Is the U.S. Environmental Protection Agency`s Revised New Source

Copyright © 2005 Environmental Law Institute®, Washington, DC. reprinted with permission from ELR®, http://www.eli.org, 1-800-433-5120.
35 ELR 10316
ELR
5-2005
NEWS&ANALYSIS
Is the U.S. Environmental Protection Agency’s Revised New
Source Review Rule Moving in the Right Direction?: A Deepened
New Source Bias, and the Need for Pursuing Sustainable Energy
Development in Air Pollution Control Law
by Inho Choi
Table of Contents
Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . 10316
I. Discussion of the NSR Program . . . . . . . . . . 10319
A. In General . . . . . . . . . . . . . . . . . . . . . . . . 10319
B. New and Modified Major Stationary
Sources. . . . . . . . . . . . . . . . . . . . . . . . . . . 10320
C. NSR Applicability . . . . . . . . . . . . . . . . . . . 10320
1. Physical or Operational Change: The Routine
Maintenance Exception . . . . . . . . . . . . . 10320
2. A Significant Net Increase in Emissions . . 10321
a. An Emissions Increase: The Actualto-Future-Actual Test. . . . . . . . . . . . . 10321
b. The WEPCO Rule: The Actual-toProjected-Future-Actual Test and Its
Extended Application . . . . . . . . . . . . 10322
c. A Significant Net Increase:
Netting . . . . . . . . . . . . . . . . . . . . . . . 10323
II. NSR Failures and the Movement to Reform
the Current NSR Program. . . . . . . . . . . . . . 10324
A. Grandfathering Under the CAA . . . . . . . . . 10324
B. Federal and State Efforts to Repeal
Grandfathering . . . . . . . . . . . . . . . . . . . . . 10325
1. Congressional Efforts. . . . . . . . . . . . . . . 10325
2. State Action . . . . . . . . . . . . . . . . . . . . . 10325
C. EPA’s Enforcement Initiative . . . . . . . . . . . 10326
D. The Overhaul of EPA’s Enforcement
Initiative . . . . . . . . . . . . . . . . . . . . . . . . . . 10327
1. The National Energy Policy Group’s Report
to the President . . . . . . . . . . . . . . . . . . . 10327
2. The DOJ’s NSR Report . . . . . . . . . . . . . 10327
3. EPA’s 90-Day NSR . . . . . . . . . . . . . . . . 10327
Inho Choi is an S.J.D. candidate and received his LL.M. in 2002 from
George Washington University Law School. He received an LL.M. in
1998, and an LL.B. in 1993 from the Chungnam National University College of Law in South Korea. He can be contacted via e-mail at
[email protected]. The author would like to express his greatest
gratitude to Prof. Arnold W. Reitze Jr., who is his mentor at George Washington University Law School and one of the nation’s leading environmental law experts, for his helpful comments on the earlier drafts of this
Article and his strong encouragement throughout its preparation. All the
remaining errors and misunderstandings are the author’s responsibility.
E. A Multi-Pollutant Trading Approach at the
Federal Level . . . . . . . . . . . . . . . . . . . . . . 10328
1. Four-Pollutant Bills . . . . . . . . . . . . . . . . 10328
2. Three-Pollutant Bills: The Bush
Administration’s Clear Skies Initiative . . 10329
III. The 2002 New NSR Rule . . . . . . . . . . . . . . 10329
A. A 10-Year Look-Back Period and the
Actual-to-Projected-Future-Actual Test . . . . 10330
B. PALs . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10330
C. The Clean Unit Exclusion . . . . . . . . . . . . . 10332
D. PCPs . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10332
IV. Another Round of Heated Debate Over the
New NSR Rules . . . . . . . . . . . . . . . . . . . . . . 10333
A. Criticisms of the New NSR Rule: Environmental
Groups’ Arguments . . . . . . . . . . . . . . . . . . 10334
B. Concerns About the Revised Routine Maintenance
Exception Rule . . . . . . . . . . . . . . . . . . . . . 10335
C. The U.S. Government Accountability Office
(GAO) Studies on Stakeholders’ Views on the
New NSR Rules . . . . . . . . . . . . . . . . . . . . . 10336
D. The Stay of the Routine Maintenance Exception
Rule and the Uncertain Future for NSR
Reform . . . . . . . . . . . . . . . . . . . . . . . . . . . 10336
E. The Legality of the New NSR Rule . . . . . . . 10336
Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10337
T
his Article analyzes the revised new source review
(NSR) rule and argues that it violates the Clean Air
Act’s (CAA’s or the Act’s)1 clean air mandate by changing
the preexisting definition of the statutory term “change” and
by extending the demand growth exclusion to all sources
and creating several NSR-exempt project-based construction activities that are applicable to existing sources, without providing meaningful procedural safeguards. This is because the new rule conflicts directly with the following requirements under the CAA’s NSR program: (1) a proposed
physical or operational change that would increase emissions or result in collateral emissions must go through NSR
preconstruction review; (2) emissions increases and de1. 42 U.S.C. §§7401-7671q, ELR Stat. CAA §§101-618.
5-2005
NEWS & ANALYSIS
35 ELR 10317
Copyright © 2005 Environmental Law Institute®, Washington, DC. reprinted with permission from ELR®, http://www.eli.org, 1-800-433-5120.
creases to be considered in NSR applicability determinations must be contemporaneous; and (3) once NSR is triggered, the stringent technology requirement, the best available control technology (BACT) or the lowest achievable
emissions rate (LAER), must be applied to the sources.
The Article argues that the revised NSR rule is moving in
the wrong direction in that it strengthens a bias against new
sources and enlarges preexisting loopholes in favor of old,
dirtier sources, which have traditionally enjoyed significant
cost advantages over cleaner, more energy-efficient sources
under the grandfathering scheme. It observes that the U.S.
Environmental Protection Agency’s (EPA’s) reliance on the
new rule’s allegedly minimal impacts on air quality and the
nation’s decade-long transition to a multi-pollutant trading
approach in air pollution control in justifying the rule
changes is untenable in view of congressional intent leading
to the enactment of NSR and the literal meaning of the term
“change.” The Article concludes with the argument that the
overriding goal in NSR reform is to create a level playing
field for all sources, whether new or old, by building
sustainability concerns into existing environmental and energy law, for example, through repealing grandfathering, the
adoption of output-based emissions standards and, possibly,
the enactment of climate change policy aimed at reducing
fossil fuel usage.
Introduction
The permitting sections of Parts C and D of CAA Subchapter I are known as the NSR program,2 whose main goal is to
protect, maintain, and improve air quality while providing
for continued economic development and meeting energy
needs.3 The CAA covers six criteria air pollutants: particulates (including particulate matter (PM) with a diameter
of 10 microns or less (PM10) and PM with a diameter of 2.5
microns or less (PM2.5)), sulfur dioxide (SO2), nitrogen oxide (NOx), carbon dioxide (CO2), ozone (O3), and lead (Pb).
It also regulates toxic air pollutants and volatile organic
compounds (VOCs). Under the Act’s clean air mandate,
EPA is to promulgate primary and secondary national ambient air quality standards (NAAQS) for each of the six criteria pollutants.4 Each state must then within three years prepare and submit to EPA for approval its implementation
plan, called a state implementation plan (SIP), for meeting
NAAQS.5 In their SIPs, the states, including tribal lands and
territories, must demonstrate the timely attainment of
NAAQS or “reasonable further progress” toward the attainment of NAAQS in all areas under their jurisdiction, using
available monitoring data and modeling analyses.6 The EPA
Administrator must promulgate a federal implementation
plan (FIP) if a state fails to submit a SIP by the statutory
deadline or if it submits an inadequate SIP, or fails to revise it
2. See id. §§7470-7492, 7501-7515. The NSR regulations are found in
40 C.F.R. §§51.165, 51.166, 52.21, 52.24, and pt. 51, app. S.
3. See, e.g., H.R. Rep. No. 95-294, at 13 (1977), reprinted in 1977
U.S.C.C.A.N. 1077, 1091 (noting that “[t]his section is proposed as a
means of assuring realization of the dual goals of attaining air quality
standards and providing for new economic growth”).
4. See 42 U.S.C. §7409.
5. See id. §7410.
6. Id. §7503(c)(2)(B). CAA §§110(a)(2)(A)-(M) provide for the basic
requirements for the SIP, which are to be used by EPA as criteria for
the approval of individual SIPs. Id. §7410(a)(2)(A)-(M); see 40
C.F.R. §51.
after EPA’s notice of disapproval.7 The EPA Administrator
can disapprove the entire SIP or part of it.8 Harsh sanctions
may be imposed on states that fail to meet the statutory
deadlines for SIP submittal or NAAQS attainment.9 On the
other hand, states have the wide discretion to choose measures to comply with NAAQS as long as they can demonstrate timely attainment to EPA and make reasonable further progress. Each state, by adopting a SIP, is empowered
to determine which sources to regulate and which pollution
control measures to employ to meet NAAQS.10 In short, the
CAA’s basic scheme for accomplishing its goals is “cooperative federalism” with distinct roles for the states and the
federal government.
However, state authority to shape air management strategies and plans has its limits. Congressional dissatisfaction
with the 1970 CAA’s performance led to the enactment of
the prevention of significant deterioration (PSD) and
nonattainment programs in 1977. The main thrust of the
PSD program is to protect and enhance the high air quality
of areas with clean air. Under the nonattainment program,
states are required to implement more stringent SIP requirements in return for more time for attainment in areas within
their jurisdiction that have failed to meet the applicable
NAAQS. States must impose emissions reduction requirements based on reasonably available control technology
(RACT) on existing major sources covered by EPA guidelines.11 New and significantly modified major stationary
sources that want to locate in PSD or nonattainment areas
must obtain preconstruction permits from state permitting
agencies. For new sources in PSD areas, this usually means
that they must go through air impact analyses and air quality
modeling at the preconstruction stage, and meet post-con7. 42 U.S.C. §7410(c)(1).
8. Id. §7410(c)(1)(B).
9. States may be subject to highway sanctions prohibiting approval and
funding by the Secretary of Transportation of highway projects, and
to stringent offset requirements under the nonattainment NSR program of at least 2 to 1. Id. §7509.
10. Id. §7407(a), declaring that
[e]ach State shall have the primary responsibility for assuring air quality within the entire geographic area comprising
such State by submitting an implementation plan for such
State which will specify the manner in which national primary and secondary ambient air quality standards will be
achieved and maintained within each air quality control region in such State.
See also, e.g., Train v. Natural Resources Defense Council, 421 U.S.
60, 79, 5 ELR 20264 (1975) (concluding that “so long as the ultimate
effect of a State’s choice of emission limitations is compliance with
the national standards for ambient air, the State is at liberty to adopt
whatever mix of emission limitations it deems best suited to its particular situation”); Union Elec. Co. v. EPA, 427 U.S. 246, 269, 6
ELR 20570 (1976) (stating that the states have “the power to determine which sources would be burdened by the regulations and to
what extent” in implementing their SIPs approved by EPA); EPA v.
Brown, 431 U.S. 99, 103, 7 ELR 20375 (1977) (per curiam). Therefore, a state can demonstrate timely attainment to EPA in its SIP by
relying more on control measures that target area sources, such as
dry cleaners and gas stations, and mobile sources than other states.
11. EPA has interpreted RACT to mean “the lowest emission limitation
that a particular source is capable of meeting by the application of
control technology that is reasonably available considering technological and economic feasibility.” The 1990 CAA Amendments incorporated this RACT requirement, as interpreted by EPA guidelines. Arnold W. Reitze Jr., Air Pollution Control Law:
Compliance and Enforcement 79 n.18 (2001).
35 ELR 10318
ENVIRONMENTAL LAW REPORTER
5-2005
Copyright © 2005 Environmental Law Institute®, Washington, DC. reprinted with permission from ELR®, http://www.eli.org, 1-800-433-5120.
struction air quality monitoring requirements.12 New
sources in nonattainment areas must obtain an offset from
other sources in surrounding areas that is equal to or greater
than the proposed increase in emissions at their facility.13
Once it is found by the permitting agency that NSR requirements apply, new or significantly modified sources must install BACT or LAER in PSD and nonattainment areas, respectively.14 This technology requirement is quite onerous
12. See 42 U.S.C. §7475(e)(3)(B) (stating that PSD regulations “shall
require an analysis of the ambient air quality, climate and meteorology, terrain, soils and vegetation, and visibility” at the proposed construction site and in nearby areas); 40 C.F.R. §52.21(m)(1)(iv),
52.21(m)(2), (m)(3).
13. 42 U.S.C. §7503(c)(1). The applicable offset ratio is different depending on the location’s nonattainment classification. CAA §182
sets out offset ratios for O3 nonattainment areas. An applicable minimum ratio is from VOCs of 1.1 in marginal areas down to 1.5 in extreme areas (1.15 for moderate areas, 1.2 for serious areas, and 1.3
for severe areas). See id. §§7511a(a)(4), 7511a(b)(5), 7511a(c)(10),
7511a(d)(2), and 7511a(e)(1). In principle, the required emissions
reductions must come from a source in the same area. Id.
§7503(c)(1). However, an exception applies when the offset is provided by a source in another attainment area with an equal or higher
nonattainment classification (2) whose emissions from this area contribute to nonattainment in the area where the new source is sited. Id.
This exception may apply in the context of transboundary pollution
in which the transport of a pollutant emitted from sources located in
upwind areas contribute to nonattainment in downwind areas. Offsets are often called emissions reduction credits (ERCs). Prior to the
1990 CAA Amendments, ERCs were still used in offsets in
nonattainment areas, bubbles and netting, and banking. See generally U.S. EPA, Emissions Trading Policy Statement; General Principles for Creation, Banking, and Use of Emission Reduction Credits,
51 Fed. Reg. 43814 (Dec. 4, 1986).
14. 42 U.S.C. §§7502, 7503. Note that an area can be in attainment for
one criteria pollutant and in attainment for another pollutant. As a result, both technology standards could apply to the same source and
the source applicant must prepare for both PSD and nonattainment
NSR, simultaneously. The nonattainment NSR requirements are
more stringent than the PSD NSR requirements. The applicant
should identify all technologies, including those listed in
RACT/BACT/LAER Clearinghouse (RBLC), in which EPA has
maintained a list of technologies on its website that have been demonstrated to be effective on similar sources. The applicant should
also consider a control technology that has successfully been applied at other source categories. U.S. EPA, Draft New Source
Review Workshop Manual: Prevention of Significant Deterioration and Nonattainment Area Permitting B.11
(1990), available at http://www.epa.gov/region07/programs/artd/
air/nsr/nsrmemos/1990wman.pdf (last visited Mar. 1, 2005) [hereinafter Draft NSR Workshop Manual]. Pollution control technologies that are being successfully applied to similar sources in foreign
countries are also potential candidates for BACT or LAER. Id. at
B.5. While BACT is determined “on a case-by-case basis, [after] taking into account energy, environmental, and economic impacts, and
other costs,” LAER is a much more demanding one, without mentioning costs and other related considerations in the relevant provision. Compare 42 U.S.C. §7479(3); 40 C.F.R. §51.166(b)(12)
(BACT), with 42 U.S.C. §7501(3) (LAER). This is mainly because
nonattainment areas must make reasonable further progress toward
attainment, which is intended to ensure that air quality in
nonattainment areas must be improved continuously, while allowing
for economic growth. BACT is an emission limitation standard,
which is set at the most stringent level that can be achieved by a similar source in industry unless it is proved by the applicant as technologically or economically infeasible. Reitze, supra note 11, at 195.
It usually requires the use of best available pollution control methods
and technologies. If emissions standards prove to be infeasible, design, equipment, work practices, operational standards, or any combination thereof can be used. 40 C.F.R. §51.166(g)(12). LAER may
also lack numerical emissions limitations because of “the technological or economic limitations on the application of measurement
methodology to a particular class of sources.” Id. §51, app. S. IV-A
n.4. (“Hereafter, the term emission limitation shall also include such
design, operational, or equipment standards.”) (emphasis in original). EPA has allowed states instead to prescribe a design, operational, or equipment standard in the permits for these sources that
in that sources are required to pursue the right mix of control
options to minimize air quality impacts to the maximum extent possible.15 It also is comprehensive because the applicant for an NSR permit must consider all possible environmental impacts on the environment a particular technology
would have, in the technology selection process.16 Thus,
NSR is quite similar to an environmental impact analysis
under the National Environmental Policy Act (NEPA) but,
because of its substantive bite, NSR can be described as
“NEPA with teeth.”17 Moreover, EPA has significant leverage over state decisions to choose NSR technology.18 As
with the O3 nonattaiment program, NSR is the product of
congressional policy judgment favoring the Act’s clean air
goal at the expense of state sovereignty and industry’s operational flexibility. It constitutes an integral part of the Act’s
PSD and nonattainment programs and is one of the most important tools in moving the nation toward attaining the goal
of clean air for all Americans.
Attempting to balance the two competing interests of environmental protection and accommodating economic and
must contain enforceable conditions on design characteristics or
equipment. The owner or operator of a major emitting facility is required to perform BACT analysis independently of its analyses of its
source impacts on ambient air quality. The CAA requires EPA to issue guidance documents on BACT/LAER technology and to revise
these documents at least every two years. 42 U.S.C. §7508. Additionally, EPA may consider incidental effects of emissions of toxic
pollutants, which are not regulated under the PSD program, in making BACT determinations. See id. §7412(b)(6) (“The provisions of
[the PSD program] shall not apply to [hazardous pollutants] under
[§112].”); see, e.g., In re North County Resource Recovery Assocs.,
2 E.A.D. 229, 1986 EPA App. LEXIS 14 (Adm’r 1986) (stating that
“the net environmental impact of such emissions is eligible for consideration in making the BACT determination”) (emphasis added);
see Draft NSR Workshop Manual, supra, at B.50-.53 (stating
that “the generation or reduction of toxic and hazardous emissions,
including compounds not regulated under the Clean Air Act, are
considered as part of the environmental impacts analysis[,]” citing In
re North County Resource Recovery Assocs., 2 E.A.D. at 229); see In
re Steel Dynamics, Inc., 9 E.A.D. 165, 189 n.29, 2000 WL 833062
(EAB Apr. 23, 2000) (stating that state permitting agencies have
“considerable discretion to evaluate HAPs emissions and potential
health impacts as part of its consideration of environmental impacts
in general[,]” citing Draft NSR Workshop Manual, supra and
In re North County Resource Recovery Assocs., 2 E.A.D. at 229).
15. In practice, EPA has employed a “top-down” approach. Under this
approach, all technologically feasible and available technology options are identified and ranked on the basis of its stringency or effectiveness, and the permit applicant has the burden to reverse the presumption in favor of the most stringent technology available, by
showing why this will not be appropriate for it. See Draft NSR
Workshop Manual, supra note 14. The NSR technology selection
process have traditionally focused primarily on end-of-pipe, postcombustion, controls. It, however, does not necessarily mean that
the chosen technology is limited to a particular pollution control
technology. It may include the use of a cleaner fuel or innovative
production processes, and, even, a combination of all available control methods that can achieve the maximum degree of emissions reductions of the pollutant subject to NSR. See 42 U.S.C. §7479(3) (requiring that an applicable emission limitation be “based on the maximum degree of reduction of the [covered] pollutant . . . through application of production processes and available methods, systems,
and techniques, including fuel cleaning, clean fuels, or treatment or
innovative fuel combustion techniques for control of each such pollutant”) (emphasis added).
16. Draft NSR Workshop Manual, supra note 14, at B.26.
17. Gregory B. Foote, Considering Alternatives: The Case for Limiting
CO2 Emissions From New Power Plants Through New Source Review, 33 ELR 10642, 10651 (July 2004).
18. See Alaska Dep’t of Envtl. Conservation v. EPA, 540 U.S. 461, 34
ELR 20012 (2004) (holding that a state permitting agency must present a reasonable justification for its BACT determination to EPA’s
satisfaction in order to meet the Act’s NSR requirements).
