Response to Comment Document for the PSD and Nonattainment NSR:
Reasonable Possibility in Recordkeeping: Final Rule



 

Response to Comment Document for the PSD and Nonattainment NSR:
Reasonable Possibility in Recordkeeping: Final Rule

By: 

Office of Air Quality Planning and Standards

U.S. Environmental Protection Agency

Research Triangle Park, North Carolina

U.S. Environmental Protection Agency

Office of Air Quality Planning and Standards

Air Quality Policy Division

New Source Review Group

Research Triangle Park, NC

December 2007



Table of Contents

  TOC \o "1-3" \h \z \u    HYPERLINK \l "_Toc185418946"  Abbreviations
List	  PAGEREF _Toc185418946 \h  ii  

  HYPERLINK \l "_Toc185418947"  Chapter 1 - Introduction	  PAGEREF
_Toc185418947 \h  1  

  HYPERLINK \l "_Toc185418948"  Chapter 2 - Percentage Increase Trigger
  PAGEREF _Toc185418948 \h  6  

  HYPERLINK \l "_Toc185418949"  2.1	General Comments on "Percentage
Increase Trigger" Option	  PAGEREF _Toc185418949 \h  6  

  HYPERLINK \l "_Toc185418950"  2.2	Applicability	  PAGEREF
_Toc185418950 \h  7  

  HYPERLINK \l "_Toc185418951"  2.2.1	Percentage of Significant
Emissions Rate	  PAGEREF _Toc185418951 \h  7  

  HYPERLINK \l "_Toc185418952"  2.2.2	Apply Percentage to Net Emissions
Increase	  PAGEREF _Toc185418952 \h  11  

  HYPERLINK \l "_Toc185418953"  2.2.3	Emissions Increases are Based on
Project Size	  PAGEREF _Toc185418953 \h  11  

  HYPERLINK \l "_Toc185418954"  2.2.4	Applicability Exemptions	  PAGEREF
_Toc185418954 \h  12  

  HYPERLINK \l "_Toc185418955"  2.2.5	Apply Percentage as Part of a
Two-Step Test	  PAGEREF _Toc185418955 \h  15  

  HYPERLINK \l "_Toc185418956"  2.3	Calculating Baseline Emissions	 
PAGEREF _Toc185418956 \h  15  

  HYPERLINK \l "_Toc185418957"  2.4	Differing Recordkeeping and
Reporting Requirements	  PAGEREF _Toc185418957 \h  18  

  HYPERLINK \l "_Toc185418958"  2.5	Need to Include Verification
Requirements or Permitting Authority Oversight Within the Permitting
Process	  PAGEREF _Toc185418958 \h  19  

  HYPERLINK \l "_Toc185418959"  Chapter 3 - "Potential Emissions
Trigger" Option	  PAGEREF _Toc185418959 \h  22  

  HYPERLINK \l "_Toc185418960"  3.1.	General Comments on "Potential
Emissions Trigger" Option	  PAGEREF _Toc185418960 \h  22  

  HYPERLINK \l "_Toc185418961"  3.1.1	Support "Potential Emissions
Trigger"	  PAGEREF _Toc185418961 \h  22  

  HYPERLINK \l "_Toc185418962"  3.1.2	Oppose "Potential Emissions
Trigger" Option	  PAGEREF _Toc185418962 \h  23  

  HYPERLINK \l "_Toc185418963"  3.2	Applicability	  PAGEREF
_Toc185418963 \h  26  

  HYPERLINK \l "_Toc185418964"  3.2.1	Sources Operating at Less Than
Full Capacity	  PAGEREF _Toc185418964 \h  26  

  HYPERLINK \l "_Toc185418965"  3.2.2	Source Emissions Increases Linked
to Project/Projected Actual Emissions	  PAGEREF _Toc185418965 \h  27  

  HYPERLINK \l "_Toc185418966"  3.3	Other Comments on "Potential
Emissions Trigger" Option	  PAGEREF _Toc185418966 \h  28  

  HYPERLINK \l "_Toc185418967"  Chapter 4 - Other Miscellaneous Comments
  PAGEREF _Toc185418967 \h  31  

  HYPERLINK \l "_Toc185418968"  4.1	Burden Associated With Recordkeeping
Requirements	  PAGEREF _Toc185418968 \h  31  

  HYPERLINK \l "_Toc185418969"  4.2	Pre-Existing
Records/Reports/Inventories Sufficient to Enforce NSR Provisions	 
PAGEREF _Toc185418969 \h  32  

  HYPERLINK \l "_Toc185418970"  4.3	Appendix S	  PAGEREF _Toc185418970
\h  36  

  HYPERLINK \l "_Toc185418971"  4.4	Utility versus Non-Utility
Requirements	  PAGEREF _Toc185418971 \h  37  

  HYPERLINK \l "_Toc185418972"  4.5	Miscellaneous Implementation Issue	 
PAGEREF _Toc185418972 \h  38  

  HYPERLINK \l "_Toc185418973"  4.6	Miscellaneous
Clarifications/Corrections	  PAGEREF _Toc185418973 \h  38  

 

Abbreviations List

BACT			Best Available Control Technology

CAA			Clean Air Act

CFR			Code of Federal Regulation

EUSGU		Electric Utility Steam Generating Unit

EPA			United States Environmental Protection Agency

FDMS			Federal Docket Management System

FERC			Federal Energy Regulatory Commission

Fed. Reg.		Federal Register

FR			Federal Register

H2S			Hydrogen Sulfide

LAER			Lowest Achievable Emission Rate

MACT			Maximum Achievable Control Technology

NAAQS		National Ambient Air Quality Standards

NSPS			New Source Performance Standards

NSR			New Source Review

OAQPS		Office of Air Quality Planning and Standards

PSD			Prevention of Significant Deterioration

PTE			Potential to emit

SIP			State Implementation Plan

tpy			Tons per year

TRE			Total Resource Effectiveness index value

TRS			Total Reduced Sulfur

Chapter 1 - Introduction

In the 2002 NSR reform rules (67 FR 80186, December 31, 2002), we
promulgated an actual-to-projected-actual methodology for major NSR
applicability determinations.  That rule provides that if a source
calculates its projected actual emissions for a project to be below
major NSR applicable significant emissions rates, the source must comply
with recordkeeping and, in some cases, reporting requirements, if there
is a "reasonable possibility" that the project would result in a
significant emissions increase (e.g., 40 CFR 52.21(r)(6)).  We did not
specifically define the term "reasonable possibility" or specifically
identify the criteria under which a "reasonable possibility" would
arise.

The U.S. Court of Appeals for the D.C. Circuit in New York v. EPA, 413
F.3d 3 (D.C. Cir. 2005) (New York) remanded for the EPA either to
provide an acceptable explanation for its "reasonable possibility"
standard or to devise an appropriately supported alternative.  In
response, on March 8, 2007 (45 FR 10445), we proposed two options.  The
options included the "percentage increase trigger" and the "potential
emissions trigger."

  

The public comment period ended on May 7, 2007.  We received
38 comment letters (excluding duplicates and not including
attachments).  Table 1 presents the list of comment letters and
attachments (including duplicates) submitted to the docket for this
rulemaking (Docket ID No. EPA-HQ-OAR-2001-0004, at
http://www.regulations.gov) in response to the proposed rule, "PSD and
Nonattainment NSR: Reasonable Possibility in Recordkeeping."  The
breakdown of commenters is: one citizen letter, 29 industry comment
letters (13 from utility companies), and 8 State/local comment letters.
 No comments from environmental groups or Tribal organizations were
submitted.

Table 1.  Comment Letters Received on Proposed Rule

Docket Identification Number	

Description/Affiliation



EPA-HQ-OAR-2001-0004-0792	

Comment submitted by M. Weniger

EPA-HQ-OAR-2001-0004-0793	Comment submitted by John A. Paul,
Administrator, Regional Air Pollution Control Agency (RAPCA), Dayton,
Ohio

EPA-HQ-OAR-2001-0004-0793.1	Comment Attachment submitted by John A.
Paul, Administrator, Regional Air Pollution Control Agency (RAPCA),
Dayton, Ohio 

EPA-HQ-OAR-2001-0004-0794	Comment submitted by Stephen Fogle,
Engineering Coordinator, Company Facilities/Environmental, Honda of
America Manufacturing, Inc.

EPA-HQ-OAR-2001-0004-0795	Comment submitted by Richard J. Sandberg,
Manager, Air Quality Permits Section, Industrial Division, Minnesota
Pollution Control Agency (MPCA)

EPA-HQ-OAR-2001-0004-0796	Comment submitted by Michael W. Stroben,
Environment, Health & Safety Policy Director, Duke Energy

EPA-HQ-OAR-2001-0004-0796.1	Comment attachment submitted by Michael W.
Stroben, Environment, Health & Safety Policy Director, Duke Energy

EPA-HQ-OAR-2001-0004-0797	Comment submitted by Makram Jaber, Hunton &
Williams on behalf of Utility Air Regulatory Group (UARG)

EPA-HQ-OAR-2001-0004-0797.1	Comment attachment submitted by Makram
Jaber, Hunton & Williams on behalf of Utility Air Regulatory Group
(UARG) 

EPA-HQ-OAR-2001-0004-0798	Comment submitted by William Neal, Chair of
LPPC Environmental Task Force and Division Manager Environmental and
Regulatory Affairs, Omaha Public Power District

EPA-HQ-OAR-2001-0004-0798.1	Comment attachment submitted by William
Neal, Chair of LPPC Environmental Task Force and Division Manager
Environmental and Regulatory Affairs, Omaha Public Power District 

EPA-HQ-OAR-2001-0004-0799	Comment submitted by Bernard Paul, Eli Lilly
and Company

EPA-HQ-OAR-2001-0004-0799.1	Comment attachment submitted by Bernard
Paul, Eli Lilly and Company 

EPA-HQ-OAR-2001-0004-0800	Comment submitted by David Friedman, Director,
Environmental Affairs, National Petrochemical & Refiners Association
(NPRA)

EPA-HQ-OAR-2001-0004-0800.1	Comment attachment submitted by David
Friedman, Director, Environmental Affairs, National Petrochemical &
Refiners Association (NPRA) 

EPA-HQ-OAR-2001-0004-0801	Comment submitted by Robert D. Bessett,
President, The Council of Industrial Boiler Owners (CIBO)

EPA-HQ-OAR-2001-0004-0801.1	Comment attachment submitted by Robert D.
Bessett, President, The Council of Industrial Boiler Owners (CIBO) 

EPA-HQ-OAR-2001-0004-0802	Comment submitted by Timothy G. Hunt, Senior
Director, Air Quality Programs, American Forest and Paper Association.

EPA-HQ-OAR-2001-0004-0802.1	Comment attachment submitted by Timothy G.
Hunt, Senior Director, Air Quality Programs, American Forest and Paper
Association.  

EPA-HQ-OAR-2001-0004-0803	Comment submitted by Chris M. Hobson, Senior
Vice President, Research and Environmental Affairs, Southern Company

EPA-HQ-OAR-2001-0004-0803.1	Comment attachment submitted by Chris M.
Hobson, Senior Vice President, Research and Environmental Affairs,
Southern Company 

EPA-HQ-OAR-2001-0004-0804	Comment submitted by Shawn Glacken, Vice
President, Environmental Policy, TXU Power 

EPA-HQ-OAR-2001-0004-0804.1	Comment attachment submitted by Shawn
Glacken, Vice President, Environmental Policy, TXU Power 

EPA-HQ-OAR-2001-0004-0805	Comment submitted by Leslie Ritts, Counsel,
The National Environmental Development Association's Clear Air Project

EPA-HQ-OAR-2001-0004-0805.1	Comment attachment submitted by Leslie
Ritts, Counsel, The National Environmental Development Association's
Clear Air Project 

EPA-HQ-OAR-2001-0004-0806	Comment submitted by Lisa A. Wilkinson, Senior
Attorney, New York State Department of Environmental Conservation 

EPA-HQ-OAR-2001-0004-0806.1	Comment attachment submitted by Lisa A.
Wilkinson, Senior Attorney, New York State Department of Environmental
Conservation 

EPA-HQ-OAR-2001-0004-0807	Comment submitted by Patrick A. Dal Porto,
American Electric Power System (AEP)

EPA-HQ-OAR-2001-0004-0807.1	Comment attachment submitted by Patrick A.
Dal Porto, American Electric Power System (AEP)

EPA-HQ-OAR-2001-0004-0808	Comment submitted by M. Gary Helm, Senior
Environmental Coordinator, Conectiv Energy

EPA-HQ-OAR-2001-0004-0808.1	Comment attachment submitted by M. Gary
Helm, Senior Environmental Coordinator, Conectiv Energy

EPA-HQ-OAR-2001-0004-0809	Comment submitted by Francis Streitman, Vice
President Environmental Affairs, Ash Grove Cement Company

EPA-HQ-OAR-2001-0004-0809.1	Comment attachment submitted by Francis
Streitman, Vice President Environmental Affairs, Ash Grove Cement
Company

EPA-HQ-OAR-2001-0004-0810	Comment submitted by Michael J. Myers,
Assistant Attorney General, New York State Attorney General

EPA-HQ-OAR-2001-0004-0810.1	Comment attachment submitted by Michael J.
Myers, Assistant Attorney General, New York State Attorney General

EPA-HQ-OAR-2001-0004-0811	Comment submitted by Andrew N. Sawula, Schiff
Hardin LLP, NiSource Inc 

EPA-HQ-OAR-2001-0004-0811.1	Comment attachment submitted by Andrew N.
Sawula, Schiff Hardin LLP, NiSource Inc

EPA-HQ-OAR-2001-0004-0812	Comment submitted by Shannon S. Broome, Air
Permitting Forum 

EPA-HQ-OAR-2001-0004-0812.1	Comment attachment submitted by Shannon S.
Broome, Air Permitting Forum

EPA-HQ-OAR-2001-0004-0813	Comment submitted by Barry R. Wallerstein,
Executive Officer, South Coast Air Quality Management District 

EPA-HQ-OAR-2001-0004-0813.1	Comment attachment submitted by Barry R.
Wallerstein, Executive Officer, South Coast Air Quality Management
District 

EPA-HQ-OAR-2001-0004-0813.2	Comment attachment submitted by Barry R.
Wallerstein, Executive Officer, South Coast Air Quality Management
District 

EPA-HQ-OAR-2001-0004-0814	Comment submitted by Frank W. Rambo, Baker
Botts, LLP, Class of '85 Regulatory Response Group

EPA-HQ-OAR-2001-0004-0814.1	Comment attachment submitted by Frank W.
Rambo, Baker Botts, LLP, Class of '85 Regulatory Response Group

EPA-HQ-OAR-2001-0004-0815	Comment submitted by Andrew N. Sawula, Schiff
Hardin LLP, Midwest Generation EME, LLC and Homer City EME LLC 

EPA-HQ-OAR-2001-0004-0815.1	Comment attachment submitted by Andrew N.
Sawula, Schiff Hardin LLP, on behalf of Basil G. Constantelos, Director,
Environmental, Health & Safety, Midwest Generation EME, LLC and Homer
City EME LLC 

EPA-HQ-OAR-2001-0004-0816	Comment submitted by Shannon Broome, on behalf
of Julie C. Becker, Assistant General Counsel, Alliance of Automobile
Manufacturers

EPA-HQ-OAR-2001-0004-0816.1	Comment attachment submitted by Shannon
Broome, on behalf of Julie C. Becker, Assistant General Counsel,
Alliance of Automobile Manufacturers

EPA-HQ-OAR-2001-0004-0817	Comment submitted by Nancy C. Wrona, Director,
Air Quality Division, Arizona Department of Environmental Quality (ADEQ)

EPA-HQ-OAR-2001-0004-0817.1	Comment attachment submitted by Nancy C.
Wrona, Director, Air Quality Division, Arizona Department of
Environmental Quality (ADEQ)

EPA-HQ-OAR-2001-0004-0818	Comment submitted by Steven B. Lomax, Manager,
Air Quality Programs, Edison Electric Institute (EEI)

EPA-HQ-OAR-2001-0004-0818.1	Comment attachment submitted by Steven B.
Lomax, Manager, Air Quality Programs, Edison Electric Institute (EEI)

EPA-HQ-OAR-2001-0004-0819	Comment submitted by Robert J. Morehouse,
Downstream and Chemical SH&E, ExxonMobil

EPA-HQ-OAR-2001-0004-0819.1	Comment attachment submitted by Robert J.
Morehouse, Downstream and Chemical SH&E, ExxonMobil

EPA-HQ-OAR-2001-0004-0820	Comment submitted by William C. Herz, Vice
President, Scientific Programs, The Fertilizer Institute (TFI)

EPA-HQ-OAR-2001-0004-0820.1	Comment attachment submitted by William C.
Herz, Vice President, Scientific Programs, The Fertilizer Institute
(TFI)

EPA-HQ-OAR-2001-0004-0821	Comment submitted by Marjorie Gail Twymon,
Advanced Environmental Specialist, Environmental Department, FirstEnergy
Corp.

EPA-HQ-OAR-2001-0004-0821.1	Comment attachment submitted by Marjorie
Gail Twymon, Advanced Environmental Specialist, Environmental
Department, FirstEnergy Corp.

EPA-HQ-OAR-2001-0004-0822	Comment submitted by Bill O(Sullivan and John
Paul, Co-Chairs, New Source Review (NSR) Committee, National Association
of Clean Air Agencies (NACAA)

EPA-HQ-OAR-2001-0004-0822.1	Comment attachment submitted by Bill
O(Sullivan and John Paul, Co-Chairs, New Source Review (NSR) Committee,
National Association of Clean Air Agencies (NACAA)

EPA-HQ-OAR-2001-0004-0823	Comment submitted by Bryan L. Brendle,
Director, Energy and Resources Policy, The National Association of
Manufacturers (NAM)

EPA-HQ-OAR-2001-0004-0824	Comment submitted by Arthur N. Marin,
Executive Director, Northeast States for Coordinated Air Use Management
(NESCAUM)

EPA-HQ-OAR-2001-0004-0824.1	Comment attachment submitted by Arthur N.
Marin, Executive Director, Northeast States for Coordinated Air Use
Management (NESCAUM)

EPA-HQ-OAR-2001-0004-0825	Comment submitted by Thomas B. Carter, Staff
Vice President, Environment, Health and Safety, Portland Cement
Association (PCA)

EPA-HQ-OAR-2001-0004-0825.1	Comment attachment submitted by Thomas B.
Carter, Staff Vice President, Environment, Health and Safety, Portland
Cement Association (PCA)

EPA-HQ-OAR-2001-0004-0826	Please refer to an updated version of this
document, which is document number OAR-2001-0004-0827

EPA-HQ-OAR-2001-0004-0826.1	Please refer to an updated version of this
document, which is document number OAR-2001-0004-0827.1

EPA-HQ-OAR-2001-0004-0827	Comment submitted by Timothy G. Hunt, Senior
Director, Air Quality Programs, American Forest and Paper Association. 
Duplicate.

EPA-HQ-OAR-2001-0004-0827.1	Comment attachment submitted by Timothy G.
Hunt, Senior Director, Air Quality Programs, American Forest and Paper
Association.  Duplicate. 

EPA-HQ-OAR-2001-0004-0828	Comment submitted by Dean Metcalf, Director,
Air and Water, Xcel Energy Environmental Services

EPA-HQ-OAR-2001-0004-0829	Comment submitted by William H. Lewis, Morgan
Lewis, Counselors at Law, on behalf of the Clean Air Implementation
Project, American Chemistry Council and American Petroleum Institute

EPA-HQ-OAR-2001-0004-0830	Comment submitted by Barry Christensen,
Manager, Air Quality, Occidental Chemical Corporation 

EPA-HQ-OAR-2001-0004-0831	Comment submitted by Pamela F. Faggert, Vice
President and Chief Environmental Officer, Dominion Resources Services,
Inc. 

EPA-HQ-OAR-2001-0004-0832	Comment submitted by Michael J. Myers,
Assistant Attorney General, Environmental Protection Bureau, New York
State Attorney General on behalf of various State Attorney Generals.

EPA-HQ-OAR-2001-0004-0832.1	Comment attachment submitted by Michael J.
Myers, Assistant Attorney General, Environmental Protection Bureau, New
York State Attorney General on behalf of various State Attorney
Generals.

Chapter 2 - Percentage Increase Trigger

We proposed two options for identifying the circumstances under which
the increase in emissions caused by a project triggers the "reasonable
possibility" recordkeeping and reporting requirements (72 FR 10445). 
The first option, our preferred option, proposed using a "percentage
increase trigger" test.  Under this proposed test, a major source must
comply with the "reasonable possibility" recordkeeping requirements if
the change’s projected actual emissions increase equals or exceeds
a percentage of the applicable NSR significant emissions rate for that 
pollutant.  Our proposed value was 50 percent, but we solicited comment
on alternative percentage values.  This chapter presents comments
received on the percentage increase trigger test.

2.1	General Comments on "Percentage Increase Trigger" Option

Comment:

Most of the comments received from industry (@0794, @0796, @0797, @0798,
@0800, @0801, @0802, @0803, @0804, @0805, @0807, @0808, @0809, @0812,
@0814, @0816, @0818, @0819, @0820, @0821, @0828, @0830, @0831) provided
support for the "percentage increase trigger" approach over the
"potential emissions trigger" test, but not all supported the proposed
50 percent of the applicable NSR significant emissions rate.  General
comments received supporting the "percentage increase trigger" option
are presented in the following paragraphs.

Most of the commenters supported the "percentage emissions trigger"
approach because it is based on anticipated actual emissions versus
potential emissions after a source change.  One commenter (@0805)
specifically expressed a preference for the "percentage increase
trigger" to the "potential emissions trigger," because they believed the
reasonable possibility test should be based on the same actual emissions
calculations and assumptions, including historical demand, that the
source used to project actual emissions increases.

One commenter (@0794) believed the percentage increase trigger option
would allow the recordkeeping flexibility that the EPA allows under the
NSR Reform rules for small projects.  Two commenters (@0804, @0821)
supported the "percentage emissions trigger" approach, believing it sets
a workable methodology for keeping records only for projects that have a
likelihood of triggering NSR rather than for every single nonroutine
maintenance project.

One commenter (@0828) supported the "percentage increase trigger"
option, noting that it takes into account the impracticality of having
to keep special records (and, for EUSGUs, having to report the data)
when anticipating no increase or only a small increase in emissions.

Several commenters (@0804, @0807, @0808, @0814, @0818, @0828, @0830)
supported adoption of a "percentage increase trigger," believing such a
trigger clearly defines the circumstances under which records must be
kept for projects that are not projected to cause a significant
emissions increase.  One of these commenters (@0814) noted that to
express the trigger as a percentage increase would allow sources to
determine with precision and certainty what the applicable standard is
in advance of a change.  Two of these commenters (@0814, @0818) believed
the "percentage increase trigger" approach comports with the Court's
recognition "that less burdensome requirements may well be appropriate
for sources with little likelihood of triggering NSR" and with the
Court's remand that the EPA needs to "explain how its recordkeeping and
reporting requirements allow it to identify such sources."  One of these
commenters (@0830) provided that the "percentage increase trigger"
option would provide sufficient information to alert inspectors to
situations when records should be available for review.

Several commenters (@0796, @0797, @0807) observed that the "percentage
increase trigger" option recognizes the importance of a causal link
between a project and some emission increase following the project in
determining whether the project could potentially be a major
modification.

One commenter (@0798) preferred the "percentage increase trigger"
because the commenter believed the "potential emissions trigger" is
overly broad and would likely require compliance with recordkeeping and
reporting in the vast majority of cases.

Response:  

Based on our evaluation and consideration of comments received, which
largely supported the "percentage increase trigger" option, we are
finalizing that option, with one change to address concerns raised by
commenters.  The sections that follow provide more specific details of
the comments concerning both the "percentage increase trigger" and
"potential emissions trigger" options.

2.2	Applicability

2.2.1	Percentage of Significant Emissions Rate

We proposed to use 50 percent of the applicable NSR significant
emissions rate for the relevant regulated NSR pollutant as the trigger
under the "percentage increase trigger" option.  However, we solicited
comment on use of other possible values, such as 25, 33, 66 or
75 percent, to trigger the recordkeeping and reporting requirements.  

A total of 19 commenters supported setting the trigger at 50 percent or
higher.  Seven specifically supported a trigger of 50 percent.  

Of the numerous comments we received in support of the "percentage
increase trigger" option, all comments that expressed support for a
specific percentage favored a percentage of 50 percent or higher. 
Commenters supported EPA's creating a "bright line" for assessing when
recordkeeping and reporting are appropriate.  Following are comments
recommending how the percentage should be applied and the percentage
value of the applicable NSR significant emissions rate that should
apply.

Comment:

One commenter (@0800) noted it appropriate for EPA to engage in a
line-drawing exercise to assess when recordkeeping and reporting are
appropriate, i.e., when there is a reasonable possibility of exceeding
applicable NSR significant emissions rates.

Commenter (@0828) believed a trigger lower than 50 percent would be
unnecessary.  The commenter noted that the applicable NSR significant
emissions rates are already relatively small, and an additional safety
net for compliance concerns is that the trigger is applied to increases
before accounting for creditable emission decreases (that is, no
netting).

Commenter (@0814) endorsed EPA's proposal to set the trigger at
50 percent of the applicable NSR significant emissions rate for each
regulated NSR pollutant.  The commenter believed that level to be
conservative not only because the applicable NSR significant emissions
rates themselves are so small but also because, unlike for NSR
applicability, the recordkeeping threshold does not count project
decreases (netting).

Several commenters (@0802, @0803, @0806, @0820, @0821) supported the
50-percent threshold, believing it to provide certainty or a bright-line
test for sources to when a reasonable possibility exists that a change
will result in a significant emissions increase when seeking to comply
and to eliminate burdensome requirements for projects that are only
expected to result in a very small increase in emissions. 

Commenter (@0820) and its member companies support a threshold of
50 percent, believing it addresses concerns associated with predicting
uncertain future events such as malfunctions.

One commenter (@0798) was concerned that a 50-percent threshold applied
to "relatively quite small" applicable NSR significant emissions rates
(ranging from 15 to 40 tpy for most pollutants) would allow only a
limited number of power plant construction projects to avoid triggering
recordkeeping and reporting requirements.  Another commenter (@0799)
recommended that if EPA chooses to adopt an emissions-based trigger as a
basis for defining reasonable possibility, it should be greater than
50 percent of any of the NSR significant emissions thresholds.  This
commenter recommended a two-step test, as discussed in section 2.2.5 of
this Response to Comments. 

Seven commenters (@0808, @0809, @0812, @0816, @0823, @0825, @0829)
recommended a threshold of 75 percent.  One of these commenters (@0816)
supported the percentage increase trigger at a 75-percent threshold in
combination with a "potential emissions trigger" test, which is
discussed in section 3.1.2 of this Response to Comments.  

 Two commenters (@0809, @0825) claimed that emissions projections,
especially for modifications, are precise enough in most cases for there
to be a reasonable degree of certainty that the projection will not be
off by more than 25 percent of the applicable NSR significant emissions
rate.  The commenters recommended a threshold of at least 75 percent
and further recommended that the threshold be established as a
rebuttable presumption, meaning a source would be able to exceed the
threshold without triggering recordkeeping requirements if the facility
is able to demonstrate a high level of certainty as to the project's
emissions increase.  As an example of a project with a high level of
certainty as to the emissions increase, the commenters provided the
example of a facility that is modifying an emissions unit in a manner
consistent with what it has done for five other identical emissions
units in the past.  In support of the rebuttable presumption approach,
the commenters noted that EPA has successfully used a similar approach
in its used oil regulations, where oil with halogen levels above a
specified threshold is presumed to be hazardous waste unless the
presumption is rebutted.

Commenter (@0812) noted that a 75-percent threshold creates a 33-percent
margin for error in emissions projections.  For example, if the
significant threshold for a pollutant is 40 tons per year, and a project
is estimated to increase actual emissions by 30 tons per year, under the
75-percent threshold it would not have a reasonable possibility of
exceeding applicable NSR significant emissions rates.  For NSR to be
triggered in this situation, the project would need to have an increase
of 10 more tons, or one-third more than the source's projection.  The
commenter believed that such a large margin of safety virtually ensures
that changes with a reasonable possibility of significantly increasing
emissions will be tracked and reported.  Two commenters (@0816, @0829)
believed a 75-percent threshold to allow a 25-percent buffer or margin
of error in the emissions forecast or a substantial buffer between the
trigger level and applicable NSR significant emissions rates.

One of the commenters (@0823) that supported a threshold of 75 percent,
however, believed that a source owner should be required to maintain the
original documentation showing the basis for the emissions increase
calculations and the determination that no "reasonable possibility"
recordkeeping or reporting provisions apply.

 

One commenter (@0808) believed that most regulated entities take great
care in determining future actual emissions in order to determine
whether a modification will cause a significant emission increase. 
Because of this, the commenter does not believe a "reasonable
possibility" exists that a modification will cause a significant
emission increase unless the projected increase is at least 75 to
80 percent of the applicable NSR significant emissions rate, and the
commenter recommends that the trigger be set at 75 percent of the
significant emissions rate.

One commenter (@0800) believed a 75-percent or 90-percent threshold
could be supported.  This commenter believed emissions decreases should
be taken into account when calculating projected actual emissions.  This
commenter (@0800) also noted that 10 to 25 percent (such as through a
90-percent or 75-percent threshold) is a substantial buffer against a
highly unlikely event that an error in estimating actual emissions would
result in emissions that trigger NSR.  The commenter stated that they
believed EPA's proposal, even if at a 75-percent threshold, is extremely
conservative on the side of triggering requirements.

Commenter (@0805) stated that the 50-percent trigger is a very
conservative value (even more so when considering sources cannot count
emission decreases against increases from the project) that could
potentially sweep a large number of insignificant changes into a
regulatory program already riddled with redundant recordkeeping.  The
commenter believed 90 percent would provide a suitable basis and urged
EPA not to adopt a value lower than 75 percent.

Two commenters (@0801, @0819) supported a trigger of 90 percent of the
applicable NSR significant emissions rate.  Commenter (@0801) believed a
50-percent threshold was too low, requiring significant accounting and
tracking of emissions for too many small changes that do not present a
reasonable possibility of a significant emissions increase. 

One commenter (@0802) recommended multiple trigger percentages and
associated recordkeeping and reporting requirements depending on
the percentage of the applicable NSR significant emissions rate.  This
commenter’s proposed alternative is discussed in section 2.4 of this
Response to Comments.

Response:

We respond to most of these comments in the preamble to the rulemaking,
including a justification for the 50-percent level.  We note further
that commenters who recommended a level above 50 percent did not
provide specific examples of changes that should not be subject to the
reasonable possibility requirements and would not be subject if the
level were higher than 50 percent, but would be subject at the
50-percent level. We did not adopt commenters’ suggestion for a
rebuttable presumption because we believe it would be difficult to
establish standards for when the rebuttable presumption would be met. 
We note that the suggestion of one commenter (@0823) fits with our final
rule's approach, under which some projects are adequately documented
with only pre-change records while other projects merit additional
recordkeeping and reporting requirements.

2.2.2	Apply Percentage to Net Emissions Increase

Comment:

Four commenters (@0800, @0802, @0812, @0816) noted that EPA recently
proposed a "project netting" approach and believed the percentage could
or even should be applied to a projected net emissions increase (that
is, counting both the emissions increases and emissions reductions
resulting from the project).  One of these commenters (@0802) added that
any other approach would result in an inconsistent and indefensible
regulatory framework.  Another commenter (@0803) recommended that EPA
allow sources the option of applying the "percentage increase trigger"
based on net emissions increases and believed that to prohibit sources
from applying the percentage increase trigger on a net emissions basis
would result in needless reporting every time a source conducts a
netting analysis.  Commenter (@0819) believed EPA should clarify that
the percentage is to be applied to the project's net emissions increase,
noting that the true representation of the project is the net emissions
impact.  Commenters urged EPA to coordinate the "reasonable possibility"
rule with an anticipated "project netting" rule, such that decreases as
well as increases are counted in step 1 of the NSR emissions analysis
and for purposes of applying the "reasonable possibility" threshold.

Response:

Under our current interpretation of the major NSR rules, sources may not
"project net" to avoid major NSR.  We specifically designed the
"reasonable possibility" recordkeeping and reporting requirements to
address the increase in emissions from a project without consideration
of contemporaneous or project emissions decreases.  Nonetheless, EPA is
currently seeking comment on revising our rules and interpretation of
major NSR to allow for project netting, 71 FR 54235, and this comment
may be relevant in the context of taking final action on that
rulemaking.

2.2.3	Emissions Increases are Based on Project Size

Comment:

A few commenters (@0796, @0803, @0816) requested that EPA clarify its
statement in the proposal preamble that "very large sources are less
likely to make changes that are covered by the 'reasonable possibility'
standard because virtually any change that a very large source makes may
be expected to increase emissions above the applicable NSR significant
emissions rates and require a major NSR permit."  72 Fed. Reg. at
10,450.  Commenter (@0803) also disputed this characterization of large
sources.  Commenter (@0812) disagreed with EPA's statement, noting that
the determination of whether a reasonable possibility exists depends on
the project rather than on the unit where it occurs.  Commenter (@0816)
noted that even changes on a large unit, e.g., to address a paint
quality issue, may not be expected to result in a significant net
emission increase.

One commenter (@0800) noted that even large refineries have numerous
small emission units and undertake many small projects that could not
reasonably be expected to trigger NSR.  The commenter noted that even
for projects involving units with larger potential emissions there are
often "small projects" for which there is no reasonable possibility of
exceeding applicable NSR significant emissions rates.

