to identify and address regulatory alternatives “which minimize any significant economic impact of the rule on small entities.” 5 U.S.C. 603 and 604. Thus, an agency may certify that a rule will not have a significant economic impact on a substantial number of small entities if the rule relieves regulatory burden, or otherwise has a positive economic effect on all of the small entities subject to the rule.

We have therefore concluded that this final rule will relieve the regulatory burden for most affected small entities associated with the major PSD and title V operating permits programs for new or modified major sources that emit GHGs, including small businesses. This is because this rule raises the major source applicability thresholds for these programs for the sources that emit GHGs. As a result, the program changes provided in this rule are not expected to result in a significant economic impact on a substantial number of small entities.

D. Unfunded Mandates Reform Act

This rule does not contain a federal mandate that may result in expenditures of $100 million or more for state, local, and tribal governments, in the aggregate, or the private sector in any 1 year. Only those few states whose permitting authorities do not implement the federal PSD and title V rules by reference in their SIPs will have a small increase in burden. These states will have to amend their corresponding SIPs to incorporate the new applicability thresholds, since the burden reducing thresholds that we are finalizing with this rule will not otherwise apply to the PSD and title V programs. Thus, this rule is not subject to the requirements of sections 202 or 205 of UMRA.

This rule is also not subject to the requirements of section 203 of UMRA because it contains no regulatory requirements that might significantly or uniquely affect small governments. As discussed earlier, this rule is expected to result in cost savings and an administrative burden reduction for all permitting authorities and permittees, including small governments.

E. Executive Order 13132—Federalism

This action does not have federalism implications. It will not have substantial direct effects on the states, on the relationship between the national government and the states, or on the distribution of power and responsibilities among the various levels of government, as specified in Executive Order 13132. These final amendments will ultimately simplify and reduce the burden on state and local agencies associated with implementing the PSD and title V operating permits programs, by providing that a source whose GHG emissions are below the proposed levels will not have to obtain a PSD permit or title V permit. Thus, Executive Order 13132 does not apply to this action.

In the spirit of Executive Order 13132, and consistent with EPA policy to promote communications between EPA and state and local governments, EPA specifically solicited comment on the proposed rule from state and local officials.

F. Executive Order 13175—Consultation and Coordination With Indian Tribal Governments

Subject to the Executive Order 13175 (65 FR 67249, November 9, 2000) EPA may not issue a regulation that has tribal implications, that imposes substantial direct compliance costs, and that is not required by statute, unless the federal government provides the funds necessary to pay the direct compliance costs incurred by tribal governments, or EPA consults with tribal officials early in the process of developing the proposed regulation and develops a tribal summary impact statement.

EPA has concluded that this action may have tribal implications. However, it will neither impose substantial direct compliance costs on tribal governments, nor preempt tribal law. There are no tribal authorities, currently issuing major NSR permits; however, this may change in the future.

EPA consulted with tribal officials early in the process of developing this regulation to allow them to have meaningful and timely input into its development by publishing an ANPR that included GHG tailoring options for regulating GHGs under the CAA. (73 FR 44354, July 30, 2008) As a result of the ANPR, EPA received several comments from tribal officials on differing GHG tailoring options presented in the ANPR which were considered in the proposal and this final rule. Additionally, we also specifically solicited comment from tribal officials on the proposed rule (74 FR 55292, October 27, 2009).

G. Executive Order 13045—Protection of Children From Environmental Health and Safety Risks

EPA interprets Executive Order 13045 (62 FR 19885, April 23, 1997) as applying only to those regulatory actions that concern health or safety risks, such that the analysis required under section 5-501 of the Executive Order has the potential to influence the regulation. This action is not subject to Executive Order 13045 because it does not establish an environmental standard intended to mitigate health or safety risks.

H. Executive Order 13211—Actions That Significantly Affect Energy Supply, Distribution, or Use

This action is not a “significant energy action” as defined in Executive Order 13211 (66 FR 28355 (May 22, 2001)), because it is not likely to have a significant adverse effect on the supply, distribution, or use of energy. Further, we have concluded that this rule is not likely to have any adverse energy effects because this action would not create any new requirements for sources in the energy supply, distribution, or use sectors.

I. National Technology Transfer and Advancement Act

Section 12(d) of the National Technology Transfer and Advancement Act of 1995 (“NTTAA”), Public Law No. 104-113, 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 action does not involve technical standards. Therefore, EPA did not consider the use of any voluntary consensus standards.

J. Executive Order 12898—Federal Actions To Address Environmental Justice in Minority Populations and Low-Income Populations

Executive Order 12898 (59 FR 7629 (Feb. 16, 1994)) establishes federal executive policy on environmental justice. Its main provision directs federal agencies, to the greatest extent practicable and permitted by law, to make environmental justice part of their mission by identifying and addressing, as appropriate, disproportionately high and adverse human health or environmental effects of their programs, policies, and activities on minority populations and low-income populations in the United States.