5-2005
NEWS & ANALYSIS
35 ELR 10319
Copyright © 2005 Environmental Law Institute®, Washington, DC. reprinted with permission from ELR®, http://www.eli.org, 1-800-433-5120.
energy needs19 has led to a highly complicated regulatory
system, which is often criticized by the regulated community as burdensome, complex, time-consuming and costly,
inflexible, and even frustrating good-faith efforts to improve environmental performance in pollution control technology.20 Given its technology-forcing nature and onerous
requirements, it is not surprising that industry has every incentive to avoid NSR by taking advantage of the weaknesses and loopholes in the NSR program. A spectrum of
stakeholders, including industry representatives, environmental nongovernmental organizations (NGOs), and state
and federal regulators, reached a broad consensus on the
need for NSR reform around the early 1990s. EPA then
embarked upon a process for NSR reform by authorizing
the formation of a subcommittee to the Clean Air Act Advisory Committee in 1993.21 Since then, especially EPA’s
new NSR enforcement initiative in the late 1990s, the NSR
program has been at the center of debate over how to reform
this system. The stated objectives of the NSR reform
seemed promising.22 In reality, however, EPA is caught in
the middle of a tug of war between industries and environmentalists that want to shape the agenda to their own interests and values.
The new NSR rule, which was promulgated in 2003, is no
exception. It creates a new controversy on its legality under
the CAA, provoking another round of heated debate. The
new rule is moving in the wrong direction because it
strengthens a new source bias and enlarges preexisting loopholes in favor of old, dirtier sources, which have traditionally enjoyed significant cost advantages over cleaner, more
energy-efficient sources under the grandfathering scheme.
Allegedly minimal impacts on air quality and the nation’s
decade-long transition to a multi-pollutant approach in air
pollution control should not be used as an excuse for relaxing preexisting rules. NSR has been the center of the U.S.
19. See Chevron, U.S.A., Inc. v. Natural Resources Defense Council,
467 U.S. 837, 851, 14 ELR 20507 (1984) (observing that in the NSR
program “Congress sought to accommodate the conflict between the
economic interest in permitting capital improvements to continue
and the environmental interest improving air quality”).
20. A relevant EPA report reads as follows:
For more than 10 years now, the Environmental Protection
Agency (EPA) has been engaged in an effort to improve the
New Source Review (NSR) Program in response to widespread concerns from stakeholders who are concerned that it
is too complex and burdensome, it introduces uncertainty in
planning, it inhibits industry’s ability to quickly make needed
changes, and it is not working as effectively as it could be to
protect air quality.
U.S. EPA, New Source Review Improvements: Supplemental Analysis of the Environmental Impact of the 2002
Final NSR Improvement Rules 1 (2002), available at
http://www.epa.gov/nsr/documents/nsr-analysis.pdf (last visited
Mar. 1, 2005) [hereinafter Supplemental Analysis of 2002 Final NSR’s Impact].
21. U.S. EPA, Notice of Public Meeting, 58 Fed. Reg. 36407 (July 7,
1993).
22. See, e.g., U.S. EPA, Prevention of Significant Deterioration (PSD)
and Nonattainment New Source Review (NSR): Baseline Emissions
Determination, Actual-to-Future-Actual Methodology, Plantwide
Applicability Limitations, Clean Units, Pollution Control Projects,
67 Fed. Reg. 80186, 80189 (proposed Dec. 31, 2002) (to be codified
at 40 C.F.R. §§51, 52) (stating that the aim of NSR reform is to “reduce burden, maximize operating flexibility, improve environmental quality, provide additional certainty, and promote administrative efficiency”).
legal system’s failure to attain the goal of sustainable energy development. The correct direction to be taken in reforming NSR should be to incorporate sustainability concerns into legal decisionmaking processes under the CAA,
e.g., through the adoption of output-based emission standards, repealing grandfathering, and/or the integration of
sustainable energy development goals into NSR permitting processes.
This Article aims to discuss the problems with NSR and
analyze the new NSR rule in detail. It argues that, while its
impact will be minimal on electric utilities, the new NSR
rule arguably violates the CAA’s “clean air” mandate because: (1) a proposed physical or operational change that
would increase emissions or result in collateral emissions
must go through NSR preconstruction review; (2) emissions
increases and decreases to be considered in NSR applicability determinations must be contemporaneous; and (3) once
NSR is triggered, the stringent technology requirement,
BACT or LAER, must be applied to the sources.
Part I describes the elements of the CAA’s NSR program
with much focus on baseline determinations and NSR applicability. It partially compares the preexisting rule with the
new NSR rule. Part II explains why NSR has not worked as
well as expected at the time of its enactment in 1977, and
discusses NSR reform moves by the U.S. Congress, some
states, and the previous and current Administrations. Part III
examines the new NSR rule and discusses what changes in
EPA’s prior position took place and the rationales for the
changes given by the Agency. Part IV discusses grave concerns expressed by environmentalists and state agencies
about the potential adverse impacts of the new rule on existing air quality. It then critically analyzes EPA’s current legal
position and arrives at the presumptive conclusion that the
new NSR rule is violative of the CAA’s clean air mandate in
view of congressional intent leading to the enactment of
NSR and the literal meaning of the statutory term “change.”
The Article concludes with the argument that the overriding
goal in NSR reform is to create a level playing field for
sources, whether new or old, by building sustainability concerns into existing environmental and energy law.
I. Discussion of the NSR Program
A. In General
The essence of the PSD and nonattainment NSR programs is
the requirement for preconstruction review. The owner or
operator planning to construct a new major stationary
source or to make a major modification to an existing major
stationary source must undergo a preconstruction permitting process. Preconstruction review is designed to select
proven modern pollution control technology as applied to
each regulated pollutant emitted from the facility, including
new emissions of a collateral pollutant. In order to obtain a
preconstruction permit, the facility must prove to the permitting agency that it would not result in a violation of
NAAQS or any applicable PSD regulations in local or
downwind areas currently in compliance with NAAQS.23
Because NSR involves a lengthy and complex process,
much attention is paid to its applicability.
23. For permit requirements, see generally 42 U.S.C. §7475(a); id.
§7503(a) (nonattainment).
35 ELR 10320
ENVIRONMENTAL LAW REPORTER
5-2005
Copyright © 2005 Environmental Law Institute®, Washington, DC. reprinted with permission from ELR®, http://www.eli.org, 1-800-433-5120.
B. New and Modified Major Stationary Sources
Under the CAA, stationary source means “any source of
an air pollutant except those emissions [from mobile
sources].”24 CAA §111(a)(3) further defines the term stationary source as “any building, structure, facility, or installation which emits or may emit any air pollutant.”25 The
threshold emission levels for qualification as a major stationary source in nonattainment areas are set at a potential to
emit (PTE) of more than 100 tons per year (tpy)26 of any pollutant subject to regulation under the CAA down to smaller
amounts depending on the area’s nonattainment classification.27 In the case of PSD areas, the threshold is 100 or 250
tpy of any regulated pollutant under the CAA, depending on
the source type.28 Under the NSR program, the amount of
emissions is calculated based on aggregating sources located on contiguous or adjacent properties that are under
common control, having the same two-digit Standard Industrial Classification code.29 Note that states have implemented minor source programs. Thus, even if new or modified sources do not qualify as major and, hence, are not subject to NSR, they can still be subject to minor source requirements imposed by states. A source’s emissions can include
fugitive emissions.30
24. Id. §7602(j). Actually, the term “major emitting facility” is used under CAA Subchapter I, Part C, Subpart 1. See id. §7479(1). Its definition is similarly worded as that of a major stationary source.
25. Id. §7411(a)(3). Roughly speaking, a stationary source is a discrete
point from which one of any regulated air pollutants under the CAA
is released, such as smokestacks. But note that fugitive emissions
may be included for the purposes of calculating emissions from stationary sources within the certain industrial categories covered by
EPA regulations.
26. Id. §7602(z).
27. See id. §§7511a(c)-(e), 7511c(b)(2) (O3); id. §7512a(c)(1) (CO); id.
§7513a(b)(3) (PM). This can be as low as 10 tpy of VOCs in an extreme O3 nonattainment area.
28. 42 U.S.C. §7479(1); 40 C.F.R. §52.21(b)(1). Usually the threshold is
250 tpy, but the 100-tpy threshold applies to a list of 28 source categories (industrial groupings such as petroleum refineries, fossil
fuel-fired steam-generated electric power plants, pulp mills, and iron
and steel mill plants). See 40 C.F.R. §52.21(b)(1)(a). See also U.S.
EPA, NSR 90-Day Review Background Paper 3 (2001), available at http://www.epa.gov/air/nsr/documents/nsr-review.pdf (last
visited Mar. 1, 2005) [hereinafter NSR Background Paper].
29. 40 C.F.R. §§51.166(h)(6), 52.21(b)(6).
30. CAA §302(j) reads:
New source is defined as any stationary source that begins construction or modification after the promulgation of
proposed regulations for a source category.31 “‘[M]odification’ means any physical change in, or change in the
method of operation of, a stationary source which increases
the amount of any air pollutant emitted by such source or
which results in the emission of any air pollutant not previously emitted.”32 These provisions are contained in the
Act’s new source performance standard (NSPS) program,
but the NSPS program has a purpose and scope that are
wholly different from the NSR program. BACT or LAER
are mass-based standards applicable only to major stationary sources, depending on the area’s air quality, is determined on a case-by-case basis, and is usually much more
stringent than NSPS. On the other hand, generally speaking,
NSPS is a national, uniform performance standard for approximately 69 categories, which does not mandate the use
of particular technologies.
C. NSR Applicability
1. Physical or Operational Change: The Routine
Maintenance Exception
The NSR program has been applicable only to major modifications that would “result in a significant net emissions increase,”33 and the NSR regulations establish significant
emissions levels, which vary by pollutant.34 Therefore, determining whether a major modification has occurred is a
two-prong test. First, there must be a physical or operational
change at the facility. Neither Congress nor EPA has protion of State Implementation Plans, 45 Fed. Reg. 52676, 52690
(Aug. 7, 1980). There are 27 source categories covered by this regulation. See id. at 52692. EPA agreed to withdraw its position as part
of a settlement with industry petitioners in Chemical Mfrs. Ass’n v.
EPA after a series of legal challenges to the rules, but the Agency
thereafter returned to the interpretation it initially adopted under the
1980 regulations, a move encouraged by the judicial opinion of the
D.C. Circuit in Duquesne Light Co. v. EPA. See No. 79-1112 (D.C.
Cir. filed Jan. 26, 1979); 698 F.2d 456, 13 ELR 20251 (D.C. Cir.
1983). Therefore, the Agency’s current position is that it will balance
all possible socioeconomic costs and benefits in determining
whether fugitive emissions should be included in the calculation of
emissions in the context of PSD NSR. In light of the court ruling in
the Alabama Power case, it can be said that EPA has the discretion to
determine whether to require a source category to consider fugitive
emissions in the definition of a major source for the purposes of PSD
and nonattainment NSR preconstruction review. For example, in
Ogden Projects, Inc. v. New Morgan Landfill Co., the U.S. District
Court for the Eastern District of Pennsylvania concluded that “fugitive emissions may not be counted unless EPA has first conducted a
rulemaking” for listing a source category, as required by §302(j).
911 F. Supp. 863, 878, 26 ELR 20843 (E.D. Pa. 1996). See also 40
C.F.R. §51.165(a)(1)(C) (nonattainment NSR). In addition, on November 27, 2001, EPA promulgated a rule, under which a source
within a category subject to the NSPS or the hazardous air pollutants
(HAPs) rule issued after August 7, 1980, is not required to include
fugitive emissions of all regulated pollutants under PSD or
nonattainment NSR for the purpose of determining whether it has a
major source status. See U.S. EPA, Change to Definition of Major
Source, 66 Fed. Reg. 59161, 59162 (Nov. 27, 2001) (codified at 40
C.F.R. §70). However, sources are still required to include fugitive
emissions of all HAPs in determining whether they are major
sources under §112. Id.
Except as otherwise expressly provided, the terms “major
stationary source” and “major emitting facility” mean any
stationary facility or source of air pollutants which directly
emits, or has the potential to emit, one hundred tons per year
or more of any air pollutant (including any major emitting facility or source of fugitive emissions of any such pollutant, as
determined by rule by the Administrator).
42 U.S.C. §7602(j) (emphasis added). However, there has been continued controversy as to whether and how fugitive emissions will be
handled under the NSR program. In Alabama Power Co. v. Costle,
the U.S. District Court for the District of Columbia (D.C.) Circuit
held that fugitive emissions must be included in determining
whether a source constitutes a major emitting facility under CAA
§302(j). 636 F.2d 323, 10 ELR 20001 (D.C. Cir. 1979). The court observed that the wording of §302(j) was controlling in the definition
of a major emitting facility in §169(1), even though whether to include fugitive emissions in calculating threshold emission levels for
major stationary sources should be determined by EPA. See id. at
369-70. EPA then issued rules that listed the source categories covered by PSD and NSPS rules that were required to consider fugitive
emissions. U.S. EPA, Requirement for Preparation, Adoption, and
Submittal of State Implementation Plans; Approval and Promulga-
31.
32.
33.
34.
42 U.S.C. §7411(a)(2).
Id. §7411(a)(4); 40 C.F.R. §60.2.
See 45 Fed. Reg. at 52676; 40 C.F.R. §52.21(b)(2)(i).
40 C.F.R. §52.21(b)(23).
5-2005
NEWS & ANALYSIS
35 ELR 10321
Copyright © 2005 Environmental Law Institute®, Washington, DC. reprinted with permission from ELR®, http://www.eli.org, 1-800-433-5120.
vided a clear definition of these terms. But EPA’s NSR regulations recognize that certain types of projects are exempt
from NSR and, among other things, allow for the exception
for routine maintenance, repair, and replacement.35
While creating certain categories of NSR-exempt activities clearly makes sense, the obscurity surrounding the question of what types of projects should be considered routine
maintenance has been highly controversial between regulators and industry. Most of the past and current EPA- or
state-initiated NSR enforcement actions have targeted industry’s strategic behavior aimed at maximizing the use of
this exemption. The routine maintenance exception has created a loophole that, if abused, could inflict significant damage on the integrity of the entire NSR program. This is especially so because a facility is not required to ask the permitting agency to determine whether the planned activity is
within the scope of the routine maintenance exemption, although EPA will decide the applicability of the exemption
on a case-by-case basis when asked to do so.
The court’s ruling in Wisconsin Electric Power Co.
(WEPCO) v. Reilly36 provides useful guidance in this regard.
Responding to the utility petitioner’s argument that its
planned replacement project was within the scope of the
routine maintenance exception, the U.S. Court of Appeals
for the Seventh Circuit ruled in favor of EPA that it was beyond the exception and therefore was covered by NSPS and
NSR, finding as a reasonable application of the relevant regulations EPA-used factors, such as the nature, extent, purpose, frequency and cost of work, for determining the applicability of the exception.37 The court did not agree with
WEPCO that the cost, magnitude, and nature of its project
were irrelevant for purposes of the routine maintenance exception to NSPS and PSD. Among others, the court regarded the following facts as decisive: (1) the project was a
“life-extension” project; (2) WEPCO admitted that a project
of such magnitude “would normally occur only once or
twice during a unit’s expected life cycle,” and it never occurred before; and (3) it would cost at least $70.5 million.38
Therefore, a strong presumption can be established from a
reading of the WEPCO decision that maintenance projects
intended to increase the life expectancy of an electric-generating unit (or other industrial units) are considered a modification (not routine), thereby triggering NSR.
In 2000, EPA’s Environmental Appeals Board (EAB)
heard a case involving life-extension projects at nine electric-generating units owned by the Tennessee Valley Authority (TVA) in Kentucky, Tennessee, and Alabama.39 This
case also involved TVA as one of the nine electric utilities
against which EPA took enforcement actions in 1999. Unlike other companies, for jurisdictional concerns, EPA issued an administrative order under §§113 and 167 of the
CAA against TVA whose failure to comply with the order
could independently lead to severe penalties. The EPA Ad35. Id. §§52.21(b)(2)(iii) (PSD), 52.24(f)(5) (nonattainment). This
exemption was added after a similar provision under the 1975
NSPS regulations.
36. 893 F.2d 901, 20 ELR 20414 (7th Cir. 1990).
37. See id. 910-13.
38. Id. at 911-12.
39. See In re Tennessee Valley Auth., No. CAA-2000-04-008, 9 E.A.D.
357, 2000 WL 1358648, 32 ELR 41231 (EAB Sept. 15, 2000) (final
order on reconsideration).
ministrator directed the EAB to reconsider the administrative order and to issue a final order. The EAB applied a
four-part test to determined whether the routine maintenance exception was applicable to the company’s projects
at issue: (1) the nature and extent of the change; (2) the
purpose of the change; (3) the frequency of the change;
and (4) the cost of the change. It ruled that none of the
TVA’s 14 life-extension projects qualified for the routine
maintenance exception, thereby violating the NSPS and
NSR requirements.40
On July 26, 2002, the U.S. District Court for the Southern
District of Indiana issued an important ruling on preliminary motions in an ongoing lawsuit involving the Southern
Indiana Gas & Electric Company (SIGECO).41 The court
held that EPA’s enforcement was not barred by the Indiana
Department of Environmental Management’s (IDEM’s)
previous determination that SIGECO’s plant upgrades constituted routine maintenance. SIGECO’s main argument
was that the IDEM’s ruling was binding on EPA as a result
of its delegation of enforcement power to the state agency.42
The court rejected that argument, however, finding that EPA
is not precluded from bringing an enforcement action, given
the broad language of §113 of the Act, which provides that
EPA is authorized to enforce “any requirement or prohibition” of “an applicable implementation plan or permit,” and
§111(c)(2) that authorizes EPA to enforce “any applicable
standard of performance.”43 It held that the doctrine of equitable estoppel does not apply unless EPA “knew the facts”
relating to a state agency’s ruling and had engaged in “affirmative misconduct.”44 This ruling was another victory for
EPA which has been engaging in legal battles with large
electric utility companies since 1999.
However, note that the definition of the term “significant” has been changed with respect to three newly created
mechanisms designed to promote the use of clean energy
technologies: plantwide applicability limits (PALs), the
Clean Unit exclusion, and pollution control projects (PCPs).
For sources choosing to use PALs or the Clean Unit exclusion, allowable emissions, instead of actual emissions, become the basis for determining whether a significant emissions increase would result. A qualifying PCP is deemed not
to result in an increase in collateral emissions if its net air
quality benefits are judged as positive.
2. A Significant Net Increase in Emissions
a. An Emissions Increase: The Actual-to-Future-Actual
Test
Once it is determined that a physical or operational change
to a major stationary source would occur, the next step is to
40. See id. pt. III.C.3. and app. A. But in 2003, the U.S. Court of Appeals
for the Eleventh Circuit struck down the administrative order on procedural grounds. See Tennessee Valley Auth. v. Whitman, 336 F.3d
1236 (11th Cir. 2003), reh’g en banc denied, 82 Fed. Appx. 220,
2003 U.S. App. LEXIS 27278, 33 ELR 20231 (11th Cir. 2003), cert.
denied sub nom. Leavitt v. Tennessee Valley Auth., 124 S. Ct. 2096
(2004). As a consequence, the EAB’s decision lost much of its
precedential value.
41. United States v. Southern Ind. Gas & Elec. Co., 2002 U.S. Dist.
LEXIS 14039 (S.D. Ind. July 26, 2002).
42. Id. at **9-11, 14.
43. Id. at **13, 15.
44. Id. at **16-17.
35 ELR 10322
ENVIRONMENTAL LAW REPORTER
5-2005
Copyright © 2005 Environmental Law Institute®, Washington, DC. reprinted with permission from ELR®, http://www.eli.org, 1-800-433-5120.
identify whether that change would produce a significant
net increase in emissions in order for the source to be subject
to NSR. The initial step for the permitting agency to take is
to determine the baseline for the actual emissions before the
change, which is compared to the projected post-change
emissions to determine if there will be an increase in emissions before and after the modification.45 Under the old rule,
the pre-change actual emissions were to be calculated based
on the average rate in tpy, actually emitted during the previous two years, if those emissions were representative of normal operations at the unit during this time.46 Therefore, the
baseline emissions meant a source’s actual emissions
shortly before the proposed modification begins. The permitting agency was allowed to use a different time period if
the source shows that it is more representative of normal operations.47 If that is the case, the calculation of actual emissions must be based on “the unit’s actual operating hours,
production rates, and types of materials processed, stored,
or combusted during the selected time period.”48 The new
NSR rule replaced the 2-year time period with a 10-year
look-back period except for electric-generation units.49
Once the baseline is determined, the following step is to
calculate projected postmodification emissions. Under the
old rule, which still applies to new sources, post-change projected emissions must be equal to the PTE, which was defined as “the maximum capacity of a stationary source to
emit a pollutant under its physical and operational design.”50 While it was relatively easy to determine a new
source’s PTE, much of the controversy over NSR applicability had centered around this issue, which was the trickiest
part of NSR implementation. The PTE, as applied by EPA,
was based on the presumption that the unit will run at full capacity (namely, 24 hours a day year-round). This was the
so-called actual-to-potential test. EPA had applied this test
to modifications to existing sources since it is presumed
that they have not begun normal operations. This test was
quite onerous for most existing sources because they usually
do not operate at their maximum capacity. Even after the
WEPCO decision, the actual-to-potential test had been applied to all sources with the exception of fossil fuel-fired
electric utilities51 until the new NSR rule was promulgated
in 2003. As will be discussed below, the actual-to-future-actual test is in place for all sources.52
45. See Letter from Francis X. Lyons, Regional Administrator, to Henry
Nickel, Counsel for Detroit Edison Company, Detroit Edison Applicability Determination: Detailed Analysis 18 (May 23, 2000), available at http://yosemite.epa.gov/r5/ardcorre.nsf/36ae8bf3212bb6b
28625650c0079f5da/dde17f64f29e6a36862568ef0067cb13/$FILE/
de_enclosure.pdf (last visited Mar. 1, 2005) [hereinafter Letter from
Francis Lyons].
46. 40 C.F.R. §§52.21(b)(21)(ii), 51.165(a)(1)(xii), 51.166(b)(21)
(pre-2002 NSR rule).
47. Id. §52.21(b)(21)(ii) (pre-2002 NSR rule). This provision was
heavily influenced by the WEPCO decision.
48. Id.
49. Id. §§52.21(b)(48), 51.165(a)(1)(xxxv), 51.166(b)(47) (2003).
50. Id. §52.21(b)(4) (pre-2002 NSR rule).
51. See 893 F.2d at 901. EPA’s continued use of the actual-to-potential
test was upheld by federal courts. See, e.g., Puerto Rican Cement Co.
v. EPA, 889 F.2d 292, 20 ELR 20259 (1st Cir. 1989).
52. 40 C.F.R. §§52.21(b)(41), 51.165(a)(1)(xxviii), 51.166(b)(40)
(2003).
b. The WEPCO Rule: The Actual-to-ProjectedFuture-Actual Test and Its Extended Application
As mentioned above, under the preexisting rule, there was
an important exception that electric-generation units, called
electric utility steam-generating units (EUSGUs), were subject to a different standard other than the PTE: the actual-to-projected-actual test that had been adopted in the
1992 regulation, known as the WEPCO rule,53 as a result of
the 1990 WEPCO ruling. In this case, the Seventh Circuit
faulted EPA for wholly disregarding past operating conditions at the facility for which an emission history could be
established, so that “a more realistic assessment of its impact on ambient air quality levels is possible, and thus is directed.”54 The court required EPA to utilize a different calculation method for an electric steam-generating unit’s
like-kind replacements of equipment if it is an established
operation. After the decision, EPA’s WEPCO rule adopted
an “actual-to-future-actual methodology” for changes at
electric utility plants except the construction of a new unit or
reconstruction of an existing emissions unit.55 Under this
formula, the premodification actual emissions are compared
to the projected postmodification actual emissions, and the
baseline emissions are calculated based on the highest
hourly emissions rate achievable in any two-year period
within a five-year period preceding the proposed change.56
For verification purposes, a utility must monitor actual
emissions after the modification and report data and information to the permitting agency for the first five years.57 In
some cases, EPA or the state agency may extend a monitoring period up to 10 years if the 10-year period is determined
to be more appropriate.58 Also, the new regulations exempted emissions increases due to demand growth. Increases in emissions caused by high market demand for
electricity may not be included in the calculation of projected-future-actual emissions.59 Understandably, however,
it is a very difficult task to distinguish between increased
emissions due to demand growth and those emissions increases from the physical or operational change. EPA conceded this problem and proposed to eliminate the demand
growth exclusion in its 1998 Notice of Availability.60 How53. See U.S. EPA, Requirements for Preparation, Adoption, and
Submittal of Implementation Plans; Approval and Promulgation of
Implementation Plans; Standards of Performance for New Stationary Sources, 57 Fed. Reg. 32314 (July 21, 1992) (codified at 40
C.F.R. §§51, 52, and 60).
54. 893 F.2d at 917 (quoting Alabama Power Co. v. Costle, 636 F.2d
323, 379, 10 ELR 20001 (D.C. Cir. 1979) (emphasis in original).
55. 57 Fed. Reg. at 32326. See Memorandum from John Seitze, Director
of Air Quality Planning and Standards, U.S. EPA, to Air Directors
Regions I-X, at 2 (July 1, 1994).
56. 57 Fed. Reg. at 32324. See also 40 C.F.R. §52.21(b)(3)(i)(b)
(pre-2002 NSR rule); Supplemental Analysis of 2002 Final
NSR’s Impact, supra note 20, at F-1.
57. 57 Fed. Reg. at 32325.
58. Id. See 40 C.F.R. §§52.21(b)(21)(v), 51.165(a)(1)(xii)(E),
51.166(b)(21)(v) (pre-2002 NSR rule).
59. 57 Fed. Reg. at 32326; 40 C.F.R. §52.21(b)(41)(ii)(c) (pre-2002
NSR rule). See 893 F.2d at 918 n.13 (observing that market fluctuations in the electricity marketplace make it difficult for utilities to use
synthetic minor permits (quoting Puerto Rican Cement Co. v. EPA,
889 F.2d 292, 298, 20 ELR 20259 (1st Cir. 1989))).
60. EPA said:
[A]ttempting to discern whether increased utilization and
emissions should be attributed to physical or operational
5-2005
NEWS & ANALYSIS
35 ELR 10323
Copyright © 2005 Environmental Law Institute®, Washington, DC. reprinted with permission from ELR®, http://www.eli.org, 1-800-433-5120.
ever, the 2003 new NSR rule extends the WEPCO rule and
the demand growth exclusion to all other industries.61 The
only difference between EUSGUs and other sources is that
a 10-year look-back period applies to the latter in calculating baseline emissions. The owner or operator of an existing source is now allowed to project future-actual emissions based on historical data on its operations during any
one of the 5 or 10 years immediately preceding the proposed change.
c. A Significant Net Increase: Netting
The final element in determining NSR applicability is that
an increase in emissions must be a net increase. Therefore,
the reviewing authority must determine if there will be a net
increase in emissions. Net emissions are determined after
considering “[a]ny other increases and decreases in actual
emissions at the major stationary source that are contemporaneous with the [proposed] change,” in addition to the projected increase in emissions from the change.62 Under the
changes versus purely independent demand-satisfying increased capacity utilization will be much more difficult in the
future, as restructuring in the electric power industry allows
electric-generating companies to compete for retail customers. As a result, the marketplace will drive electric generators
to function as any other consumer-driven industry, that is, to
ensure their ability to supply the market and collaterally to increase their revenues. In addition, as utilities respond to a
competitive market for the generation of electric power they
can no longer be expected to accurately predict their level of
operations and post-change emissions. Each physical or operational change that makes it possible for a source to efficiently increase its level of utilization, then, will likely be
pursued and turned into electricity for sale.
U.S. EPA, Notice of Availability; Alternatives for New Source Review (NSR) Applicability for Major Modifications; Solicitation of
Comment, 63 Fed. Reg. 39857, 39860 (July 24, 1998).
61. 40 C.F.R. §§52.21(b)(41)(ii)(c), 51.165(a)(1)(xxviii)(B)(3),
51.166(b)(40)(ii)(c) (2003) (demand growth exclusion).
62. Id. §52.21(b)(3) (emphasis added). The current regulations for netting was heavily influenced by the D.C. Circuit’s 1979 decision in
Alabama Power Co. v. Costle, 636 F.2d 323, 402, 10 ELR 20001
(D.C. Cir. 1979). The court stated: “The Agency retains substantial
discretion in applying the bubble concept. First, any offset changes
claimed by industry must be substantially contemporaneous. The
Agency has discretion, within reason, to define which changes are
substantially contemporaneous. Second, the offsetting changes must
be within the same source, as defined by EPA.” (emphasis added).
The use of netting was finally upheld by the Court in the famous
Chevron case. Chevron, U.S.A., Inc. v. Natural Resources Defense
Council, 467 U.S. 837, 14 ELR 20507 (1984). The Court observed
that the PSD and nonattainment NSR permit program “represented a
balance between the economic interests in permitting capital improvements to continue and the environmental interest in improving
air quality.” Id. at 851. It finally concluded that EPA’s policy decision to adopt flexible mechanisms such as netting “represent[ed] a
reasonable accommodation of manifestly competing interests and
[was] entitled to deference” in the absence of contrary clear congressional intent. Id. at 865. EPA has developed and implemented a policy to promote emissions trading, including netting. See 51 Fed. Reg.
43814. Netting is an internal trading mechanism in which increased
emissions in one point are used to offset decreased emissions in other
points as long as it is expected that there is no net increases in emissions within the entire plant. Bubble is a very similar mechanism because all individual emission sources under the control of the same
person are regarded as a single source for regulatory purposes, as if
the total emissions combined were coming from a single imaginary
outlet in the bubble. The bubble is what makes netting legal in the
first place. But it is different in that it is used by existing sources to
pursue flexibility in complying with pollution control requirements,
and that the use of bubbles is limited by the regulatory definition of a
source. See Arnold W. Reitze Jr., A Century of Air Pollution Control
preexisting rule, any increase or decrease is deemed to be
contemporaneous if it happened within the five-year period
immediately before the change actually occurs.63 Thus a
major emitting facility can net out of NSR by subtracting
any “creditable” decreases it caused to happen in the two
years of the previous five-year period.64 Sources cannot
claim a change in emissions as offsets that it is otherwise obligated to comply with under their permit conditions or other
applicable laws.65 Any increase in emissions is creditable
“only to the extent that the new level of actual emissions exceeds the old level.”66 To be creditable, any emission decrease must: (1) reflect emission reductions from the old
level of actual emissions or the old level of applicable allowable emissions, whichever is lower; (2) be “enforceable as a
practical matter” before the proposed modification actually
occurs; (3) have “approximately the same qualitative significance for public health and welfare as that attributed to the
increase from [the proposed modification]”; and (4) not result from the use of the “add-on control technology or application of pollution prevention practices” relied on by the
source in qualifying for the Clean Unit exemption.67 These
requirements are designed to prevent “paper credits” from
being used and to ensure that offset trading must represent
real progress toward attainment of NAAQS.
In 1979, EPA proposed a premodification notification requirement with regard to netting, but it was never adopted
due to objection from industry.68 As a result, currently the
owner or operator planning a change to his facility that has
the potential to significantly increase net emissions may
forego NSR completely using a netting mechanism, and is
not required to notify EPA or the state permitting agency of
it. This lack of control over netting practices has been the
target of criticism by environmentalists for creating another
significant loophole, along with the routine maintenance
exclusion, that has allegedly been taken advantaged of by
industry, especially grandfathered coal-fired electric utilities. There is little data available concerning how frequently
63.
64.
65.
66.
67.
68.
Law: What’s Worked; What’s Failed; What Might Work, 21 Envtl.
L. 1549, 1622-25 (1991).
40 C.F.R. §52.21(b)(3)(ii). See also id. §§51.165(a)(vi),
51.166(b)(3). Net emissions after the change are equal to the projected emissions increases from the baseline plus plantwide creditable increases minus plantwide creditable decreases. Note that
states may use a different time period in calculating a net emissions change.
In California, however, netting is not allowed, and NSR thresholds
can be as low as one pound per day. See Report by a Panel of the
National Academy of Public Administration (NAPA) for
the U.S. Congress and the U.S. EPA, A Breath of Fresh
Air: Reviving the New Source Review Program 31 (2003),
available at http://209.183.198.6/NAPA/NAPAPubs.nsf/9172a14f
9dd0c36685256967006510cd/ae53e82c36ab2f1985256d1800494
4b5/$FILE/Fresh+Air+Full+Report.pdf (last visited Mar. 1, 2005)
(citing interview with California air officials) [hereinafter NAPA
NSR Report].
See 40 C.F.R. §52.21(b)(3)(iii)(a)-(b).
Id. §52.21(b)(3)(v).
Id. §52.21(b)(3)(vi)(a)-(d).
U.S. EPA, Requirement for Preparation, Adoption, and Submittal of
State Implementation Plans; Approval and Promulgation of State
Implementation Plans, 45 Fed. Reg. 51923 (Sept. 5, 1979). However, EPA stressed that “owners and operators are hereby put on
notice that they should maintain sufficient records regarding contemporaneous emission increases and decreases so as to verify no
permit was required.” See 45 Fed. Reg. at 52676. EPA added a similar requirement to the 1975 NSPS regulations, but it was subsequently removed.
35 ELR 10324
ENVIRONMENTAL LAW REPORTER
5-2005
Copyright © 2005 Environmental Law Institute®, Washington, DC. reprinted with permission from ELR®, http://www.eli.org, 1-800-433-5120.
netting actually is used. According to one study, approximately 800 netting transactions took place in between 1974
and 1989.69 An estimated cost savings were somewhere between $25 million and $300 million, which mostly came
from avoiding costs associated with permitting processes
and installing modern pollution control.70
It is important to note, however, that the new NSR rule
has changed the preexisting netting policy in a significant
way by adopting the new definition of baseline emissions
(in other words, changing the prior definition of the statutory term “change”). Thus, the baseline year is not the year
when the proposed construction actually takes place, but
any year in a 10-year look-back period during which the
highest emissions were ever recorded.
II. NSR Failures and the Movement to Reform the
Current NSR Program
A. Grandfathering Under the CAA
The perceived failures of NSR are attributable to grandfathering under the CAA. Old sources were exempted because it was thought to be more economically efficient to
mandate the installation of new pollution controls at the
time existing facilities would be upgraded, rather than requiring those facilities to be retrofitted immediately.71 Congress expected many of the existing plants would soon be retired and replaced with new ones, and that future technological breakthroughs would make the costs of state-of-the-art
pollution control technologies significantly lower.72 Unlike
Congress’ expectations, however, the dichotomy between
new and existing sources has allowed grandfathered major
sources to stay operational beyond their expected life cycle
without being subject to NSPS and NSR requirements. Notably, electric utilities have kept their old coal-fired electric
units operating beyond life expectancy.73 Few new coalburning electric power plants have been constructed since
1980, and preconstruction permits for major modifications
have rarely been issued to grandfathered coal-fired power
69. Robert W. Hahn & Gordon L. Hester, Where Did All the Markets
Go? An Analysis of EPA’s Emissions Trading Program, 6 Yale J.
on Reg. 109, 133 (1989). In this article, the authors observed:
From available data it appears that netting is the most commonly used emissions trading activity by a wide margin. In
1984, the only year for which detailed data are available, an
estimated 900 sources used netting. This is about fifteen
times as often as offsets were used during the same year, and
it is far more often than bubbles have ever been used.
Id. (citation omitted).
70. Id. at 136.
71. See H.R. Rep. No. 95-294, at 185-86 (1977), reprinted in 1977
U.S.C.C.A.N. 1077, 1264-65.
72. See id.
73. Although EPA’s NSPS regulations were revised several times, old
electric power plants are still running, thereby even avoiding NSPS
requirements. See Arnold W. Reitze Jr., State and Federal Command-and-Control Regulation of Emissions From Fossil Fuel Electric Power Generation Plants, 32 Envtl. L. 369, 380-83 (2002).
Grandfathering is not limited to NSPS and NSR. Prior to the enactment of the 1990 CAA Amendments, grandfathered power plants
were allowed to disperse their emissions using tall smokestacks.
They were subject to less restrictions when compared to new units.
With the acid rain program being implemented, however, they may
not use that option, since it must reduce their SO2 and NOx emissions
to the levels set by the Act or the regulations.
plants.74 Most of the new power plants have been gas-fired
plants.75 Grandfathering facilitated gaming of the NSR program by the electric power industry which took advantage
of the routine maintenance exception and netting to forego
NSR altogether. This has given huge cost advantages to old,
dirty coal-burning power plants over oil- and gas-fired
power plants, and renewable energy facilities. It creates an
uneven playing field in the energy sector and thus frustrates
efforts to promote efficient use of energy and renewable energy development.76 It apparently contravenes the “polluter-pays” principle enunciated in various international environmental agreements, and is the most significant hurdle
for the United States to moving toward achieving the future
energy policy goal of sustainability.
Electric utilities are by far the nation’s largest polluters.
Old and energy-inefficient coal-fired power plants release
into the atmosphere significant amounts of SO2, NOx and
PM, as well as CO2, disproportionately compared to other
stationary sources, even over 30 years after the passage of
the 1970 CAA. In 1998, electric utilities were responsible
for 25% of national NOx emissions, 67% of SO2 emissions,
and 8% of PM10 emissions, respectively.77 Electricity generation was responsible for approximately 40% of national
CO2 emissions in 2001, which are believed to be associated
74. In 2000, more than 50% of electricity was generated “by coal-fired
power plants, most of which were built between 1950 and 1980.”
Byron Swift, Grandfathering, the New Source Review, and Nitrogen
Oxide—Making Sense of a Flawed System, 15 Env’t Rep. (BNA)
1538, 1538 (2000). According to EPA, of 274 PSD permits issued
since 1995, over 250 have been issued to gas turbine electric power
plants, with only 10 going to coal-fired power plants. NSR Background Paper, supra note 28, at 9.
75. In addition to burdensome NSR requirements, the main reason is
that new gas-fired turbines can produce electricity more cheaply
than new coal-fired power plants. Natural gas power plants are less
expensive and take less time to build, since they requires much lower
construction costs and are relatively small and modular, thereby
minimizing capital expenditures and maintenance and other related
costs. See Environmental Law Institute, Cleaner Power:
The Benefits and Costs of Moving From Coal Generation
to Modern Power Technologies 21 n.17 (2001). In 2000, the
Energy Information Administration projected that 92% of new
power plants were expected to be fired by natural gas during the next
20-year period. U.S. Energy Information Administration
(EIA), Annual Energy Outlook 2001 With Projections to
2020, at 73 (2000) (DOE/EIA-0383). However, construction of new
coal-fired power plants has recently been proposed in increasing
numbers. See Foote, supra note 17, at 10643. In recent years, natural
gas prices have gone up, as gas supply has become constrained due to
increased demand and limited stocks of natural gas in North America. This volatility in gas prices have negatively impacted some of
the combined-cycle gas turbines, whose competitiveness and thus
attraction for investment money rest heavily on relatively low and
stable gas prices. See Tom Woods, What Are the Prospects for
Coal?; Unless Gas Prices Stabilize, Coal Prices Will Continue Rising, Pub. Utils. Fortnightly, May 9, 2004, at 13.
76. Other factors are the low price of coal and the fact that the assets of
old power plants were fully amortized during their life span. The cost
of producing electricity in grandfathered electric power plants
ranges between 1.5 and 3 cents per kilowatt hour (kwh). See Bruce
Biewald, David While & Tim Woolf, Grandfathering and Environmental Comparability: An Economic Analysis of Air
Emission Regulations and Electricity Market Distortions
28, tbl. 5.1. (Prepared for the National Association of Regulatory
Utility Commissioners by Synapse Energy Economics. Inc., Cambridge, Mass. 1998), available at http://www.synapse-energy.com/
publications.htm (last visited Mar. 1, 2005).
77. U.S. EPA, National Air Pollutant Emission Trends:
1900-1998, at 2-2, 2-3 (2000) (EPA 454-R-00-002), available at
http://www.epa.gov/ttn/chief/trends/trends98/trends98.pdf (last
visited Mar. 1, 2005) [hereinafter 1998 National Air Pollutant
Emission Trends].
5-2005
NEWS & ANALYSIS
35 ELR 10325
Copyright © 2005 Environmental Law Institute®, Washington, DC. reprinted with permission from ELR®, http://www.eli.org, 1-800-433-5120.
with global climate change.78 The heavy use of coal for electric generation significantly contributes to acid deposition
and precipitation, ground-level O3 formation, reduced visibility in pristine areas, and global climate change.79 Moreover, in today’s deregulatory environment in the electricity
markets, there also is concern among the environmental
community that coal-fired power plants with cost advantages will be more fully utilized. This implies that the nation’s air will be dirtier and greenhouse gas (GHG) emissions will increase. Given the fact that air pollution problems root in the massive use of fossil fuels for electricity
production, regulation of CO 2 and NSR reform could
produce synergistic effects.80 Inducing the retirement of
grandfathered coal-fired power plants and promoting energy efficiency and use of renewable energy sources should
be a top priority goal. It should be pursued through legal reforms that aim to align energy production with environmental goals.
B. Federal and State Efforts to Repeal Grandfathering
1. Congressional Efforts
Several bills were proposed in Congress that would have removed the grandfathering of old coal-fired power plants. On
October 9, 1990, one U.S. Senate bill, sponsored by Sen. Joseph Lieberman (D-Conn.), would have required some fossil fuel-fired electric-generating units constructed after August 17, 1971, to be subjected to the same emissions standards those applied to new or modified units.81 Affected
units would be fossil fuel-fired steam-generating units with
the capacity of 25 megawatt hours (Mwhs) and interconnected to the interstate electrical transmission grid for the
wholesale sales of electricity.82 The bill would have allowed
78. U.S. EPA, Inventory of U.S. Greenhouse Gas Emissions
and Sinks: 1900-2001, at ES-15 (2003) (EPA 430-R-03-004),
available at http://yosemite.epa.gov/OAR/globalwarming.nsf/Unique
KeyLookup/LHOD5MJQ6G/$File/2003final-inventory.pdf (last
visited Nov. 22, 2004) [hereinafter 2003 U.S. GHG Emissions
and Sinks].
79. In the United States, 90% of coal is consumed in the electricity sector. Id. at 2-13. More than 90% of electric utility NOx and SO2 emissions came from coal-fired power plants, two-thirds of whose emissions took place during bituminous coal combustion. 1998 National Air Emission Trends, supra note 77, at 2-2. Coal combustion represented approximately 85% of CO2 emissions from fossil
fuel-fired electric generation. See 2003 U.S. GHG Emissions and
Sinks, supra note 78, at 2-4, tbl. 2-3. This is so because the carbon
content of coal per unit of energy produced is much higher than that
of petroleum or natural gas. Petroleum and natural gas contain 25%
and 45% less carbon than coal, respectively. Id. at ES-14. Coal contains carbon content of 95 teragrams CO2 equivalent per quadrillion
British thermal unit (Tg CO2 Eq./QBtu), while natural gas’ carbon
intensity is about 53 Tg CO2 Eq./QBtu. Id. at 2-14. Note that Tg CO2
Eq./QBtu is a weighted value of each fuel type’s global warming potential (GWP).
80. Regulation of CO2 requires the reduced usage of fossil fuels. This in
turn would bring “ancillary benefits” to society in the form of reduced emissions of other persistent air pollutants. See Dallas
Burtraw et al., Ancillary Benefits of Reduced Air Pollution in the
United States From Moderate Greenhouse Gas Mitigation Policies
in the Electricity Sector (Resources for the Future, Discussion Paper
No. 01-61, 2001), available at http://www.rff.org/rff/Documents/
RFF-DP-01-61.pdf (last visited Mar. 1, 2005). Strong and effective
NSR enforcement would lead to upgrading grandfathered old facilities, making them more energy efficient and thus saving energy.
Good for the environment. Good for the U.S. economy.
81. S. 2610, 105th Cong. (1998).
82. Id.
five years for those affected units to comply with its mandate, and directed EPA to develop an emissions trading
mechanism to help affected units meet its deadline.83 On
November 7, 1997, a similar bill was introduced in the U.S.
House of Representatives by Rep. Frank Pallone (D-N.J.).84
This House bill provided for a nationwide cap-and-trade
program for NOx and PM2.5.85
2. State Action
Some states have moved to eliminate the grandfathered status of old electric power plants, and/or to adopt a multi-pollutant trading strategy which may include CO2 emissions
control. On May 17, 2000, the governor of the state of Connecticut signed an Executive Order directing the Connecticut Department of Environmental Protection to develop
regulations no later than May 1, 2003, to reduce annual SO2
and NOx emissions from 61 major sources, including all fossil fuel-fired power plants, by 30-50% and by 20-30%, respectively.86 The final regulations, which were promulgated
on December 28, 2000, require covered facilities to take
NOx control measures throughout the year.87 The regulations extend the coverage of the SO2 acid rain program to 61
from 28 units. They also require the 28 units to retire some
of their SO2 allowances, which were initially allocated under Title IV.88 Emissions trading may be used to comply
with the NOx and SO2 reduction requirements.89
In 2001, Texas enacted legislation to phase out grandfathering of one-third of the state’s industrial facilities, under which those in East Texas must go through permitting by
2007, and other facilities by 2008.90 The same year, Massachusetts promulgated a regulation requiring fossil fuel-fired
boilers, including indirect heat exchangers with a nameplate
capacity of 100 Mwhs or more, to meet output-based emission rate standards that would cut NOx emissions by 50%
and SO2 emissions by 74%. This regulation covers mercury,
CO2, and fine particle emissions from power plants and employs a credit trading mechanism.91
83. Id.
84. H.R. 2909, 105th Cong. (1998).
85. Id. The introductory part of the bill expressed concerns about the anticipated harmful effects on the nation’s air quality of electricity deregulation pushed by the Federal Energy Regulatory Commission in
recent years after the passage of the 1992 Energy Policy Act, especially in nonattainment areas suffering from transboundary pollution. It said that the nation and the general public would not benefit
much from competition in the electricity marketplace “if some competitors enjoy an advantage resulting from externalization of environmental or other costs, permitting them to charge prices for electricity that do not reflect the full economic and environmental cost of
production.” Id.
86. Exec. Order No. 19 (May 19, 2000). Connecticut chose to target
these sources to comply with the 1998 NOx SIP call.
87. Regulations of Connecticut State Agencies (RCSA) §22a–174-22.
88. Id. §22a–174-19a.
89. The SO2 program is to be implemented in two phases. In Phase I, averaging is only allowed. For basic information about the final regulations, see Bureau of Air Management, Connecticut Department of Environmental Protection, The Implementation of
Executive Order No. 19: RCSA Sections 22a–174-19a & 22
(2001), available at http://www.dep.state.ct.us/air2/siprac/2001/
sec19.pdf (last visited Mar. 1, 2005).
90. S.B. 493, H.B. 356 (2001).
91. See Emissions Standards for Power Plants, 310 C.M.R. §7.29, at
http://www.mass.gov/dep/bwp/daqc/files/regs/729final.doc (last
visited Mar. 1, 2005).
35 ELR 10326
ENVIRONMENTAL LAW REPORTER
5-2005
Copyright © 2005 Environmental Law Institute®, Washington, DC. reprinted with permission from ELR®, http://www.eli.org, 1-800-433-5120.
New Hampshire’s Clean Power Strategy placed caps on
emissions of four covered air pollutants from three coalfired power plants owned by Public Service Company of
New Hampshire.92 It expects to cut 75% of SO2 and mercury
emissions and 70% of NOx emissions from baseline levels.
This strategy also required a 7% reduction of CO2 emissions
by 2010, which is the same as the reduction target for the
United States under the Kyoto Protocol.93 The state of New
Hampshire employs a cap-and-trade approach, in which the
company is allowed to comply with these requirements using credits earned by purchasing from outside sources or
banking its future emissions, and offers several other incentives designed to lower compliance costs and encourage the
timely attainment by the company of the reduction goals.94
The state of Illinois directed the Illinois EPA to issue findings about the need for controlling emissions from power
plants by September 30, 2004, and, if needed, to propose
rules containing options to reduce those emissions to be finalized by the Illinois Pollution Control Board.95 On June
20, 2002, North Carolina enacted its Clean Smokestacks bill
requiring 14 coal-fired power plants to reduce NOx emissions 77% by 2009, and SO2 emissions 73% by 2013, from
1998 baseline levels.96
C. EPA’s Enforcement Initiative
As a response to perceived failures of the NSR program discussed above, EPA mounted enforcement actions against
coal-fired power plants owned by seven large electric utilities in midwestern and southeastern regions during the
Clinton Administration.97 This enforcement initiative tar92. New Hampshire Clean Power Act, H.B. 284 (2002). See Environmental Services Department, the State of New Hampshire,
Clean Power Strategy: An Integrated Strategy to Reduce
Emissions of Multiple Pollutants From New Hampshire’s
Electric Power Plants 69-71 (2001), available at http://www.
des.state.nh.us/ard/pdf/NHCPS.pdf (last visited Mar. 1, 2005).
93. Id. at 71, tbl. 11-2.
94. Id. at 71-80.
95. Michael Bologna, Governor Signs Law Seeking to Limit Pollution
From “Grandfathered” Power Plants, 32 Env’t Rep. (BNA) 1602,
1602 (Aug. 10, 2001).
96. Air Quality/Electric Utilities Bill, S.B. 1078 (2002). This bill provides for a two-phase reduction of SO2 emissions from the power
plants, which must reduce 49% of their SO2 emissions by 2009. Key
facts of the bill are available in the official website of the Division of
Air Quality, North Carolina Department of Environment & Natural
Resources, at http://daq.state.nc.us/news/leg/stackfacts.shtml (last
visited Mar. 1, 2005).
97. Reitze, supra note 73, at 389-90; see Office of Legal Policy, U.S.
Department of Justice (DOJ), New Source Review: An
Analysis of the Consistency of Enforcement Actions With
the Clean Air Act and Implementing Regulations 13-14
(2002), available at http://www.usdoj.gov/olp/nsrreport.pdf (last
visited Mar. 1, 2005) [hereinafter DOJ NSR Report]. The initial
list included: (1) American Electric Power Company; (2) Ohio Edison and First Energy; (3) Cinergy Corporation; (4) Southern Indiana
Gas & Electric Company; (5) Illinois Power Company; (6) Southern
Company affiliates (including Alabama Power Company and Georgia Power Company); and (7) Tampa Electric Company. Id. at 14.
The same day, EPA Region IV issued an administrative compliance
order under CAA §§113 and 167 against nine coal-fired power
plants owned by the TVA. See id. 16. In September 2000, the EAB
ruled that none of the TVA’s 14 rehabilitation projects did not qualify for the routine maintenance exception under the NSR and NSPS
regulations. TVA then filed a lawsuit challenging the administrative
compliance order in the Eleventh Circuit. After several jurisdictional
issues had been resolved, the three-judge panel in the Eleventh Circuit finally held that it lacked jurisdiction to review the order since
geted the industry’s decade-old practice in which electric
utilities made component replacements incrementally for
the purpose of maintaining reliability, efficiency, and safety
of electric-generating plants. It also included enforcement
actions against the refinery, wood products, and other industries.98 EPA’s changed position was based on its 1998 NSR
guidance, which adopted a more stringent definition of
modifications.99 According to the NSR guidance, there were
two scenarios in which NSR requirements could be invoked: (1) when a stationary source exceeded an applicable
major source threshold level without obtaining a preconstruction permit; and (2) when a stationary source with a
synthetic minor permit exceeded an applicable major source
threshold level in violation of the permit limitation.100 In
both situations, violating sources would be required to undergo the NSR process. EPA’s theory of liability was threefold: (1) utility life-extension projects replacing major components of the unit are not considered routine; (2) the reduced hours of operation during interim shutdowns or curtailments are excluded when the physical construction is involved; and (3) component repair or replacement projects
that caused forced outages or deratings can always be projected to increase the utilization of the unit after the project.101 EPA intended to narrow or close a loophole in the
NSR program that it believed was being taken advantage of
by regulated industries to forego NSR using the routine
maintenance exception. While many electric utilities alleged that EPA’s new interpretation constituted a rulemaking without fair notice as required by the Administrative
the order was “legally inconsequential” and thereby did not constitute final agency action within the meaning of the Administrative
Procedure Act (APA). The panel reasoned that, notwithstanding explicit congressional intent, the statutory scheme was unconstitutionally drafted that authorizes EPA to impose penalties for failure to
comply with an administrative order without any provision for affording challengers due process rights. It concluded that, since a
mere failure to comply with an administrative order cannot be allowed to deprive any person of his property or liberty, the order at issue was not final agency action subject to judicial review. See supra
Part I.C.1. and text accompanying note 40. On December 22, 2000,
EPA sued Duke Energy in the U.S. District Court for the Middle District of North Carolina. DOJ NSR Report, supra, at 15.
98. See id. at 17-19 and 43, app. II; Christopher W. Armstrong, EPA’s
New Source Review Enforcement Initiative, Nat. Resources &
Env’t, Winter 2000, at 203, 203-04. EPA’s NSR enforcement began
in the late 1980s. The early enforcement actions were filed against
the wood products industry. The WEPCO decision was an ignition
point for enforcement actions against large electric utilities. Beginning in the mid-1990s, EPA’s Petroleum Refinery Initiative addressed possible NSR violations in the refinery industry. See DOJ
NSR Report, supra note 97 at 11-19.
99. Memorandum from Eric V. Schaeffer, Director, Office of Regulatory Enforcement, Guidance on the Appropriate Injunctive Relief
for Violations of Major New Source Review Requirements (Nov.
17, 1998), available at http://www.epa.gov/Region7/programs/artd/
air/nsr/nsrmemos/nsrguida.pdf (last visited Mar. 1, 2005).
100. See id. at 3-6. EPA said that “as part of an EPA settlement, the Consent Decree should require a minimum level of control which the
Agency believe[d] ensures BACT/LAER-equivalent emission reductions.” Id. at 3. To avoid the NSR requirements, a new source or
an existing source opting in to the old actual-to-potential test can voluntarily choose to become a “synthetic minor source” by agreeing to
a permit condition setting a federally enforceable emission limit on
the changed unit, which imposes restrictions on its operations, such
as hours of operation less than full capacity, the use of pollution controls, and changes in production. See NSR Background Paper,
supra note 28, at 6-7; Letter from Francis Lyons, supra note 45,
at 18.
101. Makram B. Jaber, Utility Settlements in New Source Review Lawsuits, Nat. Resources & Env’t, Winter 2004, at 22, 23.
5-2005
NEWS & ANALYSIS
35 ELR 10327
Copyright © 2005 Environmental Law Institute®, Washington, DC. reprinted with permission from ELR®, http://www.eli.org, 1-800-433-5120.
Procedure Act (APA),102 some of the offending power
plants facing EPA enforcement actions agreed to the installation of control equipment or implementation of process
changes that were equivalent to NSR requirements through
settlement with EPA.103
tor of the Office of Regulatory Enforcement, resigned in a
protest to the new Administration.107 In his resignation letter, he strongly criticized the Bush Administration for its
hostility to the Agency’s NSR enforcement campaign.108
2. The DOJ’s NSR Report
D. The Overhaul of EPA’s Enforcement Initiative
1. The National Energy Policy Group’s Report to the
President
EPA’s enforcement campaign was subject to a possible
change when the current Bush Administration took office.
In late January 2001, the Bush Administration convened a
National Energy Policy Development Group (NEPD) to be
headed by Vice President Dick Cheney. The NEPD submitted its 170-page report to the president on May 16, 2001.104
The NEPD recommended in its report that the president direct federal agencies to review the NSR program.105 Accordingly, President George W. Bush ordered EPA to conduct a 90-day review of the NSR regulations and asked the
U.S. Department of Justice (DOJ) to independently determine whether EPA’s enforcement campaign was consistent
with the CAA or its implementing regulations, or whether it
constituted administrative rulemaking within the meaning
of the APA.106 Shortly thereafter, Eric Schaeffer, EPA Direc102. 5 U.S.C. §§551 et seq., available in ELR Stat. Admin. Proc.
103. As of the end of 2003, the DOJ settled litigation with the following
companies: Tampa Electric Company, PSEG Fossil Limited Liability Company, Virginia Electric Power Company, Wisconsin Electric, Southern Indiana Gas & Electric Company, and Alcoa, Inc.
(electric utilities); Chevron, U.S.A., Inc., Conoco, Premcor, Navajo
Refining Company, Montana Refinery, Murphy Oil, Cenex, Ergon,
Coastal Eagle Point, Koch Petroleum Corporation, BP Exploration
& Oil Company, Motiva/Equilon/Shell, and Marathon Ashland Petroleum Limited Liability Company (refineries); Willamette Industries, Inc. and Boise Cascade Corporation (wood products companies). On October 6, 2004, EPA announced that it had reached settlement with Citgo Petroleum Corporation, one of the nation’s largest
refineries. Under the settlement, Citgo agreed to spend an estimated
$320 million to install state-of-the-art emissions control technologies, to pay a $3.6 million civil penalty, and to invest in $5 million
worth of a supplemental environmental project at one of its six refining facilities. EPA estimated that the settlement would reduce NOx
and SO2 emissions by more than 7,184 and 23,250 tpy, respectively. U.S. EPA, Press Release, U.S. Announces Clean Air Agreement With CITGO Petroleum Corp.; Petroleum Refiner to Reduce Air
Emissions at Refineries in Five States (Oct. 6, 2004), available at
http://yosemite.epa.gov/opa/admpress.nsf/b1ab9f485b098972852562
e7004dc686/db2aa8edb865a54485256f2500534222!OpenDocument
(last visited Mar. 1, 2005); see also Juliet Eilperin, Citgo Reaches
$323 Million Emissions Settlement, Wash. Post, Oct. 7, 2004,
at A12.
104. Report of the National Energy Policy Development Group,
National Energy Policy: Reliable, Affordable, and Environmentally Sound Energy for America’s Future (2001),
available at http://www.whitehouse.gov/energy (last visited Mar. 1,
2005).
105. It recommended to the president that EPA, in consultation with the
Secretary of Energy and other federal agencies, “review New Source
Review regulations, including administrative interpretations and
implementation, and report to the president within 90 days on the impact of the regulations on investment in new utility and refinery generation capacity, energy efficiency, and environmental protection.”
It also recommended that the DOJ “review existing enforcement actions regarding New Source Review to ensure that the enforcement
actions are consistent with the Clean Air Act and its regulations.” Id.
at 7-14.
106. David G. Mandelbaum, Thoughts on the Bush Clean Air “Strategy”
So Far and a Suggestion for What Might Work, 21 Temp. Envtl. L.
& Tech. J. 1, 6 (2002).