Response:

We acknowledge the commenters' points and provide clarification.  To
make the case that not all projects have equal likelihood of resulting
in a significant emissions increase, we wished to illustrate that
"reasonable possibility" of a significant emissions increase could be
correlated to project size.  Our explanation should more clearly have
been that we believe smaller projects will typically result in smaller
emissions increases, and so the likelihood of a significant emissions
increase, even if the source were to underestimate emissions by a
large percentage, would be so low as to make recordkeeping and
reporting requirements less justifiable than for larger projects.  Along
this same line of thinking, we would claim that a larger project (not
source, as the commenters pointed out) would typically result in a
projected emissions increase more likely to exceed significant emissions
rates and so more likely to trigger the recordkeeping and reporting
requirements under our "percentage increase trigger" option.  We
recognize that even very large sources annually undertake many smaller
projects that are not major modifications.  We also recognize that the
correlation between amount of emissions increase and likelihood of
significant emissions increase may but does not always hold true. 
Nevertheless, where a proposed project, regardless of the size of the
existing source, will involve only a very small increase in emissions,
the 50-percent test should allow that source to be exempted from the
monitoring and reporting requirements associated with the "percentage
increase trigger" test.

2.2.4	Applicability Exemptions

Comment:

A couple of commenters (@0801, @0805) requested that routine changes and
equipment maintenance projects be exempt from the reasonable possibility
standard.  Commenter (@0801) asked EPA to clarify that the reasonable
possibility standard does not apply to routine maintenance and repair
that would not increase emissions.  The commenter suggested that sources
could document routine changes simply with project objectives and a
technical rationale as to why the project will not cause any increase in
emissions.  The other commenter (@0805) noted that manufacturers have
special concerns about equipment maintenance projects being subjected to
the reasonable possibility standard and so asked EPA to expressly
clarify that routine maintenance projects, including but not limited to
like-kind replacements, are not included under reasonable possibility
because they are exempt from the definition of "modification." 

One commenter (@0801) suggested that EPA promulgate a simpler level of
assessment for projects not expected to affect capacity or increase
emissions.  The commenter noted that it would be far easier to
demonstrate that these types of projects do not cause an emissions
increase than to use EPA's formula to calculate a baseline, estimate
future actual emissions, determine the demand growth exclusion, and
track all of the above for every minor project.

Commenter (@0802) believed there are certain changes that this rule
should exempt.  In the case where the post-change PTE of the affected
unit or units will be less than significant thresholds, EPA's rule
should recognize that a significant emissions increase is logically
impossible.  In the case where the change is subject to minor source
permit requirements that disallow a significant emissions increase,
EPA's rule should find that such permit limits provide conclusive proof
that the change in question did not trigger NSR.  In another case, the
change will be unable to cause a significant emissions increase because
its effect on the price or other competitive characteristics of a
product will be so small. 

One commenter (@0799) recommended that the reasonable possibility
provision exempt from recordkeeping any project that is reviewed under a
State minor NSR program or a title V permit modification if either of
those processes determines that NSR is not applicable.  

Several commenters listed specific examples of changes that they
believed would trigger reasonable possibility requirements, but for
which they believed such requirements would not be appropriate.  These
include:  

(i) "a project that involves the shifting of production from an older
emissions unit, which is being phased out, to a more recently built
unit."  (Nat’l Petrochemicals & Refiners Ass’n, p. 3 n.3; Alliance
of Automobile Manuf., p. 5 n.3)

(ii) "even small changes on a large unit – e.g. to address a paint
quality issue –" (Alliance of Automobile Manuf., p. 4)

(iii) "raw material or process changes – converting from solvent-borne
to waterborne topcoats" (Alliance of Automobile Manuf., p. 8)

(iv) "any changes related to electrocoat operations" (Alliance of
Automobile Manuf., p. 8)

(v) "adding an economizer to a boiler to capture waste heat."  (Alliance
of Automobile Manuf., p. 8)

(vi) powder coating, repair coatings, sealer, adhesive, etc.  (Alliance
of Automobile Manuf., p. 8)

(vii) "changes or additions to natural gas-fired equipment such as space
heaters" (Alliance of Automobile Manuf., p. 8)

(viii) "construction of a new exercise room at a facility" (Duke, p. 2;
UARG, p. 5)

Response:

With regard to the comments expressing concern that a threshold at
50 percent of applicable NSR significant emissions rates would capture
too many small projects, including routine maintenance projects, we
respond by pointing out that the "reasonable possibility" standard
applies when a major source undergoes a physical change or change in the
method of operation.  That is, in defining "major modification," the
major NSR regulations specify that a "physical change or change in the
method of operation" excludes routine maintenance, repair, and
replacement, certain uses of alternative fuel or raw material, certain
increases in hours of operation or production rate, changes in
ownership, and certain activities associated with clean coal technology.
 (See, e.g., 40 CFR 52.21(b)(2).)  Thus, a project that is not a
"physical change or change in the method of operation" is not subject to
"reasonable possibility" recordkeeping and reporting requirements.

We appreciate the specific examples of projects provided by commenters
that they believe should not be subject to the reasonability
requirements.  However, their comments generally did not provide
sufficient information about the nature of the projects, or why the
commenters believe that such projects would not result in emissions
increases, to allow an assessment of how the reasonable possibility rule
would apply to those projects.  We note that many of the specific
examples may well be examples of routine maintenance and other
exemptions that would not be considered changes that could trigger the
reasonable possibility requirements.

We agree with the comment indicating that where the post-change PTE of
the affected unit or units will be less than significant thresholds, a
significant emissions increase is logically impossible.  Moreover, we
point out that sources with a PTE that would not result in a significant
emissions increase over the baseline actual emissions may use the PTE as
their projected actual emissions, in accordance with 40 CFR
52.21(b)(41)(ii)(d), to demonstrate that no significant emissions
increase will occur.

We appreciate the suggestions from commenters to exempt from the
reasonable possibility standard those projects that are prohibited under
minor source permit requirements from emitting at a significant
emissions rate or that are exempt from State minor NSR or title V
permit requirements.  We believe these suggestions have merit but
because we did not specifically solicit comment on this topic, we are
not revising the reasonable possibility standard to incorporate them. 
However, we will consider the recommended approaches on a case-by-case
basis within the context of approving reasonable possibility NSR SIP
revisions. 

Comments received on the adequacy of pre-existing requirements
(including State minor NSR program or title V permits) are discussed in
chapter 4 of this Response to Comments.

2.2.5	Apply Percentage as Part of a Two-Step Test

Comment:

One commenter (@0799) stated a belief that to apply a percentage
threshold to projected increase in actual emissions is an overly
simplistic expression of the probability as to whether the actual
emissions might increase above applicable NSR significant emissions
rates.  The commenter acknowledged that it makes sense that as the
emissions increase approaches major modification levels, there is a
higher probability that the project could ultimately cause a significant
emissions increase.  However, the commenter believed, the single
greatest factor that would cause a significant emissions increase would
be an increase in capacity or production rate.  The commenter explained
that if the project did not involve a change in operating capability,
such as capacity, then the probability that the change would trigger NSR
would be low.  With this in mind, the commenter recommended that EPA
apply a two-step test, such that if the source first determines that the
project would result in no increase in PTE, then no records of actual
emissions would need to be kept and the source would not proceed to the
"percentage increase trigger" test, for which this commenter endorsed a
threshold of 80 percent.

Another commenter (@0805) similarly suggested a two-step test that looks
at increase in potential and then applies a 75-percent threshold.

 

Response:

Sources that propose physical and operational changes to existing
emissions units without increasing the design capacity or production
rate could experience increases in actual emissions where the need for
preconstruction review hinges on the accuracy of the projection of
post-change actual emissions.  EPA therefore believes that such projects
should be open for consideration as part of the "percentage increase
trigger" test to determine the need monitoring and recordkeeping for the
project.  We also believe a two-step test would add an undesirable level
of complexity to the rule.  Accordingly, we declined to proceed using a
two-step approach.

2.3	Calculating Baseline Emissions

The major NSR rules allow sources to exclude, in calculating any
increase in emissions from a particular project, that portion of the
unit's emissions following the project that: (1) an existing unit could
have accommodated during the period used to establish the baseline
actual emissions; and (2) are also unrelated to the particular project,
including any increased utilization due to product demand growth.

The D.C. Circuit upheld this "demand growth exclusion" (as it is
commonly called) but remanded the reasonable possibility provision to
EPA, specifically citing as a problem the opportunity for sources to
overstate the demand growth exclusion.

Comment:

Commenters (@0806, @0810) expressed concern that as a practical matter
it is often difficult to separate demand growth emissions increases from
emissions increases that result from a project.  As commenters (@0806,
@0810) noted, EPA has recognized this difficulty as well, and commenters
pointed out that in its response to comments following the December 2002
NSR reform rules, EPA stated that it is "very important that the source
retain a record of all information available to support its initial
claim that an emissions increase predicted to occur as a result of
demand growth did not result from the physical or operational change to
an emissions unit.  This information may be required by the reviewing
authority should there be a question about the project being a major
modification."  EPA Response to Comments on NSR Rule (November 2002) at
I-5-44.

One commenter (@0806) supported an approach based on a percentage
increase trigger so long as the approach accounted for demand growth. 
The commenter recommended an approach to address demand growth, wherein
sources that rely on the demand growth exclusion to avoid NSR
applicability will more easily trigger recordkeeping and reporting
requirements.  The commenter's recommendation is such that a source
makes no subtraction for emissions due to demand growth in projecting
emissions increases for purposes of the reasonable possibility test.  If
the projected emissions increases (including demand growth-related
emissions) exceed the applicable NSR significant emissions rate, then
the commenter recommends that the source must monitor post-modification
emissions.  The commenter further recommends that such sources be
required to submit annual reports to verify compliance with the NSR
rule.

One commenter (@0813) believed that if we fail to make sources
accountable for demand growth in their calculations, we would be taking
arbitrary and capricious action.  The commenter expressed that, as
proposed, the "percentage increase trigger" option fails to correct a
fundamental defect identified by the Court of Appeals.

 

One commenter (@0810) believed the "percentage increase trigger" option
as proposed is legally flawed and would constitute arbitrary action on
EPA's part because it would not address the problem of a lack of
permitting agency oversight over determinations by facilities as to
whether emissions increases that occur after a project should be
attributed to the project or to some independent factor (such as demand
growth).  In support of our "percentage increase trigger" option, we
assumed that the only projects to avoid recordkeeping requirements would
be those with small emissions increases.  However, the commenter argued,
a source could predict a large increase in emissions yet not trigger
recordkeeping requirements if the source attributes most of the
emissions increase to demand growth.  The commenter provided an example
for a pollutant with a 40-ton significant emissions rate, where a
facility calculates a post-project emissions increase of 200 tons,
attributing 10 tons of the increase to the project and the other 190
tons to demand growth.  With a 20-ton threshold (at 50 percent of the
significant emissions rate), the facility would not have to keep records
of its determination.  The commenter claimed that it is often difficult
for facilities to accurately predict the cause of their emissions
increases and further that facilities may be motivated by financial
incentive to attribute emissions increases to demand growth rather than
to the project.  To address the problem of verifying that facilities are
properly attributing emissions, the commenter recommended that EPA
subject facilities to recordkeeping, monitoring, and reporting
requirements whenever they choose to exclude demand growth emissions. 
The commenter added that EPA could at least require recordkeeping,
monitoring, and reporting in those instances where predicted emissions
increases from the project coupled with predicted emissions increases
that the facility believes will be caused by demand growth exceed the
applicable NSR significant emissions rate, an approach supported also by
commenter @0806 as described above.

One commenter (@0824) provided that sources could easily conclude that a
significant emissions increase was not "reasonably possible," for
example, by understating projections for emissions associated with
malfunctions or by overstating the demand growth exclusion.  Another
example provided by the commenter is the inclusion of "quantifiable"
fugitive emissions as well as "emissions associated with startup,
shutdown, and malfunctions" in the calculation of baseline emissions. 
The commenter believed that although a parallel provision is included
that requires sources to include fugitive and startup, shutdown, and
malfunction emissions in calculating post-change emissions, it will be
more difficult for sources to estimate predicted emissions from these
events and for regulators to check those predictions.