EPA has concluded that it is not practicable to determine whether there would be disproportionately high and adverse human health or environmental effects on minority and/or low income populations from this rule. This rule is necessary in order to allow for the continued implementation of permitting requirements established in the statute. Specifically, without this rule, the CAA permitting programs (PSD and title V) would become overwhelmed and unmanageable by the millions of GHG sources that would become newly subject to them. This would result in severe impairment of the functioning of these programs with potentially adverse human health and environmental effects nationwide. Under this rule and the legal doctrines of “absurd results,” administrative necessity, and one-step-at-a-time, EPA is ensuring that the CAA permitting programs continue to operate by limiting their applicability to the maximum number of sources the programs can possibly handle. This approach is consistent with congressional intent as it allows PSD applicability to at least the largest sources initially, at least to as many more sources as possible, and as promptly as possible over time. By doing so, this rule allows for the maximum degree of environmental protection possible while providing regulatory relief for the unmanageable burden that would otherwise exist. Therefore, we believe it is not practicable to identify and address disproportionately high and adverse human health or environmental effects on minority populations and low income populations in the United States under this final rule.

K. Congressional Review Act

The Congressional Review Act, 5 U.S.C. 801 et seq., as added by SBREFA, 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. 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. 804(2). This rule will be effective August 2, 2010.

L. Judicial Review

Under section 307(b)(1) of the Act, petitions for judicial review of this action must be filed in the United States Court of Appeals for the District of Columbia Circuit by August 2, 2010. Any such judicial review is limited to only those objections that are raised with reasonable specificity in timely comments. Filing a petition for reconsideration by the Administrator of this final rule does not affect the finality of this rule for the purposes of judicial review nor does it extend the time within which a petition for judicial review may be filed, and shall not postpone the effectiveness of such rule or action. Under section 307(b)(2) of the Act, the requirements of this final action may not be challenged later in civil or criminal proceedings brought by us to enforce these requirements. Pursuant to section 307(d)(1)(V) of the Act, the Administrator determines that this action is subject to the provisions of section 307(d). Section 307(d)(1)(V) provides that the provisions of section 307(d) apply to “such other actions as the Administrator may determine.” This action finalizes some, but not all, elements of a previous proposed action—the Prevention of Significant Deterioration and Title V Greenhouse Gas Tailoring Rule Proposed Rule (74 FR 55292, October 27, 2009).

IX. Statutory Authority

The statutory authority for this action is provided by sections 307(d)(7)(B), 101, 111, 114, 116, and 301 of the CAA as amended (42 U.S.C. 7401, 7411, 7414, 7416, and 7601). This action is also subject to section 307(d) of the CAA (42 U.S.C. 7407(d)).

List of Subjects

40 CFR Part 51

• Environmental protection
• Administrative practice and procedure
• Air pollution control
• Carbon dioxide
• Carbon dioxide equivalents
• Greenhouse gases
• Hydrofluorocarbons
• Intergovernmental relations
• Methane
• Nitrous oxide
• Perfluorocarbons
• Reporting and recordkeeping requirements
• Sulfur hexafluoride

40 CFR Part 52

• Environmental protection
• Administrative practice and procedure
• Air pollution control
• Carbon dioxide
• Carbon dioxide equivalents
• Greenhouse gases
• Hydrofluorocarbons
• Intergovernmental relations
• Methane
• Nitrous oxide
• Perfluorocarbons
• Reporting and recordkeeping requirements
• Sulfur hexafluoride

40 CFR Part 70

• Environmental protection
• Administrative practice and procedure
• Air pollution control
• Carbon dioxide
• Carbon dioxide equivalents
• Greenhouse gases
• Hydrofluorocarbons
• Intergovernmental relations
• Methane
• Nitrous oxide
• Perfluorocarbons
• Reporting and recordkeeping requirements
• Sulfur hexafluoride

40 CFR Part 71

• Environmental protection
• Administrative practice and procedure
• Air pollution control
• Carbon dioxide
• Carbon dioxide equivalents
• Greenhouse gases
• Hydrofluorocarbons
• Methane
• Nitrous oxide
• Perfluorocarbons
• Reporting and recordkeeping requirements
• Sulfur hexafluoride

For reasons stated in the preamble, title 40, chapter I of the Code of Federal Regulations is amended as set forth below.

PART 51—[AMENDED]

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

Authority: 23 U.S.C. 101; 42 U.S.C. 7401-7671q.

Subpart I—[Amended]

2. Section 51.166 is amended:

a. By adding paragraph (b)(48);

b. By revising paragraph (b)(49)(iv); and

c. By adding paragraph (b)(49)(v).

The revisions and additions read as follows:

PART 52—[AMENDED]

3. The authority citation for part 52 continues to read as follows:

Authority: 42 U.S.C. 7401, et seq.

Subpart A—[Amended]

4. Section 52.21 is amended:

a. By adding paragraph (b)(49);

b. By revising paragraph (b)(50)(iv); and

c. By adding paragraph (b)(50)(v).

The revisions and additions read as follows:

5. A new § 52.22 is added to read as follows:

PART 70—[AMENDED]

6. The authority citation for part 70 continues to read as follows:

Authority: 42 U.S.C. 7401, et seq.

7. Section 70.2 is amended:

a. By revising the introductory text of paragraph (2) of the definition for “major source”; and

b. By adding a definition for “Subject to regulation” in alphabetical order.

The revision and addition read as follows:

8. A new § 70.12 is added to read as follows:

PART 71—[AMENDED]

9. The authority citation for part 71 continues to read as follows:

Authority: 42 U.S.C. 7401, et seq.

Subpart A—[AMENDED]

10. Section 71.2 is amended:

a. By revising the introductory text of paragraph (2) of the definition for “major source”; and

b. By adding a definition for “Subject to regulation” in alphabetical order.

The revision and addition read as follows:

11. A new § 71.13 is added to subpart A to read as follows: