  SEQ CHAPTER \h \r 1 Responses to Significant Comments on Notice of
Reconsideration for Phase 2 of the Final National Ambient Air Quality
Standard on the issues related to NOx RACT for EGUs in CAIR States.

71 FR 75902, December 19, 2006

  SEQ CHAPTER \h \r 1 

Docket Number EPA-HQ-OAR-2003-0079

U.S. Environmental Protection Agency

May 25, 2007

Response to Comments on Notice of Reconsideration for Phase 2 of the
Final National Ambient Air Quality Standard on the issues related to NOx
RACT for Electric Generating Units (EGUs) in CAIR States.

The proposal notice of reconsideration on which these comments were
received was published on December 19, 2006 (71 FR 75902).

I.  EPA received the following comments asserting that EPA’s proposed
CAIR-RACT determination would illegally substitute controls on sources
outside of ozone nonattainment areas for controls on sources within each
nonattainment area, and that reductions must occur within the
nonattainment area:

1.	EPA’s proposal to treat “region wide” CAIR reductions as
meeting RACT fails Section 172(c)(1)’s express requirement that SIPs
provide for “such reductions in emissions from existing sources in the
[nonattainment[ area as may be obtained through the adoption” of RACT.



2.	CAIR should not be equated with RACT because it does not assure that
emissions will occur in the nonattainment area as required by the Clean
Air Act.



3.	EPA cannot guarantee that CAIR emission reductions will occur in
nonattainment areas, but only somewhere between Texas and Massachusetts.



4.	CAA calls for SIPs to provide for "such reductions in emissions from
existing sources in the [nonattainment] area as may be obtained through
adoption" of RACT.



5.	EPA's proposed CAIR equals RACT determination would illegally
substitute controls on sources outside of ozone nonattainment areas for
controls on sources within each nonattainment area.



6.	Section 182 of the Clean Air Act requires application of RACT to
sources located "in the area".  



7.	The CAA mandates emissions reductions "in the area" that may be
obtained from application of "control technology" that is reasonably
available.



8.	A cap-and-trade program like CAIR which is designed to deal with
transport of a regional basis does not meet the CAA requirements for
RACT nor does it satisfy the goals underlying RACT because it does not
insure that sufficient reductions are gained from EGUs in nonattainment
areas.



9.	CAIR fails to satisfy the CAA requirement that RACT requires emission
reduction in the nonattainment area and not else where. Emissions
reduction must be from existing sources in the nonattainment area and
not other sources elsewhere.



10.	EPA conceded that CAIR "may affect…the spatial distribution of
emissions (by moving the around from one place to another.)"  While CAIR
may be effective region wide, it does not insure that controls will be
placed in a particular nonattainment area.



11.	CAIR will not necessarily provide assurances of ozone reductions in
specific nonattainment areas and those reductions are necessary. 
Therefore, commenter does not support a blanket finding that CAIR and
RACT are equivalent, since EGUs in nonattainment areas may have
significant NOx emissions if they are not meeting a minimum level of NOx
control.



12.	EPA cannot use a region-wide trading program like CAIR to satisfy
RACT because RACT reductions must be obtained in the nonattainment area.



13.	In equating CAIR with compliance with RACT, EPA would deem sources
to be in compliance with RACT even though they will not reduce emissions
in the nonattainment area.



14.	EPA's analysis continues to emphasis region-wide reduction rather
than focusing on reduction in a given nonattainment area.



15.	The CAA requires RACT to be applied in the nonattainment area.

16.	EPA continues to recognize that the spatial distribution of
emissions is affected by equating CAIR with RACT.



17.	By assuming CAIR equals RACT, EPA is removing any requirement that
an EGU located in a nonattainment reduce its air pollution, as the EGU
could buy allowances from outside the nonattainment area.



18.	The rule does not guarantee that any RACT level controls would
actually be installed in the “CAIR state” portion of the
nonattainment area.  This means that not only would EGUs located in the
non-CAIR state have to bear the cost of RACT controls, but the public
residing in the nonattainment areas would continue to suffer from the
emissions from those EGUs located in the CAIR state portion of the
nonattainment area that purchase and use allowances for compliance
instead of installing controls.   



19.	CAIR is a cap and trade program which cannot guarantee that a
reasonable level of control will be installed where most needed.



20.	Congress intended through CAA section 172(c) that a reasonable level
of pollution control technology needs to be installed on major sources
in all nonattainment areas (and Statewide within the Ozone Transport
Region for NOx).



Response:  EPA’s determination that compliance with the CAIR, in
certain circumstances, is sufficient to satisfy the NOx RACT requirement
for EGUs applies only in those areas where EPA’s December 2006
emissions analysis shows that the CAIR is projected to achieve greater
emissions reductions than application of source-by-source RACT within
the nonattainment area or OTR state.  However, EPA also believes it is
appropriate for the CAIR states, under certain conditions, to presume,
in general, that EGU NOx RACT requirements are satisfied through
implementation of the CAIR program. 

In addition, EPA reasonably concluded that compliance with the CAIR may
satisfy the NOx RACT requirement for EGUs applies in areas where EPA’s
December 2006 emissions analysis shows that the CAIR is projected to
achieve greater emissions reductions than application of
source-by-source RACT within the nonattainment area or OTR state.  EPA
does not believe that requiring source-specific RACT controls on EGUs in
nonattainment areas would reduce total NOx emissions from EGU sources
covered by the CAIR below the levels that would be achieved under the
CAIR alone.  EPA also believes that EGU source-specific RACT would
result in more costly emission reductions on a per ton basis. The
combination of EGU source specific RACT and the CAIR emissions cap would
not reduce the collective total emissions from EGUs covered by the CAIR,
but would likely achieve the same total emissions reductions as the CAIR
alone, in a more costly way.  

Further, EPA’s analysis for the CAIR shows the CAIR program will
result in EGUs installing emission controls on the maximum total
capacity on which it is feasible to install emission controls by the
2009 date.  (70 FR 22515-22225)  The CAIR regional budgets are based on
the level of emissions that can be achieved through the application of
highly cost-effective controls to EGUs in the CAIR region.  Due to
feasibility constraints, EPA required a phased approach for achieving
highly cost-effective emissions reduction.  For NOx, the first phase
starts in 2009 (covering 2009-2014); the second phase of NOx reductions
begins in 2015 (covering 2015 and thereafter). (70 FR 71621).  We also
noted in the June 2, 2003 proposed rule to implement the 8-hour ozone
NAAQS that we considered highly-cost effective controls for NOx for EGUs
and non-EGUs that were used to establish the NOx emission caps in the
NOx SIP call to constitute a greater level of control than RACT (68 FR
32839). 

	Moreover, EPA has predicted that the majority of large coal-fired
utilities will install advanced control technologies under the CAIR
because the larger and higher emitting sources offer opportunities to
obtain the most cost-effective emissions reductions.  EPA expects that
the largest-emitting sources will be the first to install NOx control
technology and that such control technology will gradually be installed
on progressively smaller-emitting sources until the ultimate emissions
cap is reached.

	Finally, EPA continues to believe that it is appropriate for all areas
covered by the CAIR to presume that compliance with the CAIR will
satisfy the NOx RACT requirements for EGUs if all required CAIR
emissions reductions in that state are achieved by EGUs only.  However,
this presumption is rebuttable and the State’s documentation of
reliance on this presumption must provide additional justification if
necessary. 

II.  EPA received the following comments asserting that compliance with
CAIR cannot satisfy RACT requirements because CAIR does not ensure that
each affected source in a non-attainment area will meet a specific
emissions limit. 

21.	EPA’s reconsideration notice provided no data or legal analysis
that would support its determination that compliance with CAIR should be
considered to satisfy RACT.



22.	What EPA failed to explain . . . is how sources in a nonattainment
area that purchase emission credits, rather than installing RACT, have
either complied with section 172(c)’s requirement that existing
sources “in the [nonattainment] area” adopt RACT, or contributed to
attainment in their own area.



23.	Each particular affected source in a non-attainment area is required
by the Clean Air Act to have the lowest emission limitation it is
capable of meeting.



24.	The Clean Air Act does not give EPA the option of requiring CAIR or
some other strategy in lieu of RACT.



25.	By deeming CAIR controls to be equivalent to RACT, EPA is seeking to
insulate uncontrolled or poorly controlled EGUs in current or future
nonattainment from cost effective controls that would qualify as RACT. 



26.	The commenter strongly disagrees with the EPA's proposal that EGU
sources complying with the CAIR would be automatically found to be
complying with NOx RACT requirements.



27.	Cap and trade programs like CAIR should be considered a complement
to, not a substitute for, RACT applicability.



28.	The public in nonattainment areas could suffer from emissions from
EGUs which purchase allowances instead of installing controls.



Response:  EPA disagrees with the commenters’ assertion that RACT
requirements in the Clean Air Act necessarily require every major source
in the specific geographic area to install controls.  See 70 FR 71656. 
To the contrary, EPA allows states to demonstrate that RACT is met by
groups of sources.  For example, the NOx Supplement to the General
Preamble, November 25, 1992 (57 FR 55625) permits states to “allow
individual owners/operators in the nonattainment area . . . to have
emission limits which result in greater or lesser emission limits so
long as the area wide average emission rates . . . are met on a
Btu-weighted average.”  The General Preamble also “encourage[s]
states to structure their RACT requirements to inherently incorporate an
emissions averaging concept (i.e., installing more stringent controls on
some units in exchange for lesser control on others).”  This approach
was based on EPA’s conclusion that it was permissible under the CAA
for individual sources to have “greater or lesser emission reductions
so long as the area wide average emissions rates” associated with a
RACT level of NOx emissions control were met.

  EPA also disagrees with the comment arguing that EPA is seeking to
insulate uncontrolled or poorly controlled EGUs in current or future
nonattainment areas from cost effective controls that would qualify as
RACT.  The final rule does not displace the RACT requirement for any
sources.  Instead, EPA is exercising its authority to interpret the
section 172, 182, and 184 RACT requirements for purposes of implementing
the 8-hour ozone standards.  For the reasons described in this section,
we believe that states can rely on EPA’s conclusion that compliance
with a CAIR FIP or SIP, meeting certain requirements, will satisfy the
EGU NOx RACT requirement in certain areas. 

Also, see the response to comment I above.

III.  EPA received the following comments asserting that EPA’s claim
that CAIR alone will achieve greater reductions than RACT is incorrect
and that EPA should have required both source-by-source RACT and CAIR:

29.	EPA's claim that CAIR alone will achieve greater reductions than
RACT is incorrect.



30.	The suggestion that CAIR will achieve greater reductions without
RACT is unsupportable.

31.	EPA can and must require RACT reduction on top of CAIR reductions.



32.	EPA's action ignores the possibility that requiring both RACT and
CAIR will produce faster RFP and earlier attainment than CAIR alone.



Response:  EPA’s technical analysis prepared for the December 2006
notice of proposed reconsideration generally shows that CAIR will
achieve greater emission reductions across the CAIR region and also in
most specific nonattainment areas and some OTR states than would be
achieved by requiring EGUs in these areas to meet a specific level of
control deemed to be RACT.  EPA also conducted a technical analysis such
that for each geographic area within the CAIR region where EPA expected
8-hour ozone RACT determinations would be required, EPA examined whether
the emissions reductions projected from the CAIR equal or exceed the
emissions reductions projected to occur from application of
source-by-source RACT.  Specifically, this analysis was conducted for
operating coal-, oil-, and gas-fired EGUs for each OTR state within the
CAIR region and for each nonattainment area in the CAIR region for which
a RACT SIP, separate from an attainment demonstration SIP, was expected
to be required.  The analysis was conducted on the basis of annual
emissions and also summer ozone-season emissions.  This analysis
illustrates that the CAIR achieves significantly greater overall
emissions reductions across the CAIR region and across the OTR than
would be achieved through the application of EGU source-by-source RACT
controls. In addition, the analysis showed that CAIR obtained equal or
greater summer ozone season emission reductions than source-by-source
RACT in 13 out of 18 specific nonattainment areas in the CAIR region and
in 3 out of 9 OTR states.  It also showed that CAIR obtained equal or
greater annual emission reductions than source-by-source RACT in 15 out
of 18 specific nonattainment areas in the CAIR region and in 6 out of 9
OTR states. The docket contains a Technical Support Document describing
the analysis.

	Contrary to one commenter’s suggestion, EPA did not claim that this
analysis shows that CAIR will achieve greater reductions without RACT. 
Instead and as explained below, the combination of EGU source-specific
RACT and the CAIR emissions cap would not reduce the collective total
emissions from EGUs covered by the CAIR, but would likely achieve the
same total emissions reductions as the CAIR alone, in a more costly way.

It is EPA’s belief that, due to the nature of regional emissions
transport, local nonattainment area emissions reductions alone will not
achieve the most effective or economically efficient impact on ozone air
quality in nonattainment areas.  EPA believes a combination of local and
broader regional reductions, such as those driven by the CAIR
requirements for EGUs, will achieve a more effective and economically
efficient air quality improvement in nonattainment areas and OTR states
than application of source-by-source RACT.

EPA also disagrees with the commenter’s assertion that EPA can and
must require source-by-source RACT reductions on top of CAIR reductions.
 As previously explained, requiring source-by-source RACT in certain
areas would mean that controls would not necessarily be placed on the
sources for which it is most cost-effective to control.  The result
would be the same emission reductions area wide, but at higher cost. 
Further, the CAIR final rulemaking established a region-wide NOx
emission cap, effective in 2009, at a level that, assuming the
reductions are achieved from EGUs only, would result in EGUs installing
emission controls on the maximum total capacity on which it is feasible
to install emission controls by that date. Thus, by design, the 2009
CAIR region-wide NOx emissions cap for EGUs represents the most
reductions that are reasonable to achieve.  Consequently, EPA does not
believe that further controls could be considered reasonably available.

Finally, as we have also previously noted, states have discretion to
define RACT to require greater emission reductions than specified in
EPA’s guidance and also to require beyond-RACT NOx reductions from any
source (including sources covered by the CAIR or NOx SIP Call programs),
and have an obligation to demonstrate attainment of the 8-hour ozone
standard as expeditiously as practicable.  In certain areas, states may
require NOx controls based on more advanced control technologies as
necessary to provide for timely attainment of the ozone standards.

IV.  EPA received the following comments asserting that States have
adopted RACT requirements for ozone precursors with costs per ton well
in excess of the controls under CAIR and that advanced controls like
selective catalytic reduction (SCR) and selective non-catalytic
reduction (SNCR) are reasonable and available:

33.	States have adopted RACT requirement for ozone precursors with costs
per ton well in excess of the controls under CAIR. 



34.	In connection with the adoption of the 1997 ozone and PM NAAQS, the
President issued a memorandum indicating EPA's agreement that control
costs of up to $10,000 per ton were within the reasonable range.  



35.	The evidence shows that standards for RACT-level control are more
stringent than provided by CAIR.  . . .  For example, well over a decade
ago, New Jersey adopted NOx RACT requirements with costs as high as
$4500 per ton . . . .In other instances, RACT reflects controls costing
$8000 to $10,000 per ton.



36.	Pointing to 12 and 14 year old agency guidance documents, EPA
suggested that it is appropriate for states to consider in their NOx
RACT determinations technologies that fall within the guidance cost
range of $160-1300 per ton of NOx removed.