Response:

The final rules require, in part, that emissions from demand growth must
be included with projected actual emissions for purposes of applying the
"reasonable possibility" test.  (For example, see final rule language at
40 CFR 52.21(r)(6)(vi)(a) and (b).)   Several commenters specifically
recommended this approach.  With regard to the comments that suggested
applying the trigger at 100 percent of the applicable NSR significant
emissions rate where demand growth is concerned, we do not believe such
a trigger would capture all of the changes that have a reasonable
possibility of a significant emissions increase, and so we are retaining
the proposed trigger of 50 percent of the significant emissions rate. 

2.4	Differing Recordkeeping and Reporting Requirements

Comment:

One commenter (@0802) believed a lesser degree of documentation is
appropriate for changes predicted to cause smaller emissions increases,
because they have a lesser chance of exceeding significant emissions
rates.  The commenter agreed with EPA's proposal to require full
recordkeeping requirements for changes predicted to cause emissions
increases that exceed 75 percent of the applicable NSR significant
emissions rates but went on to suggest that only abbreviated
recordkeeping requirements should apply for changes predicted to cause
increases between 50 and 75 percent of applicable NSR significant
emissions rates.

One commenter (@0795) believed that there should be different standards
for maintaining records created before a change and records of annual
emissions after the change.  Specifically, they believed that sources
should always be required to retain pre-change records but would use the
"percentage increase trigger" or "potential emissions trigger" proposed
by the EPA for determining whether post-change records are required. 
They argued that as a result of requiring all sources to keep pre-change
records, documentation would always exist for the enforcement authority
to determine whether a source’s determination of the possibility of a
significant increase was indeed reasonable.

One commenter (@0806) believed the reasonable possibility rule should
require all facilities to maintain pre-change records (to include
projected actual emissions as well as any emission projections that a
facility attributes to demand growth) when a facility undertakes a
non-exempt physical or operational change (excluding routine
maintenance) that is expected to increase emissions of any regulated
pollutant by any amount.  This commenter added that if a project would
exceed applicable NSR significant emission rates as a result of
including emissions due to demand growth, then the facility should
additionally be required to monitor post-modification emissions and
submit annual reports to verify compliance.

One commenter (@0824) recommended that EPA require facilities that plan
a modification to report to the State a description of the change and
the analysis conducted to determine the change was not major and did not
trigger "reasonable possibility" recordkeeping and reporting
requirements.  When "reasonable possibility" recordkeeping and reporting
requirements are triggered, the commenter added, facilities should be
required to report actual emissions for a period of at least 10 years,
to harmonize the recordkeeping and reporting requirements with the
10-year lookback period established in the 2002 NSR rules.

Response:

We carefully considered which projects warrant pre-change vs.
post-change recordkeeping and reporting, and we agree with commenters
that differing levels of requirements can be justified.  We continue to
believe that some sources who elect to project post-change actual
emissions can do so accurately, especially when only small changes would
occur.  We believe it unnecessary for sources to be required to keep
records of pre-change projected emissions increases whenever they make
any change that is expected to result in any increase in emissions.  The
Court understood that sources with little likelihood of triggering NSR
should be eligible for less burdensome requirements, and the only way
that small emissions increases can be identified is by allowing sources
to project post-change actual emissions rather than to base the
estimated emissions increase on potential emissions.  Projects with
projected increases below the 50-percent threshold, especially when
emissions from demand growth are included in projections, are, we
believe, sufficiently small that any variability or error is less likely
to be large enough for the change to increase emissions to the
applicable significance emissions rate.  The fact that the reasonable
possibility standard applies before netting further assures that the
standard will appropriately capture projects with a higher probability
of variability and/or error in projecting emissions.  We note further
that, putting demand growth aside, commenters did not provide specific
examples of projects that they believed would trigger NSR but which
would avoid the reasonable possibility requirements as proposed.

2.5	Need to Include Verification Requirements or Permitting Authority
Oversight Within the Permitting Process

Comment:

Six commenters (@0793, @0810, @0813, @0817, @0822, @0824) stated that
there needed to be some form of verification or independent check on a
source’s "reasonable possibility" applicability determinations and
calculations.

One commenter (@0793) believed that the proposed rule does not provide
an adequate mechanism for permitting agencies to assure compliance with
NSR in situations where a facility undertakes a modification, and, at
the same time, it projects emission increases due to demand growth.

One commenter (@0810) noted that prior to the 2002 NSR reform rules, EPA
only allowed a source (and only an electric utility) to use the
actual-to-projected-actual methodology for determining NSR applicability
if the source kept pre-change records, performed post-change monitoring,
and submitted reports to its permitting agency.  This commenter (@0810)
believed EPA could satisfy the Court by requiring, same as under the
WEPCO rule, that any facility that chooses to make its emissions
calculation using the actual-to-projected-actual methodology must keep
records of its calculation and perform post-project monitoring and
reporting to verify the accuracy of its calculation.

Commenter (@0813) believed the "percentage increase trigger" option
fails to remedy the fundamental defects of its original "reasonable
possibility" rule, because without records for each project, the
permitting authority will have no way to determine whether a source's
projection of emissions increases is correct.

A few commenters (@0817, @0822, @0824) believed the "percentage increase
trigger" option is flawed because it relies too heavily on the
discretion of the source in determining whether a proposed change will
require recordkeeping, without review or approval by the permitting
authority.  The commenters believe it essential that enforcement and
permitting authorities know whether sources have reasonably and
accurately exercised their judgment and therefore should have the
opportunity to review a source(s analysis to verify that the
determination is based on reasonable assumptions.  One commenter (@0824)
noted that the construct of the "percentage increase trigger" is
vulnerable because sources could easily conclude that a significant
emissions increase was not reasonably possible, for example, by
understating projects for emissions associated with malfunctions or by
overstating the demand growth exclusion.

One commenter (@0817) provided that there must be some form of
verification that the change in actual emissions is truly below
applicable NSR significant emissions rates.  Another of these commenters
(@0824) agreed with that position by stating that they believed that the
public cannot solely depend on the good faith of facility owners and
operators to substitute for clear and protective environmental laws. 
One commenter (@0822)  pointed to concerns regarding self-policing
expressed by the National Academy of Public Administration (NAPA) and
the Government Accountability Office (GAO), and the commenter also
referred to hearing testimony by former EPA Administrator Carol Browner
during which she testified that she was concerned that the NSR revisions
would "eliminate the very features of the current law that provide
transparency to the public(monitoring, record keeping, and reporting."

One commenter (@0829) noted that justification for the "reasonable
possibility" provisions, even if revised to include the "percentage
increase trigger," will continue to be based in part upon the fact that
sources already maintain adequate records for assuring that EPA and
State and local permitting authorities can effectively enforce NSR
compliance.

Two commenters (@0822, @0824) noted that a lack of transparency and
accountability is likely to result in diminished NSR compliance.

One commenter (@0822) noted that the proposal reflects subjectivity, as
demonstrated by the proposal preamble language, which states "[t]he
'reasonable possibility' requirements apply only in the case of a change
that the source considers small, in that the source believes it
increases projected emissions by only a small amount...[Reporting would
be required] only if the source believed that the emissions increase
from the project would be no more than 50 percent of the applicable NSR
significant emissions rates."  (Emphasis added by commenter.)  The
commenter stated that applicability determinations and reporting
requirements should be based on calculations (not beliefs) that can be
verifiable by permitting authorities.  

Commenter (@0824) stated their position that the
actual-to-projected-actual methodology included in the 2002 NSR rule
changes amplifies, rather than reduces, the need for independent review
by a permitting agency.  Of particular concern, the commenter wrote, are
provisions that allow sources to exclude emissions increases
attributable to demand growth when projecting whether a change will
result in a significant emissions increase.  The commenter (@0824) 
expressed concern that there is the possibility that a source could
include "quantifiable" fugitive emissions as well as "emissions
associated with startups, shutdowns, and malfunctions" in the
calculation of baseline emissions.  By allowing sources to increase
their pre-change emissions, the commenter noted, the likelihood that
post-change emissions increases will be deemed large enough to trigger
NSR through such calculations is lessened.  The commenter acknowledged
that EPA has included a parallel provision that requires sources to
consider fugitive emissions and emissions from startups, shutdowns, and
malfunctions in calculating post-change emissions, but that it will be
difficult for sources to estimate predicted emissions from these events
and difficult for regulators to check whether those predictions are
reasonable.  The commenter believed that, due to incentives to lower
emissions estimates, sources' analyses are likely to reflect consistent
underestimations of future actual emissions.  For these reasons, the
commenter believes, permitting authorities must have access to
calculations associated with baseline emissions analysis and future
emissions estimates in order to accurately determine if a facility is
complying with the "reasonable possibility" standard.

Response:

We have responded to these comments in the preamble to the rulemaking.

Chapter 3 - "Potential Emissions Trigger" Option

The second proposed option was called the "potential emissions trigger"
option.  Under this option, a source would look at the post-construction
potential emissions of the affected emissions unit(s) to determine
whether there exists a reasonable possibility that an emissions increase
exceeding the applicable NSR significant emissions rate could occur. 
Comments provided on this option follow.

3.1.	General Comments on "Potential Emissions Trigger" Option

3.1.1	Support "Potential Emissions Trigger"

Comment:

Several commenters (@0792, @0810, @0813, @0817) supported EPA’s
"potential emissions trigger" approach. 

Commenter (@0810) believed the "potential emissions trigger" would
capture a much larger universe of projects than would the "percentage
increase trigger" option, and the commenter believed to do so would
facilitate enforcement of NSR rules.  The commenter believed the
approach would address the issue of demand growth and would also create
a bright-line test that should be easy for facilities to use.  This
commenter (@0810) believed the "potential emissions trigger" approach
would require recordkeeping only for those projects that may result in
significant—not small—emissions increases.

Commenter (@0817) recommended that EPA move forward with the "potential
emissions trigger" option for determining if a reasonable possibility
exists.  This commenter further recommended that every source triggering
"reasonable possibility" recordkeeping requirements should annually
report post-change emissions to the permitting authorities, to create a
paper trail that will allow the permitting authority to compare the
source's actual emissions with the emissions projected by the source.

Commenter (@0813) supported only the "potential emissions trigger"
approach, for reasons that all projects that could result in significant
emissions increases would be subject to major NSR and that the
permitting authority will be able to determine how a source came to the
conclusion that its emissions increases would not be significant.

One private citizen (@0792) commented that they preferred the "potential
emissions trigger" to the "percentage increase trigger" option, because
it would allow fewer sources to avoid recordkeeping and reporting
requirements, also believing the "potential emissions trigger" to more
clearly lay out the requirements.

Commenter (@0824) claimed that little detail was provided on the
"potential emissions trigger" option.  If post-change potential is
determined using the actual-to-potential test, the commenter could
support this option, contingent on certain requirements.

Response:

As explained earlier, we have decided to adopt the "percentage increase
trigger" option, with one modification that takes into consideration the
effect of demand growth emissions.  We believe this option considerably
limits the number of projects that could avoid "reasonable possibility"
recordkeeping and reporting requirements, but also takes into account
the impracticality of sources having to keep records (other than what
may already be required) when anticipating only a small increase in
emissions resulting from a proposed project.  If we were to rely on the
"potential emissions trigger," sources proposing projects that would
truly have only a small emissions increase would be unable to make such
demonstration and would be unnecessarily subject to the recordkeeping
and reporting requirements. 

