							

State and Tribal Comment Summary and Response Document on EPA’s
Recommended Area Designations for the 2006 24-hour PM2.5 Designation
Recommendations

EAP-HQ-OAR-2007-0562

December 22, 2008



TABLE OF CONTENTS

  TOC \o "1-3" \h \z \u    HYPERLINK \l "_Toc217305073"  List of Tables	
 PAGEREF _Toc217305073 \h  3  

  HYPERLINK \l "_Toc217305074"  List of Figures	  PAGEREF _Toc217305074
\h  3  

  HYPERLINK \l "_Toc217305075"  List of Acronyms	  PAGEREF _Toc217305075
\h  4  

  HYPERLINK \l "_Toc217305076"  GENERAL ISSUES	  PAGEREF _Toc217305076
\h  6  

  HYPERLINK \l "_Toc217305077"  SECTION I:  2006-2008 Data	  PAGEREF
_Toc217305077 \h  6  

  HYPERLINK \l "_Toc217305078"  SECTION II:  Contributing Emissions
Score (CES)	  PAGEREF _Toc217305078 \h  6  

  HYPERLINK \l "_Toc217305079"  SECTION III:  Technical Analysis	 
PAGEREF _Toc217305079 \h  24  

  HYPERLINK \l "_Toc217305080"  SECTION IV:  Power Plant Issues	 
PAGEREF _Toc217305080 \h  35  

  HYPERLINK \l "_Toc217305081"  SECTION V:  Consideration of Other
Regulations	  PAGEREF _Toc217305081 \h  37  

  HYPERLINK \l "_Toc217305082"  SECTION VI:  Exceptional Events	 
PAGEREF _Toc217305082 \h  41  

  HYPERLINK \l "_Toc217305083"  SECTION VII:  Other General Issues	 
PAGEREF _Toc217305083 \h  42  

  HYPERLINK \l "_Toc217305084"  SECTION VIII:  Typographical Errors	 
PAGEREF _Toc217305084 \h  44  

  HYPERLINK \l "_Toc217305085"  AREA-SPECIFIC ISSUES	  PAGEREF
_Toc217305085 \h  46  

  HYPERLINK \l "_Toc217305086"  Allentown, PA	  PAGEREF _Toc217305086 \h
 46  

  HYPERLINK \l "_Toc217305087"  Birmingham, AL	  PAGEREF _Toc217305087
\h  49  

  HYPERLINK \l "_Toc217305088"  California Air Resources Board (CARB)
Multi-Area	  PAGEREF _Toc217305088 \h  56  

  HYPERLINK \l "_Toc217305089"  Chicago, IL-IN	  PAGEREF _Toc217305089
\h  57  

  HYPERLINK \l "_Toc217305090"  Chico, CA	  PAGEREF _Toc217305090 \h  62
 

  HYPERLINK \l "_Toc217305091"  Cincinnati-Hamilton, OH-KY-IN	  PAGEREF
_Toc217305091 \h  63  

  HYPERLINK \l "_Toc217305092"  Clarksville, TN-KY	  PAGEREF
_Toc217305092 \h  68  

  HYPERLINK \l "_Toc217305093"  Cleveland-Akron-Lorain, OH	  PAGEREF
_Toc217305093 \h  73  

  HYPERLINK \l "_Toc217305094"  Columbus, OH	  PAGEREF _Toc217305094 \h 
75  

  HYPERLINK \l "_Toc217305095"  Davenport-Moline-Rock Island, IA-IL	 
PAGEREF _Toc217305095 \h  77  

  HYPERLINK \l "_Toc217305096"  Dayton-Springfield, OH	  PAGEREF
_Toc217305096 \h  82  

  HYPERLINK \l "_Toc217305097"  Detroit-Ann Arbor, MI	  PAGEREF
_Toc217305097 \h  83  

  HYPERLINK \l "_Toc217305098"  Evansville, IN	  PAGEREF _Toc217305098
\h  93  

  HYPERLINK \l "_Toc217305099"  Fairbanks, AK	  PAGEREF _Toc217305099 \h
 94  

  HYPERLINK \l "_Toc217305100"  Fort Belknap Indian Community, MT	 
PAGEREF _Toc217305100 \h  99  

  HYPERLINK \l "_Toc217305101"  Grand Rapids, MI	  PAGEREF _Toc217305101
\h  99  

  HYPERLINK \l "_Toc217305102"  Green Bay, WI	  PAGEREF _Toc217305102 \h
 100  

  HYPERLINK \l "_Toc217305103"  Huntington-Ashland, WV-KY-OH	  PAGEREF
_Toc217305103 \h  101  

  HYPERLINK \l "_Toc217305104"  Imperial County, CA	  PAGEREF
_Toc217305104 \h  115  

  HYPERLINK \l "_Toc217305105"  Indianapolis, IN	  PAGEREF _Toc217305105
\h  118  

  HYPERLINK \l "_Toc217305106"  Juneau, AK	  PAGEREF _Toc217305106 \h 
119  

  HYPERLINK \l "_Toc217305107"  Klamath Falls, OR	  PAGEREF
_Toc217305107 \h  121  

  HYPERLINK \l "_Toc217305108"  Knoxville-Sevierville-LaFollette, TN	 
PAGEREF _Toc217305108 \h  121  

  HYPERLINK \l "_Toc217305109"  Logan, UT-ID	  PAGEREF _Toc217305109 \h 
122  

  HYPERLINK \l "_Toc217305110"  Louisville, KY-IN	  PAGEREF
_Toc217305110 \h  124  

  HYPERLINK \l "_Toc217305111"  Madison-Baraboo, WI	  PAGEREF
_Toc217305111 \h  137  

  HYPERLINK \l "_Toc217305112"  Milwaukee-Racine, WI	  PAGEREF
_Toc217305112 \h  138  

  HYPERLINK \l "_Toc217305113"  Muscatine, IA	  PAGEREF _Toc217305113 \h
 138  

  HYPERLINK \l "_Toc217305114"  Nez Perce Tribe	  PAGEREF _Toc217305114
\h  140  

  HYPERLINK \l "_Toc217305115"  Oakridge, OR	  PAGEREF _Toc217305115 \h 
141  

  HYPERLINK \l "_Toc217305116"  Paducah-Mayfield, KY-IL	  PAGEREF
_Toc217305116 \h  141  

  HYPERLINK \l "_Toc217305117"  Parkersburg-Marietta, WV-OH	  PAGEREF
_Toc217305117 \h  148  

  HYPERLINK \l "_Toc217305118"  Pascua Yaqui Tribe	  PAGEREF
_Toc217305118 \h  154  

  HYPERLINK \l "_Toc217305119"  Philadelphia-Wilmington, PA-NJ-DE	 
PAGEREF _Toc217305119 \h  154  

  HYPERLINK \l "_Toc217305120"  Pinehurst, ID	  PAGEREF _Toc217305120 \h
 169  

  HYPERLINK \l "_Toc217305121"  Sacramento, CA	  PAGEREF _Toc217305121
\h  171  

  HYPERLINK \l "_Toc217305122"  Saint Louis, MO-IL	  PAGEREF
_Toc217305122 \h  173  

  HYPERLINK \l "_Toc217305123"  Salt Lake City-Provo, UT	  PAGEREF
_Toc217305123 \h  177  

  HYPERLINK \l "_Toc217305124"  San Carlos Apache Indian Reservation, CA
  PAGEREF _Toc217305124 \h  191  

  HYPERLINK \l "_Toc217305125"  Youngstown, OH	  PAGEREF _Toc217305125
\h  193  

  HYPERLINK \l "_Toc217305126"  Yuba City-Marysville, CA	  PAGEREF
_Toc217305126 \h  193  

 

List of Tables

  TOC \h \z \c "Table"    HYPERLINK \l "_Toc217331174"  Table 1. 
Delaware EGUs.	  PAGEREF _Toc217331174 \h  167  

  HYPERLINK \l "_Toc217331175"  Table 2.  Moor Facility Annual Emissions
Limits.	  PAGEREF _Toc217331175 \h  167  

  HYPERLINK \l "_Toc217331176"  Table 3.  PM2.5 Modeling Results.  From
the MDNR 10-18-08 response.	  PAGEREF _Toc217331176 \h  175  

  HYPERLINK \l "_Toc217331177"  Table 4.  Days with High Concentrations
at Granite City or VFW.

	From the MDNR October 18, 2008 response.	  PAGEREF _Toc217331177 \h 
177  

 

List of Figures

  TOC \h \z \c "Figure"    HYPERLINK \l "_Toc217329864"  Figure 1. Wind
Pattern Diagram for January 11 through 13, 2004.	  PAGEREF _Toc217329864
\h  15  

  HYPERLINK \l "_Toc217329865"  Figure 2.  HYSPLIT Back Trajectory for
January 13, 2004, all Salt Lake County Monitors.	  PAGEREF _Toc217329865
\h  16  

  HYPERLINK \l "_Toc217329866"  Figure 3.  Meteorology Near the Utah
County/Salt Lake County Line, January 13, 2004.	  PAGEREF _Toc217329866
\h  17  

  HYPERLINK \l "_Toc217329867"  Figure 4.  Wind Pattern Diagram for
January 20 through 22, 2004.	  PAGEREF _Toc217329867 \h  18  

  HYPERLINK \l "_Toc217329868"  Figure 5.  24-hour Backtrajectory for
January 22, 2004, All Salt Lake County Monitors.	  PAGEREF _Toc217329868
\h  19  

  HYPERLINK \l "_Toc217329869"  Figure 6.  Meteorology Near the Utah
County/Salt Lake County Line, January 22, 2004.	  PAGEREF _Toc217329869
\h  20  

  HYPERLINK \l "_Toc217329870"  Figure 7.  Wind Pattern Diagram for
January 24 through 26, 2006.	  PAGEREF _Toc217329870 \h  21  

  HYPERLINK \l "_Toc217329871"  Figure 8.  24-hour Backtrajectory for
January 26, 2006.	  PAGEREF _Toc217329871 \h  22  

  HYPERLINK \l "_Toc217329872"  Figure 9.  Utah DOT Alpine and Bluffdale
Meteorology Station Locations.	  PAGEREF _Toc217329872 \h  23  

  HYPERLINK \l "_Toc217329873"  Figure 10.  Point of the Mountain
Surface Winds, January 21-30, 2007.	  PAGEREF _Toc217329873 \h  24  

  HYPERLINK \l "_Toc217329874"  Figure 11.  Pollution Rose for St. Clair
County, Michigan.	  PAGEREF _Toc217329874 \h  91  

  HYPERLINK \l "_Toc217329875"  Figure 12.  Position of Herriman and
Highland relative to Point of the Mountain.	  PAGEREF _Toc217329875 \h 
181  

  HYPERLINK \l "_Toc217329876"  Figure 13.  Concentration and Point of
the Mountain Surface Winds on the Peak PM2.5 Day of 2007.	  PAGEREF
_Toc217329876 \h  182  

  HYPERLINK \l "_Toc217329877"  Figure 14.  Point of the Mountain
Meteorology, January 27, 2007.	  PAGEREF _Toc217329877 \h  183  

 

List of Acronyms

 

Acronym	Definition

AQS	Air Quality System

ASOS	Automated Surface Observing Systems

BACT	Best Available Control Technology

BART	Best Available Retrofit Technology

CAIR	Clean Air Interstate Rule

CALGRID	California Photochemical Grid

CAMx	Comprehensive Air Quality Model with Extensions

CARB	California Air Resources Board

CBSA	Core Based Statistical Area

CEM	Continuous Emissions Monitoring

CENRAP	Central Regional Air Planning Association

CES	Continuous Emissions Score

CMAQ	Community Multi-Scale Air Quality Modeling

CMSA	Combined Metropolitan Statistical Area

CRPAQS	California Regional PM10/PM2.5 Air Quality Study

CSA	Combined Statistical Area

EDAS	Eta Data Assimilation System

EGU	Electric Generating Unit

ESP	Electrostatic Precipitator

FEM	Federal Equivalent Method

FGD	Flue Gas Desulfurization

FHWA	Federal Highway Administration

FRM	Federal Reference Method

GIS	Geographic Information Systems

HPMS	Highway Performance Modeling System 

HYSPLIT	Hybrid Single Particle Lagrangian Integrated Trajectory

KYOVA	Metropolitan Planning Organization for southwestern West Virginia
and southern Ohio

LADCO	Lake Michigan Air Directors Consortium

LAER	Lowest Achievable Emission Rate

LMACD	Louisville Metropolitan Air Control District

MANE-VU	Mid-Atlantic/Northeast Visibility Union

MARAMA	Mid-Atlantic Regional Air Management Association

MVEB	Motor Vehicle Emission Budget

MPO	Metropolitan Planning Organization

MSA	Metropolitan Statistical Area

NAAQS	National Ambient Air Quality Standard

NEEDS	National Electric Energy Data System

NEI	National Emission Inventory

NEPA	National Environmental Policy Act

NMIM	National Mobile Inventory Model

NOAA	National Oceanic and Atmospheric Administration

OTC	Ozone Transport Commission

PM	Particulate Matter

PMF	Positive Matrix Factorization

PSAT	Particulate Source Apportionment Technology

PSD	Prevention of Significant Deterioration

QA	Quality Assurance

RACT	Reasonably Available Control Technology

SCR	Selective  Catalytic Reduction

SIP	State Implementation Plan

SLAMS	State and Local Air Monitoring Stations

SNCR	Selective Non-Catalytic Reduction

SOFA	Separated Overfire Air

TSD	Technical Support Document

UMAERO/MM5	Urban Airshed Model for Aerosols/Mesoscale Model Version 5

VMT	Vehicle Miles Traveled

WWW	Wood-Washington-Wirt

GENERAL ISSUES

SECTION I:  2006-2008 Data

State Comment:  Consideration of 2008 Data

A number of States believe that data for one or more of the areas
violating the 2006 

24-hour PM2.5 National Ambient Air Quality Standard (NAAQS) may show
attainment based on 2006-2008 air quality data.  These States ask that
EPA consider 2006-2008 data for purposes of determining final area
designations.

EPA Response:

The U.S. Environmental Protection Agency (EPA or the Agency) understands
the States’ concerns with using 2005-2007 data in lieu of the most
recent data (2006-2008) for making designation decisions.  The Agency
recognizes that many areas in the country are making improvements in
their air quality and believes that it is important to recognize such
improvements.  In making decisions about whether an area is meeting the
24-hour PM2 5 NAAQS, EPA also believes that it is important to use the
most recent air quality data.  EPA will be unable to use 2008 air
quality data for the designation decisions that will be made in December
2008 because that data will not be final and certified by that date. 
However, all States will still have the opportunity to benefit from
improvements air quality by using 2008 air quality data to demonstrate
attainment of the 24-hour PM2.5 NAAQS.

Prior to the effective date of final 24-hour PM2.5 designations, States
will have the opportunity to provide more recent data showing that an
area is attaining the 24-hour PM2.5 NAAQS.  To do so, a State must
submit its complete, quality assured, certified 2008 air quality data to
EPA earlier than the usual June 30 deadline.  This early submittal
deadline will be February 20, 2009 – approximately 45 days prior to
the 90-day effective date of publication of the final 24-hour PM2.5
designations.  If the Agency agrees that a change of designation status
is appropriate based on 2006-2008 air quality data, it would withdraw
the nonattainment designation prior to the 90-day effective date of
final 24-hour PM2.5 designations and take a new final action designating
such areas consistent with the new 2008 data.

SECTION II:  Contributing Emissions Score (CES)

General Response to Comments Regarding the CES:

Several commenters asserted that too much emphasis had been placed on
the Contributing Emissions Score (CES) in deciding the nonattainment
area boundaries.  EPA has stated that the CES is one tool which utilizes
information from the nine factors recommended through guidance to be
used in determining potential nonattainment area boundaries.  EPA’s
intent in developing and utilizing the CES was to have it serve as a
starting point in determining the possible spatial extent of a potential
nonattainment area with additional information through the use of the
nine factors as well as other information the Federal, State or Local
air agency could provide to determine an area that was contributing to
concentrations greater than the level of the standard.  The CES was in
no way considered to be outcome determinative, but just one more
analytical tool that could help inform the final boundary decision.  

As stated in the Technical Support Document (TSD), the CES represents a
generalized approach that was implemented across the country for
consistency purposes.  The Agency strongly encouraged State and Local
agencies to submit other data and analyses to provide a more detailed
and specific characterization of the contributing emissions to the
ambient PM2.5 concentrations violating the NAAQS.  The comments EPA has
received including these additional pieces of information have played a
major role in the final decision making process.  

Other commenters stated that the CES was limited based on the data and
the weighting factors used in its calculation.  EPA recognizes that it
is difficult for any analytical tool to completely characterize the
unique and complex conditions observed in many areas across the country
using a general approach.  The CES TSD clearly states that several
factors could not be adequately accounted for such as topography,
overall emissions estimates and emissions estimates in counties with
isolated densely populated areas or large rural areas.  EPA recognized
that additional information from the State and Local agencies would be
required to better characterize these unique circumstances and
encouraged State and Local agencies to submit through the comment
process any supplemental information such as population density,
commuting patterns, more detailed or localized weather data, and
topography data.  Additional data and analyses were received to aid U.S.
EPA staff in better determining the spatial extent of areas in parts of
the country where these limitations were experienced and were considered
to help inform the final regulatory decision.  Many States, however, did
not provide additional analyses which required the Agency to use the
information it had on hand at the end of the comment period to inform
its decision.

State Comment:  Introduction and application of CES methodology

A number of States expressed concerns that information about the CES
should have been introduced to the States much sooner, and that it can
produce obscure or inaccurate results.

EPA Response:

The concepts used in the CES were introduced early in the process at the
PM2.5 Implementation and Designations Workshop held in Chicago, IL from
June 20-21, 2007.  At this workshop, the concept of an emissions score
was presented that envisioned taking into account an area’s emissions,
back trajectories, and the speciation increments.  That presentation can
be found at   HYPERLINK
"http://www.epa.gov/ttn/naaqs/pm/presents/pm2.5_designations_part3.pdf" 
http://www.epa.gov/ttn/naaqs/pm/presents/pm2.5_designations_part3.pdf . 
At this meeting EPA expressed its interest in further developing some
sort of score similar to what was developed during the designation
process for the 1997 PM2.5 NAAQS, utilizing the concepts presented
during the Chicago workshop.

EPA introduced the CES in its entirety through its modification letters
to States provided in 2008, and allowed the States to further comment on
the technique.  This provided States the opportunity to assess the
applicability of the CES to areas within their States.

California:

Although the CES is only one element in determining the nonattainment
boundary areas, a high CES suggests that a county has a high impact on
the adjacent violating county.  However, CES numbers are based on data
for entire counties.  The CES should be adjusted to reflect only those
portions of a county to be included with an adjoining nonattainment
area, such as Solano, El Dorado, and Placer Counties within the
Sacramento nonattainment area.  The higher score of Solano was
discounted based on its contribution to the San Francisco Bay Area
nonattainment area and the higher population in the western portion of
the county.  The high scores for Placer and El Dorado were based,
partially, on analysis done for the entire counties.  As noted in a U.S.
EPA Technical Document (Rizzo and Hunt, 2008), the CES methodology uses
county-based emissions inventories which may be inaccurate in counties
with large rural populations or with mountainous terrain, both of which
occur in El Dorado and Placer.  Although EPA took some of this into
account in recommending only a part of each county for inclusion in the
nonattainment area, it did not take into account the fact that the
majority of PM2.5 emissions are from residential wood burning.  These
emissions were recently found to be inaccurate (pages 17 and 18 of the
California comment letter) and a significant portion may be occurring in
the Lake Tahoe Air Basin segment of these counties.

Use of population and population growth as factors in EPA’s
decision-making was not consistent throughout the country.  Warren
County, NJ is an example of a county not included with an adjacent
violating area.  According to EPA,

“Warren County [New Jersey] ranks low in terms of population and in
population density in comparison to counties located near the violating
monitor in Northampton County, Pennsylvania.  In comparison to the two
counties that have been recommend as nonattainment for the Allentown,
PA-NJ area, Warren County’s population and population density is below
50 percent that of Lehigh and Northampton.  (U.S. EPA Response to New
Jersey, 2008)”.

Warren County’s population density is, in fact, 32 percent of Lehigh
County and 40 percent of Northampton County.  Although the Sacramento
County population is larger than the populations for counties around
Warren County, NJ, Sacramento’s population density is very similar. 
Both total populations and population densities for all surrounding
counties are below those of Sacramento County and far below EPA stated
limit of above 50 percent.

In an additional example, Hamblen County, part of the
Knoxville-Sevierville-LaFollette,

TN Core Based Statistical Area (CBSA), has a population density 44
percent of neighboring (and violating) Knox County.  Hamblen County was
designated in attainment (U.S. EPA Response to Tennessee, 2008).  There
are many other examples of counties with higher population densities
than those adjoining Sacramento, within a MSA, but not designated
nonattainment.  EPA has placed a high importance on the CES in
designating nonattainment areas.  While several counties in California
have a relatively low CES and no violating monitor, EPA has still
proposed a nonattainment designation in tandem with neighboring
violating counties.  In several other areas throughout the country,
however, counties with similar, or higher, CES are not wed to their
adjacent nonattainment counties (Table 2-3 on page 25 of the California
comment letter).  California requests similar flexibility as provided to
other areas of the country.

EPA Response:

EPA recognizes that using county level emissions can have limitations in
areas with geographically large counties.  However, the Agency also used
information from the nine factors as well as additional information from
the State to assess whether partial county designations were
appropriate.  For example, in the case of Placer County EPA has
determined it is appropriate to designate only a portion of that county
based on information such as topographical features and locations of
emission sources.  As noted above, the CES is one analytical tool that
EPA has used in conjunction with other information as appropriate.  The
CES is designed and intended to assess the relative contributions only
within a specific geographic area.  Just as it is not appropriate to
consider a county within metropolitan New York City and a county within
metropolitan Los Angeles as equal in all respects, it is also not
appropriate to compare CES values between different nonattainment areas.
 

Delaware: 

EPA has relied heavily on the CES in evaluating this factor, using CES
as an arbitrary and inadequate means to evaluate transport, and
apparently putting much weight on the CES in their analysis.  First, the
CES utilizes 2005 National Emission Inventory (NEI) emissions data,
which completely misrepresents Delaware emissions.  Second, tools such
as the CES should be considered only when more sophisticated tools like
modeling are not available.  EPA Clean Air Interstate Rule (CAIR)
modeling has already demonstrated that the PM2.5 transport problem is a
regional problem, and explicitly demonstrates that Delaware does not
significantly contribute to the Philadelphia Combined Statistical Area
(CSA) problem.  The CES indicates that New Castle County impacts the
Philadelphia area more than any other county, and the Pennsylvania CAIR
modeling indicates that the entire State of Delaware does not contribute
significantly to any part of the Philadelphia CSA (i.e., they reach
opposite conclusions)

EPA Response:

The 2005 NEI was used for the CES calculation because it was the most up
to date inventory EPA had at the time.  EPA believes that the inventory
is sufficiently accurate and appropriate for this purpose because it
reflects information submitted by States concerning emission sources
within their boundaries.  In addition, although not directly involved
with the CES calculation per se, EPA invited States to submit further
information regarding emission reductions since 2005 to be considered as
part of the designation process.  The CES was intended to be a screening
tool to provide an initial evaluations of areas suitable for inclusion
within a nonattainment boundary.  Additional information such as that
recommended by EPA in its guidance document was taken into account to
better refine the final decision.

The purpose of the CAIR modeling was to assess the contribution of
interstate transport under Section 110 (a)(2)(D) of the Clean Air Act.  
The purpose of the CES was to provide an initial evaluation of counties
possibly contributing to the violating monitor concentrations within the
Philadelphia CSA.  These two tests were implemented to assess two
different problems, the first being long-range interstate transport as
opposed to the second, which was the contribution from a county to the
nearby area.  Under section 107(d), EPA must designate the nearby areas
that contribute to violations of the NAAQS.  Interstate or long range
transport is addressed under other provisions of the CAA.  

A CES value is interpreted as the magnitude of contribution a county has
on any violating county within an area.  This may mean that a high CES
value may indicate that a county is affecting itself if it also contains
a violating monitor as is the case with New Castle, DE.  EPA considered
many sources of information when determining whether to include New
Castle, DE as part of the Philadelphia CSA including its inclusion in
the current annual PM2.5 nonattainment area and its close proximity to
the city of Philadelphia.  

Iowa:

The CES methodology is incapable of resolving scales finer than the
county level, utilizes data during periods without a violation of the
24-hour PM2.5 NAAQS, and is based upon unrepresentative speciated data. 
The boundaries should be determined based upon the analyses utilizing
data corresponding to the nonattainment design values, 2005 – 2007. 
County-scale data as aggregated in the CES analysis is insufficient to
resolve the impacts of the sources adjacent to the monitors and results
in arbitrary county boundaries.

EPA Response:

The Agency used county level emissions data because that was the finest
resolved data available.  This level of detail was sufficient for the
analytical purposes in part because PM2.5 and its precursors generally
transport across distances greater than individual counties.  However,
where county sizes varied this may be a limitation of the CES as
acknowledged in the technical support document.  In such situations, the
Agency considered additional sources of information as recommended in
its guidance.  

For this purpose, the CES was appropriate for use as a screening tool
with additional data refining the area to better reflect specific
conditions to both Davenport and Muscatine.  For both Davenport and
Muscatine, 2005-2007 data were used.  This has been clarified in the CES
TSD which now lists those areas for which 2005-2007 data were used as
opposed to 2004-2006.  In generalizing the CES so it could be
implemented across the country, default data were used to represent the
speciation concentrations in the violating area based on the area’s
general location within the country.  

In the case of Muscatine, EPA recognizes that using default data may not
specifically represent the actual speciation profile for the area.  The
default values represented the average profile for the region containing
the violating monitor which EPA felt served as a valid substitution for
lacking speciation data.  For this case, the Agency relied on data from
the State to better clarify the specific circumstances associated with
both the Davenport and Muscatine areas to make its final designation
decision.  

Kentucky:

The Commonwealth was surprised to learn that EPA had employed the use of
a “contributing emissions scoring” process to evaluate counties for
emissions contributions to an area attainment problem.  At no time
during States’ development of recommendations did EPA offer
information concerning this methodology.  Further, EPA did not afford
the States the opportunity to provide input on the appropriateness of or
the science behind this methodology.  Information on this methodology
only became available in August 2008, after recommendations were done. 
This approach was revealed in EPA’s comments on Kentucky’s
recommended PM2.5 Boundary Designation December and June submittals. 
Taking this approach, especially at such a late date, is not only
contrary to boundary guidance provided to States by EPA, but insults the
established designation process which allows States to use their
thorough knowledge of the monitoring network and local and regional
circumstances to make those designations.  A brief description and a
link to the CES TSD, which consists of over 1,000 pages of information,
were provided in EPA’s comments.

EPA should provide all documentation and a step-by-step process report
that shows the calculations of the CES for each county in Kentucky that
EPA proposes to be nonattainment.  The development of the CES appears to
involve a large number of assumptions, data, and calculations, and it is
not practicable to expect Kentucky to be able to verify the accuracy of
each step without the supporting documentation and processes.  Due to
the number of errors contained in the August 19, 2008 letter, Kentucky
feels that it is imperative to review the process of developing the CES
for the areas proposed for nonattainment by EPA.

EPA Response:

The concepts utilized in the CES were presented at the Regional and
State designations meeting held in Chicago in June 2007.  The CES served
to provide an initial evaluation of what potential  areas are
contributing.  It incorporated forms of information recommended in 
EPA’s guidance to preliminarily determine contributing counties which
potentially affected the ambient PM2.5 concentrations at violating
monitors on high days.  EPA has given the State the opportunity to
comment on the CES through the 120 day letter process which the State
has availed itself of here.  

The CES technical support document stated that the technique had several
limitations which the States would supplement with additional
information based on the results from the 9-factor analysis.  Thus, the
CES score acted as one tool that was not used to determine the final
outcome but as an initial piece of information from which additional
data from the 9-factor analysis and other analyses submitted to EPA by
the States could be used to making a more informed decision.  

Michigan:

The State maintains that the CES calculation is a complicated and
obscure analysis of air pollution concentrations across Michigan. 
Ambient values clearly are the best representation of actual
environmental conditions in areas of concern.  The CES calculation used
to determine the daily PM2.5 nonattainment areas is much more complex
than the version used previously for the annual PM2.5 designations.  The
Michigan Department of Environmental Quality (MDEQ) is concerned about
the extensive use of assumptions and weighting factors, all of which
culminate in propagation of error in the final result.

EPA Response:

Michigan’s theory is that EPA should only use monitoring data to
promulgate designations.  Obviously, violating monitors indicate where
there violations.  However, to evaluate nearby areas contributing to
those violations, EPA believes that it is necessary to evaluate other
forms of information.  Therefore, EPA has recommended consideration of
different types of information as laid out in the guidance.  The CES is
intended to be another method to evaluate several of these types of
information.  

The CES aggregates data from some of the types of information
recommended in EPA’s guidance.  The tool provides an estimate of the
contribution a particular county has on a violating county.  It does not
attempt to characterize air pollution concentrations across an area.  It
utilizes air pollutant concentrations at violating sites to determine
which days to examine more closely and use a series of weighting factors
to determine which counties could possibly contribute to the violation. 


The limitations of the CES have been discussed in the TSD.  The Agency
has given States the opportunity to submit additional information
specific to their areas to make a better informed decision.  Although
the CES influenced the Agency’s decision, it served as one tool out of
many whose information was utilized to determine potential nonattainment
area boundaries.  

Utah: 

The State indicates for the CES that since there is no documentation of
the actual analysis itself, but only conceptual descriptions of the
process at the web link provided by the Agency, one assumes this was an
analysis done for all of the nonattainment areas in the country.  The
State continues with concerns with the validity of the CES with respect
to the Wasatch Mountains as a topographical issue.

EPA Response:

EPA notes that the CES scores that were provided for every proposed
nonattainment area across the nation do not provide definitive answers,
but do provide a useful screening analysis for the areas.  The EPA does
not disagree that the CES screening analyses have limitations especially
for areas in the western U.S.  The CES information was an additional
analytical tool in EPA’s technical analysis.  The following is
extracted from section 3 “Limitations” of the CES reference
document:

“Topography also affects the CES outcome.  For example, mountain
ranges can act as barriers to transported emissions as well as
potentially limiting the size of the region considered when calculating
the urban increment.  The CES cannot adequately account for the effects
of mountainous terrain which would essentially split a county into
different parts, each having their own potentially unique effect on the
violating county.  A more in-depth discussion of these limitations is
contained within the “Methodology Details" section of this document. 
These potential deficiencies can be overcome on an area-by-area basis by
using supplemental information, such as population density, commuting
patterns, more detailed or localized weather data, and topography
data.”

In addition to the information provided in the Agency’s letter were
web links to specific documentation regarding the development of our
9-factor analysis.  EPA acknowledges that the information was
voluminous.  However, this information helped to inform EPA’s
decisions for the Wasatch Front.

Utah:

The State expressed concerns about using an annual emission inventory
data (in tons per year) rather than an episodic inventory for the
9-factor analysis.  The State is convinced that the wrong inventory data
were used in the CES evaluations, and that EPA drew incorrect summary
information from the annual emission inventory data.

EPA Response:

EPA utilized monthly total emission county level estimates that were
aggregated to represent daily seasonal estimates for the cold and warm
seasons used in the CES.  The Agency believes that these estimates
provide a good basis for generally assessing which counties possibly
contribute to a nonattainment area.

Utah:

EPA ran and provided the results of backwards trajectories from the
National Oceanic and Atmospheric Administration’s (NOAA) Air Resource
Laboratory’s Hybrid Single Particle Lagrangian Integrated Trajectory
(HYSPLIT) model for selected high PM2.5 days (inversion conditions). 
Unfortunately, specific model run details were not included in EPA’s
response.  In order to properly interpret the results of the
trajectories, it is critical to know the following:

What meteorological data set was used?

What meteorological data set resolution was used?

What vertical motion scheme was used?

What trajectory start height was used?”

EPA Response:

EPA has already provided this information through the CES TSD.  The data
set used was the Eta Data Assimilation System (EDAS) data set from the
NOAA archives.  It has a 40 km resolution.  The vertical motion scheme
was the model default and four trajectory start heights were used as
detailed in the TSD.

Utah:

All of the available meteorological data sets on the HYSPLIT website
have horizontal resolutions far too coarse to accurately depict a
near-surface air parcel’s trajectory in a cold pool.  The coarse
resolution of the meteorological data does not capture the
topographically driven micro- and meso-scale features of the wind field.
 As mentioned in the pollution rose diagram section, topography controls
the wind during quiescent ridge dominated synoptic conditions.  For this
reason, the backwards trajectories produced by EPA must be removed as a
technical justification for any argument put forth.

EPA Response:

EPA concurs with much of the above statement, but believes that
“topography controls the wind during quiescent ridge dominated
synoptic conditions” is an overstatement of the case.  EPA’s
analysis of the surface wind data collected by the Utah Department of
Environmental Quality (DEQ) indicates that there are often relatively
uniform movements of air across the basin during inversion episodes,
which appear to overlay any micro- and meso-scale features of the wind
field caused by local topography.  This data set is completely
independent of the data used by NOAA to create the EDAS meteorology
files.  It represents an independent assessment of the HYSPLIT results. 
This regional motion derived from Utah DEQ data appears to be adequately
represented by the low speed regional movements seen in the HYSPLIT
analyses on inversion days.  Currently, the HYSPLIT results do appear to
have some value, and so do not need to be removed from consideration as
suggested by Utah.

To evaluate the appropriateness of conclusions reached using HYSPLIT
back trajectories, EPA obtained surface meteorological data collected by
Utah DEQ across their monitoring network from EPA’s Air Quality System
(AQS).  Using 1-hour average wind speed and direction for all of the DEQ
met stations (excluding Logan in the Cache Valley); the Agency ordered
the data from north to south.  The monitors used, in north to south
order are:

Brigham City

Harrisville

Washington Terrace

Antelope Island

Syracuse 

Badger Island 

Bountiful 

Salt Air 

North Salt Lake*

Beach*

Hawthorne

Magna

West Valley City

Cottonwood 

West Jordan 

Tooele City*

Herriman*

Highland 

Lindon

N.  Provo 

Spanish Fork



* Indicates monitor did not operate in all years in the period 2004
through 2008

EPA plotted hourly wind speed and direction from these monitors during
the periods used in the HYSPLIT trajectory analyses shown in Figure 2. 
EPA then color coded hourly average wind directions, where blue
indicated wind directions originating from the north (northwesterly,
northerly or northeasterly winds), and salmon represents times with
southerly winds (winds from the southwest, south or southeast).  Winds
from the east or west remained unshaded.

Rather than showing only winds controlled by local topography during
cold pool inversion patterns, the resulting diagrams show basin scale
uniformity in wind direction for much of the basin over much of the
inversion periods.  In addition, for specific monitoring days targeted
by the HYSPLIT back trajectories, the diagrams show wind directions over
the basin consistent with the trajectories generated by HYSPLIT.  Thus
the HYSPLIT results and conclusions are in fact consistent with
independent local wind measurements.

Figure 1 shows the three days used in the HYSPLIT trajectory ending at
midnight on January 13, 2004 (shown in Figure 2).  EPA used the 3-day
HYSPLIT trajectory ending on January 13, 2004 to illustrate a monitoring
day in which winds were blowing from the south, and emissions from Utah
County were seen in the HYSPLIT trajectory to move north into Salt Lake
County.  The third day of the diagram in Figure 1 shows southerly winds
across nearly the DEQ’s entire meteorology network for the morning
hours, and southerly winds in Utah County all day, consistent with the
HYSPLIT trajectory for the day.

 

Figure   SEQ Figure \* ARABIC  1 . Wind Pattern Diagram for January 11
through 13, 2004.

	

Figure   SEQ Figure \* ARABIC  2 .  HYSPLIT Back Trajectory for January
13, 2004, all Salt Lake County Monitors.

Figure 3 shows the meteorology data from the Utah Department of
Transportation (DOT) monitor located just north of Point of the Mountain
along I-15.  The monitor shows variable light winds all day, but with
frequent winds from the south through the Point of the Mountain gap
before noon.

 

Figure   SEQ Figure \* ARABIC  3 .  Meteorology Near the Utah
County/Salt Lake County Line, January 13, 2004.

The second wind pattern diagram in Figure 4 shows the three days used in
the HYSPLIT trajectory ending at midnight on January 22, 2004.  Figure 5
shows this three-day HYSPLIT trajectory to illustrate a monitoring day
in which winds were generally blowing from the north.  The third day of
the wind diagram in Figure 4 shows winds from the north across nearly
the DEQ’s entire meteorology network in the middle of the day,
consistent with the HYSPLIT trajectory.  The Bluffdale UDOT wind data
for the day is shown in Figure 6.  The monitor was not reporting prior
to 9:15 AM, at which time flow through Point of the Mountain was from
the south (downvalley from Utah County to Salt Lake County).  Consistent
with the Utah DEQ network and HYSPLIT, flow in the afternoon hours is
up-valley, flowing south from Salt Lake County into Utah County.

 

Figure   SEQ Figure \* ARABIC  4 .  Wind Pattern Diagram for January 20
through 22, 2004.

Figure   SEQ Figure \* ARABIC  5 .  24-hour Backtrajectory for January
22, 2004, All Salt Lake County Monitors.

 

Figure   SEQ Figure \* ARABIC  6 .  Meteorology Near the Utah
County/Salt Lake County Line, January 22, 2004.

The third day of HYSPLIT trajectories EPA utilized were those ending on
January 26, 2006 (figure 8).  This day was again a southerly wind
dominated day.  It was selected because it included more southwesterly
flow than other trajectories, indicating the potential for flow of
Tooele Valley emissions toward Salt Lake City and Ogden, but represents
a day with much more complex wind patterns in the HYSPLIT simulation. 
Again, as shown in figure 7, the measurements of the Utah DEQ
meteorology network are consistent with the HYSPLIT trajectory, showing
pronounced flow from the south across the basin prior to midnight on
January 25 through midday on January 26, 2006.  Figure 8 shows the
24-hour backtrajectory from HYSPLIT.

 

Figure   SEQ Figure \* ARABIC  7 .  Wind Pattern Diagram for January 24
through 26, 2006.

Figure   SEQ Figure \* ARABIC  8 .  24-hour Backtrajectory for January
26, 2006.

Wind data from the Bluffdale Utah DOT monitor for January 26, 2006 is
almost uniformly from the south.  While the Quality Assurance (QA)
status indicator for this data in the MesoWest database is “OK”, the
monitor seems to show a very high level of high winds from due south or
south-southeast during January of 2006, so the data appears to reflect a
bias in the instrument toward southerly winds.  Therefore, that data is
not used here.

With respect to the wind pattern diagrams in Figures 1, 4 and 7 and the
Utah DOT wind measurements at Point of the Mountain, to a greater or
lesser extent, all show some degree of diurnal winds across the Salt
Lake and Utah Lake basins; surface winds across the network are more
often southerly at night and in the morning, with northerly winds most
likely in the afternoon.  This would indicate that the two basins are
acting as a connected valley system.  Generalized basin level winds can
be seen well beyond any degree to which local topography is controlling
winds at any given monitor location.

Subsequent to 2004-2006 (the period over which HYSPLIT trajectories were
calculated for every PM2.5 exceedance day), UDOT meteorology data is
available from the Alpine station along I-15 just south of Point of the
Mountain.  The station location is shown in figure 9.  The Alpine
monitor, on the south entrance to the gap at Point of the Mountain
generally shows much more intense winds than does the UDOT Bluffdale
monitor north of Point of the Mountain.  As an example, figure 10 shows
the winds recorded by the Alpine monitor on January 21-30, 2007.  This
period included the highest PM2.5 days in both Utah County and in Salt
Lake County during 2007.  Winds throughout the period are dominated by
southerly (down valley) winds from Utah County to Salt Lake County. 
During the peak days in the period, some degree of up-valley flow occurs
in the middle of several days.  The dominance of downvalley flow at this
site is typical during the inversion periods since March of 2006,
indicating that consistent and significant flow is the usual pattern;
some lesser degree of northerly winds is also typical, indicting the
likelihood of 2-way transport.

 

Figure   SEQ Figure \* ARABIC  9 .  Utah DOT Alpine and Bluffdale
Meteorology Station Locations.

Figure   SEQ Figure \* ARABIC  10 .  Point of the Mountain Surface
Winds, January 21-30, 2007.

SECTION III:  Technical Analysis

State Comment:  2005 NEI Inadequate

Delaware:

EPA used the 2005 NEI as a basis for evaluating this factor.  The 2005
NEI is not a high quality inventory, as it was not developed in
conjunction with State agencies, nor has it been quality assured by
States.

EPA Response:

EPA believes the 2005 NEI Version 1 was the highest quality emission
inventory available during the 2006 24-hour PM2.5 designations process. 
Most of the point and non-point emissions in the 2005 NEI Version 1 were
taken directly from the 2002 NEI Version 3 which was developed in
conjunction with State agencies and underwent thorough QA, including QA
by State agencies.  New emission measurements and estimates were
included in the 2005 NEI Version 1 for Electric Generating Units (EGUs),
mobile sources, and fires because EPA had more recent data and/or models
for these sources (e.g., Continuous Emission Monitoring (CEM) data for
EGUs, new versions of the MOBILE model for mobile sources).  While it is
true that these new emissions data were not developed in conjunction
with State agencies, nor quality assured by States, EPA believes using
these data more accurately characterize emissions than the alternative
of using an older version of the NEI (e.g., 2002 NEI Versions 3) for
these sources.

State Comment:  Determination of contributing areas

Indiana:

Indiana firmly believes that due to the localized influence (urban
excess) of PM2.5, nonattainment boundaries for the 24-hour PM2.5
standard should be limited only to counties that possess a three-year
average ambient monitor-based design value above the standard.  However,
if the only violating monitor within the county is a source-oriented
site, the boundary should be limited to the township in which the
monitor resides.  This is consistent with U.S. EPA designations under
the PM10 standard.  

EPA Response:

CAA Section 107(d)(1)(A)(i) provides that nonattainment areas must
include “any area that does not meet (or that contributes to ambient
air quality in a nearby area that does not meet)” the applicable
NAAQS, therefore to limit nonattainment boundaries to only those
counties with monitored-based design values above the standard ignores
the role of contribution from nearby areas mandated by the Act.  In most
nonattainment areas, speciation data indicate significant contributions
from sulfur dioxide and nitrogen oxide sources throughout broader
metropolitan and contributing areas..  

With regard to source-oriented monitors, those in Indiana are in areas
that also have other monitors violating the standard (for example in
Illinois).  EPA has found that even source-oriented monitors reflect
substantial contribution from the metropolitan area and beyond.  EPA
does not find the comparison to PM10 appropriate because areas with high
PM2.5 concentrations are more affected by secondary formation.  PM10 and
PM2.5 are different particles, that are often formed from emissions from
different types of sources that may require different control
strategies.  PM2.5 can transport greater distances than PM10.  PM10
nonattainment area boundaries often date back nearly 20 years.

Indiana:

Recognizing that EPA is obligated to consider a county’s contribution
to a downwind monitored violation of the standard, sound evidence must
exist that demonstrates that a county actually contributes to the
downwind violation in order for the upwind county to be designated
nonattainment.  Such evidence should not only be supported by a
culpability demonstration, the culpability of the upwind county, or
portion thereof, must be significant.  Counties that measure air quality
below the standard, and are not proven to be significantly culpable for
a downwind violation, should be designated attainment.  Counties for
which monitoring data does not exist, and are not proven to be
significantly culpable for a downwind violation, should be considered
unclassifiable and designated attainment.

EPA Response:

EPA appreciates Indiana’s acknowledgement that EPA must consider the
contribution of areas as part of its decision-making on nonattainment
area boundaries.  However, Indiana suggests application of a burden of
proof that is not required under section 107 of the Act which is
inconsistent with the role of designations as distinct from attainment
planning.  In many respects, the role of designations is to define the
planning area for the State to use in planning for achieving air quality
standards.  While EPA must make judgments of nonattainment area
boundaries based on the best available evidence, the expectation is that
the promulgation of designations will trigger more thorough analyses of
impacts and potential controls of sources as part of the process of
formulating and adopting an attainment plan.  Indiana seems to seek a
requirement for analyses of impacts of sources typical of an attainment
plan analysis in order to determine what area should be subject to such
an attainment planning requirement.  Such an approach would be
inconsistent with the structure of the Act.  Furthermore, Indiana in
effect seeks for the Agency to be bound to designate potentially
contributing areas as attainment even if the best available evidence
indicates that the area contributes to the violations, merely because
the available evidence is insufficient to meet the burden of proof that
Indiana seeks for the Agency to apply.  Again, such an approach would be
inconsistent with the mandates and structure of the Act.  The Agency
construes “contribution” to allow a case-by-case approach in
determining what level of contribution warrants inclusion in a
nonattainment area.  Each area has unique facts which must be analyzed. 
Rather than adopting specific materiality requirements or bright-line
tests, the Agency reasonably developed an analytical approach to
evaluate whether emissions in a particular area contribute to violations
in another nearby area.  Nothing in Section 107(d) requires the
culpability demonstration suggested by Indiana.  The definition of
“nonattainment” specifically includes any area that “contributes
to” violations of NAAQS in a nearby area, not only those that
literally cause such violations.  Ambient PM2.5 at every monitor
reflects the cumulative impacts of many types of emissions from many
sources, near and far, that result in primary and secondary formation of
particles.  Imposing a culpability analysis would frustrate the goal of
the statute by narrowing its application.  The Agency’s analytical
approach was designed to evaluate contribution in a qualitative way by
considering the facts and circumstances of each area.

Indiana:

Since there are multiple species that comprise PM2.5, and the portion of
specie contribution varies by season and geographic location, relying
solely on precursor emissions data weighted evenly by species is
inappropriate and does not constitute a culpability demonstration. 
Additionally, assuming that mobile source contributions driven by
population density substantiate culpability for a collar county is
erroneous and contradicted by speciation and source apportionment data
available to EPA.

EPA Response:

This comment misrepresents EPA’s approach.  The Agency applied
different weights to the emissions of different pollutants that directly
or indirectly contribute to PM2.5 concentrations, explicitly in
computation of its contributing emission score and also in its
examination of underlying information.  The contributing emission scores
were computed by first estimating the urban increment contributions of
each major PM2.5 component, estimated separately for each metropolitan
area and separately for warm versus cold months.  The relative
percentages for each component were then used as weighting factors to
give the highest score to counties that had high emissions of the most
important pollutants.  Since this approach appears to match the approach
that Indiana recommends, it appears that Indiana endorses the
contributing emission score as a tool for judging relative contributions
of different counties.

Indiana also appears to misrepresent the approach EPA took regarding
mobile source emissions.  In general, one of the most important
indicators of contribution by a county is the emissions within that
county.  Commonly, these data strongly indicate contribution of relevant
counties, and information regarding the location of mobile sources and
their emissions had little influence on EPA’s judgment.  Nevertheless,
since some mobile source control programs are applied based on the home
station of the vehicle, one factor in judging the contribution of a
county is the number of vehicles that are housed in the county.  For
example, many vehicles registered in Lake County, Indiana, incur much of
their emissions in Cook County, Illinois; these vehicles represent Lake
County sources that are contributing to air quality problems in Cook
County.  Mobile sources are important contributors to PM2.5
concentrations, and population data can be an important indicator of the
home station of these sources.  Mobile source use in an area, as
reflected in information like the number of drivers and vehicle miles
traveled (VMT) in an area, is relevant to determining whether an area
contributes to violations in nearby areas.  Similarly, the number of
commuters in a county who drive to another county, the percent of total
commuters in each county who commute to other counties, and their
destinations is important to the analysis.  Information about commuting
patterns may indicate where emissions from mobile sources might actually
be occurring, and therefore the degree to which drivers based in one
area are contributing emissions to another area.

Indiana:

In its October 2, 2008 letter, Indiana included an evaluation of fine
particulate matter nonattainment area boundaries.  Indiana provided this
information in Attachment D of its letter in response to the Agency’s
August 2008 recommended designations.

EPA Response:

Indiana focused on information it believes justifies excluding selected
counties or partial counties from the respective nonattainment areas. 
Nevertheless, the information that Indiana provided in Attachment D is
similar to information that EPA has reviewed.  Therefore, this
information does not justify a change in the boundaries of the pertinent
nonattainment areas.

State Comment:  Relevance of jurisdictional boundaries and consistent
boundaries for PM2.5, PM10, and other nonattainment areas

California:

One of U.S. EPA’s goals in designating nonattainment areas in
California was to achieve a degree of consistency with existing ozone
and PM10 nonattainment areas.  Application of this goal in California
led to differences between the State’s recommended nonattainment areas
and EPA’s proposed designations.  EPA expanded many of the State’s
recommended PM2.5 nonattainment areas boundaries to match 8-hour ozone
nonattainment area boundaries.  However, we do note areas throughout the
country where EPA proposed PM2.5 nonattainment area designations are not
consistent with existing 8-hour ozone nonattainment area boundaries. 
Examples are shown in Table 2 on page 3 of the California comment
letter.

Some of the areas in Table 2 were excluded based on the nature of the
pollutant.  PM2.5 is comprised of both primary and secondary components;
the primary being more localized.  The California Air Resources Board
(CARB) requests that EPA recognize the technical basis for different
boundaries for regional ozone and localized PM2.5.

EPA Response:

EPA looked at the violating monitors and sources in nearby areas
potentially contributing to the violations.  As described above,
residential woodburning and transportation were identified as the
principal sources based on speciation data and, therefore, nonattainment
areas were intended to capture population centers, major transportation
corridors, and stationary sources in the area.  The Agency revised the
eastern boundary of the Sacramento, Sutter/Yuba and Butte nonattainment
areas.  Rather than extending the boundary to the ridgeline of the
Sierra Nevada Mountains or the county line, which was consistent with
the ozone nonattainment boundaries where they existed, the eastern
boundaries were redefined based on the topography of the Sierra Nevada
Mountains and the height of the winter time inversion layer. 
Considering this physical boundary, which also resembles the California
Air Basin boundary for the Sacramento Valley, EPA defined the eastern
boundaries with township and range lines to capture this topographical
boundary, as well as the major population centers and transportation
networks.

Delaware:

EPA indicated that New Castle County historically has been part of the
Philadelphia nonattainment area for ozone and PM2.5, and that Delaware,
Pennsylvania, and New Jersey have a long history of working
cooperatively with ozone and particulate matter (PM) attainment
planning.  This statement by EPA is not in the proper context of factor
8.  While these States do work cooperatively together, no cooperative
air planning effort to date has occurred as a result of being in a
common nonattainment area.  These States have historically worked
together only as part of larger efforts, like the Mid-Atlantic/Northeast
Visibility Union (MANE-VU) and the Ozone Transport Commission (OTC). 
Not a single control measure has been developed as a cooperative effort
amongst these States outside of MANE-VU and OTC context.

EPA Response:

Delaware has conceded that it has worked with the other States in the
Philadelphia area as part of MANE-VU and the OTC.  Therefore, as stated
in its August 18, 2008 letter, EPA does not anticipate that including
New Castle County as part of the Philadelphia nonattainment area for the
2006 PM2.5 NAAQS will be an undue burden on Delaware.

Delaware:

In addition, no CSA scale efforts are necessary relative to planning for
the 2006 PM2.5 standard.  This is because the New Castle County
non-attainment problem is separate and distinct from the Philadelphia
problem, and the only commonality is transport, which is regional, not
local, in nature.  Factor 8 supports New Castle County as a separate
non-attainment area from the rest of the Philadelphia CSA, and the
continuation of larger regional efforts to develop control strategies
and address transport.

EPA Response:

EPA has determined that emissions from New Castle County contribute to
downwind fine particulate concentrations in the Philadelphia CSA.  New
Castle County’s SO2 emissions are the highest of all the counties in
the Philadelphia-Wilmington area.  The predominant wind direction in the
Philadelphia-Wilmington area is from the southwest.  The violating
monitors in the Philadelphia-Wilmington area, in Delaware, Chester,
Camden, and Philadelphia Counties, are downwind of both northern and
southern New Castle County.  High PM2.5 days at the Chester County
monitor are from the south-southwest, southeast, and south-southeast. 
Winds from the southeast and south-southeast would pass through New
Castle County.  Therefore, CSA-scale planning is necessary.

Indiana:

When designating areas under the annual PM2.5 standard, EPA relied on
guidance and criteria established for the 1-hour ozone and carbon
monoxide NAAQS.  Section 107(d)(4)(A)(iv) is the only citation of the
CAA that references a Metropolitan or Consolidated Metropolitan
Statistical Area (MSA or CMSA) boundary as the presumptive boundary for
a nonattainment area.  However, not only is this citation limited to
ozone and carbon monoxide areas, it is limited to areas classified as
Serious, Severe, or Extreme under Subpart 2 of the CAA.  EPA relied on
Subpart 1 of the CAA when it designated areas under the annual PM2.5
standard and no areas were classified.  Section 107(d)(4)(B) of the CAA
defines a nonattainment area for PM10 based on a violation of the
standard, suggesting that the boundary should be limited to the
jurisdiction where the violation occurs.

EPA Response:

Contrary to the commenter’s claim, EPA did not rely on guidance and
criteria established for the 1-hour ozone and carbon monoxide NAAQS in
promulgating PM2.5 designations.  Instead, EPA issued separate guidance
for designations for PM2.5 and conducted PM2.5-specific evaluations
(including use of a weighted emission score that was not used for ozone)
to determine appropriate boundaries for PM2.5 nonattainment areas.  As
explained in EPA’s guidance, PM2.5 arises from a combination of
distant sources and urban scale sources (including sources very nearby
and sources elsewhere in metropolitan areas).  Ozone also arises from
both distant sources and urban scale sources, and EPA’s guidance on
PM2.5 designations notes the similarity of approaches for designating
the two pollutants.   This similarity in approaches does not mean that
EPA is simply applying criteria contained in section 107(d)(4)(A)(iv)
(applicable to ozone designations for areas with specified
classifications) in promulgating PM2.5 designations.  To the contrary,
EPA acknowledged in its guidance for the 2006 PM2.5 designations that
when determining boundaries in urban areas for the 1997 annual PM2.5
NAAQS, it had applied a presumption that they should be based on
Metropolitan Statistical Area (MSA) or Consolidated MSA (CMSA)
boundaries.  However, we specifically stated that “[f]or the PM2.5
24-hour NAAQS, EPA is establishing no such presumption.”  The Agency
did note that it anticipated that the same boundaries established for
the annual PM2.5 standard, where that presumption might have been
applied, may also be appropriate for implementing the 24-hour PM2.5
NAAQS because the same sources of emissions may also be contributing to
violations of the 24-hour standard based on emissions and air quality
data and meteorology.  

EPA continues to believe that C/MSA boundaries may be relevant and
appropriate to consider in making PM 2.5 designations because of the
distances that PM2.5 and its precursors can be transported.  Therefore,
“nearby areas” for PM2.5 are not only those in the immediate
vicinity of a violating monitor.  Like ozone, a significant fraction of
PM2.5 particles results from secondary formation of particles in the
atmosphere due to emissions of different precursor chemicals from a
variety of nearby and distant sources.  EPA’s analysis indicated that
the local component of particles at a violating monitor were likely to
result from a combination of sources throughout the urban area.  Thus,
it was logical to analyze C/MSAs as potential boundaries to ensure
inclusion of relevant sources in evaluating nonattainment area
boundaries.  

When Congress enlarged the presumptive boundaries for ozone and carbon
monoxide nonattainment areas to the C/MSAs, EPA had already designated
areas for ozone and carbon monoxide.  Section 107(d)(4)(A) explicitly
required that those already designated areas be enlarged to full C/MSA
boundaries by operation of law.  By making this change, Congress
presumably recognized that artificially small nonattainment areas could
exclude sources that should be subject to SIP requirements.  While
Section 107(d)(4)(A) does not directly apply to PM2.5 designations, it
provides a logical basis for EPA to suggest a comparable analysis for
PM2.5, which, like ozone, is a pollutant that results from a broad range
of urban sources.  If Congress intended to prohibit States or EPA from
using C/MSAs as appropriate geographic starting points for evaluating
potential contributions to PM2.5 violations, Congress could have done
so.  Silence in the CAA on the subject does not prohibit using C/MSAs as
potential boundaries, but rather leaves to EPA the task of determining
how to draw appropriate nonattainment boundaries.  Thus EPA considered
it appropriate to analyze counties within the C/MSA boundary as well as
a ring of counties outside the C/MSA to determine potential contribution
and therefore inclusion in the nonattainment area. Additionally, the
fact that EPA declined to classify areas for PM2.5 when designating
areas under the annual PM2.5 standard and relied in its implementation
rule instead on Subpart 1 (CAA Section 171 – 179B) has no bearing on
whether EPA has the discretion to consider any potential group of
counties, such as those within the C/MSAs, for analysis purposes. 
Further, EPA is not promulgating the 2006 PM2.5 designations under
Section 107(d)(4)(B) but instead under Section 107(d)(1).  Therefore,
the provisions contained in Section 107(d)(4)(B) do not limit EPA’s
authority to designate nearby contributing areas, otherwise mandated by
the CAA, nor otherwise limit its analysis of potentially contributing
counties.

Indiana:

Both the annual and 24-hour PM2.5 standards are a revised version of the
PM standards.  Therefore, Indiana firmly believes that PM2.5
designations should be consistent with designations under the PM10
standard.  Since U.S. EPA is relying on the nonattainment boundaries for
the annual PM2.5 standard to serve as the presumptive boundaries for the
24-hour PM2.5 standard, and the presumptive boundaries would be more
representative of ozone as the pollutant, U.S. EPA must reconsider the
criteria and presumptive boundaries prior to proceeding with final
designations under the 24-hour PM2.5 standard.  

EPA Response:

As explained above, EPA did not apply any presumption that annual
nonattainment area boundaries would serve as the presumptive boundaries
for the 24-hour standard, only that we anticipated that the annual PM2.5
boundaries, which were based upon a number of relevant factors, might
also be appropriate for the 24-hour PM2.5 boundaries.  EPA is deciding
the boundary of each 24-hour PM2.5 nonattainment area on its own merits.
 EPA’s decision to exclude Dubois County from the Evansville 24-hour
PM2.5 nonattainment area, in contrast to its inclusion in the annual
PM2.5 nonattainment area, illustrates the fact that EPA is making
independent decisions on the boundaries of nonattainment areas for the
2006 air quality standards based on current information about potential
contributions to violating monitors.  

Nevertheless, in most cases, areas that contribute to violations of the
annual standard also contribute to violations of the 24-hour standard. 
For both standards, violations in the eastern United States generally
reflect significant quantities of sulfate, nitrate, and organic
particulate matter.  These violations result from a range of sources,
including sources at great distances (most notably power plants,
yielding regionally distributed sulfate and nitrate concentrations),
sources within a few kilometers, and sources emitting either particles
or precursors or both within the broader metropolitan area. 
Consequently, the requirement of the Act, dictating that nonattainment
areas shall include nearby contributing areas as well as the area
actually observing a violation, is to include much if not all of the
metropolitan area in the nonattainment areas for both standards.  It was
thus appropriate for EPA to analyze counties within and just outside of
C/MSA boundaries to determine whether such areas were contributing to
monitored violations.

Preexisting PM10 nonattainment boundaries generally have little, if any,
relevance to what boundaries would be appropriate for PM2.5.  The PM10
NAAQS involves a different standard, with different indicator particles,
different concentrations levels, resulting from different sources that
may require different control strategies.  These boundaries were
generally promulgated almost 20 years ago, for a completely different
NAAQS. 

Indiana

Section 107(d)(4)(B) of the CAA defines a nonattainment area for PM10
based on a violation of the standard, suggesting that the boundary
should be limited to the jurisdiction where the violation occurs.  Both
the annual and 24-hour PM2.5 standards are a revised version of the PM
standards.  Therefore, Indiana firmly believes that PM2.5 designations
should be consistent with designations under the PM10 standard.  

EPA Response:

Section 107(d)(4)(B) explicitly applies to designations for PM10, not to
designations for any other form or constituent of particulate matter. 
Therefore, this section does not govern designations for PM2.5.  Given
that most designations for PM10, including all the designations for PM10
in Indiana, took effect pursuant to section 107(d)(4)(B)(i) (reflective
of pre-existing Group I areas), not section 107(d)(4)(B)(ii), and given
that no Group I areas exist for PM2.5, the Act cannot reasonably be
interpreted to provide for promulgating PM2.5 designations pursuant to
section 107(d)(4)(B).  

The comment regarding designations for PM10 is not germane and
misinterprets this provision of the Act.  Section 107(d)(4)(B) applies
explicitly to designations for PM10, and the Act does not provide for
this section also to apply to other pollutants such as PM2.5.  In
addition, section 107(d)(4)(B)(ii) need not be interpreted to dictate a
different legal structure than section 107(d)(1) (“Designations
generally”).  For PM10, just as for PM2.5, the planning areas that
became the nonattainment areas included the violating areas but also
included contributing areas.  For example, the nonattainment area in
Northwest Indiana included the entirety of four cities in Lake County,
notwithstanding the fact that monitoring data indicated violations in
only a portion of that area.  Although in Indiana the PM10 designations
were promulgated under subsection (i) rather than subsection (ii) of
Section 107(d)(4)(B), EPA promulgated PM10 designations elsewhere that
reflected its view that PM10 nonattainment areas are to include
contributing areas as well as violating areas.  The difference between
the size of most PM10 nonattainment areas and the size of most PM2.5
nonattainment areas thus reflects differences in the typical size of the
contributing area and not to any legal differences as to whether
contributing areas are to be included.  Thus, PM10 which consists of
larger particles that more rapidly settle to the ground will generally
be caused by sources closer to a violating monitor, while PM 2.5 which
consists of smaller particles that can be transported greater distances
will more likely see contributions from throughout a metropolitan area. 
In fact, the distribution of PM2.5 bears more similarity to the
distribution of ozone than it does to the distribution of PM10, which is
the primary reason that the PM2.5 nonattainment areas generally resemble
the size of ozone nonattainment areas more than they do the PM10
nonattainment areas.

Iowa:

The use of arbitrary political boundaries to delineate the extent of the
proposed nonattainment boundaries is not consistent with EPA’s
guidance.  In the June 8th, 2007, “Area Designations for the Revised
24-Hour Fine Particulate National Ambient Air Quality Standards”
guidance document, EPA stated that the Metropolitan Statistical Area
presumptive boundary for areas violating the annual standard would not
apply to areas violating only the 24-hour standard.  In effect, no
presumptive boundaries are to be assumed for any nonattainment area in
Scott or Muscatine Counties.  The EPA has given no technical
consideration to sub-county boundaries.  All technical analyses
completed by EPA, for example the use of the CES analysis, have relied
upon data aggregated to the county level.  In effect, this approach has
elevated the importance of the jurisdictional/political boundary factor
above the other eight factors.  This gives rise to the conceptual model
of which sources and what conditions lead to PM2.5 formation in a given
area.

EPA Response:

EPA emphasizes that it did not rely on boundary presumptions in its
120-day letter to Iowa, nor did EPA use it to consider the
jurisdictional boundaries factor to override other important factors to
be more significant than others.  To the contrary, the 120-day letter
included EPA’s application of all nine factors to the area in stating
its intent to establish the county boundaries.  The letter also stated
that EPA was considering other boundaries, including a partial county
boundary, and sought information on whether different boundaries would
be appropriate.  Finally, and most significantly, EPA’s promulgation
does establish a partial county boundary for both areas, based on a
case-specific application of the nine factors and information provided
by the state.  However, EPA does not agree that the relatively small
area covered by the partial city boundaries recommended by Iowa is
supported by the available information.

West Virginia:

EPA’s analysis correctly identifies relevant State and county
jurisdictions.  But that analysis fails to provide any justification
whatsoever for adding part of Mason County to the nonattainment area. 
Rather, the text notes that the ozone nonattainment area includes Cabell
and Wayne County (WV) and Boyd County (KY).  It is puzzling that EPA
ostensibly performed a similar 9-factor analysis for both ozone and
PM2.5 and reached different conclusions, even though NOX is a common
precursor.  It is also troubling that, prior to the 2005 annual PM2.5
designations, EPA had generally based nonattainment boundaries on county
and Metropolitan Statistical Area boundaries rather than arbitrarily
adding “islands” such as the portion of Mason County.  EPA provides
no support for how such an isolated tax district bears any relevant
jurisdictional relationship to the remainder of the nonattainment area. 
Moreover, its inclusion within the nonattainment area presents an
additional and unnecessary burden on KYOVA because it must be addressed
under transportation conformity requirements even though no air quality
benefits result.

EPA Response:

Mason County is not part of KYOVA, the metropolitan planning
organization for Cabell and Wayne Counties and the City of Huntington in
West Virginia, and Lawrence County and the City of Ironton in Ohio. 
However, EPA believes that including a portion of Mason County in the
Huntington-Ashland nonattainment area will not pose an undue burden on
KYOVA.  In fact, under the 1997 PM2.5 NAAQS, KYOVA has successfully
completed its transportation planning with the inclusion of the Graham
Tax District in Mason County.

EPA included a portion of Mason County in the Huntington-Ashland area
because it determined that emissions from two large sources in that tax
district currently contribute to violations of the 2006 24-hour PM2.5
NAAQS in the Huntington-Ashland area.  Mason County, which is adjacent
to Cabell County, has two large electric generating units. 
Meteorological data shows that on one-third of the days in 2004-2006
with monitored 24-hour PM2.5 values greater than 35 and 40 µg/m3 in
Cabell County, winds came from the north-northeast.  This data suggests
that emissions from the north-northeast, i.e., from Mason County, WV
likely contribute to the 2006 24-hour PM2.5 NAAQS violations.   However,
since Mason County did not rank highly for other factors EPA concluded
that including only the Graham Tax District in Mason County in the
nonattainment area would capture the bulk of emissions in that county
that are contributing to the monitored violations.

West Virginia:

EPA identifies the State boundary between Ohio and West Virginia as the
major jurisdictional boundary.  Further, the Agency encourages the two
States to work collaboratively to reduce emissions.  The West Virginia
Department of Environmental Protection (DEP) agrees that cooperation
will facilitate meeting the air quality standard.  EPA goes on to note
that the former 8-hour ozone nonattainment area included Wood County
(WV) and Washington County (OH).  However, EPA provides absolutely no
support for the inclusion of any portion of Pleasants County based upon
jurisdictional considerations.  DEP continues to pursue litigation
concerning EPA’s inclusion of the Grant Tax District (Pleasants, WV)
in the PM2.5 annual standard designations.  We strongly believe that EPA
has inappropriately proposed inclusion of this district in the 24-hour
PM2.5 Parkesburg-Marietta nonattainment area.  DEP encourages EPA to
reconsider and exclude the entire county of Pleasants from the
nonattainment designation.

Pleasants County is not part of the Parkersburg MSA.  Therefore, it is
not within the jurisdiction of the Wood-Washington-Wirt Interstate
Planning Commission (WWW), the transportation planning organization. 
Moreover, its inclusion within the nonattainment area presents an
additional and unnecessary burden on WWW because it must be addressed
under transportation conformity requirements even though no air quality
benefits result.