37.	EPA's RACT determination for EGUs is 14 years old and does not
reflect the current state of the art.  For example New Hampshire has
required EGUs to install SCR and SNCR which give a level of control
beyond what EPA assumed 14 years ago.

38.	NOx RACT should now be post-combustion controls such as SNCR or SCR.
 Both are reasonable and available.



39.	EGU sector can make reasonably cost effective emission reduction up
to the $4500/ton threshold.  OTR states are implementing controls
costing the level of investment in many other sectors.  





Response:  EPA believes the assumptions in its technical analysis
regarding the controls that would be considered RACT (if RACT were to be
applied on a source-by-source basis) are correct.  EPA’s views on NOx
RACT were set forth in the “NOx Supplement to the General Preamble,”
November 25, 1992 (57 FR 55620).  In that document, EPA determined that
in the majority of cases, RACT will result in an overall level of
control equivalent to specified maximum allowable emission rates (in
pounds of NOx per million Btu) for certain specified electric utility
boilers.  Section 4.6 of the NOx Supplement to the General Preamble (57
FR 55625) noted in part, “In general, EPA considers RACT for utilities
to be the most effective level of combustion modification reasonably
available to an individual unit.  This implies low NOx burners, in some
cases with overfire air and in other instances without overfire air;
flue gas recirculation; and conceivably some situations with no control
at all.”  The assumptions in EPA’s technical analysis are consistent
with this guidance.

	In making a general determination of what controls are representative
of RACT, EPA does not necessarily recommend the level of stringency
which is imposed by the most stringent State.  However, EPA does not
restrict States from imposing costs which the states determine are
necessary to meet the ozone NAAQS.  Some states are willing to impose
higher cost controls than others.  However, if all states chose to
impose stricter controls by 2009 there could be shortages of labor and
materials.  EPA’s analysis shows that CAIR produces the maximum level
of control that is feasible by 2009 on a region wide basis.

V.  EPA received the following comments regarding whether the 1992 and
1994 agency guidance documents to which EPA refers are outdated and not
consistent with RACT controls being imposed by States today:

40.	EPA's NOx Supplement (November 25, 1992) says the RACT for utilities
is the most effective level of combustion modification reasonably
available to an individual unit.  However the NOx Supplement encourages
States to allow emissions averaging at EGUs to meet RACT such that
individual sources in nonattainment areas are permitted to have emission
limits with greater or lesser emission reductions as long as area-wide
emissions rates are met.  The CAIR cap and trade is consistent with the
NOx Supplement.



41.	The 1992 and 1994 agency guidance to which EPA refers are outdated
and not consistent with RACT controls being imposed by States today.



42.	1992 EPA guidance made clear the EPA viewed determination of RACT to
require a unit-by-unit inquiry.



43.	Many new controls are under study that can reduce NOx emissions at a
fraction of the cost assumed in the CAIR rulemaking process, and are
well within the RACT level of control.



44.	One commenter noted that there are many new controls being studied
that can reduce NOx emissions at a fraction of the cost assumed in the
CAIR rulemaking.  These new controls, which would fall under RACT, are a
refinement of existing combustion control technologies, along with
injection of an inexpensive reagent in the boiler. 



45.	EPA makes certain assumptions regarding what control level is
considered RACT based on EPA guidance that is several years old.  States
have been increasingly more stringent in their RACT requirements and
have imposed higher costs that reflected in the CAIR rule with its
estimated $900/ton control cost.  If EPA’s analysis does not truly
reflect RACT, then the analysis could be wrong.



Response:  For EPA’s technical analysis, EPA assumed that RACT is
represented by combustion controls for EGUs defined as: 1) low NOx
burners with overfire air for wall-fired units; and 2) low NOx
coal-and-air nozzles with close–coupled and separated overfire air for
tangentially-fired units.  For oil and gas steam EGUs, the RACT-level of
control was assumed to be 0.20 pounds of NOx per million BTU for
tangentially-fired gas or oil burning and 0.30 for wall-fired gas or oil
burning.  This level of control is consistent with EPA’s past NOx RACT
guidance [see “NOx Supplement to the General Preamble,” November 25,
1992 (57 FR 55620)].  EPA still considers the combustion modification
guidance from the early 1990’s to express what constitutes NOx RACT
for EGUs considering technical feasibility and cost.  As EPA’s CAIR
technical analysis has shown, and as previously noted, CAIR requires the
installation of NOx controls on the maximum capacity on which it is
feasible to install such controls by 2009.  Therefore, additional
controls are not “reasonably available.”  In addition, EPA does not
restrict individual states from going beyond the level of control
expressed as RACT in EPA guidance documents if states want to impose a
higher level of control and are willing to impose extra costs on the
industry in their state in order to achieve compliance with the ozone
NAAQS.  Thus, EPA believes higher levels of control (represented by SCR
and SNCR) could be considered as beyond RACT, although states would be
free to impose these levels of control if needed to reach attainment. 
The fact that some states chose to go beyond RACT in order the meet air
quality standards within their state does not necessarily mean that this
level of control should be considered RACT for all EGUs.  In cases where
states are relying on the presumption that compliance with CAIR
satisfies certain NOx RACT requirements and additional information is
presented during the public comment period, states will necessarily
consider the additional data in reviewing what control obligation is
consistent with RACT.

VI.  EPA received the following comments saying that compliance with
CAIR cannot be said to satisfy RACT requirements because EPA’s
technical analysis shows that at least some nonattainment areas would
achieve greater emission reduction with implementation of source-by
source-RACT alone that with CAIR alone, and the following comment
agreeing with EPA that this analysis demonstrates that CAIR achieves
greater reductions across the region:

46.	EPA's technical analysis determined that at least some nonattainment
areas would achieve greater annual and seasonal emissions with
implementation of RACT without CAIR.  In these areas CAIR implementation
would not be "equivalent" to RACT for EGUs.



47.	EPA's technical support analysis clearly demonstrates that the
implementation of CAIR achieves greater overall emissions reductions
across the CAIR region than would be achieved through the implementation
of source-specific RACT controls.



48.	EPA's technical analysis, which looked at implementation of CAIR as
compared with implementation of RACT without CAIR,  shows that at least
some nonattainment areas would achieve greater annual emission
reductions with implementation of source-by-source RACT alone, rather
than with CAIR alone.



Response:  In this final action we are determining that compliance with
CAIR satisfies NOx RACT requirements for EGUs only where the technical
analysis conducted for the December 2006 proposal, shows that CAIR is
projected to achieve greater emissions reductions in the specific
nonattainment area or OTR state than application of source-by-source
RACT.  As explained above, other areas may still rely on the presumption
that compliance with CAIR rules satisfies NOx RACT requirements in
certain circumstances.  This presumption is rebuttable and the thus
State may choose to provide supporting analyses and will have to respond
to any comments received during the comment period challenging the
presumption.

VII.  EPA received a comment which argues that EPA’s Technical Support
Document for Phase 2 of the Final Rule to Implement the 8-Hour Ozone
NAAQS – Notice for Reconsideration (‘analysis’) is inadequate to
base the rule on since it emphasizes  region-wide reductions rather than
reductions in a given nonattainment area:

49.	The commenter argues that EPA’s analysis could result in EPA
“forfeiting” 3,000 tons per year of NOx reductions in the summer
ozone season in nonattainment areas.  On the other hand, other
commenters emphasize that CAIR achieves greater overall emissions
reductions across the CAIR region than would be achieved through the
implementation of source-specific RACT controls.



Response:  The supplemental technical analysis that EPA performed for
this reconsideration is explained in the December 2006 reconsideration
notice at 70 FR 75909.  Even though this analysis contained several
conservative assumptions, the analysis still showed that for both the
OTR and the nonattainment areas that EPA believed would be required to
submit RACT SIPs, CAIR emissions reductions, taken as a group, are
overall significantly greater reductions than reductions obtained from
source-by-source RACT for EGUs.  This analysis also showed that most
nonattainment areas achieved lower emission or ozone levels with CAIR
alone versus RACT alone, in some cases significantly lower.  For
example, for reduction of summer time ozone season emissions in 18
specific non-attainment areas in the CAIR region, CAIR achieved lower
emissions in 15 out of the 18 areas.   Taken as a group, the CAIR
reductions for these nonattainment areas was almost three times that of
source-specific RACT alone.  In the areas where the conservative
analysis did not show this to be true, the analysis showed only slightly
larger emission reductions in the specific geographic area from the
application of RACT on each source than from the area wide CAIR
emissions reductions.  In addition, as noted in the final Phase 2 Rule,
a state has discretion to define RACT to require greater emission
reductions than specified in EPA’s guidance and also to require
beyond-RACT NOx reductions from any source (including sources covered by
the CAIR or NOx SIP Call programs), and has an obligation to demonstrate
attainment of the 8-hour ozone standard as expeditiously as practicable.
 In certain areas, states may require NOx controls based on more
advanced control technologies as necessary to provide for attainment of
the ozone standards.

	Further, EPA disagrees with the comment to the extent it suggests that
this analysis provides inadequate support for the rule.  This analysis
showed that for both the OTR and the nonattainment areas that EPA
concluded would be required to submit RACT SIPs, CAIR emissions
reductions, taken as a whole, are overall significantly greater than
reductions obtained from source-by-source RACT for EGUs. Further, EPA
acknowledges that the conservative analysis did not show that CAIR
emissions reductions are greater than reductions obtained from
source-by-source RACT for all such nonattainment areas and OTR states. 
As explained above, the determination that compliance with CAIR
satisfies RACT requirements for EGUs does not apply unless the technical
analysis shows that annual and ozone-season CAIR emissions reductions
are greater than reductions obtained from source-by-source RACT in the
relevant nonattainment area or OTR state.  For areas not subject to this
determination, states may rely on the presumption that compliance with
CAIR rules satisfies certain NOx RACT requirements.  This presumption is
rebuttable and the thus State may choose to provide supporting analyses
and will have to respond to any comments received during the comment
period challenging the presumption. 

EPA also disagrees with both the analysis and the conclusions of the
commenter who argued that EPA would “forfeit” 3000 tons per year of
NOx Reductions in the summer ozone season in nonattainment areas.  The
commenter appears to have derived the 3000 ton number by adding numbers
in the technical analysis from the tables for OTR states and from the
table for nonattainment areas, some of which are within those OTR
states.  This double counts NOx reductions in nonattainment areas that
are also in OTR states and thus overstates the total NOx reductions
projected to occur from the application source-by-source RACT.  Further,
the commenter appears to have only summed the differences between CAIR
reductions and the reductions from source-by-source RACT for some of the
nonattainment areas and some of the OTR states in the analysis.  Summing
the differences for the whole OTR, not just selected states, shows that
CAIR provides overwhelmingly better emissions control than source
specific RACT (20,410 tons reduction for CAIR versus 2,440 tons
reduction for RACT for the whole OTR).  EPA believes that this latter
calculation is the most useful way to look at this data.  In addition,
the sum of ozone season emission reductions for all the ozone
nonattainment areas in the CAIR region as a group (some of which are
outside the OTR) indicates that the total reductions on an area wide
basis achieved by CAIR over RACT are even greater (94,140 tons of ozone
season emission reductions for CAIR versus 5,980 tons for RACT).  

VIII.  EPA received the following comments asserting that an effective
attainment strategy requires both area wide programs like CAIR and
nonattainment area specific programs such as source-by-source RACT on
EGUs: 

50.	EPA's technical support document is not a fair comparison because we
do not believe RACT should be adopted in lieu of CAIR.  The comparison
should be between the effectiveness of CAIR alone and the effectiveness
of CAIR plus RACT.

51.	An effective attainment strategy requires both regional rules like
CAIR that are designed to deal with transported pollution and
requirements to reduce emissions locally, such as RACT on EGUs.



52.	EPA should adopt the Ozone Transport Commission's (OTC's) approach
to cap-and-trade programs in which phase one of the NOx reduction
program was based on application of RACT, and phases two and three
involved progressively more stringent caps under a cap-and-trade
framework.



53.	EPA should adopt the OTC’s approach to cap-and-trade programs
where it is assumed that RACT was applied first. 



Response:  The supplemental technical analysis prepared by EPA for the
December 2006 reconsideration proposal was designed to analyze whether
compliance with a SIP or FIP meeting the requirements of CAIR may
satisfy the RACT requirement for certain EGUs.  Thus, it was appropriate
for EPA to compare the reductions under CAIR alone with the reductions
that would be achieved by another possible method of satisfying RACT
requirements (i.e. the application of source-by-source RACT controls). 
The comparison that the commenter suggests should have been prepared
would not have shed light on the question the analysis sought to answer,
namely whether compliance with CAIR satisfies the nonattainment program
requirement in question.  In addition, as noted above, by design, the
2009 CAIR region-wide NOx emissions cap for EGUs represents the most
reductions that are reasonable to achieve.   Further, as explained in
the reconsideration notice, source-specific control requirements layered
on top of the overall allowance-based emissions cap might affect the
temporal distribution of emissions or the spatial distribution of
emissions but would not affect total allowed emission in the CAIR
region.  EPA expects that, under the CAIR trading programs the
largest-emitting EGU sources (and those with the most cost effective
reductions available) will be the first to install NOx control
technology.  If states were to require smaller-emitting EGU sources in
nonattainment areas to meet source-specific RACT requirements, they
would likely use labor and other resources that would otherwise be used
for emission controls on larger sources and the cost of achieving the
regional reductions would be greater on a per ton basis.

IX.  EPA received a comment  that EPA’s determination that compliance
with CAIR rules, in some circumstances, satisfies NOx RACT requirements
for EGUs will create inequality between CAIR states and bordering
non-CAIR states.

54.	EPA’s determination creates an inequity where the geographic
boundary of a nonattainment area crosses state lines from a CAIR state
into a non-CAIR state.  In the CAIR state portion of the non-attainment
area, EPA would allow compliance with CAIR rules to satisfy NOx RACT for
EGUs while in the non-CAIR state portion of the nonattainment area NOx
RACT for EGUs would still be a source-specific requirement.



Response:  All nonattainment areas are subject to the RACT requirements
in CAA section 172(c)(1).  However, each air pollution control agency
determines for each source or source-category in the state, the specific
controls that constitute RACT.  Thus, the precise requirements applied
to ensure that the RACT requirement is satisfied may differ from source
to source and state to state.  Since sources in non-CAIR states are not
subject to rules implementing the CAIR emission reduction requirements,
those states naturally could not rely on compliance with those rules to
show that RACT has been satisfied.  The fact that the non-CAIR states
must use a different method to show that the same RACT requirement has
been met, however, does not show an inequity between states.  Further,
none of the nonattainment areas covered by the determination that
compliance with CAIR rules satisfies certain NOx RACT requirements (i.e.
those for which our technical analysis shows that CAIR provides equal or
greater annual and ozone-season emissions reductions than
source-by-source RACT) lie across the boundary of two states, one of
which is a CAIR state and the other of which is a non-CAIR state.

X.  EPA received the following comments indicating that it is improper
for EPA to determine that compliance with the CAIR may satisfy certain
NOx RACT requirements because the purposes of the CAIR and the RACT
requirements are different:

55.	The purpose of CAIR was to address the effects of emissions from
upwind States on the ability of downwind States to reach attainment of
the NAAQS.



56.	CAIR was intended to address emissions from EGUs that were found to
contribute significantly to nonattainment in downwind states, while RACT
is intended to reduce emissions within a nonattainment area.



57.	The purpose of RACT is not the same as the purpose served by CAIR. 
CAIR is designed to reduce transport across state lines.



58.	EPA has stated that the phase 2 8-hour ozone implementation rule is
not intended to address interstate transport, which is the purpose of
CAIR.



59.	EPA has previously indicated that CAIR is designed to address the
transport of pollutants, not nonattainment.



60.	CAIR is not intended as an attainment strategy.



61.	CAIR, a cap-and-trade program, is not a “control technology” in
any sense as intended by section 172(c).





Response:  We find the attempt by commenters to characterize CAIR as a
strategy to address only regional pollution transport as overly
simplistic.  The EPA analyses for CAIR show that there are significant
air quality benefits projected for individual nonattainment areas as a
result of NOx reductions across the multi-state CAIR region.  The Clean
Air Act does not prevent states from properly crediting measures that
achieve multiple objectives (e.g. regional transport and local
nonattainment).  Moreover, CAA section 110(a)(2)(D) requires SIPs to
contain adequate provisions to assure that sources in the state do not
contribute significantly to nonattainment in any other state.  The CAIR
rule is an integral element in meeting the states’ section 110
attainment obligations.  Accordingly, it is reasonable to incorporate
this consideration in determining what measures qualify as RACT.  Even
though the CAIR may have been initially designed to get regional
reductions, if it produces the most reductions that are feasible, it can
also represent RACT for subject areas.

XI.  A comment, which was offered in support of the proposition that
CAIR served a different purpose than the RACT requirement, says the
agency’s modeling for CAIR did not include or account for in-state
emissions.

62.	The Agency’s modeling for CAIR looked at the impacts of emissions
from one state affecting another state, and not the impact of emissions
within a state on that state’s nonattainment area or areas.





Response:  The comment seems to imply that certain in-state emissions
were not included in the CAIR modeling analysis.  This is untrue.  In
developing CAIR, EPA used modeling to determine which states
significantly contributed to downwind states’ nonattainment problems
(to address the requirement of CAA section 110(a)(2)(D)).  This
“significant contribution” air-quality modeling did not separately
assess the impact of in-state emissions on nonattainment areas, but
contrary to petitioner’s assertion, the modeling did not exclude
in-state emissions.  Moreover, EPA did not rely on this “significant
contribution” modeling in our analysis of whether the CAIR would
provide similar ozone reductions as the application of source-by-source
RACT.  Thus, commenters reliance on the fact that the amount of
contribution from in-state emissions was not a criteria for determining
whether or not the contribution from an upwind state was significant, is
misplaced.

Further, EPA also conducted modeling to project future concentrations
throughout the modeling domain.  This modeling focused on identifying
the air-quality impacts of the CAIR emission reductions and thus did not
distinguish between in-state and out-of-state emissions. The
“Technical Support Document for the Final Clean Air Interstate Rule
– Air Quality Modeling” (EPA-OAR-2003-0053-2123), cited in the
December 2006 federal register notice, presented the results of this
modeling of the expected air quality impacts of CAIR emission
reductions, as well as the aforementioned modeling to assess interstate
contributions.    

XII.  EPA received the following comments saying that to allow
compliance with CAIR to satisfy NOx RACT requirements in certain
circumstances would be to fail to impose RACT on EGUs in nonattainment
areas, and that this might force states to impose more costly controls
on non-EGU sources in order to reach attainment:

63.	EPA ignores the impact on other non-EGU sources within nonattainment
areas which may be required to impose more costly controls to make up
for lost reductions due to the failure to impose RACT on EGUs.



64.	If EGUs fail to install RACT controls, Illinois may have to require
non-EGUs to install additional controls which may be much more
expensive.



65.	EGUs are usually the largest and most easily controlled NOx sources
in a non-attainment area and are usually the most cost effective to
control.  The CAIR equals RACT determination may result in adoption of
draconian controls at smaller (non-EGU) emission sources that will have
harsh local economic consequences in a community.



66.	EPA ignores that by failing to impose RACT on EGUs in nonattainment
areas, States may be forced to impose more costly controls on non-EGU
sources to reach attainment in a nonattainment area.



67.	Since States may require NOx controls based on more advanced control
technologies to provide for attainment of the ozone standards, there is
no basis for NRDCs arguments that the Phase 2 rule improperly forces
states to impose controls on any particular sources or source categories
to achieve attainment.



Response:  EPA disagrees with the commenters’ assertion that EPA’s
determination that compliance with CAIR may satisfy NOx RACT
requirements for EGUs constitutes “failure to impose RACT on EGUs.” 
 Nothing in the final rule displaces the RACT requirement for EGUs.
Further, CAIR will achieve widespread SO2 and NOx emission reductions
from EGUs and will provide significant air quality benefits for ozone
and PM2.5 nonattainment areas.  In developing attainment SIPs and
identifying Reasonably Available Control Measures (RACM), states will
need to consider additional cost-effective and reasonable controls on
all sources to reach attainment as expeditiously as practicable.  States
must also consider the economic feasibility of implementing a given
control measure, and EPA has determined that CAIR will result in EGUs
installing controls on the maximum total capacity on which it’s
feasible to do so by 2009.  Further, as EPA has noted before, a state
has discretion to define RACT to require greater emission reductions
than specified in EPA’s guidance and also to require beyond-RACT NOx
reductions from any source (including sources covered by the CAIR or NOx
SIP Call programs), and has an obligation to demonstrate attainment of
the 8-hour ozone standard as expeditiously as practicable.  In certain
areas, states may require NOx controls on EGUs based on more advanced
control technologies as necessary to provide for attainment of the ozone
standards.  In addition, EPA acknowledges that in order to achieve
attainment as expeditiously as practicable, some states may need to
adopt control measures for some sources which cost more per ton than the
controls on EGUs, but which are still considered to be reasonable and
cost-effective.  Because of facility-specific factors (e.g. costs in the
geographic area and the facility’s ability to sustain the cost), EPA
does not believe it would be inappropriate to establish a threshold of
control effectiveness (e.g. dollars per ton) based on control of EGUs
and apply this threshold to all source categories.

XII.  A commenter argues that states such as Illinois may be forced to
require additional emission reductions, including application of RACT
controls within their nonattainment areas, that must be achieved earlier
than CAIR reductions.

68.	Additional controls on non-EGU sources will be very costly and  EGUs
are usually the largest and most easily controlled NOx sources in a
nonattainment area.  More specifically, there are 15 coal-fired boilers
in two ozone nonattainment areas in Illinois, none of which have
installed SCRs.  EPA projects that only two of those units will install
SCRs in response to CAIR.  However, based on that projection, the
Chicago area will not meet the 8-hour standard by 2010.



Response:  This comment suggests that a RACT determination for a source
category could be considered invalid just because it results in
relatively less control on one source category compared to another.  EPA
disagrees since the two categories may be sufficiently dissimilar so as
to render a comparison irrelevant.  RACT represents only such technology
as is reasonably available, not all controls that may be necessary to
attain as expeditiously as practicable.  The State is required to
demonstrate attainment as expeditiously as practicable and has the
discretion to choose in its public process how to apportion
responsibility for emission reductions to meet that requirement.

XIII.  EPA received the following comments concurring with its
determination that CAIR will likely achieve the same emission reductions
as RACT:

69.	The commenter concurs with EPA's determination that CAIR reductions
meet the 8-hour ozone SIP RACT requirements.

70.	The Agency determination is reasonable because CAIR will likely
achieve the same emissions reductions as source specific RACT in a more
cost efficient way.

  

71.	CAIR is based on the most efficient methods of controls reasonably
available to EGUs and requires that collective emissions meet the
statewide cap.



72.	Commenter supports EPA's determination that CAIR is equivalent to
RACT for NOx.



73.	EPA's technical analysis shows that CAIR achieves substantially
greater overall emissions reductions across the CAIR region than would
be achieved through application of EGU source-by-source RACT controls.



74.	EPA's technical analysis shows that CAIR achieves greater overall
emissions reductions - in some cases much more - through CAIR than
through application of EGU source-by-source RACT controls.





Response:  EPA agrees that CAIR will achieve the same or lower emissions
over the CAIR region than source-by-source RACT controls and that it
will achieve these emission reductions in the most cost effective
manner. 

XIV.  EPA received the following comments addressing the contention in
EPA’s analysis that CAIR will result in EGUs installing controls on
the maximum total capacity on which it is feasible to do so by 2009:

75.	CAIR will result in EGUs installing controls on the maximum total
capacity on which it is feasible to do so by 2009.  Further controls
will be installed by 2015.



76.	EPA's contention that CAIR 2010 reductions are as fast as achievable
because of an insufficient supply of boilermaker's labor to install
additional emission controls by 2010 is contradicted by a 2004 analysis
conducted by Institute of Clean Air Companies (ICAC) which concluded
that labor is available to install 2015 CAIR levels of reductions by
2010.  Therefore, EPA's implication that RACT requirements on EGUs in
the CAIR regions would not achieve more reductions than those achieved
by CAIR by 2010 is incorrect.



Response:  EPA considered a number of analyses related to boilermaker
labor availability provided by various commenters, including the 2004
Institute of Clean Air Companies analysis, when it prepared the Clean
Air Interstate Rule (CAIR) which was published May 12, 2005 (70 FR
25162).  EPA prepared its own technical analysis as part of the CAIR
development, and decided as a result of its analysis that the dates in
the final CAIR of January 1, 2009 for phase I for NOx controls, January
1, 2010 for phase II SO2 controls and 2015 for phase 2 controls for both
NOx and SO2 were appropriate based on projected labor availability.  EPA
believes it has set the 2009 CAIR NOx cap at a level that, assuming all
the reductions are achieved from EGUs, would result in EGUs installing
emission controls on the maximum total capacity on which it is feasible
to install emission controls by that date.  The EPA analysis, titled
“Boilermaker Labor Analysis and Installation Timing”, March 2005,
has been placed in the docket for the CAIR, docket number
EPA-HQ-OAR-2003-0053, document number EPA-HQ-OAR-2003-0053-2092.  This
issue is also discussed in the preamble to the CAIR under the heading
“Schedule for Implementing SO2 and NOx Emissions Reductions
Requirements for PM2.5 and Ozone” starting at 70 FR 25215.  EPA
concluded that its analysis rather than the ICAC analysis of feasibility
is correct.  The bottom line is that EPA’s analysis shows that the
amount of additional NOx emissions control that will be obtained under
the CAIR in 2015 is infeasible to obtain in 2009, when RACT emission
reductions under the 8-hour ozone NAAQS must be implemented.  Indeed,
EPA has determined that CAIR requires the installation of NOx controls
on the maximum capacity on which it is feasible to install controls by
2009.  Thus, in that time frame, controls beyond CAIR cannot be
considered “reasonably available.”

XV.  EPA received the following comments arguing that compliance with
CAIR cannot be said to satisfy NOx RACT requirements because the
economic test for CAIR is different from the economic test for RACT:

77.	The economic test for CAIR controls is different from RACT.  CAIR
controls need to be "highly cost effective" while RACT requires that
they be "economically feasible."  



78.	The economic test for CAIR controls is different from RACT.  CAIR
controls need only be "highly cost effective" while RACT requires that
they be "economically feasible."  Thus more controls are feasible under
RACT.



79.	While CAIR uses "highly cost effective" as its metric, RACT requires
an economically feasible level of control.  Thus more controls pass the
economic test under RACT than under CAIR.





Response:  EPA believes that the emission reductions achieved by CAIR,
while still highly cost effective, also represent the level of control
that is economically and technologically feasible as RACT for EGUs in
states that achieve all their emission reductions from EGUs.  The CAIR
final rulemaking established a region-wide NOx emissions cap, effective
in 2009, at a level that, assuming the reductions are achieved from
EGUs, would result in EGUs installing emission controls on the maximum
total capacity on which it is feasible to install emission controls by
that date.  Further, EPA does not believe that requiring source-specific
RACT controls on EGUs in nonattainment areas would reduce total NOx
emissions from EGU sources covered by the CAIR below the levels that
would be achieved under the CAIR alone.  The most expensive controls
available, which might be chosen for BACT or LAER, are not always
justified as RACT.  States have the flexibility to require such controls
as RACT in their discretion or to adopt higher levels of control as part
of their attainment strategy if they find such controls are reasonable
and necessary to achieve attainment of the ozone NAAQS as expeditiously
as practicable.

XVI.  EPA received the following comments addressing the time frames for
installation of controls for CAIR and controls for RACT requirements:

80.	The 2009 CAIR region wide NOx emissions cap for EGUs represents the
most reductions that are reasonable to achieve from EGUs in the CAIR
region by that date.



81.	The timeframe of CAIR and RACT are different.  Implementation of
RACT is "as expeditiously as practicable, but no later than the first
ozone season or portion thereof which occurs 30 months after the RACT
SIP is due.  CAIR sources have until 2010 and 2015 to put on controls.



Response:  EPA is requiring the CAIR NOx reductions in two phases.  For
states affected by the annual NOx emission reduction requirements, the
first phase begins on January 1, 2009 (not in 2010 as one commenter
mistakenly states).  For states affected by the ozone season NOx
emission reduction requirements, the first phase starts May 1, 2009
(again, not in 2010 as one commenter mistakenly states).   In general,
EPA expects that the largest-emitting EGU sources will be the first to
install NOx control technology and that such control technology will
gradually be installed on progressively smaller-emitting EGU sources
until the ultimate cap is reached.  EPA believes it has set the 2009
CAIR NOx cap at a level that, assuming the reductions are achieved from
EGUs, would result in EGUs installing emission controls on the maximum
total capacity on which it is feasible to install emission controls by
that date.

XVII.  EPA received the following comments arguing that states are not
free to require beyond-CAIR controls on EGUs since the law in many
states prohibits state air agencies from being more stringent that
federal law:

82.	EPA's statement that states are free to require more of EGUs as RACT
ignores the law or policy in many states and localities that prohibit
clean air agencies from being more stringent than federal law.  A survey
of NACAA's members found that 26 state agencies (of 50 respondents) and
9 local agencies (of 42 respondents) reported being precluded from
adopting more stringent requirements than the federal government.



83.	The CAIR equals RACT determination removes state authority and
obligation to impose NOx RACT requirements for some of the largest NOx
sources in their nonattainment areas.



Response:  There is nothing in the CAA or federal law that prohibits
state air agencies from imposing requirements more stringent than
federal law.  EPA recognizes, nonetheless, that some states have
voluntarily chosen to adopt such limits.  All states, regardless of
whether such limits have been adopted, are required by section 172 of
the CAA to attain the ozone NAAQS as expeditiously as practicable. 
Thus, requirements that are determined by the state to be necessary to
attain as expeditiously as practicable with reasonably available control
measures, are in fact required by federal law and cannot be considered
more stringent than federal requirements.  In this action, EPA has
decided that it will accept a determination that NOx RACT for EGUs is
satisfied by compliance with rules implementing CAIR in a state that
achieves all CAIR emission reductions from EGUs and where EPA’s
technical analysis presented in the December 16, 2006 notice of
reconsideration shows that CAIR will achieve greater or equal annual and
ozone-season emissions reductions than source-by-source RACT in the
relevant nonattainment area (or for section 184 requirements, the
relevant OTR state).  If a state chooses to rely on this determination,
it will not be required to perform NOx RACT analyses for sources in the
relevant nonattainment area or OTR state that are subject to a CAIR NOx
trading program.  Nonattainment areas and OTR states that cannot rely on
this determination, may still initially presume that CAIR will satisfy
the NOx RACT requirements if all CAIR reductions are achieved by
EGU’s.  Under this presumption, states are free to conduct
case-by-case RACT determinations at their discretion.  Further, the
requirement to attain the NAAQS as expeditiously as practicable is
distinct from the analysis of what specific emission controls are deemed
RACT for a particular source.  Thus, all states have discretion to
require beyond-RACT NOx reductions if necessary to comply with the
requirements of CAA section 172.