3.1.2	Oppose "Potential Emissions Trigger" Option

Comment:

Eighteen commenters (@0796, @0797, @0798, @0802, @0803, @0804, @0807,
@0809, @0811, @0814, @0815, @0818, @0820, @0821, @0825, @0828, @0829,
@0831) did not support the adoption of the "potential emissions trigger"
option. 

Commenters (@0802, @0807, @0828) believed the "potential emissions
increase" trigger to be overly conservative and believed it would
present an unnecessary recordkeeping burden on the regulated community,
because the actual-to-potential test would result in recordkeeping
requirements for many modifications that have no reasonable possibility
of causing a significant emissions increase.  One of these commenters
(@0802) noted that the burden is magnified by the difficulty in
distinguishing between changes that are and are not routine maintenance,
repair and replacement.

One commenter (@0818) believes that the "potential emissions trigger"
option is unworkable because all major industrial facilities operate
below their full potential.

Commenters (@0809, @0825) objected to the "potential emissions increase"
trigger, believing it to provide no realistic possibility for a source
facing a minor modification to avoid maintaining unnecessary paperwork. 
The commenters pointed out that the New York court recognized that it
was appropriate to develop a less burdensome requirement for sources
with little likelihood of triggering NSR. 	

Commenter (@0798) believes the "potential emissions trigger" approach is
overly broad and unnecessarily stringent without providing any
significant environmental gain.  This commenter noted that such overly
broad trigger would specifically impose an advance notification
requirement on electric utilities, and the commenter believes to do so
would overly burden permitting authorities, unnecessarily slow down
efficient administration of the permit review process, and cause
unnecessary delays for construction projects.

Commenter (@0803) believed the "potential emissions trigger" conflicts
with EPA's stated intent, because it essentially eliminates the
"reasonable possibility" standard.  The commenter noted that the trigger
described as "post-change potential to emit" is unclear but found the
option to be unnecessarily stringent and a complete elimination, rather
than a clarification, of the "reasonable possibility" standard, whether
the trigger is simply PTE or is instead the difference between
post-change PTE and baseline actual emissions.  The commenter (@0803)
believed that sources will only opt for the actual-to-projected-actual
test for projects that would trigger NSR under the actual-to-potential
test, such that recordkeeping and reporting obligations are effectively
imposed every time a source would need to rely on the
actual-to-projected-actual test, regardless of whether a change has a
"reasonable possibility" of exceeding significant thresholds in the
future.

Similarly, commenters (@0809, @0825) did not believe the "potential
emissions trigger" advances the goal identified by the EPA and endorsed
by the New York court of reducing the burden on sources with little
likelihood of triggering NSR.  The commenters pointed out that to adopt
the "potential emissions trigger" option would simply be restating that
if a source documents that it does not have a significant increase based
on the actual-to-potential methodology, it is not subject to the
"reasonable possibility" standard.

Commenters (@0811, @0815) believed the "potential emissions trigger"
option, assuming it refers to a comparison of pre-project actual
emissions with post-project potential emissions, would have the same
practical effect as deleting the reasonable possibility standard.  The
commenters noted that this approach would yield no real benefit while
imposing substantial additional burden on both sources and permitting
authorities.

Commenter (@0814) opposed the "potential emissions trigger" option,
believing it would result in sources' having to develop and maintain
records of insignificant projects, and regulatory agencies would be
flooded by reporting of projects that clearly are not subject to NSR. 
The commenter did not believe the New York court could have intended
such a result.

Commenter (@0821) did not support the "potential emissions trigger"
option, believing it would require recordkeeping and reporting for
virtually every project undertaken at major industrial sources without
differentiation as to whether the project could result in an emissions
increase.

Two commenters (@0820, @0828) did not support the "potential emissions
trigger" because it could require recordkeeping for projects for which
there was no significant increase in emissions.

Commenter (@0829) categorically opposed the "potential emissions
trigger" option.  The commenter noted that many sources initiate
thousands of projects each year that under the Court's interpretation of
"any physical change" are potentially subject to NSR review, and the
vast majority of these projects will have little or no impact on a
source's emissions.  However, for a substantial number of these
projects, the difference between actual emissions and potential
emissions, both before and after the project is undertaken, would be
greater than applicable NSR significant emissions rates.  Requiring that
detailed pre-change analyses be conducted for all of these projects and
that determination of emissions increases associated with them be made
on an annual basis would create an astronomical recordkeeping burden for
sources, the commenter believed, and benefits would be nonexistent. 
Accordingly, the commenter believed, reliance on an
actual-to-projected-actual trigger rather than an actual-to-potential
trigger is both necessary and appropriate.

Commenters (@0800, @0812, @0816) recommended that EPA include the
"potential emissions trigger" as an alternative test (that is, in
addition to the "percentage increase trigger") that a source may use to
support a finding of no "reasonable possibility" of exceeding applicable
NSR significant emissions rates.  One commenter (@0800) noted that a
project cannot trigger NSR if the potential emissions do not exceed the
applicable NSR significant emissions rate.  This commenter (@0800) added
that to include the "potential emissions trigger" approach as an option
for sources that do not wish to go through the exercise of calculating
emissions against the actual emissions trigger threshold has the
advantage of being administratively simple.

One commenter (@0822) recommended an approach beyond the "potential
emissions trigger" option proposed.  The commenter supported use of an
actual-to-potential test as a trigger for recordkeeping, but the
commenter further recommended that preconstruction analysis of the
baseline and projected actual emissions (pre-change records) be kept for
all projects. 

Response:

As noted above, in the final rule we have adopted the percentage
emissions trigger.

3.2	Applicability 

3.2.1	Sources Operating at Less Than Full Capacity

Comment:

Commenters (@0796, @0797 [and supported by (@0821, @0804)], @0807) urged
EPA to reject the "potential emissions trigger" option because it would
effectively result in recordkeeping and reporting for all projects at
the vast majority of sources, regardless of whether these projects had
any possibility of increasing emissions.  The commenters believed that
in all but the most extraordinary circumstances, post-project potential
emissions of an industrial source would exceed its baseline emissions by
more than applicable NSR significant emissions rates, noting that the
vast majority of industrial sources subject to NSR programs typically
operate at less than their full potential in any given year.

Commenter (@0805) noted that an actual-to-potential comparison almost
always exceeds applicable NSR significant emissions rates but has little
relationship with operation of a source after a modification is made,
and the commenter believed the "potential emissions trigger" option
would draw a potentially large number of projects that have no
possibility of exceeding applicable NSR significant emissions rates into
recordkeeping and reporting requirements.

Commenters (@0811, @0815) noted that sources rarely operate at maximum
emission capacity, and so every project basing NSR applicability on the
actual-to-projected-actual test would trigger recordkeeping requirements
if the actual-to-potential test is the recordkeeping trigger.  The
commenters noted that EUSGUs and compressor stations rarely if ever emit
their PTE in any 12-month period.  Commenter (@0811) typically operates
its compressor stations at between 40- and 60-percent utilization on an
annual basis.  Under the "potential emissions trigger," the commenters
noted, almost all projects at an EUSGU (@0811, @0815) or compressor
station (@0811) would have a "reasonable possibility" of causing a
significant emissions increase.  The commenters believed that to impose
recordkeeping and reporting obligations for so many projects by adopting
a meaningless "potential emissions" threshold would yield no benefit,
because most projects should not be reasonably expected to result in an
actual emissions increase.  The commenter believed that such
recordkeeping and reporting obligations would be extremely burdensome
for both owners/operators of regulated sources and permitting
authorities.

Commenter (@0818 [and supported by @0804)] noted that power plants and
industrial plants require periodic maintenance, repair and replacement
of components, meaning it is not possible for these facilities to
operate at 100 percent of capacity.  The commenter believed the
"potential emissions trigger" option would result in recordkeeping and
reporting requirements for virtually every project that is undertaken at
every major stationary source in the nation, which would render the
"reasonable possibility" standard meaningless.

Commenter (@0819) did not support the "potential emissions trigger"
option.  The commenter noted that potential emissions are always greater
than actual emissions, to allow for operating flexibility and to
establish a compliance margin.  Especially in large facilities, the
commenter noted, an actual-to-potential emissions calculation can result
in a rate greater than the applicable NSR significant emissions rate
even for projects that have minimal emissions impact or that actually
reduce emissions.

Response:

As we indicated above, we agree with the commenters who claimed that the
"potential emissions trigger" option would result in the unnecessary
application of the recordkeeping and reporting requirements on projects
whose emissions were not likely to result in significant emissions
increases.  The "percentage increase trigger" option enables sources
proposing projects with small emissions increases to avoid the
recordkeeping and reporting requirements.

3.2.2	Source Emissions Increases Linked to Project/Projected Actual
Emissions 

Comment:

One commenter (@0802) believed reporting requirements should be keyed to
the likelihood that a change will in fact cause a significant emissions
increase (be based on actual emissions) and not to some artificial
formula.  Commenters (@0796, @0797 [and supported by @0821 and @0804],
@0807) objected to the "potential emissions trigger" because, unlike the
"percentage increase trigger," it does not recognize the importance of a
causal link between a project and some resulting emissions increase. 
Commenters @0706 and @0797 [and supported by @0821 and @0804] observed
that in previous explanations by EPA on the NSR statutory and regulatory
causation requirement, the initial "focus [is] on the effect of any
nonroutine changes on the operating characteristics of the unit during
the representative baseline period."  If a change does not affect these
characteristics, the commenters noted, there is no possibility that it
would result in an emissions increase. 

One objection to the "potential emissions trigger" option raised by
commenters was that the emissions increase according to the
actual-to-potential test is a hypothetical value that does not correlate
to whether there is risk that a change will result in significant
emissions (commenter @0805) or to the actual-to-projected-actual
applicability test (commenters @0811, @0815).

Commenter (@0819) stated that the "potential emissions trigger" does not
relate to the actual impact of the project, objecting to the "potential
emissions trigger."

Response:

Again, we agree that too many projects would be unnecessarily subject to
the recordkeeping and reporting requirements if we relied on the
"potential emissions trigger" option.

3.3	Other Comments on "Potential Emissions Trigger" Option

Comment:

Several commenters (@0793, @0806, @0810, @0811, @0815, @0824) commented
that either little detail (e.g., regulation language) was provided on
the "potential emissions trigger" option or it was unclear how it was to
be implemented (e.g.,  how a project's post-change potential emissions
would be determined).

 

Commenter (@0806) noted that the proposal was not clear as to how a
project's post-change potential emissions would be determined under the
"potential emissions trigger" option, pointing out that post-change
potential emissions could be solely based on the PTE of the emissions
source after the modification or could instead refer to a comparison of
baseline actual emissions with the PTE of the emissions source after the
change.  Commenters (@0811, @0815) pointed out that post-change
potential emissions could also be interpreted to mean the difference
between pre-change and post-change potentials to emit (i.e., a
potential-to-potential test).

Commenter (@0806) found it difficult to ascertain how EPA would
implement the "potential emissions trigger" option, because we did not
propose rule language or examples.  Commenters (@0806, @0810) urged EPA
not to adopt this approach without providing a more detailed explanation
of the approach and specific regulatory language to implement it.

Comment:

Several commenters who supported adoption of the "potential emissions
trigger" option provided recommendations on the notification,
recordkeeping and reporting requirements that should be adopted with
this option.

One commenter (@0793) suggested specific recordkeeping and reporting
requirements to augment this option.  If the EPA adopts the
actual-to-potential applicability test, the commenter believes EPA
should require that the facility do the following:

Document the source modification and calculated emissions (both
potential and projected actual).

Notify the appropriate State/local permitting agency of the calculated
nonapplicability and submit a plan for "monitoring" emissions from the
source.