EPA Response:

Pleasants County is part of the Parkersburg-Marietta-Vienna MSA, which
also includes Wood and Wirt Counties in West Virginia and Washington
County in Ohio.  However, Pleasants County is not part of the WWW, which
is the metropolitan planning organization for Wood County, West Virginia
and portions of Washington County, Ohio.  EPA believes that including a
portion of Pleasants County in the Huntington-Ashland nonattainment area
will not pose an undue burden on KYOVA.  In fact, under the 1997 PM2.5
NAAQS, WWW has successfully completed its transportation planning with
the inclusion of the Grant Tax District in Pleasants County.

EPA included the Grant Tax District portion of Pleasants County in the
Parkersburg-Marietta nonattainment area because emissions from three
large sources in the Pleasants County, just over the border from Wood
County, contribute to violations of the 2006 24-hour PM2.5 NAAQS in the
Parkersburg-Marietta area.  These sources are the Pleasants Power
Station, the Willow Island Power Station, and the Cabot Corporation Ohio
River Plant.  Due to these large sources, Pleasants County has the
highest emissions in the West Virginia portion of the
Parkersburg-Marietta-Vienna MSA.  Its SO2 emissions are nearly 10 times
higher than those of Wood County, which contains the violating monitor. 
Furthermore, Pleasants County’s NOX emissions are more than twice
those of Wood County and its total PM2.5 emissions are close to two
times higher than Wood County’s.  Therefore, EPA determined that it is
appropriate to only include the portion of Pleasants County containing
these sources, namely the Grant Tax District, in the
Parkersburg-Marietta nonattainment area.

State Comment:  Consideration of topography

Kentucky:

The response for the Geography/topography factor was inconsistent
throughout the entire response.  EPA determined in each area that there
were “no geographical or topographical barriers significantly limiting
air pollution transport within its air shed.” However, despite the
common factor, the answer alternated between each factor as to whether
or not the absence of topographical and geographical barriers
“contributed to violations in the area” or “did not play a
significant role in the decision-making process.” The lack of a
geographical/topographical barrier should either benefit an area or
cause a problem.

EPA Response:

The geography/topography analysis evaluated the physical features of the
land that might have an effect on the air shed and, therefore, on the
distribution of PM2.5 over the area in question.  The nonattainment
areas in Kentucky do not have any geographical or topographical barriers
significantly limiting air-pollution transport within its air shed. 
Therefore, this factor did not play a significant role in the
decision-making process.  By “contributed to violations in the
area,” EPA, in this case, was conveying that no topographical features
interfered with or cut off the transport of relevant pollutants from
nearby emissions sources.  Kentucky is correct, however, that the
absence of topographical features impeding contribution can support the
conclusion that such contribution occurs.

State Comment:  Consideration of Counties and Time Intervals for
9-Factor Analysis (KY)

EPA used a list of counties including some that were outside of the MSA
for certain factor analyses, and would exclude these same counties in
different factor analyses.  It seems inconsistent to make comparisons
with one population of statistical data, and then continue comparisons
with a different population of statistical data.  This comment also
applies where EPA used certain time intervals for one factor analysis
and a different time interval in another factor analysis.

EPA Response:

For the CES methodology, EPA ran calculations on counties extending out
a large distance from each violating monitor.  The CES tables shown in
EPA’s intended designation letters may not include all counties
considered in the 9-factor analysis, but they include those counties
having factors which indicate they should be candidates for a
nonattainment status.  In evaluating each of the nine factors for each
potential nonattainment area, our intended designation letters showed
the counties for which the Agency intended to designate nonattainment
for the 2006 24-hour PM2.5 NAAQS, and in many cases, other surrounding
counties where a comparison seemed relevant.  EPA’s exclusion of
particular counties from a particular factor data table does not mean
those counties were excluded from our analysis, only that we considered
those counties less relevant for that factor.  With regard to the time
intervals considered for each factor, note that the Agency utilized the
latest available data sets for our analyses.

State Comment:  Documentation of VMTs

Regarding the VMTs that EPA has utilized in their analysis, Kentucky has
been unable to document the basis for the development of this data.  EPA
has provided the spreadsheet containing the VMT numbers used in the
analysis, but to date has been unable to provide Kentucky with the
supporting documentation in the development of these numbers.  Since the
numbers used in Kentucky’s previous recommendation submittals were
developed by area agencies responsible for transportation conformity, we
must insist that the numbers in our previous submittals are more
reflective of the transportation sector in these areas.

EPA Response:

EPA consistently used county-level population and VMT growth estimates
for all areas of the country.  For 2005 a full VMT database at the
county, roadway type, and vehicle type level of detail was developed
from Federal Highway Administration (FHWA) information.  For States and
local areas that submitted VMT data that were incorporated in the 2002
NEI, the 2002 NEI VMT data were grown to 2005 using growth factors
developed from the FHWA data, and these grown VMT data replaced the
baseline FHWA-based VMT data.  The resulting VMT database prepared for
2005 include data for all 50 States, the District of Columbia, Puerto
Rico, and the Virgin Islands for each of the 12 Highway Performance
Modeling System (HPMS) functional roadway types and the 28 MOBILE6
vehicle classes, for a total of 336 records per year per county.  The
data were prepared in the NMIM National County Database BaseYear VMT
table format.  At this point, States had opportunity to submit 2005
estimates that replaced EPA's estimates for 2005.

SECTION IV:  Power Plant Issues

State Comment:  Power plant counties

Several States believe that EPA should reconsider its determination that
counties and/or townships with significant power plants should be
designated as nonattainment based on the cause or contribute rationale. 
The existence of an electric generating unit in a township should not be
the sole factor in determining nonattainment.  At a minimum, EPA should
consider the size of the power plant and the emission controls it has
implemented.

EPA Response:

The fundamental statutory requirement governing the designations process
is section 107(d) of the Act, which requires the EPA Administrator (or
the Administrator) to designate each area that violates the NAAQS and
any nearby area that “contributes” to the violation.  In the PM2.5
designations process, EPA used 2005 county-level data from the National
Emissions Inventory in evaluating potential contributing counties for
area-specific technical analyses.  These data reflected emissions and
control strategies implemented by sources in the State through 2005.  

In its August 2008 letters responding to State recommendations, EPA
recognized that certain power plants or large sources of emissions in a
potential nonattainment area may have installed emission controls or
otherwise significantly reduced emissions since 2005 and that this
information may not be reflected in the 2005 data.  EPA noted that it
would consider additional information on emission controls in making
final designation decisions.  In cases where specific plants have
installed emission controls since 2005 or plan to install such controls
in the near future, EPA requested additional information on:

the plant name, city, county, and township/tax district

identification of emission units at the plant, fuel use, and megawatt
capacity

identification of emission units on which controls will be installed,
and units on which controls will not be installed

identification of the type of emission control that has been or will be
installed on each unit, the date on which the control device became /
will become operational, and the emission reduction efficiency of the
control device

the estimated pollutant emissions for each unit before and after
implementation of emission controls

whether the requirement to operate the emission control device will be
federally enforceable by December 2008, and the instrument by which
federal enforceability will be ensured (e.g.  through source-specific
State Implementation Plan (SIP) revision, operating permit requirement,
consent decree)

Some commenters claimed that EPA intends to include partial county areas
simply due to the existence of a power plant in the area.  EPA did not
include these areas simply due to the existence of a power plant, but as
a result of a reasoned evaluation of information indicating the likely
contribution by the source to PM2.5 concentrations at the nearby
violating monitor.  This information included emissions data, ambient
air quality speciation monitoring data, meteorological data, information
about current and future planned emission controls, and related
geographic information.  

A number of States provided comments to EPA with specific additional
information on power plants, and they requested that EPA remove such
facilities from the list of areas intended for inclusion as
nonattainment.  Some commenters noted that certain facilities have plans
for emission controls to be installed and operational on a date later
than the December 2008 date on which the designations were finalized,
and they have requested that EPA consider whether a facility contributes
to a violation based on projected lower future emissions levels after
future installation of controls.  EPA interprets section 107(d) as
requiring the designation of areas which currently (i.e.  through 2008)
emit pollutants contributing to violations of the 24-hour standard. 
Therefore, in making decisions about contributing areas, EPA did not
take into consideration emission controls that would be implemented
after the December 2008 date for the Agency’s final action on
designations.  EPA recognizes that implementation of future emission
controls will provide important improvements in air quality and they
should be included in the State implementation plans for these
designated nonattainment areas.  These emission reductions should be an
important element of the State plan for attaining the PM2.5 standards,

In the event that emission controls have been installed and are
operational at a facility prior to December 2008, or the facility
significantly reduced emissions due to another reason (e.g., the shut
down of certain units at the plant), EPA considered two related issues. 
One issue is whether the emission control requirement on the source is
federally-enforceable.  Generally, a requirement is considered to be
federally-enforceable if 1) it has been adopted by the State and
approved by EPA as part of the SIP; 2) it has been approved as part of a
federally-enforceable Title V operating permit or new source
review/prevention of significant deterioration (PSD) permit; or 3) it is
required as part of a consent decree, and the consent decree includes a
provision for the emission reduction requirement to be included in the
SIP or federally-enforceable permit upon fulfillment of the consent
decree.  EPA has determined that only federally-enforceable emission
controls in place prior to December 2008 should be considered in this
process.

A second issue is whether, after the implementation of emission controls
prior to December 2008, the facility continues to emit pollutants at
levels which still contribute to fine particle concentrations at a
nearby violating monitor.  For example, certain large facilities may
continue to emit several thousand tons or more per year of SO2 or NOX
even with operation of emission controls.  To assess contributions from
large sources, EPA considered a number of factors, including emissions
data, meteorological information, and ambient air quality data.  Each
area was considered based on the specific facts for that situation.  EPA
does not believe, however, that the presence of an emission control
device alone should be used as a basis to exclude a source that is
otherwise deemed to be contributing to the violations.  In the
Agency’s final action, a number of partial county areas with large
power plants were included in nonattainment areas based on the
Administrator’s judgment that such areas were contributing to nearby
violations.  

SECTION V:  Consideration of Other Regulations

State Comment:  Consideration of CAIR-related emissions reductions

Kentucky:

Although at this point in time it is uncertain what reductions can be
attributed to CAIR or its replacement, it is certain that some manner of
control similar to CAIR will be implemented.  This will further reduce
these emissions since the Selective Catalytic Reduction (SCR) should be
operated year-round instead of only during the ozone season. 
Kentucky’s NOX SIP Call regulations remain in effect and, if CAIR
continues to be in limbo, it will cover the 2009 allocation timeframe. 
In addition, non-EGUs in Kentucky will also be required to put on Best
Available Retrofit Technology (BART) controls, which will further
achieve PM air quality improvements.  To designate counties
nonattainment because they have a power plant in them would place
additional hardships on the county and would be counterproductive since
the EGUs in the entire region will be mandated by EPA’s CAIR rule to
significantly control their PM precursor emissions without being
designated nonattainment.

EPA Response:

As required by section 107 of the CAA, EPA determined violations of the
2006 24-hour fine particulate NAAQS based solely on the most recent
three complete years of certified monitoring data.  EPA set boundaries
based on available data and analysis that best represents present-day
conditions, and not on projected emissions reductions that may occur
after area designations are finalized.  Section 107 requires EPA to
designate all areas within three years of promulgation of a new NAAQS
and to designate as nonattainment any area that violates or contributes
to an area that violates a NAAQS.  The statute is written in the present
tense and thus EPA believes it can only base designations on controls
that are in place and federally enforceable by the effective date of
final designations.

States may account for other projected emissions reductions in their
State SIPs for achieving attainment, where appropriate.  States will
also be able to take advantage of any air quality benefits that may have
already occurred from recently enacted rules by submitting the 2008
monitoring data to EPA by February 20, 2009.  EPA will review 2008
monitoring data submitted by this date, and will modify designation
areas if appropriate prior to the effective date of the final
designations.  EPA notes that compliance with CAIR (or any other rule
based upon section 110(a)(2)(D) or section 126) does not negate the
obligation of a source to comply with nonattainment area requirements. 
If a source is contributes to local nonattainment, it may well need to
be evaluated for additional controls necessary for local attainment
needs, as well as reductions to eliminate interstate transport.

State Comment:  Anticipated reductions in PM2.5 levels due to recent
low-sulfur gasoline and diesel rules

Kentucky:

In many areas, EPA based potential nonattainment designations on the
supposition that population, commuter traffic, or VMT played an
important role in determining potential impacts on PM2.5 levels within
an MSA.  It is not feasible to designate a county as nonattainment if
the only reason an area has been included was due to these
population-based factors.  With national controls being implemented that
would address this contribution, including these counties as
nonattainment would place additional, burdensome planning requirements
on these local areas for no useful purpose.  Due to the Tier 2 Vehicle
and Low Sulfur Gasoline program, which became effective beginning in
2006, average national gasoline sulfur levels dropped 90 percent.  The
new Low Sulfur Diesel Rule, which began to be phased-in beginning in
2007, along with new clean engines operating requirements will reduce
NOX emissions by 50 percent, and reduce PM emissions by more than 90
percent.  The implementation of these new federal rules will
significantly decrease the fine particulate contribution in and from
areas impacted by population and transportation factors.

Michigan:

National rules are reducing emissions from both mobile and non-road
mobile sources throughout the region.  The Michigan Department of
Environmental Quality (MDEQ) believes that decreased emissions in the
area will ensure that the Grand Rapids metropolitan area will attain and
continue to maintain the daily PM2.5 standard.

Ohio:

Beginning in 2004, refineries began phasing in a new sulfur level for
gasoline due to the new federal standard for fuel.  This standard
requires the average sulfur level to be no greater than 30 parts per
million (ppm).  This represents a 10-fold reduction where average
national levels in 2002 were 300 ppm.  Also beginning in 2006, a new
requirement for ultra low sulfur diesel fuel (15 ppm) began phasing in. 
As with gasoline, this represents an enormous decrease from the 380 ppm
average measured in 2002.  These sulfur reductions are a key contributor
to large scale vehicular emission reductions in S02, which in turn, will
continue to have a positive impact on reducing PM2.5 emission.

EPA Response:

As explained above, consistent with section 107, EPA determined
violations of the 2006 24-hour fine particulate NAAQS based solely on
the most recent three complete years of certified monitoring data.  EPA
set boundaries based on available data and analysis that best represents
present-day conditions, and not on projected emissions reductions that
may occur after designation areas are finalized.  As noted, EPA does not
believe the statute allows consideration of future controls that have
not produced emissions reductions by the effective date of the
designations.

States may account for other projected emissions reductions in their
SIPs for achieving attainment, where appropriate.  States will also be
able to take advantage of any air quality benefits that may have already
occurred from recently enacted rules by submitting the 2008 monitoring
data to EPA by February 20, 2009.  EPA will review 2008 monitoring data
submitted by this date, and will modify area designations if appropriate
prior to the effective date.

State Comment:  Anticipated reductions in PM2.5 levels due to State or
local rules and control strategies

Iowa:

Iowa is making “substantial progress” on the control strategies for
the sources which are most responsible locally for the violations
(Blackhawk Foundry, Grain Processing, and Muscatine Power and Water),
which will be in place on an “expedited timeline” which is likely to
achieve attainment for both areas by 2014.  In addition, voluntary
measures are ongoing in the Quad Cities which will continue to improve
air quality.

EPA Response:

EPA recognizes the efforts by the state and the stakeholders to achieve
emissions reductions as soon as possible.  EPA also encourages both
states to continue efforts to assist the community in implementing
voluntary measures.  However, under section 307(d) of the Clean Air Act,
in designating nonattainment areas, EPA must designate, as
nonattainment, areas that are violating a NAAQS and areas that are
contributing to the violations.  This requirement does not authorize EPA
to rely on the ongoing efforts to develop future control strategies in
determining nonattainment boundaries.  EPA’s decision must be based on
the statutory requirements, and EPA’s promulgation of the
nonattainment designations for these areas, as described in detail in
the TSDs, meets the statutory requirements.

We also note that the state’s submission includes a “mitigation
plan” provided by Blackhawk Foundry.  The plan states that the
company’s intended control strategy, when fully implemented, “will
significantly reduce PM2.5 concentrations”.  However, the document
also concludes that, after implementation of the “preferred
alternatives” identified in the company’s control strategy, the
“combination of the predicted concentration” from the foundry, and
the background concentration, “could result in a predicted exceedance
of the PM2.5 NAAQS”.  (“PM2.5 Emissions Mitigation Plan”, Oct. 10,
2008, at p. 3.)  This information provides additional justification for
establishing a boundary that ensures emissions from other sources in the
area are considered in developing the attainment demonstration control
strategies for the Muscatine and Davenport areas.

Wisconsin:

Since 2007, the State has promulgated four significant rules. 
Collectively, when fully implemented in 2015, these four rules will
reduce SO2 and NOX emissions in the State by 82 percent and 62 percent
respectively below 2005 levels.

EPA Response:

Wisconsin also noted several emission reduction rules that it has
recently promulgated.  Full implementation of the rules should result in
significant emission reductions throughout the State.  The emission
reductions resulting from current programs and future planned reductions
will help improve the air quality in Wisconsin.  EPA commends Wisconsin
for it’s emission reduction efforts.  Still, for the fine particulate
designations, EPA must use the current information for making its
designations.  As explained above, section 107 requires designations to
be based on current conditions.  The uncertainty of future emission
reductions and new or expanded sources would not allow EPA to make an
accurate determination of the nonattainment area if these assumptions
were used.  Thus, EPA considers the emission controls in place for
designations.  Also, with EPA allowing early submission of 2008
monitoring data prior to the effective date of the designations, the air
quality improvement from current emissions reduction programs will be
factored into the final effective designations.  Wisconsin can include
the planned future emission reductions as it develops the area
attainment plans.