On January 15, 2002, the DOJ published its NSR report. The
DOJ’s NSR report almost exclusively focused on the routine maintenance exception. It asked: (1) whether the enforcement actions constitute “a substantive change in EPA’s
interpretation of the CAA and its regulations that would require APA-compliant notice-and-comment rulemaking”;
and (2) whether, despite a lack of administrative rulemaking, EPA’s interpretation of the routine maintenance exception is “reasonable” in light of the Act and its implementing
regulations, and prior guidance documents.109 Based on a
reading of the applicable case law, the DOJ found administrative rulemaking procedures unnecessary because it believed that EPA’s legal position in the enforcement actions
against large electric utilities could be categorized as being
interpretive and “did not constitute a departure from a prior
authoritative interpretation of ‘routine maintenance.’”110
Moreover, the report emphasized that EPA was entitled to
Chevron deference111 in its interpretation of the CAA, and
that it deserved utmost deference as announced in Bowles v.
Seminole Rock & Sand Co.112 in the interpretation of its own
implementing regulations.113 It finally concluded that it
would continue to pursue the enforcement actions pending
in federal courts.114
In this regard, on October 24, 2002, the Southern District
of Indiana held that EPA’s interpretation did not constitute
a rulemaking in violation of the APA.115 Therefore, there
exists an authoritative judgment that EPA’s enforcement
actions are not unlawful. However, the reason why the pace
of the litigation has been slow thus far is that EPA has to
prove facts which often spanned more than two decades.
Most of the reviewing courts did not enter summary judgment for EPA.116
3. EPA’s 90-Day NSR
For its part, EPA finalized its 90-day review of the NSR pro107. See Steve Cook, Departing EPA Official Issues Broadside at Bush
Administration’s Clean Air Programs, 33 Env’t Rep. (BNA) 462
(Mar. 1, 2002).
108. See Letter of Resignation, Eric V. Schaeffer, Director Office of
Regulatory Enforcement (Feb. 28, 2002), available at http://www.
Mindfully.org/Reform/2002/Resignation-SchaefferEPA28feb02.
htm (last visited May 25, 2004).
109. See DOJ NSR Report, supra note 97, at 24.
110. Id. at 25-33.
111. See 467 U.S. 837, 842-43, 14 ELR 20507 (1984).
112. See 325 U.S. 410 (1945).
113. See DOJ NSR Report, supra note 97, at 35-36. It argued that the
Christensen case decided by the Supreme Court in 2000 reaffirmed
the Seminole Rock’s holding. DOJ NSR Report, supra note 97, at
36; see Christensen v. Harris Country, 529 U.S. 576, 588 (2000).
114. DOJ NSR Report, supra note 97, at 39-40.
115. United States v. Southern Ind. Gas & Elec. Co., 2002 WL 31427523
(S.D. Ind. Oct. 24, 2002); see also United States v. Ohio Edison Co.,
276 F. Supp. 2d 829, 33 ELR 20253 (S.D. Ohio 2003).
116. See, e.g., United States v. SIGECO, 2003 WL 21024595 (S.D. Ind.
Apr. 17, 2003).
35 ELR 10328
ENVIRONMENTAL LAW REPORTER
5-2005
Copyright © 2005 Environmental Law Institute®, Washington, DC. reprinted with permission from ELR®, http://www.eli.org, 1-800-433-5120.
gram in June 2002.117 EPA’s final report borrowed largely
from the findings of its pervious study of NSR118 and comments from various stakeholders, including members of the
general public. It addressed the impacts of the NSR program
on capital investment in the energy sector, especially for
electric utilities and refineries. This issue had important implications for the adequacy and reliability of the nation’s energy supplies. EPA’s findings were twofold. For new power
plants and refineries, EPA found that
the NSR program has not significantly impeded investment in new power plants or refineries. For the utility industry, this is evidenced by significant recent and future
planned investment in new power plants. Lack of construction of new greenfield refineries is generally attributed to economic reasons and environmental restrictions
unrelated to NSR.119
For existing power plants and refineries, EPA concluded
that
the NSR program has impeded or resulted in the cancellation of projects which would maintain and improve reliability, efficiency and safety of existing energy capacity. Such discouragement results in lost capacity, as well
as lost opportunities to improve energy efficiency and reduce air pollution.120
EPA’s findings appeared to favor more to industry’s dominant view that the NSR program “discourage[s] investment
in both preserving and maintaining utility and refinery generating capacity as well as in improving energy efficiency
and expanding capacity.”121 Furthermore, whereas it said
there is no question that the NSR program has made a significant contribution to improving the nation’s air quality,122
EPA stated that
[it] also believes, however, that for particular industry
sectors the benefits currently attributed to NSR could be
achieved much more efficiently and at much lower cost
through the implementation of a multipollutant national
cap and trade program. In particular, the President’s
Clear Skies initiative is a much more certain and effective way of achieving emissions reductions from the
power generation sector.123
There may be some truth in these findings in view of the
time delays and costs associated with the NSR process itself. Perhaps, NSR may have “failed to accommodate adequately industries with short product cycles and large-scale
batch production, affecting them in ways that may reduce
their competitiveness.”124 But the findings and industry
comments seemed to reaffirm the common understanding
that the NSR program has not worked as intended at the time
of its enactment. They also implied that the CAA’s NSR
scheme may have been gamed by some industries. The low
number of NSR permits issued to old, dirtier electric power
units and cost disparities between grandfathered and new
power plants dictate this conclusion. Therefore, it is one
thing to say that NSR reform is needed, and it is another to
argue that NSR itself is to blame for its alleged failures to induce clean energy development. The challenge is how to redesign the regulatory structure in a way that distinguishes
good- and bad-faith players and rewards the former. There
also is a need to level the playing field for alternative energy
resources, which, thus far, have been disadvantaged under
the current regime.
E. A Multi-Pollutant Trading Approach at the Federal
Level
1. Four-Pollutant Bills
117. U.S. EPA, New Source Review: Report to the President
(2002), available at http://www.epa.gov/nsr/documents/nsr_report_
to_president.pdf (last visited Mar. 1, 2005) [hereinafter EPA 90Day NSR].
118. See NSR Background Paper, supra note 28. An interagency
group composed of several federal agencies and a private consulting firm, ICF Consulting Inc., participated in the preparation of
this preliminary report. See EPA 90-Day NSR, supra note 117,
at 2-3.
119. Id. at 1.
120. Id.
121. See id. at 8-21. However, it did not make any definitive findings supported by fresh data and rigorous analysis. EPA based its findings
more on general perceptions among industry than on hard evidence.
Its NSR Background Paper did find that “capital expenditures for
air pollution control as a percentage of total capital expenditures on
new plant construction are significantly lower than those expenditures on existing plants.” NSR Background Paper, supra note 28, at
18. It noted that it could not answer the question of “whether or not
NSR had affected the economic behavior of new plant owners or developers.” Id. at 21. It then illustrated a number of factors that may
contribute to cost increases, such as the costs of pollution costs, and
time delays, complexity and regulatory uncertainty, commonly associated with NSR. Id. at 21-23. It did state, however, that these costs
were difficult to quantity, and referred to comments and studies that
argued economic factors, not environmental regulations, are decisive in making siting and expansion decisions. Id. at 24. As for the refinery industry, it found that pollution control costs constituted a
small portion of capital investment. Id. at 41-42. Then EPA rather
summarily concluded that NSR has impeded or resulted in the cancellation of projects that would maintain or improve reliability, efficiency or safety of existing power plants and refineries.
122. It conceded the difficulty of quantifying the benefits in the report,
but its NSR Background Paper did estimate that as a result of NSR
4.1 million tons of all regulated air pollutants per year were avoided
There have been legislative efforts to introduce multi-pollutant bills primarily targeting the electric utility industry.
This move is inspired largely by the relatively successful
performance of the acid rain program under the 1990
CAA.125 On the other hand, it derives in part from the widespread recognition that it is much more economical to concentrate regulatory energy and efforts on the electric utility
industry often characterized by inefficiencies. The so-called
four-pollutant bill, called the Clean Power Act, was introduced in the Senate to mandate reductions in SO2, NOx, mercury, and CO2 emissions from electric power generators using a cap-and-trade approach on a pollutant-by-pollutant
basis.126 This Senate bill would require the electric-generation industry to cut 75% of its SO2 and NOx emissions, 90%
of its neurotoxin mercury emissions, and 20% of its CO2
emissions, respectively, and calls for implementation of
policies such as strengthened efficiency standards for buildings and appliances, and incentives for development of renewable energy sources.127 But it is now being stalled in the
123.
124.
125.
126.
127.
or 1.4 million tons of criteria pollutants per year. It noted that 90% of
those reductions were thought to be from electric utilities. EPA 90Day NSR, supra note 117, at 9.
Id. at 2 (emphasis added).
NAPA NSR Report, supra note 64, at 100-03.
See 42 U.S.C. §§7651-7651o.
S. 556, 107th Cong. (2001).
Id.
5-2005
NEWS & ANALYSIS
35 ELR 10329
Copyright © 2005 Environmental Law Institute®, Washington, DC. reprinted with permission from ELR®, http://www.eli.org, 1-800-433-5120.
face of opposition to the inclusion of CO2 as one of the covered pollutants.128
2. Three-Pollutant Bills: The Bush Administration’s Clear
Skies Initiative
President Bush has pushed the Clear Skies Initiative (threepollutant bill),129 and, on February 27, 2003, S. 485 and H.R.
999, known as the Clear Skies Act of 2003, were introduced
in both houses of Congress, which would create a cap-andtrade program for SO2, NOx, and mercury for electric utilities, with a possibility of future downward cap adjustments,
and deadlines of 2008, 2010, and 2018.130 This bill would
exempt all changes at existing sources, called “affected
units,” from NSR requirements, and would require that new
sources meet NSPS rather than BACT or LAER. A stationary source qualifies as an affected unit if it satisfies the requirement that its operational changes not “increase the
maximum hourly emissions of any air pollutant achievable
at the unit during the last five years.”131 Therefore, the proposed bill tried to amend the current law concerning NSR
preconstruction review. EPA’s recent NSR reform efforts
have been based on this bill’s basic scheme.132 On the other
hand, the Clear Skies Initiative is designed to reduce carbon
intensity by encouraging electric power plants to develop
clean coal technologies with regard to CO2 emissions and by
supporting other programs to enhance energy efficiency and
to develop renewable energy resources and clean fuels.133
Carbon intensity is “the ratio of [GHG] emissions to economic output,” and President Bush’s plan aims to reduce
carbon intensity by 18% in the next 10 years.134 This effort
could be promising, given the fact that energy-related CO2
emissions contribute over 80% of national GHG emissions.135 But it appears that the plan will not greatly help the
nation to achieve the Kyoto Protocol target, or even the stabilization goal of the Climate Change Convention, because
it does not impose any legal obligations on industry to reduce CO2 emissions, and because the U.S. economy is ex128. This bill was reintroduced in 2003. S. 366 and H.R. 2042, 108th
Cong. (2003). This bill had a “birthday” provision that would require
existing power plants to meet NSR and NSPS requirements 30 years
from either the date of the plant began operation or 10 years after passage of the bill, whichever is later. For updated information about
congressional activity, see Pew Center on Global Climate Change,
108th Congress Proposals, at http://www.pewclimate.org/what_s_
being_done/in_the_congress/108th.cfm#Clean_Coal (last visited
Mar. 1, 2005).
129. See Steve Cook, Bush’s Clear Skies Plan Moving Slowly in Both
Houses; Senate Markup Planned, 34 Env’t Rep. (BNA) 2009
(2003).
130. S. 485 and H.R. 999, 108th Cong. (2003).
131. See U.S. EPA, Summary of the Clear Skies Act of 2003, at 4-5
(2003), available at http://www.epa.gov/air/clearskies/CSA2003
shortsummary2_27_03_final.pdf.
132. See, e.g., Hearings Before the Clean Air Subcomm. of the Comm. on
Environment and Public Works (Apr. 8, 2003) (testimony of Christine Todd Whitman, Administrator, EPA), available at http://www.
epa.gov/air/clearskies/testimony.html (last visited Mar. 1, 2005).
133. See News Release, White House, Global Climate Change Policy
Book (Feb. 2002), available at http://www.whitehouse.gov/news/
releases/2002/02/climatechange.html (last visited Mar. 1, 2004).
134. Id.
135. U.S. EPA, Inventory of U.S. Greenhouse Gas Emissions
and Sinks: 1990-2000, at 2-1 (2003) (EPA 236-R-02-003), available at http://yosemite.epa.gov/oar/globalwarming.nsf/UniqueKey
Lookup/SHSU5BMQAR/$File/2002-inventory.pdf (last visited
Nov. 22, 2004).
pected to continue to grow. In its 2002 report to the United
Nations, the U.S. Department of State projected that GHG
emissions would increase by 42.7% until 2020.136 It is expected that political support would increase for addressing
global warming through regulation of CO2 as an air pollutant, but, as of now, the lack of political enthusiasm remains a
significant obstacle to establishing a nationwide global
warming strategy.
After a series of energy bills had been defeated in Congress, EPA announced that it would pursue a regulatory approach until a White House-sponsored three-pollutant bill is
passed.137 On March 15, 2005, it promulgated regulations,
which would adopt an emission trading mechanism for utility SO2, NOx, and mercury emissions.138 Under the finalized
rules, the current SO2 cap emissions would be further tightened and NOx control would become more stringent.139
However, the exclusion of CO2 may increase the longterm costs of CO2 control by postponing the issue of regulation of CO2 to a future day, and could have chilling effects on voluntary efforts by the private sector and some
states to reduce CO2 and other GHG emissions.
III. The 2002 New NSR Rule
On December 31, 2002, EPA promulgated new NSR rule,
which took effect on March 3, 2003.140 The same day, it published a proposed rule for changing a regulatory definition
for the “routine maintenance, repair, replacement exemp136. U.S. Department of State, Climate Action Report 2002, at 73
(2002), available at http://yosemite.epa.gov/oar/globalwarming.
Nsf/content/ResourceCenterPublicationsUSClimateActionReport.
html (last visited Mar. 1, 2005).
137. Press Release, U.S. EPA, New Power Plant Rule to Achieve Largest
Emission Reductions in a Decade (Dec. 4, 2003), available at http://
yosemite.epa.gov/opa/admpress.nsf/b1ab9f485b098972852562
e7004dc686/17302e197330932585256df200686549?Open
Document (last visited Mar. 1, 2004).
138. U.S. EPA, Rule to Reduce Interstate Transport of Fine Particulate
Matter and Ozone (Interstate Air Quality Rule); Revisions to Acid
Rain Program; Revisions to the NOx SIP Call, 70 Fed. Reg. __ (Mar.
15, 2005) (codified at 40 C.F.R. §§51, 72, 73, 74, 77, 78, and 96);
U.S. EPA, Standards of Performance for New and Existing Stationary Sources: Electric Utility Steam-Generating Units, 70 Fed. Reg.
__ (Mar. 15, 2005) (codified at 40 C.F.R. §§60, 63, 72, and 75).
139. Eric Pianin, EPA Aims to Change Pollution Rules: Utilities Could
Buy Credits From Cleaner-Operating Power Plants, Wash. Post,
Dec. 5, 2003, at A2. The final rules are based on proposed rules dated
January 30, 2004. Under the proposed interstate air quality rules, a
cap on SO2 emissions would be further tightened in the eastern half
of the United States beyond the current level required under the
Act’s acid rain program. 69 Fed. Reg. at 4617. This was possible by
applying a new PM2.5 standard in view of the fact that SO2 emissions
are main precursors to fine particle pollution and regional haze. And,
more areas and sources would have to comply with more stringent
NOx control requirements under the new eight-hour O3 standard.
EPA proposed to find that NOx emissions from sources in 25 states
and the District of Columbia significantly contribute to the
nonattainment of the new eight-hour O3 NAAQS in downwind areas. Id. at 4570. Since NOx emissions are also precursors to the formation of fine PM, NOx reduction requirements would be imposed
throughout the year in areas found to significantly contribute NOx
emissions to PM2.5 NAAQS nonattainment downwind. See id. at
4633. In the final rules, EPA made relatively minor changes to the
proposed rules by reducing the number of covered states subject to
new fine particle-related SO2 and Nox reduction requirements based
on new modeling results, by adding such new features as opt-in requirements, and by deciding to apply new NOx reduction requirements one year earlier. For a concise summary of the final rules, see
U.S. EPA, Basic Information (2005), available at http://www.
epa.gov/cair/basic.html.
140. 67 Fed. Reg. at 80186.
35 ELR 10330
ENVIRONMENTAL LAW REPORTER
5-2005
Copyright © 2005 Environmental Law Institute®, Washington, DC. reprinted with permission from ELR®, http://www.eli.org, 1-800-433-5120.
tion” to NSR.141 After reviewing the comments received on
proposed changes to the routine maintenance exemption,
EPA issued a final rule in October 2003.142
A. A 10-Year Look-Back Period and the Actual-toProjected-Future-Actual Test
Generally speaking, the December 31, 2002, rule was based
on the 1996 NSR reform proposal.143 Its key provisions can
be summarized as follows. The actual-to-projected-futureactual test is extended to all source categories. The new rule
did not adopt the potential-to-potential test proposed by
some members of the electric utility industry.144 Sources
have the option to choose between the preexisting actualto-potential test and the new actual-to-future-actual test.145
Sources choosing the former are not subject to recordkeeping requirements that otherwise apply to sources using the
actual-to-future-actual test.146 The actual-to-future-actual
test allows source owners to project future actual emissions
based on projected capacity and usage, historic trends and
emissions from the unit before the modification, and other
emissions factors, during any one of the 5-10 years immediately preceding the proposed change.147
Different time periods for determining the baseline emissions apply: a 10-year look-back period for all industries,
except for EUSGUs that is still subject to the WEPCO rule’s
five-year period. Sources, except EUSGUs, may use any
consecutive 24-month period in the past 10 years.148 They
may consider “the utilization rate of the equipment, fuels
and raw materials used in the operation of the equipment,
and applicable emission factors.”149 However, past emissions that are not allowed under “the most current legally
enforceable limits” applicable to the changed unit must not
141. Id. at 80290.
142. U.S. EPA, Prevention of Significant Deterioration (PSD) and
Nonattainment New Source Review (NSR): Equipment Replacement Provision of the Routine Maintenance, Repair, and Replacement Exclusion, 68 Fed. Reg. 61248 (Oct. 27, 2003). The purpose of
this new rule is to maintain and improve safety, reliability, and efficiency. Id. at 61252. It contains two key features, which all aim to
simplify the applicability of the routine maintenance exception,
thereby offering regulatory certainty. First, it established the cost
threshold at 20% of the cost to replace the entire process unit. To
qualify, replacement equipment must be “functionally equivalent,”
which means it serve the same function and basic design parameters
as the old one, such as heat input and fuel consumption. Id. Second,
activities whose cost are below an annual maintenance allowance to
be set by EPA on an industry-by-industry basis would be exempt
from NSR. Id. Because of a lot of controversy about this second option, EPA did not finalize it in issuing the final rule. Id. (“We have
decided, for now, not to take final action on the proposed annual
maintenance, repair and replacement allowance approach.”). As a
result of this final rule, many of the changes to existing sources that
would otherwise trigger NSR under the CAA will be out of NSR altogether. EPA’s position on the routine maintenance exception is
well reflected in its pronouncement of NSR recommendations,
which had been published immediately before the 2002 new NSR
rule was proposed. See U.S. EPA, New Source Review, New
Source Review Recommendations 4-6 (2002).
143. U.S. EPA, Prevention of Significant Deterioration and Nonattainment New Source Review; Proposed Rules, 61 Fed. Reg. 38249
(proposed July 23, 1996) (to be codified at 40 C.F.R. §§51 and 52).
144. 67 Fed. Reg. at 80189.
145. Id.
146. Id.
147. Id.
148. Id. at 80196.
149. Id. at 80195.
be included.150 For EUSGUs, the baseline emissions are the
average emission rate, in tpy, of any regulated pollutant during any two years within the five-year period shortly before
the proposed construction begins.151
The new rule not only retained the demand growth exclusion, which had been applied to electric utilities as part of
the WEPCO rule, but also extended the exclusion to all
sources. Therefore, sources “could exclude emissions resulting from increased utilization due to demand growth that
the unit could have accommodated before the change” in
calculating projected future emissions or establishing
PALs.152 This was in stark contrast to EPA’s position in its
1998 Supplemental Notice.153 Furthermore, the source
owner is not subject to rigorous reporting requirements.
B. PALs
PALs can only be established through a public comment
process.154 A PAL is one option that sources can use to avoid
NSR. It establishes a plantwide emissions cap for any regulated pollutant, which is similar to a bubble concept.155 As
long as a PAL is not exceeded, a source can increase its
emissions without triggering NSR, thereby affording maximum flexibility.156 If PALs are set in an environmentally
friendly manner and can actually be successful in giving incentives for source owners to install state-of-the-art pollution control equipment or processes, they could make a significant contribution to achieving the dual goals of NSR reform: environmental protection and economic growth. EPA
used the term “baseline actual emissions” instead of “actual
emissions” that it had used in its 1996 NSR reform proposal.157 While in its 1998 notice it had said that it considered requiring facilities to consider contemporaneous emission decreases and increases, EPA stated that “there is no
need [ ] to quantify contemporaneous emissions increases
and decreases for individual emissions units.”158
150.
151.
152.
153.
154.
155.
156.
157.
158.
Id.
Id. at 80189.
Id. at 80192.
See 63 Fed. Reg. at 39860-61 (stating that the exclusion “ignore[d]
the realities of a deregulated electric power sector,” and that its
“self-implementing and self-policing” mechanism created enforcement problems).
67 Fed. Reg. at 80206. The applicant must go through all applicable
procedural requirements, under the state’s minor NSR permit program or the Title V operating program. “Where the PAL is established in a major NSR permit, major NSR public participation procedures apply.” Id. at 80208.
As a practical matter, PALs can be established for more than one pollutant. See id.
Id. at 80206.
See id. n.26.
Id. at 80206-07; see also 63 Fed. Reg. at 39863 (pointing out that
PALs could function as another loophole for avoiding NSR if the
contemporaneity requirement as dictated by the Alabama Power
court is not applied). EPA noted:
We believe that the concept of contemporaneity, as articulated
in Alabama Power and as set forth in the regulations governing
the major NSR program, does not apply to PALs. The PAL
program differs in certain important respects from our current
regulations and from the 1978 regulations at issue in Alabama
Power. The Alabama Power court was not presented with the
PAL approach for determining whether there was an increase
in emissions and did not consider whether the principles it set
forth in its opinion would apply to such an approach.
67 Fed. Reg. at 80215.
5-2005
NEWS & ANALYSIS
35 ELR 10331
Copyright © 2005 Environmental Law Institute®, Washington, DC. reprinted with permission from ELR®, http://www.eli.org, 1-800-433-5120.
A PAL is the sum of the baseline actual emissions plus an
amount equal to the applicable significant level for the covered pollutant.159 The formula for establishing a PAL is
complicated. The calculation of the baseline emissions distinguishes between existing units with more than a two-year
operating history and new units with less than a two-year operating history.160 The first step is to calculate the baseline
actual emissions based on the average emission rates, in tpy,
of existing units for the covered pollutant that existed during
any consecutive 24-month period chosen by the applicant
within the 10-year period immediately preceding the
change.161 The second step is to add emissions equal to PTEs
of existing and new units that were constructed since the
24-month period and then to subtract the emissions of any
units that was “permanently shut down” or dismantled since
that time period.162 Baseline emissions must reflect currently available technology requirements.163 A PAL’s
facilitywide emissions cap should consider fugitive emissions to the extent it is quantifiable.164 The permitting
agency must include a PAL in a federally enforceable permit.165 It must specify in the permit that a reduced PAL level
would be imposed at the time any applicable federal or state
requirements that it is aware of prior to issuing the permit.166
A PAL will be valid for 10 years.167 It would be renewed
for another 10 years if there is a timely request for renewal
and the reviewing authority approves it.168 At renewal time,
the PAL must be revised in consideration of newly applicable requirements.169 Where a source’s PTE has declined below the PAL, the new PAL must be readjusted at a level that
does not exceed its PTE.170 Besides these requirements, in
general, the reviewing authority has great discretion to
choose the new PAL. It can approve the application without
any adjustments to the original PAL if the sum of the baseline actual emissions plus an amount equal to the applicable
significant level for the covered pollutant is equal to or
greater than 80% of the PAL level.171 If it is less than 80%,
159. 67 Fed. Reg. at 80208. The 1996 NSR reform proposal used the language “a reasonable operating margin less than the applicable significant emissions rate.” 61 Fed. Reg. at 38265. The final rule has instead chosen “the applicable significant amount” as specified in the
NSR regulations or the CAA. For explanation on the part of EPA, see
67 Fed. Reg. at 80218-19.
160. 67 Fed. Reg. at 80218-19.
161. Id. EPA said that “you will have broad discretion to select any consecutive 24-month period in the last 10 years to determine the baseline actual emissions.” Id.
162. Id. at 80208-09. For EUSUGs, however, a different formula is applied. Whether nor not a shutdown is considered permanent is decided on a case-by-case basis considering all relevant facts and circumstances. The foremost consideration is the intention of the owner
or operator of the unit at issue. There is a rebuttable presumption that
“[s]hutdowns of more than 2 years, or that have resulted in the removal of the source from the State’s emissions inventory,” are permanent. Id. at 80209 n.30.
163. Id. at 80209.
164. Id. at 80208.
165. Id.
166. Id. at 80209.
167. Id.
168. See id. at 80209-10. At least 6 months prior to, but not earlier than 18
months from, the expiration date of the PAL, the facility owner or
operator must submit a complete application. Id. at 80209.
169. Id.
170. Id.
171. Id.
the authority may create a new PAL level that is more representative of the source’s actual emissions, or taking into account other relevant factors.172 Despite emissions increases
exceeding the plantwide emissions cap, the PAL may be adjusted upward without triggering NSR requirements if it is
demonstrated that the owner is unable to reduce emissions
levels below the PAL even with BACT-equivalent technology being applied to units that have a PTE greater than the
applicable significant level.173 Otherwise, all exceedances
above the PAL must go through NSR.174 Readjustment decisions during the 10-year term are largely at the discretion
of the reviewing authority with some exceptions.175
Any monitoring system must be “based on sound science
and must conform to generally acceptable scientific procedures for data quality and manipulation.”176 Any monitoring
system contained in the permit must satisfy the minimum requirements as required by the rule.177 Monitoring systems
must be able to precisely quantify the emissions from each
unit on a 12-month rolling basis.178 But this does not mean
that the use of a continuous emissions monitoring system
(CEMS) or other rigorous monitoring requirements is mandated. The source may employ emission factors to monitor
actual emissions at each unit.179 Sources must use “current
emissions or other current direct measurement data.”180 The
reevaluation of the data must occur “at least once every 5
years” for the PAL term, using “a performance evaluation
test or other scientifically valid means [ ] approved by the reviewing authority.”181 Then the final rule provided for minimum recordkeeping requirements.182 The facility owner
must submit a semiannual emissions report to the authority,
172.
173.
174.
175.
Id.
Id. at 80210.
Id.
Id. The final rules explained the reasons for mandatory adjustments:
(1) To correct typographical/calculation errors made in setting the PAL or to reflect a more accurate determination of
emissions used to establish the PAL; (2) to reduce the PAL if
the owner or operator of the major stationary source creates
creditable emissions reductions for use as offsets; or (3) to revise a PAL to reflect an increase in the PAL.
176.
177.
178.
179.
Id.
Id. at 80211.
See id. at 80212-13. See also 40 C.F.R. §60, app. B.
67 Fed. Reg. at 80211. Compliance with the PAL is determined
based on a consecutive 12-month period, rolled monthly. Id. at
80214.
The monitoring system must be one of the following methods or any
combination thereof:
(1) Mass balance for processes, work practices, or emissions
sources using coatings or solvents; (2) Continuous Emissions
Monitoring System (CEMS); (3) Continuous Parameter
Monitoring System (CPMS) or Predictive Emissions Monitoring System (PEMS) with Continuous Emissions Rate Monitoring System (CERMS) or automated data acquisition and
handling system (ADHS), as needed; or (4) emission factors.
Id. at 80211.
180. Id.
181. Id.
182. Id. at 80213. Emissions data during periods of startup, shutdown,
maintenance, and malfunction must be collected even though they
may not be considered part of the emissions in determining compliance with the PAL. The reviewing authority has the discretion to approve different monitoring for various operating conditions for each
unit. However, the facility owner is still subject to the same minimum monitoring requirements. Id.
35 ELR 10332
ENVIRONMENTAL LAW REPORTER
5-2005
Copyright © 2005 Environmental Law Institute®, Washington, DC. reprinted with permission from ELR®, http://www.eli.org, 1-800-433-5120.
and must record “all periods of deviation, including the date
and time that a deviation started and stopped and whether
the deviation occurred during a period of startup, shutdown,
or malfunction.”183
C. The Clean Unit Exclusion
The Clean Unit exclusion applies to units that installed
BACT or LAER through recent NSR.184 A unit can still
qualify as a clean unit if it is demonstrated that its emissions
control level is comparable to BACT or LAER.185 The new
NSR rule adopted a new applicability test. The Clean Unit
status gives its owner operational flexibility to make any
change to the designated clean unit without triggering NSR
if a change to the unit does not alter the emissions limitations
or work practice requirements imposed in the permit in conjunction with BACT or LAER, or physical or operational
characteristics that formed the basis of the BACT or LAER
determination, or if it does not result in a significant net
emissions increase in violation of the CAA.186 The comparability requirement is satisfied when the candidate pollution control technology is BACT or LAER chosen for other
similar sources in the RACT/BACT/LAER Clearinghouse
(RBLC), or when it is demonstrated on a case-by-case basis
that it is “substantially as effective” as BACT or LAER.187
The Clean Unit status will be valid up to 10 years.188 The
new applicability test was a departure from the 1996 NSR
reform proposal, which stated that the new rule would base
the Clean Unit status on the unit’s pre-change hourly potential emissions rate.189 In 1996, EPA proposed that there
would be three routes to be used for the Clean Unit designation. The first and second would be through major or minor
NSR, which took place within the last 10 years.190 As for the
third, to be comparable BACT or LAER, the pollution control technology’s performance level must be: “(1) the average of the BACT or LAER for equivalent sources over a recent period of time (such as 3 years); or (2) [ ] within some
percentage (such as 5 or 10) of the most recent, or average of
the most recent, BACT or LAER levels for equivalent or
similar sources.”191 For the units within the third category,
the 1996 proposed rule provided that the Clean Unit status
would last for five years.192
As a result of these changes, stringency required for qualifying technologies is greatly reduced, and thus it would be
much easier for major sources to use the clean unit option.
Furthermore, its longer term (10 years other than 5 years)
would have potential to offset much of the beneficial effects
that the Clean Unit exclusion might otherwise have. Units
that have gone through major NSR automatically qualify.193
183.
184.
185.
186.
187.
188.
189.
190.
191.
192.
193.
Id.
Id.
Id.
Id.
Id.
Id.
Id.
Id.
Id.
Id.
Id.
at
at
at
at
at
at
at
at
at
80213-14.
80189.
80190.
80189-90.
80222.
80190.
80222.
80222-23.
80223.
Other units must go through a SIP-approved permitting
process.194
A unit may requalify for the Clean Unit status after the
10-year period, subject to the above-mentioned applicability test. In other words, it must go through major NSR or a
SIP-approved permitting process once again and meet technology requirements for pollution control to be adopted at
the unit, reflecting advances in technology and changes to
the existing unit during the effective period.195 Where the
unit’s location has been reclassified as a nonattainment area
during the term, it must install LAER or LAER-comparable
pollution control at the time of expiration to requalify.196
The required emissions reductions under the Clean Unit
exclusion are not allowed to be used for netting purposes or
as offsets.197 But those reductions below the emissions limitation that qualifies the unit as a clean unit can be used in a
netting analysis or as offset credits if the general requirements applicable to netting or offsetting are met.198 In principle, the Clean Unit exclusion is pollutant-specific with the
exception that “simultaneous Clean Unit status [may be
granted] for other pollutants at those emissions units that are
sufficiently controlled to independently qualify as ‘clean’
for each pollutant.”199 The Clean Facility exclusion, which
was proposed in the 1996 proposed rule, was omitted.200
D. PCPs
The new NSR rule extended the utility-specific PCP exclusion to all types of sources.201 Listed PCPs are automatically
exempted from NSR if there is no violation of a NAAQS or
any of the PSD requirements, such as PSD increments and
visibility.202 PCPs that are not listed must pass the “environmentally beneficial” test on a case-by-case basis.203 The
PCP exclusion offers flexibility while giving incentives for
sources to install modern pollution control. Therefore, its
success depends in large part on the effectiveness of the
technology selected.
194. Id.
195. Id. This means that they are subjected to the control technology determination, air quality review, public participation and other requirements under state-administered permit programs, which are designed to ensure that no violation of a NAAQS or any of the PSD requirements (increments or visibility) would not occur.
196. Id. at 80226. Note that the new rule stated:
However, we will not necessarily require you to meet an additional investment test to re-qualify for Clean Unit status for
the same controls. That is, unless the controls used to establish Clean Unit status are no longer BACT/LAER or comparable, there will be no requirement for an investment to requalify for Clean Unit status.
Id.
Id. at 80227.
Id. at 80228.
Id.
If adopted, the clean facility exclusion would have exempted from
NSR requirements major stationary sources that have undergone
NSR for the entire source within the last 10 years. 61 Fed. Reg. at
38258.
201. 67 Fed. Reg. at 80233.
202. Id. at 80190. There is a rebuttable strong presumption that listed
PCPs and technologies, and other standards are environmentally
beneficial. See id. at 80233-34.
203. Id. at 80190. For non-listed PCPs and technologies, the reviewing
authority must consider the case-specific factors and employ a public notice-and-comment process. Id. at 80234.
197.
198.
199.
200.
5-2005
NEWS & ANALYSIS
35 ELR 10333
Copyright © 2005 Environmental Law Institute®, Washington, DC. reprinted with permission from ELR®, http://www.eli.org, 1-800-433-5120.
One of the concerns about the PCP exclusion is that PCPs
may result in emissions increases of any collateral pollutant,
which triggers NSR under the CAA. The 1996 proposal embodied the “primary purpose” test, which said that the primary function of a PCP is to reduce pollution.204 It also
stated that the listed add-on technologies and switch to less
polluting fuels may qualify as a PCP that would be presumed to be environmentally beneficial.205 Other PCPs not
listed must be environmentally beneficial, and new add-on
technologies must be demonstrated in practice.206
The new rule eliminated the primary purpose test and expanded the scope of listed PCPs and add-on technologies.207
It included as potential PCPs energy efficiency projects,
the replacement, reconstruction, and modification of existing pollution control equipment, and work practice standards.208 Upgraded or rebuilt control equipment must
achieve a more stringent level of emissions reductions than
the original one in terms of input- or output-based emissions
rate or must have the same level of performance, provided
that it is more energy-efficient.209 It clarified that non-air
pollution impacts would not be considered in the environmentally beneficial determination.210 It limited the applicability of the PCP exclusion to existing sources.211
One of the most significant changes to the 1996 NSR proposal was that the environmentally beneficial test is conducted based on the determination as to whether a PCP
would have net environmental benefits.212 Net environmen204.
205.
206.
207.
208.
209.
210.
Id. at 80232; see 61 Fed. Reg. at 38261.
67 Fed. Reg. at 80232; see 61 Fed. Reg. at 38260.
67 Fed. Reg. at 80232; see 61 Fed. Reg. at 38261.
67 Fed. Reg. at 80233.
Id.
Id. at 80234.
Id. EPA explained:
[C]ross-media tradeoffs are difficult to compare, so it is difficult to weigh their importance in appraising the overall environmental benefit of a PCP. We solicited comments in the
proposal on how to compare cross-media pollution, but we
received no suggestions on how to design such a system. As a
result, we have determined that it is inappropriate to consider
non-air impacts when considering whether projects, activities, or work practices qualify for the PCP Exclusion.
Id. at 80236.
211. Id. at 80235. EPA reasoned:
Installing or implementing a project on an existing source is
more likely to improve the environment than is the construction of a new source, since one can reasonably expect a PCP
to reduce overall emissions, barring a considerable utilization
increase. New sources, however, introduce new emissions to
the air without reducing existing emissions, and consequently should be as clean as possible. Furthermore, new
emissions units are among the major capital investments in
industrial equipment, which are the very types of projects that
Congress intended to address in the NSR provisions when
such projects result in an overall emissions increase from the
major stationary source. Thus, when emissions from a new
source exceed the significant level, they are subject to NSR,
and all emissions that are generated from the new project
should be addressed in the major NSR permit evaluation for
the major stationary source.
Id.
212. 61 Fed. Reg. at 38262 (soliciting comment on whether applicants for
the PCP exclusion should be required to consider cross-media impacts on any of the applicable CAA requirements; stating that only
“de minimis” increases in the emissions of a collateral pollutant
tal benefits result when the emissions reductions of the primary pollutant(s) are anticipated to outweigh any potential
increases in collateral pollutants.213 The new actual-to-future-actual test is applied to the calculation of any collateral emissions increase.214 PCPs are available both in PSD
and nonattainment areas.215 Only where any collateral pollutant contributing to nonattainment increases by a significant amount as a result of the PCP can the offset requirement apply.216
Another significant change was that the applicant for the
PCP exclusion is not required to conduct air quality modeling if he determines no air quality-related values (AQRVs)
exists in a nearby Class I area that could be impacted by expected collateral emissions increases.217 The applicant can
make this determination after checking information, which
is publicly available on the Internet about whether any
AQRVs such as visibility have been identified for that area
by the federal land manager.218 Even if an AQRV exists that
have been identified by the federal land manager, the applicant also is not required to conduct a modeling analysis if
there is no likely correlation between the AQRV and the pollutants emitted as a result of the PCP, including the case
where collateral emissions will not increase by a significant
amount.219 Then the applicant is merely required to submit
such determination to the reviewing agency.220 In general,
the reviewing authority has the discretion to request more
specific information about adverse impacts on AQRVs in
nearby Class I areas and, if it determines it as necessary, to
require the applicant to conduct air quality modeling.221
These changes aim to streamline the PCP process for providing major sources with incentives to undertake environmentally beneficial projects.222
IV. Another Round of Heated Debate Over the New
NSR Rules
The new NSR rules provoked uproar among many stakeholders. Nine northeastern states brought suit in the U.S.
213.
214.
215.
216.
217.
218.
219.
220.
221.
222.
could be exempted from the application of the environmentally beneficial test).
67 Fed. Reg. at 80232.
Id.
Id. 80237.
Id. EPA said, however, that “a less than significant emissions increase may be subject to a State’s minor NSR requirements.” Id.
Id.
Id.
Id.
Id.
Id.
In this regard, EPA said:
The new, broader PCP Exclusion will ensure equitable treatment of all source categories and remove any disincentive for
companies that wish to install pollution control and pollution
prevention projects, to the extent allowed by the CAA . . . .
Despite today’s rule revisions addressing a broader array of
pollution control and pollution prevention projects at a larger
variety of sources, we feel that the rule’s procedures are less
complex than and are clearer than the WEPCO PCP Exclusion and the July 1, 1994 policy guidance. We are satisfied
that the final PCP Exclusion best achieves the goals of minimizing regulatory burden and reducing procedural delays for
projects that ensure net overall environmental protection.
Id. at 80233.
35 ELR 10334
ENVIRONMENTAL LAW REPORTER
5-2005
Copyright © 2005 Environmental Law Institute®, Washington, DC. reprinted with permission from ELR®, http://www.eli.org, 1-800-433-5120.
Court of Appeals for the District of Columbia (D.C.) Circuit
seeking an order staying the new proposed rule on December 31, 2002, the day it was promulgated.223 They contended
that “the changes to the preexisting NSR regulations deviated from CAA requirements and also that EPA’s
rulemaking process was procedurally flawed.”224 On March
6, 2003, however, the D.C. Circuit refused to grant the motion for the stay.225
A. Criticisms of the New NSR Rule: Environmental
Groups’ Arguments
A coalition of environmental groups also filed a petition for
reconsideration with EPA. Environmental groups argued
that EPA must begin a new rulemaking process for three
reasons: (1) EPA had relied on “materials and recommendations that were developed after the 1996 and 1998 comment periods,” thereby precluding informed public comment; (2) changed circumstances after 1998 justified new
rulemaking, such as electricity deregulation and a better understanding of the public health and environmental effects
of power plant emissions; and (3) the December, 31, 2002,
NSR rule violated the substantive provisions of the CAA,
and EPA failed to adequately explain why the deviations in
the new rule from the 1996 and 1998 proposals were warranted.226 Environmental groups challenged almost every
aspect of the final rule, alleging that it creates too many
loopholes and, hence, would seriously compromise the environmental protection goal of the CAA’s NSR program.
First, they argued that EPA’s business cycle rationale for
the 10-year look-back period is unfounded in light of the
study it relied on,227 and that basing the calculation of the
baseline actual emissions on a source’s highest emissions
rate in any year within a 10-year period would nullify the
simple mandate of the CAA, which requires NSR whenever
a proposed change is expected to result in an emissions increase, and violates the contemporaneity requirement as articulated by the D.C. Circuit in the 1979 Alabama Power
Co. v. Costel228 case. According to environmental groups,
allowing netting transactions in addition to a 10-year baseline period and the use of a different time period for each
regulated pollutant confirmed the conclusion that EPA’s
methodology violates the CAA’s requirement that NSR be
based on contemporaneous emissions increases and decreases as part of the proposed change.229 Also, they warned
that allowing inclusion of fugitive emissions in baseline
emissions calculation would inflate baseline emissions, on
the one hand, and overestimate projected emissions, on the
other hand, making it easy for existing sources to escape
NSR.230 Environmental groups contended that the new
NSR rule lacks meaningful limitations on the discretion of
223. David Mastroyannis-Zaft, EPA’s Revised New Source Review Regulations Take Effect, 30 Ecology L.Q. 805, 805 (2003).
224. Id. at 805-06.
225. New York v. EPA, No. 02-1387, 2003 U.S. App. LEXIS 19029
(D.C. Cir. Mar. 6, 2003).
226. Earthjustice, Petition for Reconsideration 1-8 (2003), available at http://www.earthjustice.org/backgrounder/documents/NSR
petition.pdf (last visited Mar. 1, 2005).
227. Id. at 16.
228. 636 F.2d 323, 10 ELR 20001 (D.C. Cir. 1979).
229. Earthjustice, supra note 226, at 21-25.
230. Id. at 9.
a permitting authority to approve the source’s quantification of fugitive emissions.231 This concern may be addressed by applying conservative assumptions about quantifiable fugitive emissions. But verification systems will
vary in terms of stringency from state to state. Given the fact
that fugitive emissions are extremely difficult to quantify,
inclusion of fugitive emissions will likely create enormous
enforcement problems or big loopholes, depending on the
will or the financial resources of a state or local air quality
management agency.
A recent report, published by the Environmental Integrity
Project (EIP), shows that large quantities of fugitive emissions in “upset” conditions are being released from regulated stationary sources.232 This report analyzed upset reports submitted by 57 facilities in 5 states, which include
California, Louisiana, Ohio, Pennsylvania, and Texas.
These facilities include oil refineries, chemical plants, natural gas-fired power plants, and one carbon black plant.233 Of
these facilities, relatively accurate information about fugitive emissions was available only with regard to 37 facilities
from Texas and Louisiana.234
The EIP found that these facilities, in 2003, emitted fugitive emissions in an amount that is many times greater their
reported 2002 annual emissions. Of six natural gas plants,
four released significant amounts of VOCs and SO2.235 Ten
of the 18 refineries included in the study had annual emissions of at least one pollutant, SO2, CO, or VOCs, that were
more than one-quarter of reported emissions.236 Chemical
plants and the one carbon black plant emitted significant
quantities of VOCs and CO during upsets. Benzene and butadiene, toxic air pollutants subject to regulation under CAA
§112, were released in massive amounts from some of the
chemical plants.237 VOC and CO emissions from the carbon
black plant were 85 and 8 times greater the reported emissions, respectively.238
In overall, these 37 facilities released 63,411,603 pounds
of air pollutants in 2003, which included 167,133 pounds
of benzene and 142,754 pounds of butadiene.239 More
than one-half of these emissions were CO emissions, and
the other one-half were split almost equally by VOCs and
SO2 emissions.240
This report demonstrates two things. First, many stationary sources may have gamed upset provisions under the
CAA and facility-specific permit variances under state
SIPs.241 Second, most states have not yet developed a highly
developed fugitive emissions reporting system. Even in
case a relatively reliable reporting system is in place, such as
one in Texas, regulated sources underreported their fugitive
emissions. Therefore, environmental groups’ argument
231. Id. at 8-9.
232. EIP, Gaming the System: How Off-the-Books Industrial
Upset Emissions Cheat the Public Out of Clean Air (2004)
available at http://www.environmentalintegrity.org/pubs/EIP_upsets_
report_FULL.pdf (last visited Mar. 1, 2005).
233. Id. at 5.
234. Id. at 5, 20-21.
235. Id. at 7-8; see also id. at 8, fig. 2.
236. Id. at 8-9; see also id. at 8, tbl. 2.
237. Id. at 9.
238. Id.
239. Id. at 5, 9; see also id. at 6-7, tbl. 1.
240. Id. at 5, fig. 1.
241. See id. at 13-17.
5-2005
NEWS & ANALYSIS
35 ELR 10335
Copyright © 2005 Environmental Law Institute®, Washington, DC. reprinted with permission from ELR®, http://www.eli.org, 1-800-433-5120.
raising serious concerns about inclusion of fugitive emissions in the calculations of baseline emissions and future
projected actual emissions should be paid attention to by
EPA and state and local air officials.
Second, environmental groups contended that EPA failed
to reasonably explain why it had decided to retain the demand growth exclusion and even to extend it to other industries, because it did not come up with a new rationale for justifying the significant departure from its prior position.242
Specifically, the lack of procedural safeguards “would place
an unduly large burden on [state] permitting authorities,”
deny the public access to emissions data, and transform
NSR into post-construction review.243 Also, state-by-state
variations in NSR enforcement would weaken the effectiveness of the NSR program.244
Third, they argued that EPA’s decision not to subject
PALs to the contemporaneity requirement violated the CAA
and were arbitrary and capricious because of the Agency’s
failure to explain the reason for the change to prior proposals.245 Under the final rule, the term of a PAL is 10 years.
And, despite the bubble concept underlying PALs, a source
owner is allowed to raise the PAL level without undergoing
NSR, provided that the existing major emissions units currently subject to a BACT or LAER requirement that was imposed within the last 10 years are not contributing to the
emissions increase.246 Automatic renewal is granted to the
source whose average emissions in any 2 years within the
preceding 10 years are at least 80% of the PAL. Environmental groups argued that, combined with a 10-year lookback period used for setting a PAL, this would allow for netting during an extended period of time, deprive the public
of a meaningful opportunity to participate in establishing
PALs, and does not create any incentive to install modern
pollution control.247 They also alleged that the final rule
permits the use of alternate monitoring without providing
any meaningful check on the discretion of a state permitting authority and includes emissions factors with inherent
inaccuracies as one of the four monitoring approaches, despite the fact that “a PAL necessitates superior monitoring”
to be effective.248
Fourth, environmental groups criticized EPA for choosing the net emissions benefits test to be used for determining
a qualifying PCP in violation of the statutory requirement
that NSR be triggered whenever a proposed physical or operational change would “result[ ] in the emission of any air
pollutant not previously emitted.”249 More specifically, they
argued that EPA had not provided an opportunity for public
comment on the inclusion of the replacement or reconstruction of an emissions unit as a PCP, or offered any new justification for the departure from its prior position that, however
beneficial, “major capital investments in industrial equipment are the very types of projects that Congress intended to
address in the new source modification provisions.”250
242.
243.
244.
245.
246.
247.
248.
249.
250.
Earthjustice, supra note 226, at 27-31.
Id. at 31-42.
Id. at 33-34.
Id. at 46-55.
Id. at 55-58.
Id. at 58-67.
Id. at 70-85.
Id. at 112-21.
Id. at 121-35.
Finally, environmental groups contended that, unlike the
prior proposals, the final rule sets the effective term of the
Clean Unit exemption at 10 years and allows the exemption
to be renewed for another 10 years without the unit being
subject to new technology requirements.251 They basically
argued that EPA’s chosen method for the Clean Unit designation “flatly contravenes the statutory requirement[s] [ ]
that a modification be determined based on changes that increase emissions,”252 and that the chosen technology should
be BACT or LAER, the most stringent one of its kind.253
B. Concerns About the Revised Routine Maintenance
Exception Rule
Many stakeholders expressed concerns about EPA’s proposed rule for the routine maintenance exception, because it
would allow sources to avoid NSR indefinitely by making
changes to their facilities in an incremental manner, thereby
resulting in increased pollution. They argued that EPA’s categorical approach failed to consider “the large diversity of
industries and situations,” because even sources with the
same industry have different maintenance needs “based on
such factors as age, prior maintenance history, intensity of
use, raw materials used in production processes, climate,
and local labor costs.”254 They also alleged that the cost
threshold and the minimum annual budget allowance to be
established by EPA violated the statutory requirement that
NSR be conducted whenever emissions increases would result from a proposed change.255 According to them, states,
especially downwind states, would face more difficulty ensuring compliance with NAAQS, since “the proposed per se
exemptions deprive states of one of the strongest tools they
have in controlling emissions: federally uniform restrictions
on modifications to existing sources.”256 This is because
cost-based exclusions may “allow sources to operate indefinitely without implementing state-of-the-art control technology.”257 Furthermore, the cost-based approach would
not eliminate uncertainty altogether, in that even an activity
which falls within the annual budget but is in essence not a
routine maintenance is considered a major modification.258
Establishing annual maintenance allowances “invites manipulation of expenditure.” Sources would likely engage in
creative, but bad, accounting practices in an attempt to
spread costs in a multiyear period.259 This in turn will lead to
conflict and litigation over the cost calculation, and thereby
create another administrative complexity and resulting
costs and time, which might otherwise be spent on other
more important legal issues.260
In June 2002, EPA promised to reconsider the rules. On
October 27, 2003, however, it promulgated the final regula251.
252.
253.
254.
255.
256.
257.
258.
259.
260.
Id. at 94-96.
Id. at 94.
Id. at 96-104.
Victor B. Flatt et al., Let the People Speak: Notice-and-Comment
Rulemaking (Lessons From the Controversial New Source Review
Proposal of the Clean Air Act), 34 ELR 10115, 10119, 10122 (Feb.
2004).
See id. at 10119-20, 10124, 10126-27.
See id. at 10129-30.
Id. at 10123.
See id. at 10119.
See id. at 10122.
See id. at 10125.
35 ELR 10336
ENVIRONMENTAL LAW REPORTER
5-2005
Copyright © 2005 Environmental Law Institute®, Washington, DC. reprinted with permission from ELR®, http://www.eli.org, 1-800-433-5120.
tion for the routine maintenance exception without much of
a change.261 The only exception was that EPA decided not to
finalize the annual budget allowance proposal.262 EPA also
decided to implement the NSR rule as finalized on December 31, 2002.
C. The U.S. Government Accountability Office (GAO)
Studies on Stakeholders’ Views on the New NSR Rules
In response to objections to the new rules, Congress commissioned the GAO to conduct surveys of key stakeholders’
views on the revisions. In a October 2003 report, the GAO
found that federal and state enforcement officials and other
stakeholders were concerned that some of the revisions, especially the proposed changes to the routine maintenance
rule, would negatively impact ongoing federal and state enforcement actions and settlement negotiations with industry.263 It also found that certain provisions in the December
2002 final rule would limit the public’s access to emission
data, because “[u]nder the rule, fewer facility changes may
trigger NSR and thus the need for permits and related requirements to notify the public about changes and to solicit
comments—unless state and local air quality agencies have
their own permit and public outreach rules.”264 Furthermore, the GAO pointed out that the lack of clarity about the
definition of “reasonable possibility” and its “self-policing”
nature in NSR determinations with regard to the demand
growth exclusion “could potentially hinder enforcement
and monitoring activities.”265
The October 2003 GAO report prompted some congressional members to ask EPA’s Inspector General to investigate the Administration’s claims that the new regulations
would not affect the ongoing NSR enforcement actions.266
A study conducted by a Rockefeller Family Fund project
and Council of State Governments found that changes to the
preexisting NSR rules could lead to an almost combined 1.4
million tons more of persistent air pollutants in 12 northeastern states.267
In another report published in February 2004, the GAO
gave a survey result, which showed that a majority of the
state officials expected the new rules to increase air emissions.268 According to the report, some state officials were
concerned that the revisions would rather complicate their
jobs and thereby increase their workloads, since weakened
NSR enforcement would give fewer options for states to
comply with NAAQS.269
D. The Stay of the Routine Maintenance Exception Rule
and the Uncertain Future for NSR Reform
On December 24, 2003, the D.C. Circuit ordered the stay of
the routine maintenance rule.270 However, the court once
again denied the motions for the stay of the December 2002
NSR rule on the grounds that “[p]etitioners [had] not demonstrated sufficient changed circumstances to justify revisiting the [May 6] order.”271 On July 1, 2004, EPA formally
issued an administrative stay,272 and announced that it
planned to reconsider three issues concerning the routine
maintenance rule: (1) whether the rule as originally finalized is allowable under the CAA; (2) the legal basis for selecting the 20% cost threshold; and (3) a simplified procedure for incorporating a FIP into SIPs to accommodate
changes to the NSR rules.273 Many observers expected the
result of the presidential election to decide the fate of the
NSR rules, and several observers have predicted that the
U.S. Supreme Court will ultimately hear the NSR debate
given splits in the various circuits.274 Because President
Bush was reelected last November, it is expected that EPA
will continue to push its NSR initiative, and that the Court
will ultimately resolve the ongoing controversy surrounding the new NSR rule in coming years.
E. The Legality of the New NSR Rule
The new NSR rule will not have much national impact until
states with an approved NSR program complete the SIP revisions and implement their revised NSR regulations upon
EPA’s approval. Currently, the new NSR rule has been implemented in 11 states that do not have an approved NSR
program and other some states with a delegated NSR program.275 Hence, it is somewhat too early to tell how the new
NSR rule will change the behavior of regulated industry in
any significant manner.
When reading its brief submitted recently to the D.C. Circuit, EPA’s legal position hinged primarily on two grounds.
First, relevant CAA provisions are ambiguous such that
EPA is entitled to Chevron deference for the new definition
of the statutory term “change.”276 Second, the environmental impacts of the new rule would be positive or zero, since it
would give regulated sources the incentive to engage in en270. New York v. EPA, 2003 U.S. App. LEXIS 26520 (D.C. Cir. Dec. 24,
2003).
271. Id.
261. 68 Fed. Reg. at 61248.
262. Id. at 61252 (“We have decided, for now, not to take final action on
the proposed annual maintenance, repair and replacement allowance approach.”).
263. U.S. GAO, New Source Review Revisions Could Affect Utility Enforcement Cases and Public Access to Emissions Data 15-21 (2003)
(GAO-04-58), available at http://www.gao.gov/new.items/d0458.
pdf (last visited Mar. 1, 2005).
264. Id. at 21-25.
265. Id. at 25-26.
266. 2 Studies Contradict EPA on New Rules; Changes to Boost Pollution, They Say, Associated Press, Oct. 23, 2003, at A2.
267. See id.
268. U.S. GAO, Key Stakeholders’ Views on Revisions to the New Source
Review Program 13-23 (2004) (GAO-04-274), available at
http://www.gao.gov/new.items/d04274.pdf (last visited Mar. 1,
2005).
269. Id. at 23-24.
272. U.S. EPA, Prevention of Significant Deterioration (PSD) and
Nonattainment New Source Review (NSR): Equipment Replacement Provision of the Routine Maintenance, Repair, and Replacement Exclusion; Stay, 69 Fed. Reg. 40274 (July 1, 2004) (codified at
40 C.F.R. §§51 and 52).
273. Id. at 40278.
274. Darren Samuelsohn, Clean Air: Court Schedule Indicates NSR
Rules’ Fate Rests on Next Presidential Administration, Greenwire, Feb. 26, 2004, Air, Water & Climate, Vol. 10, No. 9, available
at LEXIS, Nexis Library, Greenwire File.
275. Mastroyannis-Zaft, supra note 223, at 809.
276. See Brief for the United States and EPA at 69-72, New York v. EPA,
No 02-1387 (D.C. Cir. Dec. 31, 2002), available at http://www.epa.
gov/nsr/documents/respondentbrief.pdf (last visited Mar. 1, 2005).
EPA relied heavily on the Chevron and Alabama Power cases. With
respect to PALs, EPA argued that the Alabama Power case gives it
the authority to set a plantwide emissions cap. According to EPA,
deciding on the length of the contemporaneous period is also a “matter [ ] left to its discretion.” See id. at 93-95.
5-2005
NEWS & ANALYSIS
35 ELR 10337
Copyright © 2005 Environmental Law Institute®, Washington, DC. reprinted with permission from ELR®, http://www.eli.org, 1-800-433-5120.
vironmentally beneficial projects without the fear of being
subject to NSR.277 Adverse environmental impacts on existing air quality that could otherwise result from rule changes
would be reduced to a minimum by implementation of a
number of safeguards provided under the new rule.278
Throughout the brief, EPA stressed the fact that it has the
duty to balance the CAA’s clean air goal and economic
growth needs in implementing the Act’s NSR program.279
Put differently, the reason for revising the prior rule was motivated by the Agency’s belief that it deprived existing
sources of operational flexibility to meet increased market
demand or failed to give them incentives to invest in pollution control technologies or energy efficiency projects,
which would otherwise have produced air quality improvements. EPA made it clear that the primary goal of the new
NSR rule is not to unduly inhibit economic growth, by stating that “the purpose of the NSR provisions is not to compel
emissions reductions from existing sources, but to limit
emissions increases resulting from physical or operational
changes.”280 In other words, in EPA’s view, the new NSR
rule aims to return the previous state of affairs tilting toward
environmental considerations to the right balance between
air quality protection and economic development needs.
I would like to respond to EPA’s position by making two
points. First, even if it is conceded that EPA has the authority
to interpret the NSR provisions in order to balance the
NSR’s two equally important goals, the overriding goal of
the CAA is to improve air quality on a continual basis for the
benefit of the general public. The NSR program has functioned as one of the valuable tools for accomplishing that
noble goal. Throughout the Act’s history, Congress has continued to increase the Act’s stringency by adding new pro277. See id. at 65-66, 76-82. For example, in justifying its argument that
the Clean Unit exclusion would produce air quality benefits, EPA
cited the result of its review of a flexible permit pilot program, which
shows that five of the six participating facilities reduced their emissions beyond emissions limits contained in their permits. Id. at 109
and n.61.