Maintain records of emissions for 5 years for a modified source and
10 years for a new unit.

As part of its annual compliance certification, the company should
certify that actual emissions have not increased beyond the applicable
NSR significant emissions rate which would have subjected the source to
the requirement for a permit.

If at any time actual emissions increase to a point whereby a permit
requirement is triggered, the company should immediately submit
appropriate applications and undergo NSR/PSD analysis, including BACT or
LAER, as appropriate.

One commenter (@0803) suggested that each source that triggers the
"reasonable possibility" recordkeeping requirements using the "potential
emissions trigger" should annually report post-change emissions to the
appropriate State or local permitting authorities.

Another commenter (@0822) recommended two kinds of reporting
requirements.  First, for any change, the source’s preconstruction
analysis of the baseline and projected actual emissions should be
available on file and upon request provided to the permitting authority.
 Second, when any modification triggers applicable NSR significant
emissions rates using the actual-to-potential test, post-change
emissions impacts of the project should be reported annually to the
appropriate permitting authority.  If the total changes in emissions
could be significant without excluding emissions as a result of demand
growth, then monitoring and recordkeeping should be required.  The
commenter further recommended that actual annual emissions be reported
to the permitting authority for a period of 10 years for new units and
5 years for projects that involve modification of existing units.

Commenter (@0806) believed the reasonable possibility rule should
require all facilities to maintain pre-change records (to include
projected actual emissions as well as any emission projections that a
facility attributes to demand growth) when a facility undertakes a
non-exempt physical or operational change (excluding routine
maintenance) that is expected to increase emissions of any regulated
pollutant by any amount.

Response:

We appreciate the commenters' consideration and suggestions regarding
recordkeeping and reporting in connection with potential emissions.  We
note that many of the elements of recordkeeping and reporting
recommended by the commenters are already required under our existing
rules and that they remain under our new rule to clarify the reasonable
possibility standard.  We also acknowledge the suggestions of commenters
that differing levels of recordkeeping and reporting requirements may be
justified, an approach that fits with our final rule.  This approach is
also discussed in section 2.4 of this Response to Comments.  

We decided to adopt the "percentage increase trigger" option instead of
the "potential emissions trigger" option, in part because of concerns
that the "potential emissions trigger" approach would require
recordkeeping and reporting for projects that have virtually no
possibility of resulting in an emissions increase greater than the
significant thresholds.

Chapter 4 - Other Miscellaneous Comments

4.1	Burden Associated With Recordkeeping Requirements

Comment:

One commenter (@0797) [and supported by @0821 and @0804] stated that a
regime requiring recordkeeping and reporting for every activity that is
arguably a "physical change or a change in the method of operation" at a
major source, no matter whether there is any reasonable possibility that
such activity would increase emissions, would be extremely burdensome
and unworkable.

Commenter (@0801) expressed concern that under the proposal,
recordkeeping could be required for thousands of trivial projects that
annually occur on a site, even if they do not cause emissions increases.
 Commenter (@0801) asked EPA to clarify that the reasonable possibility
standard does not apply to minor improvement projects not designed to
change the process in a manner that could increase emissions.  The
commenter noted the present difficulty of distinguishing between
activities that are "physical changes or changes to the method of
operation" and the many small projects that are exempt from NSR because
they are "routine maintenance, repair and replacement."  The commenter
goes on to state that documentation of the many small and
environmentally inconsequential changes made at a facility could create
an enormous burden on existing facilities if no distinction is made
between various types of changes.  This commenter also stated that
because of the many records already available, there is no justification
for requiring additional recordkeeping for environmentally insignificant
changes.

Commenter (@0802) stated that any workable requirement to document the
emissions consequences of changes must have a de minimis level below
which the requirements do not apply, because otherwise every little
change would require documentation, no matter how remote its chances of
producing a significant emissions increase.  The commenter believed that
to document such small changes is disproportionately more difficult, as
well as less environmentally important, than documenting big changes. 
The commenter explains that in the NSR context, these documentation
burdens would be magnified by the difficulty of distinguishing between
activities that are and are not "routine maintenance, repair and
replacement." 

Commenter (@0817) expressed concern with the approach of the "reasonable
possibility" standard, finding the recordkeeping requirement to be an
example of "negative applicability" recordkeeping.  The commenter noted
that facilities are being asked to keep additional ongoing records to
demonstrate that an air rule is not applicable, creating another
regulatory burden on industry for rules that don't even apply.  While
the commenter supports documentation of applicability determinations,
the commenter doesn't support ongoing requirements to document
nonapplicability.  The commenter provided as examples of the types of
nonapplicability documentation that the Agency should not require:
requiring hourly records of emissions under NSPS to verify
nonapplicability, and requiring periodic process vent TRE calculations
under MACT to demonstrate a vent is below the threshold for controls.

Commenter (@0802) noted that some forest products industry sources emit
small quantities of H2S and TRS.  Those pollutants are regulated by
NSPS, which means a significant increase in their emissions can trigger
NSR.  The commenter pointed out that applicable NSR significant
emissions rates for these pollutants are very low, only 10 tpy.  The
commenter stated that to require sources of these pollutants to document
very low levels of emissions increases would not be justified by any
environmental benefits it would achieve.

Response:

We are not adopting a "reasonable possibility" test that would require
recordkeeping and reporting for every project, and we have clarified
that the reasonable possibility standard does not apply to projects that
are considered "routine maintenance, repair and replacement."  We
believe our adoption of a percentage increase trigger at 50 percent
appropriately balances the value of records to ensure compliance against
recordkeeping and reporting requirements that would be overly
burdensome.  Commenters did not provide examples of specific projects
that would be subject to the reasonable possibility requirements under
the 50-percent trigger but that should not be.  We further note that
comments regarding NSPS and MACT are outside the scope of this
rulemaking.

 

4.2	Pre-Existing Records/Reports/Inventories Sufficient to Enforce NSR
Provisions

Comment:

One commenter (@0793) believed EPA's claim that State and local minor
NSR programs can serve as backup programs in providing adequate
protection was dubious.

One commenter (@0799) argued that existing records are likely to
identify projects that could trigger major NSR.  This commenter believed
projects likely to increase actual emissions above major NSR levels are
those involving new equipment additions or modifications to existing
equipment that will increase capacity and PTE.  The commenter
acknowledged that while some of those projects could escape permitting,
many of them are subject to State minor NSR review.  The commenter noted
that a resulting minor NSR permit may include conditions to assure NSR
does not apply.  This commenter added that major sources subject to NSR
rules are also required to report emissions and that reviewing
authorities can use such reports to identify upward trends in emissions
and identify sources that may have misjudged NSR applicability.  This
commenter further noted that major sources are required to submit
emissions information and other emissions-related information as part of
title V periodic reports and annual compliance certifications, and such
information could be used to identify projects that caused significant
emissions increases.  Moreover, the commenter stated, the annual
compliance certification requires sources to identify their compliance
status for all Clean Air Act requirements, including NSR, for the
previous year.  If a source did not properly obtain NSR permits for a
project, the commenter added, it would have to certify intermittent
compliance with NSR requirements or risk criminal enforcement action for
false certification.

Commenter (@0803) believed pre-existing recordkeeping and reporting
requirements under other programs will provide EPA with a sufficient
basis for determining compliance.  Commenter (@0801) believed sources
provide under existing law a large volume of data clearly sufficient to
demonstrate regulatory compliance by sources, for purposes such as minor
source permitting, emission inventory updates, emission fee assessments,
title V monitoring, recordkeeping, and reporting, and title V
compliance certifications.  The commenter noted that the purpose of
maintaining the additional records is to determine whether a project
that was expected to yield an insignificant emissions increase actually
resulted in emissions that exceeded applicable NSR significant emissions
rates.  The commenter noted that, while the records to make this
determination would use essentially the same data that sources already
use to track compliance status, the reasonable possibility tests would
involve additional calculations and thus sources would effectively have
to keep two sets of records. This commenter urged EPA to further
streamline the recordkeeping requirements for compliance assurance
purposes if promulgating a reasonable possibility standard that subjects
many more projects to recordkeeping. 

Commenter (@0802) claimed that the entire history of environmental
audits and enforcement investigations has shown that capital investment
requests, work orders, State permit applications, and similar documents
allow reviewers from corporate compliance staffs and from oversight
agencies to readily identify a "physical change or change in the method
of operation."  The commenter did not think New York v. EPA called for
formalizing the long-established approach to identifying such changes.

Commenter (@0802) noted that some sources spend tens of thousands of
dollars and hundreds of hours each year preparing detailed annual
emissions inventories for submittal to title V permit agencies.  The
commenter added that, even in the days before title V requirements took
full effect, the history of NSR enforcement will show that enforcement
agencies were fully able to identify emissions increases based on
production records, emission factors, and State emission monitoring
requirements.

Commenter (@0805) believed EPA should have addressed the recordkeeping
and reporting obligations that sources already have in order to present
a case for why additional rulemaking is not needed to prevent NSR
circumvention.  The commenter noted that sources subject to NSR
regulations already provide emissions information to include title V
operating permit applications and federal and State preconstruction
permit applications that detail the emissions equipment in a plant.  The
commenter added that major sources are also subject to periodic
monitoring and recordkeeping requirements, and large emissions equipment
is subject to additional monitoring and recordkeeping under the
compliance assurance monitoring regulations.  Monitoring compliance data
include logs of operations, visible emissions and instrumental opacity
readings, stack test reports, analytically generated mass balances, and
strip charts from continuous direct emissions and parametric monitors. 
This commenter added that CAA records and other publicly available
information on production levels and growth in various industrial
sectors can be utilized by authorities to determine if unexplained
actual emissions increases are occurring at a source.  In fact, the
commenter noted, this publicly available data provided EPA enforcement
officials with the information to initiate dozens of NSR enforcement
investigations throughout the 1990s and is regularly used by
nongovernmental organizations with monitoring reports and emissions
inventories to monitor compliance with NSR and other CAA requirements. 
Once an EPA CAA investigation is launched, the Agency has extremely
broad authority under section 114 of the CAA to obtain business records
pertaining to emissions from a facility.

Commenter (@0805) noted that emissions inventories are required for
sources in federal nonattainment areas under the Clean Air Act,
acknowledging that not all reports will pinpoint without additional
information from the source which equipment caused the increase.

Commenter (@0814) noted that the clarity provided by the "percentage
increase trigger" option would aid the Agency in investigating
after-the-fact to determine whether the standard was triggered, and the
commenter claimed that records kept by sources for reasons other than
NSR compliance, such as for the reporting of emissions and numerous
business records, can supply information for that determination.

Commenter (@0819) stated that facilities already have sufficient records
to support NSR applicability decisions and respond to regulatory agency
requests.  The commenter listed examples of records that facilities
maintain that relate to NSR applicability determinations and the process
units potentially subject to NSR: capital project development and
funding requests; annual emission inventories provided to regulatory
authority; internal emission trends analysis; State and local regulatory
authority minor source construction permits (often with permit limits);
MACT and NSPS recordkeeping and reporting; operating data included in
process computers; and periodic and continuous monitoring data.

Commenter (@0806) disagreed with EPA's contention that EPA has numerous
means of enforcing the NSR provisions against a facility that is not
required under the reasonable possibility rule to keep records of its
emissions calculations, and commenters (@0806, @0810) pointed out that
the D.C. Circuit rejected similar arguments in New York.  Commenter
(@0810) pointed out that the Court noted EPA had failed to explain how
emissions reported under title V can be traced to a particular physical
or operational change.  With respect to EPA’s contention that EPA has
enforcement tools available to compel facilities to disclose information
about projects after the fact, commenter (@0810) claimed that EPA's
abilities to obtain information "after the fact" do not negate the
Court's statement that without paper trails, enforcement authorities
have no means of discovering whether a source's determination was
reasonable. 