Wisconsin has recently enacted several rules to reduce emissions of a
variety of pollutants.  Reductions in fine particulate emissions coupled
with reductions in the precursor emissions, which form fine particulate
in the atmosphere, will lead to improved air quality.  The large
reductions in two key precursors predicted by the State are impressive. 
Wisconsin expects a 62 percent reduction in nitrogen oxide and an 82
percent cut in sulfur dioxide emissions from 2005 levels in 2015.  These
are sizeable emission reductions.  Wisconsin can take advantage of the
reductions that have already occurred by submitting early 2008
monitoring data for areas that have made enough improvement to meet the
air quality standards based on 2006 – 2008 data.  The upcoming
emissions reductions will benefit areas with ambient concentrations
remaining above the fine particulate standards by continuing the
improvement toward reaching the air quality standards.

SECTION VI:  Exceptional Events

State Comment:  Exceptional Events

Indiana:

EPA needs to make a documented, reasoned determination concerning
Indiana’s exceptional events submittal for all PM2.5 data used to
determine compliance with the 24-hour PM2.5 NAAQS.  The repeated
conclusion that “Indiana did not fully establish a causal connection
to the event and failed to meet the ‘but-for’ test” does not
contain adequate information to understand EPA’s conclusion and what
specifically would need to be provided to change EPA’s determination. 
The fact that EPA concurred with every exceptional event where the
24-hour concentrations were below 35 (g/m3, but rejected every event
where the 24-hour concentration was above 35 (g/m3 appears arbitrary,
especially when the rejected episodes all relate to significant
wildfires that were widely reported in the news at the time as causing
widespread air quality issues (which we indeed saw in our monitoring
data).  

EPA Response:

EPA has evaluated Indiana’s submittal in accordance with the criteria
and procedures that it established in its exceptional event rule.  EPA
provided documentation of its “reasoned determination” regarding
claims for the 2004-2006 period with its letter to Indiana dated August
19, 2008, and EPA is providing documentation regarding claims for 2007
in the TSD.  Under the Exceptional Event rule, Indiana has the
obligation to demonstrate that each of several criteria are met, most
notably that there would have been no exceedance or violation “but
for” the exceptional event.  Indiana cannot shift this burden of proof
to EPA.  Nevertheless, in order to evaluate the Indiana Department of
Environmental Management’s (IDEM) exceptional event claims
comprehensively, EPA thoroughly reviewed information submitted by IDEM
as well as other additional information not included in IDEM’s
exceptional event demonstration.  In some cases, the supplemental
information led to concurrence on IDEM’s exceptional event claims, and
in other cases, the totality of the evidence did not support EPA
concurring.  Documentation of EPA’s review, including additional
sources of information, is included in the TSD.

Indiana has noted a pattern in EPA’s concurrences and non-concurrences
with claims for

2004-2006 events, but Indiana is incorrect in inferring that this result
is arbitrary or somehow predetermined.  An examination of patterns of
EPA’s findings regarding Indiana claims for 2007 exceptional events
clearly demonstrates that, for example, EPA neither concurred with every
claim for concentrations above 35 (g/m3 nor rejected every claim below
35 (g/m3.  Instead, EPA examined each claim for each day and each
monitoring site in accordance with the criteria and procedures
established  in the exceptional events rule, and EPA has provided
documentation of its rationale in each case.

News reports of significant wildfires causing air quality issues can be
used as supporting evidence in establishing causal relationship between
the measurement and event claimed to have affected air quality.  News
accounts can also help corroborate quantitative evidence; however, such
reports are rarely determinative of the occurrence of an exceptional
event.  Additionally, for the exceptional event claims from 2004, 2005,
and 2006 for which EPA did not concur, IDEM did not provide any news
reports or stories in the exceptional event demonstration.

SECTION VII:  Other General Issues

State Comment:  States’ authority to impose additional necessary
emissions reduction measures

Alabama:

The Alabama Department of Environmental Management (ADEM) has the legal
authority to impose reduction measures in any area of our State, as
necessary, to attain the NAAQS, regardless of which areas are formerly
designated as nonattainment.  Accordingly, in our December 2007
submittal, the only counties that we recommended be designated
nonattainment were those with monitoring data exceeding the 24-hour
PM2.5 NAAQS.  High background PM2.5 levels present in the entire eastern
United States and emissions generated in the local area around the
violating monitors are the major causes of the elevated PM2.5
concentrations.  Accordingly, air pollution controls necessary to
mitigate regional background levels must be regional or national in
nature.  To repeat, the State-specific reduction measures which will be
needed for attainment are likely to occur only in the area covered by
our proposal for a partial-county area.  In the unlikely event that
further State-specific reduction measures are needed outside the
boundaries of our proposal, ADEM has the necessary authority to impose
them.

EPA Response:

Section 107(d) directs that “…any area that does not meet (or that
contributes to ambient air quality in a nearby area that does not meet)
the national primary or secondary ambient air quality standard for the
pollutant” be designated as nonattainment.  EPA is legally bound to
carry out this directive, and therefore cannot base designation
decisions on a State’s ability or inability to impose targeted
measures either within or outside a nonattainment area that may result
in a violating monitor attaining the standard at some point in the
future.  Nonattainment designations must be based on current air quality
regardless of whether polluting emissions are local, regional or
transported from much further away.  All violating and nearby
contributing areas must be designated nonattainment consistent with the
statute.  States and EPA will then address the sources of pollution in
preparing implementation plans and other measures to address regional
transport.

State Comment:  County-level designations in Western States not
appropriate (CA)

California:

One of the primary issues that must be addressed when discussing the
boundaries of a nonattainment area in California is the large size of
California counties versus other States.  The average area of a
California county is 2,822 square miles, yet the average county size in
the United States is 622 square miles.  Alaska and Arizona are the only
States with larger average county size (Table 1 on page 5 of the
California comment letter).  The average California county is over 4½
times the average U.S. county; many as large, if not larger, than entire
States.  In many cases, California counties contain one or two urbanized
regions and large stretches of sparsely populated areas.

Much of the 9-factor analysis utilized by U.S. EPA to determine PM2.5
nonattainment areas is based at the county level.  This presents some
unusual challenges for California.  For instance, applying county-wide
VMT statistics to a large California county misrepresents differences
that may exist in VMT urban and rural areas in that county, or between
two widely separated urban areas in the same county.  Throughout this
submittal, we offer alternative approaches to analyzing the nine factors
when county size presents a particular problem.  This problem is most
evident in Imperial County where the three main urban areas represent
only one percent of the county (in square miles) recommended as a
nonattainment area.  The remaining 99 percent of the county is sparsely
populated.

EPA Response:

In designating nonattainment areas for the 2006 24-hour fine particulate
matter NAAQS, EPA considered a partial-county boundary for a violating
or contributing county on the basis of:  1) The existence of a power
plant as the major source of emissions in a county, with the county
ranking low for other factors; 2) Topography, where topography is
isolating an airshed within the county; or 3) Factors other than
topography that demonstrate a unique situation in which the sources
captured by the partial-county boundary account for nearly all of the
total fine particle mass on exceedance days; or international emissions
are a significant contributor, and the domestic sources captured by the
partial-county boundary account for nearly all of the domestic
contribution to the total fine particle mass on exceedance days.  Under
this analysis EPA considered the large size of some counties in the west
and did designate partial counties in such areas where the area-specific
data demonstrated that the partial county accounted for nearly all of
the sources contributing to the violating monitor.

While EPA used county-wide statistics, we also scrutinized population,
traffic, and stationary sources of emissions data at a smaller scale,
using spatial analysis and geographic information systems (GIS).  We
used publicly available data sources including the U.S.  Census for
population, the FHWA’s Freight Analysis Framework, and the NEI.  In
all of the areas where EPA’s boundary recommendations were larger than
California’s recommendations, the majority of the county's population,
traffic, and stationary sources were included in the boundary we
recommended.  Furthermore, in many cases we used township and range
boundaries, rather than county boundaries, as a more appropriate means
to define nonattainment area boundaries in Imperial, Butte, Yuba, Yolo,
Placer and El Dorado Counties.

State Comment:  Economic hardship considerations

Michigan:

The 2006-2008 three-year daily average shows only monitors in Wayne
County and St. Clair County in violation of the standard.  The Michigan
Department of Environmental Quality (MDEQ) continues to believe that
designating a huge nonattainment area in response to violating monitors
in a small area is heavy-handed and unnecessary.  It serves to
economically penalize the entire region at a time when the State has
been economically distressed for the past eight years and is struggling
to revive.

Ohio:

The consequences of having a nonattainment designation status will have
a significant economic impact on the townships as well as the entire
counties discussed previously, which are already experiencing severe
fiscal problems, high rates of unemployment, population migration, and
the other social dislocation associated with such economic distress. 
Given these circumstances and the absence of data which solidly
demonstrate that such nonattainment designations are appropriate, I
strongly encourage US EPA to reconsider its original findings.

EPA Response:

EPA is required by the CAA to designate all violating and nearby
contributing areas nonattainment without regard to potential economic
impacts.  EPA used the data it had available in conjunction with State
recommendations and public input to analyze on a case-by-case basis
which areas should be included in each nonattainment area.  However, EPA
does not believe that a nonattainment designation necessarily and
significantly limits economic growth.  The requirements for new sources
in nonattainment areas are similar to those in attainment areas.  In
both areas, new sources must be well controlled and must appropriately
address the impact of the new source on air quality.  While companies
will blame environmental restrictions for facility location selections
there are many other factors such as market access, skilled labor
access, and various cost factors that affect the location decision. 

SECTION VIII:  Typographical Errors

Alaska:

On page 1 of EPA’s technical analysis for Juneau there is a
typographical error.  The second sentence starts “This technical
analysis for Fairbanks…” and should read “This technical analysis
for Juneau…”

EPA Response:

EPA has corrected this error in its final Technical Support Document.

Connecticut:

We note one inaccuracy on page 3 of EPA’s response to Connecticut’s
recommended designations.  The response refers to the 1997 PM2.5 NAAQS
nonattainment boundary for the New York City metropolitan area as
including Hartford and Fairfield Counties in Connecticut.  Rather, the
New York City metropolitan area included New Haven and Fairfield
Counties in Connecticut.  We hope you will correct the record in this
respect to avoid any confusion in the future.

EPA Response:

EPA has corrected this error in its final Technical Support Document.

Idaho:

Figure 9 on page 19 of the December 2007 submittal has an incorrect
image.  Please substitute the corrected page 19 attached to this
document.

EPA Response: 

EPA has corrected this error in its final Technical Support Document.

Kentucky:

The ‘2005 VMT’ data should be labeled ‘millions’ (annual) miles
not 1000’s.  This error occurs throughout the letter in the Traffic
and Commuting Patterns Table.

EPA Response:

EPA has corrected this error throughout its final Technical Support
Document.

AREA-SPECIFIC ISSUES

Allentown, PA

New Jersey:

The “urban increment” in EPA’s technical analysis, which only uses
the “high” days (top 5 percent of days), does not take into account
the direct impact of a major point source such as the Portland
Generating Station, which is situated directly next to Knowlton
Township.  The Portland Generating Station is the single largest source
of both SO2 and carbon emissions in the Knowlton area.

EPA Response:

The “urban increment” identifies the non-regional contribution of
PM2.5 on high days to violating monitors.  For the Allentown, PA area,
the violating monitor is located in Northampton, PA.  The “urban
increment” is incorporated into the CES, which takes into account
emissions data, meteorological data, and air quality monitoring
information to provide a relative ranking of counties that contribute to
a monitored violation of a PM2.5 NAAQS within a nonattainment area.  The
CES metric was not meant to be an exclusive way for consideration of
data for these factors.

The CES method places primary emphasis on characterizing those days when
ambient PM2.5 concentrations at a monitor are near or exceed the
standard.  The CES specifically considers the influence of emissions,
including data from major point sources, and meteorology on days with
high 24-hour fine particle concentrations.  The information is
particularly relevant for a 24-hour standard, for which individual days
affect attainment status, unlike an annual standard, which considers air
quality throughout the year.  The CES examines all days in an area that
are above the 95th percentile or the level of the NAAQS, whichever is
lower.

EPA considers the CES to be a reasonably realistic model of the actual
processes that affect the contribution linkage between a monitored
violation in a violating county and emission sources in a nearby
potentially contributing county.  The CES was not designed to assess the
impact of individual sources on areas not identified as violating
through monitored data.  

New Jersey:

Any emissions (primary and secondary) from the Portland Generating
Station would easily put the Knowlton area over the 35 ug/m3 standard
based on an air quality monitor in Phillipsburg, with a design value of
34 ug/m3.  

EPA Response:

Current regulations for determining violations of the fine particle
NAAQS require that violations  be determined on the basis of complete,
quality-assured ambient air quality data at a monitor in 

the area.  The air monitor in Phillipsburg, Warren County measured 34
ug/m3.  Consequently, this county is not currently violating the PM2.5
NAAQS notwithstanding emissions from the Portland Generating Station. 
EPA strongly advises New Jersey to place an air quality monitor in the
vicinity of Knowlton Township, to support the State’s determination
that Warren County (or portions thereof) is violating the 24-hour
standard.  Without such monitoring data EPA cannot designate the area
nonattainment based on a violation, and for reasons explained in the TSD
EPA also does not have data to support designating Warren County
nonattainment based on contribution.

 

New Jersey:

New Jersey indicated that they disagreed with EPA’s assessment that
the Delaware River does not significantly influence the air shed.  Since
Knowlton Township and the Portland Generating Station are located in the
river valley, limited air mixing emissions from the Portland Generating
Station would directly impact Knowlton Township.

EPA Response:

EPA agrees with New Jersey that there would be limited air mixing within
the Delaware River Valley.   However, without monitoring data to
contradict the existing data for Warren County EPA can not determine
that Warren County is currently violating the PM2.5 NAAQS.  EPA
continues to strongly advise New Jersey to place an air quality monitor
in the vicinity of Knowlton Township, to support the State’s
determination that Warren County (or portions thereof) is violating the
24-hour standard.  

New Jersey:

EPA is using the State line between Pennsylvania and New Jersey as the
boundary for the nonattainment area.  The State line does not prevent
SO2 and PM2.5 emissions from the Portland Generating Station from
impacting Knowlton citizens.

EPA Response:

Although emissions from Pennsylvania may impact Knowlton Township in
Warren County, the current air monitor in Warren County is currently not
in violation of the fine particle NAAQS.  As stated previously, current
regulations for determining violations of the fine particle NAAQS
require that violations be determined on the basis of complete,
quality-assured ambient air quality data at a monitor in the area.

 

New Jersey:

The CAA does not specifically require monitoring data when designating
or redesignating an area.  The regulations cited by EPA, 40 CFR
50.13(c), which references Appendix N, does not specifically provide
that monitoring data is the only data that can be used in support of a
redesignation or designation.  Further, U.S. EPA’s June 8, 2007
designation guidance memo should not limit States from considering other
scientifically valid approaches to understanding air quality in an area.
 New Jersey agrees that quality assured air quality monitoring data
should be used whenever available.  In the case of Knowlton, the use of
air quality modeling data to supplement the region’s air monitors is
appropriate given the geographical complexities of the area.  

New Jersey stated that according to Section 171(2) of the CAA, the term
‘nonattainment data’ means for any pollutant area which is shown by
monitored data or which is calculated by air quality modeling (or other
means determined by the Administrator to be reliable) to exceed any
NAAQS.

New Jersey provided technically sound and consistent modeling results
based on two different models which, when used with regional air
monitoring data, indicate violations to the 2006 24-hour PM2.5 NAAQS.

EPA Response:

EPA agrees that section 107(d) does not explicitly state that only
monitoring data can be used for promulgation of designations, and EPA is
considering other information as appropriate to evaluate areas that may
be contributing to violations.  However, EPA regulations relevant to the
NAAQS and monitoring for the NAAQS do describe at length the
requirements (which are based on monitoring) for determining whether a
violation of the PM2.5 NAAQS has occurred.  EPA’s guidance for these
designations likewise explained that for an area to be “violating”
the NAAQS, it must have at least one FRM with data establishing such
violation.  For contributing areas, other information can be relevant.

 percentile 24-hour concentration, as determined in accordance with
appendix N of this part, is less than or equal to 35 μg/m3.

Appendix N to Part 50-Interpretation of the NAAQS for PM2.5 explains the
data handling conventions and computations necessary for determining
when the annual and 24-hour primary and secondary NAAQS for PM2.5
specified in 50.13 are met.  While monitoring requirements necessary for
determining NAAQS compliance are discussed in detail in Appendix N,
requirements for the use of modeling to demonstrate whether the NAAQS
has been exceeded are not specified.  Thus, EPA believes that monitoring
information is necessary in order to substantiate a nonattainment, or
attainment, designation made by EPA.  Therefore, EPA is designating
“violating” areas based upon the presence of an air monitor meeting
regulatory requirements to show that a violation, as defined in the
regulations for the NAAQS, has occurred.