278. EPA argued that only a small percentage of sources, 3% of total
emissions, might be able to take advantage of a higher baseline using
the new methodology used for calculating baseline emissions. See
id. at 78-79. EPA reiterated its statements in the preamble of the 2002
NSR rule. First, the demand growth exclusion is implementable, and
state minor NSR and Title V operating permit programs will be successful in incorporating into permits monitoring requirements adequate to prevent its abuse. See id. at 95-101. Second, in order to use
the Clean Unit exclusion, sources that have not gone through NSR
must comply with all the requirements of the NSR program. Furthermore, renewal is not an automatic one, and to requalify, an existing
clean unit must meet all the criteria as if it applied for the Clean Unit
exclusion for the first time. The public has a right to participate in the
process for designating a clean unit under the state minor NSR or the
Title V operating permit program, except for units that have recently
gone through NSR. At renewal time, an existing unit can requalify
only if it complies with the requirements applicable in the area reclassified as nonattainment or more severe nonattainment. See id. at
112-16. Third, to qualify for the PCP exclusion, a source may install
one of the listed pollution control technologies presumed to be environmentally beneficial, and in its application must provide detailed
information showing the environmental benefits of a control technology it plans to use. If the proposed project does not belong to one
of the presumptively beneficial PCPs, it must be approved by the reviewing authority in a state minor NSR or a Title V operating permitting process that provides for public notice-and-comment before the
applying source begins construction. A qualifying source has an ongoing legal duty to operate its PCP in a manner that continues to produce net environmental benefits, while minimizing emissions of collateral pollutants. See id. at 122-26.
279. Id. at 28, 75, 76, 95, 112.
280. Id. at 73-74, 75, 94-95 (emphasis in original).
gram requirements or by tightening preexisting standards
and requirements that it deemed necessary to move the nation toward meeting clean air goals. There is no doubt that
congressional intent in enacting the NSR program in 1977,
was to bring more and more sources into its coverage over
time. Congress did not anticipate that NSR would become
an end run game played by grandfathered sources, in particular those in the energy industry. As a result, contribution to
air quality improvements from industry has come mainly
from new sources or existing sources’ compliance with requirements under other programs under the CAA.
Moreover, EPA’s new definition of the term “change” is
against its common sense understanding. The primary purpose of NSR should be to improve existing air quality. Any
imaginative interpretation of the term “change” cannot go
too far so as to violate this simple mandate. It is therefore
clear that the enactment of the NSR program was the nation’s
choice to emphasize clean air goals rather than economic
growth. To paraphrase, the purpose of the NSR program is to
compel emission reductions from existing sources whenever
they propose a change that will increase their actual emissions in a way that adversely impacts existing air quality.
Thus, the first prong of EPA’s legal position is untenable.
Second, EPA’s argument that the new NSR rule would
lead to air quality improvements because, in its view, more
existing sources are expected to have incentives to invest in
clean energy technologies cannot withstand analysis. Experience with implementation of a variety of environmental
statutes tells us that industry responds to the economics
rather than act on its environmental awareness. More often
than not, it is clear market signals, usually in the form of unambiguous statutory or regulatory mandates, that have succeeded to motivate regulated sources to become cleaner or
more energy efficient. Under the new rule, existing sources
can avoid NSR more easily than in the past. It is hard to believe that they will have incentives to install state-of-the-art
pollution control technologies or inherently cleaner energy
technologies that they had little incentive to use under the
previous rule. The opposite will be more true. Given this,
PALs, the Clean Unit exclusion, and PCPs will likely be
underutilized. Otherwise, their frequent use will lead to
more and more existing sources escaping NSR for an extended period of time. If it is assumed that their use will be
subject to adequate public scrutiny, as EPA argues, industry
will find it cumbersome and hence not worthy of pursuit to
opt in to such mechanisms. Furthermore, EPA seems to neglect the fact that new sources, which have been the driving
force for technological innovation, will experience more of
an economic disadvantage under the new NSR rule. The notion of a level playing field or environmental comparability
hardly found its way into the new rule.
Conclusion
To summarize, the new NSR rule is another example of
EPA’s failure to reconcile two conflicting goals: environmental protection and economic development. While allegedly providing increased simplicity and flexibility to industry, the rules aggravate the problems with the preexisting
rules such as a bias against new sources and, most importantly, compromise the clean air goals of the NSR program.
It appears that the issuance of the new NSR rules is outside the scope of EPA’s delegated authority under the CAA.
35 ELR 10338
ENVIRONMENTAL LAW REPORTER
5-2005
Copyright © 2005 Environmental Law Institute®, Washington, DC. reprinted with permission from ELR®, http://www.eli.org, 1-800-433-5120.
The Act contains three simple but clear requirements in its
NSR program: (1) a proposed physical or operational
change that would increase emissions or result in collateral
emissions must go through NSR preconstruction review;
(2) emissions increases and decreases to be considered in
NSR applicability determinations must be contemporaneous; and (3) once NSR is triggered, the most stringent technology requirement, BACT or LAER, must be applied to the
source. In issuing the new NSR rules, EPA seemed to be so
preoccupied with one of the two NSR goals, i.e., easing the
regulatory burdens on industry, that it forgot the clear mandates of the CAA.
Although it has becomes less important in controlling
emissions of large coal-fired power plants, which has
incrementally been brought under the multi-pollutant trading approach, NSR is still a valuable tool that can be used as
a backstop to impose more stringent requirements on
grandfathered plants. Moreover, NSR enforcement represents one of the most significant options available for states
to utilize in compelling other source categories, such as refineries, smelters, wood products industries, to update existing pollution control. EPA should reconsider the December
2002 NSR rule. NSR reform should not result in a rollback
of preexisting rules. Rather, it should be aimed at improvements in the nation’s air quality and laying the foundation
for creating the clean energy path.
In pursuing these goals, it is advisable for the nation to listen to the following recommendations of the National Academy of Public Administration in its EPA-commissioned
report, which, among others, include: (1) repealing grandfathering; (2) retaining NSR for new sources; (3) the continuation of vigorous NSR enforcement; (4) the replacement of
NSR for existing source with a compulsory three-tier system (cap-and-trade for industrial sources with reliable monitoring records, cap-and-net, or unit-cap for other sources);
and (5) requiring all new and existing sources to regularly
report their emissions data to regulatory agencies and the
public in order to enhance accountability.281 To add to these
recommendations, netting also should not be allowed. In the
alternative, reporting requirements should be imposed on
those sources seeking credits for contemporaneous net
emissions decreases at their facility. Another alternative approach to NSR is to adopt output-based emissions standards
and to put more of an emphasis on energy efficiency and
conservation goals in NSR permitting processes, in order to
facilitate sustainable energy development.282
EPA has justified changing the preexisting NSR rule on
the basis of the findings of its 90-day NSR report to the president, which said that the old NSR rule impeded the energy
development and environmentally beneficial projects of existing power plants and refineries, while its adverse impacts
281. See NAPA NSR Report, supra note 64, at 133-37.
282. See Foote, supra note 17. In this article, the author, who is currently
EPA’s Assistant General Counsel in the Air and Radiation Division,
argued that the NSR program can be implemented to promote clean
energy technologies and energy conservation using a hierarchical
approach, under which permitting agencies apply: (1) conservation;
(2) renewable energy; (3) energy efficiency; and (4) add-on control
technologies in descending order, with some help from states’ integrated resource planning (IRP) and demand-side management
(DSM) policy. See id. at 10648, 10657-62. Notably, he argued that
CO2 emissions, which EPA has concluded is not an air pollutant,
should be considered in NSR permitting decisions. See id. at
10662-69.
on new sources were minimal. Even if it is assumed that the
findings were correct, EPA seems to overlook the fact that
its new NSR rule would have the effect of skewing the preexisting unlevel playing field toward old, grandfathered
sources more than in the past, however insignificant it might
be. In other words, EPA stands on the wrong foundation.
Because the status quo has been changed in a way that would
aggravate intersource and regional disparities in emission
reduction requirements, the new NSR rule may not deliver
its goal of economic efficiency unless it provides for additional mechanisms that would zero out all the advantages it
may give to existing sources, in order to maintain the preexisting status quo. The fundamental solution for resolving the inequities in NSR implementation should be to repeal grandfathering and to adopt uniform output-based
emission standards that are equally applicable to new and
existing sources.
When implementing environmental and energy law, sustainable development takes the form of environmental comparability. Environmental comparability generally refers to
a general policy approach that is designed to fully internalize the negative externalities of energy production and consumption.283 Its central strategy is to incorporate sustainability concerns into the current law under which all sources
can compete on an equal footing with one another solely on
the basis of environmental performance, regardless of their
age.284 Available policy tools to this end include, among
other things, the phaseout of grandfathering, the adoption of
output-based emissions standards, and providing subsidies
for accelerating the commercial deployment of cleaner,
more energy-efficient sources such as renewable energy.285
They essentially symbolize sticking to the principle of sustainable development in the energy law field.
Environmental regulation discriminates against new and
cleaner energy sources. In most instances, it imposes on new
sources more stringent emission reduction requirements on
a percentage, input basis, even though these sources oftentimes are much cleaner than old, grandfathered sources because of the inherent nature of fuels or combustion technologies actually used.286 “In 1996, coal plants had average
283. See generally David R. Wooley, Environmental Comparability, 12
Nat. Resources & Env’t 276 (1998).
284. Id. at 279.
285. For a discussion of various policy tools, see Bruce Biewald et al.,
supra note 76, at 46-55.
286. Technology-based standards take the form of “emission rate” standards, which are typically expressed in terms of the amount of emissions of a regulated pollutant based on heat input, such as pounds per
million Btu (lbs./mm Btu). More demanding emission rates are required of clean fuel-burning sources under the name of BACT. Establishing emission reduction requirements on a percentage, input
basis penalizes new clean fuel-burning sources in two respects. First,
clean fuel-burning sources usually employ highly energy-efficient
fuel combustion technologies. For example, the maximum thermal
efficiency of state-of-the-art combined-cycle, natural gas-fired
plants is nearly 60%, whereas the most energy-efficient coal-burning technology currently in dominant use has a thermal efficiency of
33% at best. See Steven Ferrey, The New Rules: A Guide to
Electric Market Regulation 4 (2000). Thus, input-based emission standards disregard energy efficiency aspects, producing the
practical effect of rewarding old, energy-inefficient energy sources,
which are primarily grandfathered coal-fired power plants. Second,
clean fuel-burning sources embodying energy-efficient technologies are subject to percentage reduction requirements. Though being
less polluting, more energy efficient, these sources must install expensive modern post-combustion control equipment whose efficiency gains are questionable in terms of effectiveness in pollution
control compared to incurred investment monies. See Swift, supra
5-2005
NEWS & ANALYSIS
35 ELR 10339
Copyright © 2005 Environmental Law Institute®, Washington, DC. reprinted with permission from ELR®, http://www.eli.org, 1-800-433-5120.
emission rates that were [thirty] or [fifty] times higher than
new gas units with low-NOx combustion and SCR controls.”287 Midwest electric utilities’ average NOx emissions
may be 10 times higher than those in the Northeast region
even on a per Mwh basis.288
Although implementation of both the CAA’s NOx acid
rain program289 and EPA’s NOx 1998 SIP call290 have had the
effect of addressing these disparities in emissions reduction
requirements to some extent, they fall short of achieving the
environmental comparability goal. The SIP call’s emissions
limit of 0.15 pounds per million British thermal units291 is
“still roughly [ten] to [twenty] times less stringent than typical NOx emission rates required of new natural gas combined-cycle units in nonattainment areas.”292
Moreover, both CAA’s SO2 allowance trading and the
EPA’s NOx Budget Trading (NBP) program293 designed to
implement its 1998 NOx SIP call each allocate SO2 and NOx
allowances based on historical operating data such as fuel
input.294 This allocation method penalizes new and cleaner
287.
288.
289.
290.
291.
292.
293.
294.
note 74, at 1539 (observing that new gas-fired plants subject to rigorous NSR control requirements would have to incur the cost of $2,500
to $10,000 per ton of NOx removed, while grandfathered, coalfired plants could reduce NOx emissions “at prices as low as $300
per ton”).
Tim Woolf & Bruce Biewald, Electricity Market Distortions Associated With Inconsistent Air Quality Regulations, Elec. J., Apr. 2000,
at 44.
Ellen Roy, The Uniform Generation Performance Standard: Connecting Electric Industry Restructuring and Air Quality Improvement, Elec. J., Jan./Feb. 1998, at 59.
42 U.S.C. §7651f. Section 407 subjects affected coal-fired units with
SO2 reduction requirements under other provisions of the acid rain
program to the emissions rate standards for NOx. These units must
achieve new emission rates set by the EPA Administrator, which
were to be implemented at the same pace as the SO2 emission limitations under Subchapter IV. See id.
Relying on its authority under CAA §110(k) to force states to amend
their SIPs upon a finding of “significant contribution” to another
state’s NAAQS attainment or maintenance, called a SIP call, EPA in
1998 made a NOx SIP call against 22 eastern states and the District of
Columbia. See id. §7410(k)(5); U.S. EPA, Finding of Significant
Contribution and Rulemaking for Certain States in the Ozone Transport Assessment Group Region for Purposes of Reducing Regional
Transport of Ozone; Final Rule, 63 Fed. Reg. 57356 (Oct. 27, 1998)
(codified at 40 C.F.R. §§51, 72, 75, and 96).
In its 1998 NOx SIP call, EPA made a finding that “highly cost-effective” controls on large electric-generating units included both combustion and post-combustion control techniques that could be used
to achieve an emission rate of 0.15 lbs./mm Btu based on the unit’s
1995-1996 average utilization rate. Id. at 57378. In fact, the chosen
emission rate standard is the NSPS for NOx emissions applicable to
coal-fired units that has existed since 1979. See 40 C.F.R. §60.44a.
Woolf & Biewald, supra note 287, at 47.
The NOx SIP call gave the states the flexibility to choose which
sources to target and mix of control measures needed to achieve the
required NOx emissions reductions. Its most significant feature,
however, was its requirement that the states allocate a budget for fossil fuel-fired electric steam-generating units, with the option to participate in an EPA-administered regional cap-and-trade program.
EPA’s NBP was launched on May 1, 2003, in eight northeastern
states and the District of Columbia acting under the Ozone Transport
Commission (OTC), thereby replacing the OTC’s NOx trading program with respect to these states. Beginning on May 31, 2004, 11
other states subject to the NOx SIP call joined the program. See U.S.
EPA, NOx Budget Trading Program: 2003 Progress and
Compliance Report (2004) (EPA 430-R-04-010), available at
http://www.epa.gov/airmarkets/cmprpt/nox03/noxreport03.pdf
(last visited Mar. 1, 2005).
Under CAA Subchapter IV, SO2 allowances were allocated using
historic fuel input and other operating data from 1985 through 1987.
See 42 U.S.C. §§7651c(a)(2). On the other hand, new facilities be-
sources in two ways. First, grandfathering of allowances
constitutes “scarcity rents” for large existing sources, because new sources have to purchase allowances in the trading market to begin operation. Second, input-based allowance allocation is another form of discrimination against
new and cleaner, more energy-efficient sources, because
these sources need more allowances under an input-based
allocation formula than under an output-based one.
Nine northeastern states and the District of Columbia under the Ozone Transport Commission (OTC) NOx cap-andtrade program295 or other states opting-in to EPA’s NBP
have not set aside allowances for new sources, or even if
they did, the number of set-aside allowances were “not large
enough to cover all the new power plants seeking to enter
the market.”296 These inequities can be addressed adequately by establishing a cap-and-trade program, which incorporate the following three features: (1) the adoption of an
output-based allowance allocation method; (2) periodical
updating of allowance allocations; and (3) the application of
the same emission rate to new and existing sources.297 The
way in which the method for making allowance allocations
on an output basis, called the Uniform Generation Performance Standard, can be implemented as follows. First, the
relevant authority calculates the amount of total allowable
emissions. Second, it establishes the same emission rate,
which is applicable to all new and existing sources on a per
kilowatt hour or Mwh basis, “by dividing the cap by the expected generation for that region over a set period of
time.”298 Additionally, an output-based cap-and-trade program can be designed to allow renewable energy sources to
directly participate in trading. In this way, more efficient
sources, regardless of their age and the chosen energy technology, can gain a competitive advantage. This in turn provides strong incentives for generation sources to become
cleaner, more energy-efficient.
ginning operation after December 31, 1995, must purchase allowances in EPA-administered auctions or from existing sources who
have allowances to sell in the secondary market. Id. §7651d(g)(3)(4). Under EPA’s NBP, a state’s baseline inventory for large electric-generating units is based on “the higher of the 1995 or 1996
ozone season heat input values.” 63 Fed. Reg. at 57407. Covered
large nonelectric-generating units, which are defined as nonutility
industrial boilers and turbines units with a capacity greater than
250 mm Btu per hour or with NOx emissions greater than one ton
per day, are required to achieve a 60% reduction of their preexisting NOx emissions. Id. at 57378, 57415. But note that participating states have the discretion to apportion allowances between
covered units.
295. Under CAA §§176A and 184, northeastern states comprising the
OTC, except Virginia, agreed to implement a regional NOx capand-trade program in 1994 and finalized a model rule for NOx allowance trading in 1996, which would be implemented by participating
states beginning in 1999. See 42 U.S.C. §§7506a, 7511c; Memorandum of Understanding Among the States of the Ozone Transport
Commission on the Development of a Regional Strategy Concerning the Control of Stationary Source Nitrogen Oxide Emissions
(Sept. 27, 1994); Laurel J. Carlson, Northeast States for Coordinated Air Use Management/Mid-Atlantic Regional
Air Management Ass’n NOx Budget Model Rule (1996),
available at http://www.epa.gov/airmarkets/otc/otcrule.zip (last
visited Mar. 1, 2005); OTC, NOx Budget Program: 1999-2002
Progress Report 5 (2003) (EPA 430-R-03-900), available at
http://www.epa.gov/airmarkets/otc/otcreport.pdf (last visited Mar.
1, 2005).
296. Woolf & Biewald, supra note 287, at 45-46.
297. Id. at 47.
298. Roy, supra note 288, at 57.
35 ELR 10340
ENVIRONMENTAL LAW REPORTER
5-2005
Copyright © 2005 Environmental Law Institute®, Washington, DC. reprinted with permission from ELR®, http://www.eli.org, 1-800-433-5120.
It is important to note, however, that implementation of
output-based allowance allocation or NSPS proves difficult
in practice. At one time, EPA pursued output-based NSPS
for NOx control applied to electric utility boilers built, modified, or reconstructed after July 9, 1997, but it withdrew the
proposal after the D.C. Circuit’s vacatur.299 Massachusetts,
one of the OTC states that had agreed to implement the OTC
NOx cap-and-trade program in 1994, adopted an outputbased allocation formula in November 1997. But it still set
different emission rate standards for four different categories of affected sources, and new sources were allocated allowances based on their permit limits.300 Worse, new
299. U.S. EPA, Revision of Standards of Performance for Nitrogen Oxide
Emissions From New Fossil Fuel-Fired Steam-Generating Units;
Revisions to Reporting Requirements for Standards of Performance
for New Fossil Fuel-Fired Steam-Generating Units, 63 Fed. Reg.
49442 (Sept. 16, 1998); Lignite Energy Council v. EPA, No.
98-1525, 1999 U.S. App. LEXIS 26263, *2, 30 ELR 20279 (D.C.
Cir. 1999).
300. Roy, supra note 288, at 62.
sources had to surrender up to 50% of their allowances if
they left over unused allowances.301
Climate change policy may be the most effective way of
promoting clean energy development, given that there are
currently no commercially available carbon capture and sequestration technologies. Thus, it has the effect of restricting fossil fuel usage. Because old, dirtier sources usually use
more carbon-intensive fuels, it rewards cleaner, more efficient energy sources, without relying on output-based standards. Put differently, well-designed climate change policy
has the potential to send price signals to energy producers
and consumers that adequately reflect environmental externalities of energy-related products and activities.
Most of these and other reform proposals require action
on the part of Congress. It will take time to gain political currency. The United States badly needs the leadership of both
chambers of Congress to clear the way for achieving the
goal of sustainable development.
301. Id.