XVIII.  EPA received a comment saying that CAIR gives less incentives
for control technology development than RACT.

84.	RACT is a technology based program expected to evolve over time as
more effective control technologies are developed at lower costs and
CAIR is a static requirement solely based on implementing a two-phased
NOx emissions cap developed in 2005.  As a result, no further reductions
would be required under CAIR into the future, while the application of
source specific RACT requirements would reflect advances in technology
and may require additional reductions in nonattainment areas.  EPA’s
proposal provides an incentive not to install such advanced and less
costly technologies in favor of purchasing cheaper allowances.  



Response:  CAIR provides a strong incentive to develop more efficient
and more economical controls.  If a source can discover how to reduce
its emissions in a more cost effective way and to a greater extent, it
may sell the resulting emission credits on the open market.  One of the
benefits of a cap-and-trade program is that it provides incentives to
lower emissions in the most cost effective way.  This encourages
innovation and development of improved control techniques.  CAIR is
implemented in two phases, with a first phase emissions cap for NOx
applying in 2009 and a second more stringent cap applying from 2015 on.
In addition, as generation grows, the cap actually becomes tighter in
terms of the emission rate, so we would expect allowance prices to
increase over time giving sources a continuous incentive to keep
re-evaluating whether there are advances in controls they should
consider.  Furthermore, under a system focused on source-by-source RACT
determinations, companies have the incentive to come in under their RACT
determinations and argue why controls can't be installed.  Once a limit
is established, companies have no incentive to beat the limit.  The
trading provisions provide the opposite incentive, encouraging companies
to reduce as much as possible.  A State’s determination of what level
of control constitutes RACT should account for information about
improved control methods that are brought to its attention.

XIX.  EPA received the following comments defending its proposal to
allow compliance with CAIR to satisfy NOx RACT requirements in certain
circumstances:

85.	Although sections 172(c)(1) and 183(b)(2) provide that RACT shall
apply "in the area" nothing in the text of either provision mandates
installation of emission control equipment at each individual source "in
the area".

86.	Congress's choice of the phrase 'reasonably available' bespeaks its
intention that the EPA exercise discretion in determining which control
measures must be implemented.

87.	EPA's 1979 definition of RACT could not control interpretation of
the Act's RACT provisions as they were revised by Congress in the 1990
amendment to the Act.



88.	EPA's NOx General Preamble (57 FR 55620, Nov. 25, 1992) concludes
that it is "permissible under the statute for individual sources to have
greater or lesser emissions reductions so long as the area wide average
emission rates associated with a RACT level of NOx emission controls
[are] met."



89.	It is consistent with the Act for EPA and states to determine that
compliance with an area-wide emission trading program may constitute
RACT in lieu of source-by-source emission control requirements.

90.	EPA properly concluded that NOx reductions may be obtained from CAIR
affected EGUs through adoption of emission control technology that is
"reasonably available" for those EGUs in light of CAIR.



Response:  EPA agrees with these comments which support the position EPA
took in the Phase 2 Rule to implement the 8-hour ozone national ambient
air quality standard, and in the notice announcing our decision to
reconsider this issue.

XX.  EPA received a comment that States are required to adopt control
measures necessary to meet the ozone national ambient air quality
standard (NAAQS).

91.	The petition for reconsideration ignores the point that, entirely
apart from what emission controls are deemed RACT, states must require
emission controls as necessary to attain the NAAQS as expeditiously as
practicable.  Thus a state has discretion to require beyond-RACT NOx
reductions.



Response:   This comment reiterates the statement made in the December
16, 2006 notice of reconsideration that a state has discretion to
require beyond-RACT NOx reductions from any source (including sources
covered by the CAIR or NOx SIP Call programs), and has an obligation to
demonstrate attainment of the 8-hour ozone standard as expeditiously as
practicable.  In certain areas, states may require NOx controls based on
more advanced control technologies as necessary to provide for
attainment of the ozone standards. 

XXI.  EPA received the following comments regarding its interpretation
of the term “reasonable”:

92.	The commenter argues that EPA's theory that the term "reasonable" is
ambiguous ignores the statutory language which only speaks to RACT, with
the term reasonably modifying the word available.  It is not reasonable
for EPA to interpret reasonable to apply in one manner for EGUs and a
wholly different manner for other sources. 



93.	One commenter argues that EPA attempts to stretch §172 (c)’s
definition of “reasonable,” when it states that “EPA believe that
the term ‘reasonable’ in RACT may be construed to allow
consideration of the air quality impact of required emissions reduction
from a region-wide cap and trade program such as CAIR.  



Response:  EPA disagrees with the commenter’s assertion that EPA
interprets the term “reasonable” to apply in one manner for EGUs and
in a different manner for other sources.  Section 172(c)(1) of the CAA
requires that nonattainment plans shall provide for the implementation
of all reasonably available control measures as expeditiously as
practicable.  EPA has previously stated that reasonable control measures
can include area wide averaging programs. (See NOx Supplement to the
General Preamble, November 25, 1992 (57 FR 55620).  EPA’s
determination that the term “reasonable” in RACT may be construed to
allow consideration of the air quality impact of required emissions
reduction from a region-wide cap-and-trade program such as CAIR is
consistent with past practice and appropriate for the reasons explained
in this notice.

Further, in determining a level of control which EPA recommends as RACT,
EPA studies a variety of sources and controls and determines what level
of control is applicable in the industry across a wide variety of
sources at a reasonable cost.  States are free to tailor this RACT
guidance to the particular situation confronting individual sources in
that state.  Each permitting agency determines for each source or
source-category in the state, the specific controls that constitute
RACT.  Thus, the precise requirements applied to ensure that RACT is met
may differ from source to source and source-category to source-category.

EPA’s determination that, in certain circumstances, compliance with
CAIR will satisfy the RACT requirement for EGUs in most CAIR states,
does not, as petitioner suggests, reinterpret the term RACT as it
applies to EGUs.  Instead, EPA has determined that the existing RACT
requirement is satisfied by compliance with a rule implementing the CAIR
requirements if and only if a state achieves all its reductions from
EGUs and the EPA’s technical analysis presented in the notice of
reconsideration shows that CAIR will achieve greater or equal reductions
for annual and ozone-season emission reductions than source-by-source
RACT in the relevant nonattainment area or OTR state.  If a state
achieves all of its CAIR emission reductions from EGUs then the
emissions of other source categories in the state are not controlled by
the CAIR rule.  Thus, it would be impossible for EPA to make a similar
determination that they have met their RACT requirements through
compliance with CAIR. 

XXII.  EPA received the following comments on the effects of the recent
D.C. Circuit Court of Appeals decision in South Coast AQMD v. EPA:

94.	The commenter believes that the recent D.C. Circuit of Appeals
decision in South Coast AQMD v. EPA has a direct impact on issues being
considered in this reconsideration notice.  This decision emphasized
limits on EPA's authority where the CAA provision are unambiguous.  The
commenter believes the CAA §172(c)(1) provisions related to RACT are
unambiguous.



95.	Commenter believes that the recent D.C. Circuit Court of Appeals
decision in South Coast AQMD v. EPA rules against EPA for trying to use
the rulemaking process to nullify applicable provisions of the CAA meant
to limit the EPA's discretionary powers.  The commenter thinks this
decision would mean that EPA cannot simply use its discretion to replace
source-specific provisions of the CAA with trading programs such as CAIR
that were designed for other specific air quality goals.