EPA does not believe that its June 8, 2007 designation guidance memo
limits States from considering other scientifically valid approaches to
understanding air quality in an area.  Current Agency regulations
indicate the requirements for designating “violating” areas. 
EPA’s designation guidance memo identifies the current regulations
which are applicable for identifying NAAQS violations, specifically
using data from Federal Reference Method (FRM) and Federal Equivalent
Method (FEM) monitors that are sited and operated in accordance with 40
CFR Part 58, and indicates the procedures given in 40 CFR Part 50
Appendix N for using these data to determine whether a violation has
occurred.

EPA agrees that the modeling performed by New Jersey does provide
valuable information for understanding air quality in Warren County. 
Based on the air modeling performed by New Jersey which shows the
potential to exceed the PM2.5 NAAQS using New Jersey’s modeling
inputs, EPA strongly recommends that New Jersey place an air quality
monitor in the area of impact to demonstrate a possible violation of the
PM2.5 NAAQS.

New Jersey states that Section 171(2) of the CAA defines the term
‘nonattainment data’ as

“any pollutant area which is shown by monitored data or which is
calculated by air quality modeling (or other means determined by the
Administrator to be reliable) to exceed any NAAQS.”

New Jersey is citing a pre-1990 version of the CAA that has since been
modified.  Section 171(2) of the CAA, which defines nonattainment areas,
no longer refers to air quality modeling.  The provision now refers to
areas “designated ‘nonattainment’ with respect to that pollutant
within the meaning of section 107(d).  This change was accompanied by
revisions to section 107 itself which eliminated provisions for
designations for the SO2 NAAQS that directed EPA to designate areas
based on future projected attainment status, i.e., through a process
that would have necessitated modeling the future ambient level at FRM
monitors.  This provides further evidence that it is not appropriate to
base determinations of violations for purposes of designation on
modeling data alone, especially in the face of valid monitoring data for
the same county showing attainment.

Birmingham, AL

Alabama:

Alabama argued that the existence of

“high background levels present in the entire eastern United States
and emissions generated in the local area around violating monitors are
the major causes of the elevated PM2.5 concentrations.”

As a result, Alabama argued that regional control strategies are
necessary to address the regional background levels and that “local”
emissions sources are being analyzed by the Department for Reasonably
Available Control Technology (RACT) level controls, i.e., nonattainment
area plan requirements.

EPA Response:

EPA disagrees with the State’s argument for several reasons.  First,
it is not disputed that areas in the Eastern U.S., including Birmingham,
are subject to regional transport of pollution.  However, that regional
transport does not comprise the entire amount of ambient PM2.5 measured
at the violating monitors in the Birmingham area.  The purpose of the
designation process is to identify those “nearby” areas that contain
emissions sources or activities that are contributing to the violations.
 It is not appropriate for EPA to exclude nearby areas that are
contributing, merely because there is also transport from more distant
areas.  Moreover, the CAA provides other tools, most notably section
110(a)(2)(D) and section 126, to address interstate transport.  

Second, EPA does not agree with Alabama’s view of what constitutes
“local” emissions sources in the context of designations.   As with
regional background, these “local” sources identified by Alabama are
not the exclusive source of ambient PM2.5 levels in the Birmingham area.
 Under section 107(d), EPA must identify both violating and nearby
contributing areas.  Because PM2.5 and PM2.5 precursor emissions can be
transported across great distances, the evaluation must take into
account sources that contribute in an area based upon the facts and
circumstances of that specific area.  In its guidance EPA recommended
various forms of information that are relevant for evaluating
contribution, and the Agency has used a number of relevant analytical
tools in evaluating the Birmingham area, as explained in the TSD for
this area.  

Third, the State’s argument with respect to “regional” and
“local” emissions is that contribution is limited to the source or
sources that “cause” the violation at a violating monitor, i.e.,
those sources add the last increment of ambient PM2.5 above the level of
the standard.  EPA does not agree that section 107(d) contemplates that
“contributes” should be read to mean “causes” in the sense meant
by Alabama.  EPA believes that section 107(d) requires a designation of
nonattainment for those areas that “contribute” to the violation,
i.e., that have emissions sources or activities that cumulatively add to
the ambient PM2.5 at the violating monitor.  This reading of the statute
is consistent with the purpose of nonattainment area designations, which
is to provide the boundaries for those areas within which States should
evaluate all sources for controls in the context of the nonattainment
area SIP development, to impose controls as appropriate on such sources,
and to provide for expeditious attainment of the NAAQS.  EPA believes it
is not consistent with the statute to truncate that analysis by
excluding nearby areas with sources that are also contributing to
violations in the area.

Alabama:

In its October letter, Alabama also reiterated its arguments that EPA
should only designate a smaller area nonattainment, premised upon the
argument that “specific reduction measures which will be needed for
attainment are likely to occur only in the area covered by our proposal
for a partial county area.”  

EPA Response:

EPA disagrees with this approach to designations.  In essence, Alabama
argues that the State and EPA should predetermine which sources should
be controlled as a basis for the designation.  EPA believes that this is
in direct contravention of section 107(d), which obligates EPA to
designate as “nonattainment” those violating areas and those nearby
areas that are “contributing” to the violations.  The full
evaluation of sources within a nonattainment area, including the
determination of which sources of direct PM2.5 and PM2.5 precursors
should be controlled, to what degree they should be controlled, and when
they should be controlled in order to provide for expeditious attainment
of the NAAQS, will occur during the development of the nonattainment
area SIP for this area.  EPA considers it inappropriate to truncate that
analysis and the requirements of the CAA for nonattainment area SIPs
through the creation of artificially smaller boundaries that exclude
areas that contribute to the violations.

Alabama:

The overwhelming majority of NOX and SO2 emissions in Walker County are
due to a single utility located in the county (Alabama Power Company –
Plant Gorgas).  Additional controls have been installed at this power
plant since 2005.  Flue gas desulfurization (FGD) scrubbers have been
installed on units 8, 9 and 10.  Therefore, over 80 percent of the power
generation capacity at Gorgas is being scrubbed.  

The overwhelming majority of NOX and SO2 emissions in Shelby County are
due to a single utility located in the county (Alabama Power Company –
Plant Gaston).  A scrubber will be installed on Gaston Unit 5 in 2010. 
Unit 5 is the largest unit at Plant Gaston.  

EPA Response:

See the overview response to comments on power plant issues, located at
the beginning of Section IV.  

The CAA requires that a nonattainment area must include not only the
area that is violating the standard, but also nearby areas that
contribute to the violation.  Thus, for each monitor or group of
monitors that indicate violations of a standard, EPA is establishing
nonattainment boundaries that cover a sufficiently large area to include
both the area that violates the standard and the areas that contribute
to the violations.  EPA’s designations guidance for the 2006 PM2.5
standard suggests that the same boundaries established for implementing
the annual PM2.5 standard may also be appropriate for implementing the
24-hour PM2.5 NAAQS in areas where both standards are violated.  Walker
County (partial) and Shelby County are part of the existing Birmingham
nonattainment area.  EPA finds that it is appropriate to include the
same counties in the nonattainment area for the 24-hour standard since
these counties have high emission levels and air quality data indicate
that the same PM2.5 components that are the biggest contributors to
PM2.5 mass on an annual average basis in this area (e.g., sulfate from
SO2 emissions) are also key components of PM2.5 mass on the highest
days.  

In addition, Plant Gorgas in Walker County and Plant Gaston in Shelby
County have emissions that are not fully controlled.  Two units at Plant
Gorgas do not have scrubbers.  For the first half of 2008, emissions
from the plant totaled 14,000 tons SO2 and 10,000 tons NOX.  At Plant
Gaston, while a scrubber will be installed on the largest unit in 2010,
four other units will remain without scrubbers.  The 2007 emissions from
these four units were substantial, more than 75,000 tons SO2 and 12,000
tons NOX.  

While Alabama anticipates installation and operation of additional
controls in the future, EPA must judge whether current emission levels,
supported by current regulations that provide assurance of continued
operation of existing controls, contribute to existing violations.  EPA
continues to believe that emissions from Walker County (partial) and
Shelby County contribute to violations of the 24-hour PM2.5 NAAQS in
Jefferson County.

The State of Alabama makes the following points as rationale for
excluding Walker County from the Birmingham nonattainment area:

Alabama:

Walker County has a very low population density and VMT compared to
Jefferson County.

EPA Response:

Population and population density were one type of information that EPA
used nationwide to determine the appropriate boundaries for PM
nonattainment area.  Population data gives an indication of whether it
is likely that population-based emissions might contribute to violations
of the 24-hour PM2.5 standard.  EPA agrees that Walker County has a
relatively smaller population size and density than Jefferson County. 
Of the CSA population, about 6 percent (69,980) reside in Walker County,
while about 56 percent (656,014) reside in Jefferson County.  The
population density per square mile is 87 in Walker County and is 584 in
Jefferson County.  Of the CSA vehicle miles traveled, about 5.5 percent
(797) are in Walker County, while about 58.7 percent (8,545) are in
Jefferson County.  

Low population density and VMT compared to the remainder of the areas
under consideration for the nonattainment area are among the reasons
that EPA determined that it would be appropriate to include only a
portion of Walker County, rather than the entire County.  EPA believes a
partial nonattainment area designation remains appropriate for Walker
County in view of the large amount of PM2.5 and PM2.5 precursor
emissions from Plant Gorgas and in evaluation of other information such
as the CES and the pollution rose for this area which indicate that the
emissions from this portion of Walker County are contributing to
violations of the NAAQS in adjacent Jefferson County.

Alabama:

The wind infrequently blows from the direction of Walker County towards
Jefferson County on days with high PM2.5 concentrations.

EPA Response:  

EPA disagrees with Alabama’s conclusion that winds blow infrequently
from the direction of Walker County towards Jefferson County on days
that are relevant to violations in this area.  The pollution rose in
EPA’s TSD for the Birmingham area reflects both the high number of
days on which winds blow from the direction of the County and the high
ambient levels on those days throughout the year.  Therefore, EPA
believes that elevated PM2.5 levels at the violating monitors originate
from multiple directions in the Birmingham area, including from the
direction from Walker County.

In addition, EPA’s CES score confirms the high degree of contribution
from Walker County to Jefferson County compared to all other counties in
the area except Shelby.  The CES methodology takes into account the
influence of emissions and meteorology on days with high 24-hour fine
particle concentrations and also incorporates the seasonal composition
of fine particle mass and source receptor distance relationships.  Even
though the majority of emissions from all of Walker County emanate from
one source in a small part of that County, the CES score for the County
is 17, which reflects the significance of the level of fine particle
transport to the violating ambient monitor.  The CES scores for other
counties in the area are 100 for Jefferson County, 39 for Shelby County,
and 5 or less for all other counties.  The scores suggest that
particulate matter emissions from Walker County and Shelby County and
the average wind trajectories on high PM fine days have a significant
impact on PM fine levels at the violating ambient monitor in Jefferson
County.

The State of Alabama makes the following points as rationale for
excluding Shelby County from the Birmingham nonattainment area:

Alabama:

The monitor in Shelby County measures attainment of the 24-hour PM2.5
standard, along with five of the eight Jefferson County monitors.  Data
from monitors located between Shelby County and the non-attaining
monitors in Jefferson County do not suggest that Shelby County is
significantly contributing to nonattainment in Jefferson County; rather,
the data shows an obvious local emissions impact around the violating
Birmingham monitors.

EPA Response:

The presence of an attaining monitor in Shelby County does not establish
whether emissions activity in Shelby County is contributing to
violations in adjacent Jefferson County.  EPA believes that whether
Shelby County itself contributes to the violations in Jefferson County
must be evaluated using the types of information that EPA has utilized
in the TSD for this area.  Likewise, EPA does not agree that the
existence of one or more attaining Jefferson County monitors located
between Shelby County and the violating monitors in Jefferson County
establishes whether or not Shelby County is contributing.  The amount of
emissions in Shelby County, the proximity of Shelby to the violations in
Jefferson and meteorological data indicating transport of those
emissions from Shelby towards Jefferson County suggest that the
emissions from Shelby County combine with the emissions from Jefferson
County sources (and others) and cumulatively result in the violations at
the monitors in Jefferson County.  

The pollution rose in the TSD for this area confirms that winds blow
from the direction of Shelby towards Jefferson on days with high ambient
PM2.5 in Jefferson County.  The CES score for Shelby was 39, which
places it second only to Jefferson County in the Birmingham area, and
thus ranks its relative contribution to violations in Birmingham higher
than all other surrounding counties.  This information supports EPA’s
conclusion that Shelby is contributing, as that term is intended in
section 107(d) of the CAA.

Alabama:

Alabama argued that only 73 percent of the MSA population resides in
Jefferson County.

EPA Response:

As stated above, population and population density were only one form of
information that EPA evaluated nationwide to determine the appropriate
boundaries for nonattainment areas for the 2006 24-hour PM2.5 NAAQS.  In
conducting our analysis EPA referred primarily to CSA, as defined in OMB
BULLETIN NO. 07-01 (Update of Statistical Area Definitions and Guidance
on Their Uses), December 18, 2006.  According to 2005 population data,
of the Birmingham CSA population, about 14.7 percent (171,373) reside in
Shelby County, while about 56 percent (656,014) reside in Jefferson
County.  The population density per square mile is 212 in Shelby County
and is 584 in Jefferson County.  Although Shelby County does have
smaller population than Jefferson, this does not negate the other types
of information that establish the contribution of Shelby to violations
in Jefferson, such as the particularly high emissions of SO2 and NOX,
the relatively high level of commuting between these counties, and the
lack of topography or other barriers to transport of pollutants from
Shelby to Jefferson.  EPA believes this data continues to support a
nonattainment area designation for Shelby County.

Alabama:

Jefferson County's VMT is approximately five times that of Shelby
County's.

EPA Response:

EPA agrees.  Of the total CSA vehicle miles traveled, approximately 11.3
percent (1,640) of the VMT in 2005 were in Shelby County, while about
58.7 percent (8,545) of the VMT are in Jefferson County.  EPA notes that
this amount of VMT reflects a significant amount of NOX and VOC
emissions, and that these emissions do contribute to violations in this
area, albeit less than the emissions resulting from mobile sources
within Jefferson County itself.  However, EPA’s assessment of this
single form of information does not preclude it from determining that
Shelby County contributes to increased ambient PM2.5 levels in the
Birmingham area, based upon this and other information.  

Alabama:

The bulk of the SO2 and NOX point source emissions from Shelby County
can be attributed to Alabama Power‘s Plant Gaston.  However, several
control measures have been installed and will not be reflected in any
previous inventory. 

EPA Response:

EPA agrees that Alabama Power Company – Plant Gaston is the most
significant individual source of NOX and SO2 emissions in Shelby County.
 The emissions from this plant are approximately 130,000 tons per year
of SO2 and 20,000 tons per year of NOX as well as a large quantity of
direct PM2.5.  These emissions make this source among the largest in the
Birmingham area, and these emissions are of the types that are well
represented in the speciated data for ambient PM2.5 in this area.  These
emissions occur in a county immediately adjacent to Jefferson County,
and only about 28 miles from the violating monitors in Jefferson County.
 Thus, these emissions occur at a location that is well within the range
for transport of PM2.5 and PM2.5 precursors.  Other analytical tools
used by EPA, such as the CES score and the pollution rose for this area
also confirm that emissions from this plant contribute to violations in
the Birmingham area.  The emission units at Plant Gaston are not
controlled by SO2 scrubbers, although the largest of the facility’s
five units is scheduled for a scrubber to be installed by 2010.

Alabama argues that the installation of scrubbers on one unit of this
source in 2010 is a reason to exclude Shelby County from the Birmingham
nonattainment area.  EPA disagrees.  Section 107(d) obligates EPA to
designate as nonattainment “any area that does not meet” the NAAQS,
and any nearby area that “contributes” to those violations.  Given
the present tense formulation of this provision, EPA does not believe it
is appropriate to base designations on projections of future violations
or future contributions, but rather upon the facts at the time of the
designation.  

With respect to the overall level of control at this source, EPA notes
that not all of the units at this source are fully controlled and that
additional emissions reductions may be necessary and appropriate for
this source.  The determination of what level of control for this source
should be required will be addressed in the nonattainment area plan
developed for the Birmingham area, and it would be inappropriate for
Alabama and EPA to prejudge that decision in the context of the
designations.  Based upon the amount and type of emissions, and the
location of this source, EPA has concluded that it continues to
contribute to violations in the Birmingham area.

The State of Alabama makes the following points as rationale for
designating only a portion of Jefferson County as nonattainment, which
is a change from its original recommendation.

Alabama:

Alabama revised its original recommendation that EPA designate all of
Jefferson County as nonattainment, on the grounds that the area should
include “only the local airshed for the monitors violating the 24-hour
PM2.5 standard.”  