Commenter (@0824) disagreed with EPA's statement in its proposal that
States will have access to documents to verify information necessary to
recreate the "reasonable possibility" calculation.  The commenter
contended that in many instances, State personnel do not have access to
the information or adequate skills to recreate the information necessary
to determine compliance with the rule.

Commenter (@0822) encouraged EPA to require sources to include, in
title V operating permits, demonstrations of continuing compliance with
recordkeeping requirements.  For example, according the commenter,
routine title V recordkeeping requirements and emissions statements
should be unit-specific and sufficient to provide a screening trigger
for a more detailed investigation of emissions, if necessary.  Further,
in its annual compliance certification, a facility that has modified
within the last 5 years (or constructed a new unit within the last
10 years) should certify that actual emissions have not increased
beyond the applicable NSR significant emissions rate using the
actual-to-potential test.  If actual emissions increase beyond
applicable NSR significant emissions rates, the commenter went on to
say, such a change should be considered a violation of the requirement
to obtain a major NSR permit, and the permitting authority should have
the authority to take enforcement action and require the source to
reduce emissions below the applicable NSR significant emissions rate or
require the source to apply for the appropriate NSR permit, undertake
air quality analysis, and meet BACT/LAER requirements.

Commenter (@0829) believed sufficient records already exist for
enforcement of the NSR permitting requirements, and the commenter
believed the Court made erroneous assumptions that led it to conclude
that inadequate information would be available in the future to enforce
the actual-to-projected-actual NSR applicability provisions.  Contrary
to the Court's assumption, the commenter noted, EPA has relied little on
application of the actual-to-potential test in its non-utility
enforcement actions and has only brought enforcement actions where it
could demonstrate that an increase in actual emissions resulted from a
project.  Commenter (@0829) expressed preference that EPA respond to the
Court's remand by simply retaining the existing "reasonable possibility"
provisions without including a new trigger for recordkeeping and by
explaining why the Agency will have adequate information to bring
enforcement actions for NSR violations.  Among the requirements that
already generate information:

State minor source permit requirements require facilities to obtain
permits for changes that have associated emissions increases that often
are substantially below the NSR significant thresholds.  The result of
these requirements is that permitting authorities obtain emissions
information and have opportunities to request additional information if
there is any concern that a change might trigger NSR permitting.

States require the submission of annual emissions inventory information
which provides an indication of whether emissions increases are
occurring at a facility, and thus they will be able to request
additional information if emissions trends suggest that NSR permitting
might be called for.

Numerous NSPS, MACT, nonattainment emissions requirements and other
emissions limitations include requirements that necessitate compliance
with emissions control, reporting and other requirements if emissions
are increased from existing sources or new sources are installed.

Civil and criminal penalties establish a deterrent to discourage
noncompliance with all of the foregoing clean air requirements.

In addition to the broad range of air requirements, the commenter
(@0829) went on to note, facilities routinely maintain operational
records that provide an additional source of information for determining
whether plant changes might trigger NSR permitting.  These include
production records, capital appropriation requests, purchase records,
sales records, and myriad other operation-related records.  The
commenter wrote that EPA obtained vast amounts of such information from
corporate records in connection with its NSR enforcement initiative by
issuing requests under section 114 of the Act.  To the commenter's
knowledge, EPA has never been unable to pursue NSR enforcement or
present sufficient proof of NSR claims due to the lack of adequate
information for determining applicability of the NSR requirements.

Commenter (@0819) encouraged EPA to finalize the reasonable possibility
rule in a timely manner, believing that otherwise permitting authorities
will develop their own basis for recordkeeping with a resulting
patchwork of different requirements.  The commenter noted that they
recently received a permit from one State that, despite the fact that
the project is not expected to have an emissions increase, is requiring
the facility to keep hourly records for 15 years.  While there is no
guarantee a final EPA rule will eliminate this type of requirement, the
commenter added, it would establish a basis that permitting authorities
can point to.

Response:

We respond to the principal points made by commenters in the preamble to
the rulemaking.  We note that comments on changes EPA could make to its
title V permit procedures are beyond the scope of this rulemaking.

4.3	Appendix S 

Comment:

Commenter (@0822) expressed disappointment that EPA included the
"percentage increase trigger" option as an interim interpretation of
"reasonable possibility" in its final Appendix S (nonattainment NSR)
rule.  This commenter expressed that EPA failed to provide a comment
period on this recordkeeping requirement before promulgating it in a
final rule, which the commenter believes violates the Administrative
Procedure Act.  

Similarly, another commenter (@0824) objected to use of that trigger and
noted a failure to provide public review and comment.  They stated that
EPA must afford a comment period on the recordkeeping requirements and
make changes to Appendix S as appropriate.

Response:

The comment period that followed proposal of this rulemaking served to
provide the public an opportunity to comment on "reasonable possibility"
recordkeeping and reporting as it related to 40 CFR Part 51 Appendix S.
 Thus, the revision of Appendix S promulgated with this final
rulemaking included adequate opportunity for notice and comment. 

4.4	Utility versus Non-Utility Requirements

Comment:

Several utility commenters (@0796, @0797, @0807) wanted utilities to
have same requirements as non-utilities.  These commenters provided that
there is no logical reason for recordkeeping and reporting differences
to exist between utility and non-utility companies, especially given
that the recordkeeping and transparency requirements of the utility
industry are extensive, irrespective of NSR, as a result of the heavy
regulation of the industry by EPA as well as by other agencies (e.g.,
FERC, State public utility commissions or similar bodies).

Response:

For this rulemaking, we considered recordkeeping and reporting
requirements in the context of our 2002 NSR rules.  As we explained in
the preamble to our 2002 NSR rules (67 Fed. Reg. at 80,204), we believe
it appropriate for reviewing authorities to have information on
construction and modification activities at EUSGUs readily available, in
view of the disproportionate amount of emissions generated by EUSGUs
compared to other industry sectors.  Thus, the "reasonable possibility"
final rule continues to require that, prior to construction, utilities
must submit reports of pre-change records, while non-utility projects
have no pre-change reporting requirements.

4.5	Miscellaneous Implementation Issue

Comment:

One commenter (@0824) stated that to adopt either of EPA's proposed
options would limit States' ability to enforce key provisions of the NSR
program as they apply to existing sources by limiting States' ability to
monitor modifications of existing process equipment.

Response:

Contrary to the commenter’s concerns, the States remain free to adopt
more stringent interpretations of the reasonable possibility standard.

4.6	Miscellaneous Clarifications/Corrections

Comment:

One commenter (@0830) requested EPA to provide further clarification of
situations in which records must be kept for 10 years versus 5 years.

Response:

We believe the existing regulations are clear as to the duration for
which sources are required to keep records.  For example, the
requirement for monitoring emissions, and calculating and maintaining a
record of annual emissions under 40 CFR 52.21(r)(6)(iii) is to last for
a period of 5 years following resumption of regular operation after the
change, unless the project involves increasing the design capacity or
PTE of the existing emissions unit.  In the latter case, the
requirements must be applied for 10 years following resumption of
operation.

Comment:

One commenter (@0803) found an error in the example of Table 1 in the
proposal.  Where we have 90 and 40 tons for Project 2, we should have
instead used, for example, 80 and 30 tons.  If projected actual
emissions were 40 tons, major NSR would be triggered.

  

Response:

The commenter is correct in the finding that the second example in
Table 1 illustrates a situation where the proposed project’s increase
in actual emissions would trigger major NSR, because the 40-tpy increase
equals the significant emissions rate for the hypothetical pollutant. 
Thus, the example is not appropriate for illustrating a situation where
a source determines that the change would not trigger major NSR but
there is a "reasonable possibility" that the change may significantly
increase emissions.  We do not believe, however, that this error
resulted in a misunderstanding of the way in which the proposed
"reasonable possibility" test would be carried out.

 

Comment:

One commenter (@0803) disagreed with a statement in the preamble
describing the "WEPCO rule."  EPA stated that prior to 1992, no sources
were required to keep records of projected emissions under major NSR
because only the actual-to-potentials test was used.  72 Fed. Reg. at
10,448/1.  The commenter pointed out that, by authorizing the
actual-to-actual test for all modifications to existing units, the WEPCO
rule officially eliminated the need to determine whether normal
operations have begun at existing EUSGUs.  The commenter further pointed
out that the rule merely clarified a test that had been in existence and
utilized well before 1992.  Another commenter (@0817) asked EPA to
correct the statement in the preamble that the PTE test was the basis
for determining emissions prior to the 2002 NSR reform rule and should
state that the historic test is an actual emissions test; the commenter
provided background information.  Another commenter (@0829) similarly
asked EPA to correct the statement and provided background information.

Response:

We acknowledge that our statements were simplistic in attempting to
condense a long history.  However, we do not believe that the statements
included with the brief history in the background section of our
proposed rulemaking would cause commenters to misunderstand the intent
of the proposal or to prevent anyone from adequately commenting on the
proposed changes.  Accordingly, the issue raised by commenters is
outside the scope of this rulemaking.  Additional history (by no means
complete) on applicability tests is provided in the preamble to our 2002
NSR rules (67 Fed. Reg. at 80,191-80,194).

 

Comment:

One commenter (@0811) noted that we discussed the "potential emissions
trigger" in relation to emissions from a "project" rather than from a
source or an emissions unit, and the commenter requested that EPA
clarify what is meant by "emissions from a project."

Response:

Only one commenter expressed concern about purported ambiguity in this
term.  Throughout the NPRM we referred to "projects at existing
emissions units," "a change or a project," or "the change’s projected
actual emissions increase."  Each of these phrases describes a physical
or operational change at one or more existing emissions unit that
results in an increase in the actual emissions at the unit(s).  We do
not believe that there was sufficient ambiguity to cause commenters to
misunderstand the intent of the proposal or to prevent anyone from
adequately commenting on the proposed changes.

Comment:

One commenter (@0798) argued that the Court's remand was limited to
recordkeeping requirements and thus did not apply to reporting
requirements, and the commenter interprets this to mean EPA is not
required to impose the "reasonable possibility" standard to reporting
requirements that apply to EUSGUs.

Response:

We believe that the Court remanded to us the reasonable possibility
standard, and not just the recordkeeping component of the standard.  It
is true that in the "Conclusion" section of the opinion, the Court
stated that it was remanding "the recordkeeping provisions," New York v.
EPA, 413 F.3d 3, 44 (D.C. Cir. 2005), but we read that statement to
refer to the reasonability standard itself.  In the part of the opinion
in which the Court discussed the reasonable possibility standard, the
Court stated, "…EPA needs to explain how its recordkeeping and
reporting requirements allow it to identify such sources," id. at 34, a
statement that we believe indicates clearly that the Court did not limit
the remand to the recordkeeping component. 

 Under the actual-to-projected-actual methodology, a source may opt to
use potential to emit as its projected actual emissions.  See, e.g., 40
CFR 52.21(b)(41)(ii)(d).

 In this rulemaking, the terms "we," "us," and "our" refer to the EPA
and the terms "you" and "your" refer to the owners or operators of major
stationary sources of air pollution and to State, local, and tribal air
pollution authorities.

 PAGE   

 PAGE  39 

1 – Introduction

  PAGE  31 

2 – Percentage Increase Trigger

3 – Potential Emissions Trigger

4 – Miscellaneous Comments