96.	By vacating EPA's phase 1 rule, the South Coast v. EPA Court
decision has thrown into disarray EPA's classifications of many 8-hour
ozone non-attainment areas creating great uncertainly about what areas
must adopt what RACT requirements.  This will affect the June 15, 2007
RACT SIP submittal date which EPA proposes.



97.	EPA's analysis of which EGUs would be subject to RACT requirements
does not reflect the December 2006 decision by the U.S. Court of Appeals
for the D.C. Circuit, South Coast AQMD v. EPA which indicated that EPA
incorrectly classified areas with design values of 0.09 ppm and above as
subpart 1 areas rather than subpart 2.  These areas, under subpart 2,
would be classified as moderate areas and thus subject to RACT, whereas
under subpart 1, they would have only been subject to RACT if they
requested a extension in their attainment dates or if they are located
in the Ozone Transport Regions (OTR).  About a dozen areas that are not
in the OTR would qualify as subpart 2 areas with EGUs subject to RACT.



98.	Commenter submits that the issues that are subject to the EPA action
are not affected by the Court of Appeals recent ruling in South Coast
Air Quality Management District v. EPA, No. 04-1200 D.C. Cir. Dec 22,
2006, and there is no need for further delay as a result of that ruling.



Response:  EPA disagrees with the comments to the extent they suggest
EPA is seeking to replace the RACT requirement with CAIR.  The final
rule does not displace the RACT requirement for any sources.  EPA also
disagrees with the comments to the extent they suggest that EPA’s
interpretation of the RACT requirements in sections 172(c)(1), 182(f)
and 184(b) is inconsistent with the South Coast AQMD v. EPA decision. 
Further, on March 22, 2007, EPA filed a petition for panel rehearing of
the South Coast AQMD v. EPA decision and thus the full impact of that
decision cannot yet be assessed.  At this time, EPA is unable to
determine which areas, if any, in addition to those included in the
analysis will be required to submit separate RACT SIPs.   However, as
discussed in the preamble to the final notice of reconsideration,
region-wide emissions reductions from the CAIR are projected to be
significantly greater than reductions that would be projected to occur
from application of source-by-source RACT, such that the possible
movement of areas designated in the phase 1 rule as subpart 1 to subpart
2 area designations is not expected to alter the conclusion that the
CAIR achieves greater emission reductions in the region than
source-by-source RACT.  In addition, as previously discussed, EPA is
limiting the scope of its determination that compliance with the CAIR
satisfies NOx RACT requirements.  This determination applies in areas
where EPA’s emissions analysis in the December 16, 2006 notice of
reconsideration shows that the CAIR will achieve greater or equal annual
and ozone-season emissions reductions than source-by-source RACT.

XXIII.  EPA received the following comments opposing the proposed
extension of the EGU NOx RACT SIP submittal deadline:

99.	Commenter submits that the issues that are subject to the EPA action
are not affected by the Court of Appeals recent ruling in South Coast
Air Quality Management District v. EPA, No. 04-1200 D.C. Cir. Dec 22,
2006, and there is no need for further delay as a result of that ruling.



100.	EPA should not extend the submittal date for RACT SIPS. But should
finalize a rule that requires RACT at EGUs and expedite SIP submittal.



101.	EPA has no authority to extend the due date for RACT SIPs for EGUs
to June 15, 2007.  This dead-line should be June 2006 according the CAA.



102.	The commenter opposes postponing the submission date for RACT SIPs
for EGUs until 6 months following the date that other stationary source
RACT SIPs are due.  If CAIR = RACT holds, this time is not needed. On
the other hand, if EPA determines that source by source RACT must be
done for every source (including EGUs) then there should be extensions
for all SIP RACT determinations.



Response:  Section 182 does not explicitly provide that RACT SIPs must
be submitted a certain number of months after an area is designated
nonattainment for the 8-hour ozone NAAQS.  EPA interprets the comment to
suggest that the final rule contains requirements similar to the VOC
RACT requirements in section 182(b)(2)(C), which must be submitted to
the Administrator by two years after November 15, 1990 (the date of
enactment of the CAA Amendments of 1990).  Therefore, the argument goes,
the RACT SIPs must similarly be submitted within two years of the
nonattainment designation, or June 15, 2006.  In the final Phase 2 Rule,
we determined that because some states might rely on the submittal of
SIP revisions meeting the CAIR to also satisfy RACT for some sources, it
was reasonable to extend the RACT submittal date to September 15, 2006
to correspond to the required date for submitting CAIR SIPs.   This date
has now passed, and for the reasons explained in section III.B.1 of the
preamble to the final notice of reconsideration, EPA does not believe it
would be appropriate to finalize this rule with a retroactive deadline.

XXIV.  EPA received the following comments supporting the extension of
the EGU NOx RACT SIP submittal deadline at least until June 15, 2007 and
some argued a longer extension may be necessary given the uncertainties
regarding classifications created by the decision in South Coast AQMD v.
EPA:

103.	Commenter supports the proposed extension to at least June 15, 2007
for states to submit SIPs addressing NOx RACT.



104.	The commenter believes that extending the deadline for RACT SIPs
for EGUs at least until June 15, 2007 is appropriate. The commenter
points out that South Coast v. EPA court ruling may necessitate an even
longer delay in submitting RACT SIPs until EPA resolves matters
concerning area classifications, attainment dates and emission control
requirements.



Response:  The RACT SIP submittal date in the final rule reflects
EPA’s recognition that the South Coast AQMD v. EPA decision has
created some uncertainty about which areas, by virtue of their
classification, would be required to address RACT requirements and in
what time frame.

  Since RACT is a technology requirement prescribing year-round
controls, it is appropriate to consider how participation in both CAIR
trading programs (annual and seasonal) will affect annual emissions of
NOx and to compare that to how RACT will affect annual emissions of NOx.

  40 CFR 51.912 (c)(1) (promulgated in the Phase 2 Rule) provides that
for a subpart 1 area “. . . that submits an attainment demonstration 
that requests an attainment date 5 or less years after designation for
the 8-hour NAAQS,  the state shall meet the RACT requirement by
submitting an attainment demonstration SIP demonstrating that the area
has adopted all control measures necessary to demonstrate attainment as
expeditiously as practicable.”  Thus, these areas are not required to
submit RACT SIPs separate from their attainment demonstrations. 
However, a state must submit a RACT SIP separate from an attainment
demonstration SIP for the following areas:  under 40 CFR 51.912 (a),
subpart 2 moderate and above areas; and under 40 CFR 51.912 (c)(2),
subpart 1 areas with attainment dates beyond 5 years after designation.

 Technical Support Document For Phase 2 Of The Final Rule To Implement
The 8-Hour Ozone National Ambient Air Quality Standard –Notice Of
Reconsideration; NOx RACT For EGUs In CAIR states--Supplemental
Technical Analysis. (Docket ID No. EPA-HQ-OAR-2003-0079, item1044.2)

   The determination for OTR states is separate from the determination
for nonattainment within the OTR states, i.e., this determination
applies to areas in these OTR states other than (a) moderate and above
subpart 2 areas and (b) subpart 1 areas that request an attainment date
more than 5 years after designation for the 8-hour NAAQS.  This means
that an OTR state can get a determination that CAIR equals RACT within
the State, but a particular nonattainment within the State may not get
this determination based on the results of the technical analysis.

   The determination for OTR states is separate from the determination
for nonattainment within the OTR states, i.e., this determination
applies to areas in these OTR states other than (a) moderate and above
subpart 2 areas and (b) subpart 1 areas that request an attainment date
more than 5 years after designation for the 8-hour NAAQS.  This means
that an OTR state can get a determination that CAIR equals RACT within
the State, but a particular nonattainment within the State may not get
this determination based on the results of the technical analysis.

 PAGE   

 PAGE   33 