EPA Response:

EPA disagrees with Alabama’s view that the violating monitors in
Birmingham are located in a “local airshed” that is distinct from
the remainder of the Birmingham area.  EPA has considered nonattainment
areas boundaries for smaller areas (including for partial counties),
when there is clear factual evidence that the designated area is in a
separate airshed from the immediately surrounding areas, e.g., when the
area is separated by mountain ranges that clearly impede air flow
between the areas.  Given the topography and the meteorological evidence
in the Birmingham area, EPA sees no factual basis for a credible claim
that the violating monitors in this area are in an airshed separate from
the remainder of Jefferson County, or from Walker County and Shelby
County.

Alabama:

The data from the Jefferson County monitors, along with an air quality
study commissioned in 2005 by the Alabama Department of Environmental
Management (ADEM) and the Jefferson County Department of Health,
indicates the clear existence of a local emissions influence on the
violating monitors.  This study concluded that there is a well-defined
local source influence in addition to a regional component of the annual
PM2.5 concentrations measured at the Wylam and North Birmingham
monitors.  Only three of the eight Jefferson County monitors violate the
24-hour standard, all of which will be captured by the partial area that
we are proposing for nonattainment designation.

EPA Response:

EPA agrees that “local” emissions sources have an influence on the
violations in the Birmingham area.  Alabama does not contend, however,
nor could it, that all of the ambient PM2.5 at the violating monitors in
this area are the result of one or more specific sources in the near
vicinity of the monitors.  EPA believes that the violations in
Birmingham are the result of the cumulative impacts of emissions from
various sources in Jefferson County and nearby Walker and Shelby
Counties, in addition to emissions from sources that are not in nearby
areas or are in other States.  As a result, under section 107(d) EPA
must include these nearby areas within the boundaries of the
nonattainment area for Birmingham.

Alabama:

In support of its recommendation that EPA designate only a portion of
Jefferson County nonattainment, Alabama also referred to the results of
a study it commissioned from Envair “to investigate the source(s) of
particulate matter pollution in and around the NBHM and WYL monitors.”
 According to Alabama, “monitoring data clearly indicated a strong
local influence on the high particulate matter concentrations.”

EPA Response:

EPA does not agree that the Envair study supports a position that only
“local” sources are contributing to violations of the NAAQS in this
area.  Even the portions of the study quoted by the State in its October
2, 2008, letter indicate that contribution is the result of emissions
across a broader area than that recommended by Alabama.  In addition to
a “large regional influence” and a “significant influence of local
sources,” the Envair study notes that the ambient levels are
“supplemented by a general urban component” of emissions. 
Similarly, in addition to the “local” stationary sources that
Alabama would evidently like to encompass within a much smaller
designated area, the Envair study notes that there is a “complementary
contribution of emissions from local transportation sources, including
motor vehicles and railroads.”  EPA recognizes the study indicates
that there is a strong local influence from certain industrial
facilities in the Birmingham area.  However, while the study does
implicate some of the very large local sources as significant
contributors, it does not say those are the only sources.  Total fine
particle mass on the high concentration days in Birmingham indicates
that other metropolitan-area sources such as transportation and power
plants also contribute.  Finally, the Envair study was developed to
analyze the area for purposes of the 1997 annual PM2.5 NAAQS, so it does
not necessarily translate to what may be appropriate for the 2006 24
hour PM2.5 NAAQS for this area.

California Air Resources Board (CARB) Multi-Area

CARB:

The State of California requested that EPA modify the proposed
nonattainment boundaries to be consistent with California’s original
recommendations for the City of Calexico, Sacramento County, City of
Chico, and the combined cities of Yuba City/Marysville.  The State and
EPA are in agreement regarding the boundaries for the South Coast Air
Basin, San Joaquin Valley Air Basin, and San Francisco Bay Area.  The
State of California provided additional information to document the
extent of international transport causing the localized impacts in
Imperial County, and the localized impact of wood smoke in the other
areas at issue.

EPA Response: 

EPA appreciates the additional information submitted by the CARB in
their letter dated October 15, 2008.  As CARB pointed out in their
letter, States and Tribes need to submit recommendations identifying
where violations are occurring and including nearby areas that contain
emission sources that contribute to the observed violations.

In response to the information submitted by CARB, EPA has modified our
previous recommendations for Imperial County, Butte County, Yuba/Sutter
Counties, and the Sacramento area (including Yolo, Solano, Placer and El
Dorado Counties).  EPA originally recommended a larger nonattainment
boundary for these areas but has reduced their size in light of further
information and analysis provided in CARB’s October 15 letter.  While
EPA has reduced the size of these areas, they still include the
violating monitors and the contributing sources in nearby areas and they
are still larger than those recommended by the State of California.

Several factors led EPA to recommend a larger PM2.5 nonattainment area
than recommended by California.  A significant consideration in
expanding the nonattainment areas recommended by California was that the
State relied on future mobile source controls at a statewide level to
address NOX emissions and, therefore, discounted mobile sources as an
important consideration in their analysis.  During the designation
process, as required by section 107 of the CAA, EPA determined
violations of the 2006 24-hour fine particulate NAAQS based solely on
the most recent three complete years of certified monitoring data.  EPA
set boundaries based on available data and analysis that best represents
present-day conditions, and not on projected emissions reductions that
may occur after area designations areas are finalized.  Section 107
requires EPA to designate all areas within three years of promulgation
of a new NAAQS, and to designate as nonattainment any area that violates
or contributes to a nearby area that violates a NAAQS.  The statute is
written in the present tense and thus EPA believes it can only base
designations on controls that are in place and federally enforceable by
the effective date of final designations.  Based on speciation data
provided by CARB, organic carbon and nitrates were identified as the
major components of PM2.5 where violations occurred, which were
attributed to residential woodburning and mobile sources, respectively. 
Therefore, EPA’s nonattainment areas were intended to capture the
extent of these particular sources which could be contributing to
observed violations of the PM2.5 standard.

While EPA agrees with CARB that there are more localized impacts from
residential woodsmoke emissions, as explained above, it does not agree
that the contribution of mobile sources can be discounted based on the
prospect of future controls.  EPA views mobile source emissions as a
significant component of regional PM2.5 levels in the Sacramento Valley
and elsewhere.  It appears that the combination of this regional
pollution and local woodsmoke emissions in certain areas leads to
violations of the PM2.5 standard, particularly during stagnant
conditions.  Considering this, EPA looked at the location of the
violating monitors and the sources of both woodsmoke and nitrates to
determine the appropriate boundaries, which are larger than what
California recommended.

Chicago, IL-IN

The State of Indiana

Indiana:

The monitors in the Chicago area located closest to the Indiana State
line … that should be more directly impacted by emissions from Lake
and Porter County sources, are monitoring attainment of the 24-hour
standard.”  Indiana feels that if it contributed to the violations in
Illinois, the monitors near the State boundary would record higher
concentrations.  Indiana feels the violating monitors in Illinois are
“most likely affected by localized emission sources.

EPA Response:

The fine particulate concentration in an area and thus the monitors
measuring the concentration are influenced by emissions from a variety
of sources.  The concentration is the result of complex atmospheric
chemistry that forms fine particulate from precursors and can change the
distribution of directly emitted particulate.  Fine particulate
pollution is the result of emissions from local, urban, and
regional-scale emissions.  The emissions from a county do not have to
cause a violation to be considered as contributing to a violation. 
Emissions from the area counties can increase the PM2.5 concentration
across the area which adds to regional background concentration. 
Emissions from some sources or clusters of sources can further raise the
local PM2.5 concentration which results in a monitor or some area
monitor recording a violation.  A county does have to cause the “last
microgram or few micrograms” that cause a monitor to record a
violation.  It just needs to add to the total concentration to be
considered contributing.  Lake and Porter Counties, Indiana have the
second and fourth most direct PM2.5 emissions in the 10-county
nonattainment area.  Lake County, Indiana has the second highest
emissions in the area of the two main precursors, SO2 and NOX.  The SO2
and NOX emissions from Porter County are higher than the emissions from
other Chicago area counties.  The level of PM2.5, SO2, and NOX emissions
from Lake and Porter Counties, Indiana are high enough to contribute to
the violations in the Chicago nonattainment area.    

Indiana:

Indiana requested Lake and Porter Counties be designated as a
nonattainment area separate from the Illinois portion of the Chicago
area.  Indiana notes that seven of the 18 PM2.5 monitors in the area
violated the PM2.5 standards for 2005-2007, the most recent complete
period.  Indiana feels that “it is unnecessary to extend the
restrictions of a nonattainment area outside the Chicago area.

EPA Response:

EPA agrees that it is not necessary to extend the nonattainment area
beyond the Chicago metropolitan area.  In fact, EPA designated six
counties within the Chicago combined statistical area as attainment
including Jasper, LaPorte, and Newton Counties in Indiana.  Indiana’s
comments does suggest it feels that Lake and Porter Counties, Indiana
are not a part of the Chicago area.  EPA has information that shows
these counties are an integral part of the Chicago area.  Commuting data
shows that nearly 42,000 workers commute from Lake County, Indiana into
Illinois and more 6,500 worker commute to Illinois from Porter County. 
Over 13,000 workers go into Lake and Porter Counties from Illinois.  The
Chicago PM2.5 (1997 standards) and the ozone nonattainment areas both
include Lake and Porter Counties, Indiana.  Lake and Porter Counties are
also included the Chicago combined statistical area as defined by the
Census Bureau.  EPA has concluded that Lake and Porter Counties in
Indiana are an integral part of the Chicago area and thus should not be
treated as a separate nonattainment area.  Illinois and Indiana should
work together to develop a plan to bring the entire area into
attainment.

Indiana:

Indiana provided information on emissions from Lake and Porter Counties.
 It listed the PM2.5 and NOX emissions from the Indiana portion of the
Chicago area and compares them to the emissions from the entire area. 
Indiana noted that “most of the area in Lake and Porter Counties is
urban” and that the two counties emit 66.3 percent of NOX and 75.3
percent of PM2.5 emissions in the Indiana portion of Chicago.  Indiana
included a chart with the 2002, 2010, 2020, and 2030 emissions of PM2.5
and NOX for the Illinois portion, the Indiana portion, and the Chicago
area.  The chart shows the Indiana portion responsible for about 15
percent of the Chicago area emissions in 2002.  That amount decreases in
the future projections.

EPA Response:

EPA agrees that most of the emissions in the Indiana portion of the
Chicago area are from Lake and Porter Counties.  That is why EPA
included Lake and Porter Counties in the Chicago nonattainment area and
designated Jasper, LaPorte, and Newton Counties as attainment.  It is
not clear where the Indiana emissions data provided by the State is
from.  EPA used emission date from the 2005 National Emissions Inventory
is assessing areas for the designations under the 2006 PM2.5 standards. 
The 2005 emissions from Lake and Porter Counties of PM2.5 and NOX are
much higher than the 2002 Indiana portion totals the State provided. 
Indiana did not list SO2 emissions on it chart.  EPA also concluded that
the PM2.5, NOX, and SO2 emissions from Lake and Porter Counties
contribute to the violations being monitored in the Chicago area.  The
future reductions projected on the table are impressive.  As it is a
projection, EPA cannot use this information in designation decisions.   
 

Indiana:

Indiana stated that, “Lake and Porter Counties are subject to the most
stringent group of emission controls within the State of Indiana.  This
collection of permanent and enforceable controls is equally as stringent
as those applied elsewhere in the Chicago MSA and in some cases, more
stringent.

EPA Response:

EPA commends Indiana for applying strict controls to its sources in Lake
and Porter Counties.  Reducing emissions throughout the Chicago area
helps improve the air quality and move the area closer to attaining the
air quality standards.

Indiana:

Indiana notes that vehicles registered in Lake and Porter Counties are
subject to reformulate gasoline and vehicle inspection and maintenance
programs.  Indiana tests all 1976 model year and newer vehicles in its
program.  Indiana states that Illinois does not test pre-1996 model year
vehicles in its program.  Indiana feels that the Illinois program misses
“high-emitters” by exempting pre-1996 vehicles.

EPA Response:

The Illinois inspection and maintenance program and others around the
nation discontinued testing pre-1996 model year vehicles because there
are few of these vehicles left in operation.  Even though the vehicles
do emit more, the decreasing number of vehicles remaining on the road
means that the emission reductions from testing older vehicles are
small.  A large amount of mobile source emissions come from sources
other that cars and light trucks, such as heavy trucks.  The Lake and
Porter Counties mobile source emissions from non-road sources had higher
PM2.5 and SO2 and similar NOX emissions as the on-road inventory (2005
NEI data).

Indiana: 

NOX emissions within Northwest Indiana area are projected to decline by
almost 42 percent between 2005 and 2020.

“[The] [e]mission reduction benefits from federal rules are factored
into the changes.  …Further, due to implementation of the NOX SIP Call
across the eastern Unites States, fine particles and precursors for fine
particles emissions entering this area are also decreasing.”

EPA Response:

As Indiana stated, the benefits of federal rules cutting emissions of
PM2.5 and its precursors are beginning to benefit the Chicago area and
other areas throughout the nation.  National emission reduction program
help decrease the regional background concentration of PM2.5.  However,
such future projected emissions reductions do not negate the obligation
under section 107(d) for EPA to designate as nonattainment those areas
that are violating, or contributing to violations of, the 200 PM2.5
NAAQS based on the most recent three years of data.  Congress amended
section 107(d) to eliminate provisions that previously directed EPA to
base designations for selected NAAQS based on projected future
attainment status (e.g., for SO2).  Section 107(d) is now phrased in the
present tense, and EPA interprets this provision to require designations
based upon current facts, not some projected future facts.  To the
extent that current federal programs will result in near term emission
reductions, these will be reflected in the nonattainment area SIPs
developed for these areas, and thus will play an important role in
helping these areas to attain the NAAQS as expeditiously as practicable.
 They are not, however, a basis for ignoring current violations or
contribution to violations.

Indiana: 

Over 80 percent of Lake County’s workforce is employed within the
county.  Over 60 percent of Porter County’s workforce is employed
within the county, and the majority of those employed outside the county
commute to Lake County, and not Chicago.  Therefore, the portion of
commute traffic with Chicago from Lake and Porter Counties is
insignificant.  Indiana adds, “gasoline-powered commuter vehicles are
insignificant emitters of sulfates, which is the driving precursor for
violating sites in Northeast Illinois.”

EPA Response:

The commuting data helps show a link between the Illinois and Indiana
portions of the Chicago area.  Commuting data shows that nearly 42,000
workers commute from Lake County, Indiana into Illinois and more 6,500
worker commute to Illinois from Porter County.  Over 13,000 workers go
into Lake and Porter Counties from Illinois.  There are also nearly
22,000 Porter County workers commuting into Lake County, Indiana.  The
number of workers commuting between the Illinois and Indiana portions of
the area is large even if the percentages are not because of the large
populations of the area counties.  EPA does not agree that SO2 is  the 
“driving precursor” in an area because the direct PM2.5 emissions
and all PM2.5 precursor emissions in an area cumulatively add to the
ambient PM2.5 concentration in a violating area.  The PM2.5
concentration is not the result of one precursor solely.  Reductions of
direct emissions or in any precursor will benefit the air quality.  
Details about appropriate control strategies for any area, including the
evaluation of which sources of which pollutants to control, consistent
with the requirements of section 172, in order to attain the NAAQS
expeditiously is among the questions the State and EPA will resolve
during the development of the nonattainment area SIP for such area.    

Indiana:

Indiana further analyzed the days that the monitoring levels in the
Chicago area were substantially higher than the next highest monitor
value in the region on that date, as shown below.  Indiana then compared
those days to wind data to determine the direction of prevailing winds
during those days.  …  The comparison for the highest monitor value
days for the Chicago monitors show winds from different directions with
winds predominately from the east, south, west, and southwest.  Based on
this sample of high fine particle monitor value days, while emissions
from all surrounding areas may have small impacts, it is evident that
there is no significant impact from Northwest Indiana.  … Further
analysis of high fine particle days included a back trajectory analysis.
 A back trajectory measures the winds at different heights in the
atmosphere to determine from what locations pollutants may be picked up
and transported to an area.

EPA Response:

Indiana examined the meteorology on only three days, October 27, 2004,
February 3, 2005, and January 23, 2006.  The conclusion that emissions
from the Indiana portion of the area do not contribute to violations
cannot be reasonably reached using such a small data set.  As the PM2.5
concentration in an area involves complex atmospheric reactions of the
precursor pollutants that blend with direct PM2.5 emissions, one cannot
say the wind drew emissions from a downwind area to create the
concentration being measured at a monitor.  Therefore, EPA does not
consider Indiana’s hypothesis to be proved true from the limited
sample it analyzed.   

Indiana:

Culpability analysis prepared and submitted by Indiana to EPA on April
3, 2008 demonstrates that not only are Lake and Porter counties not
culpable for monitored violations in northeast Illinois, counties
designated as attainment in southeast Wisconsin are more culpable to
measured levels in northeast Illinois than are Lake and Porter counties.
 Analysis of trajectory information for four days in which
concentrations in Chicago were significantly higher than elsewhere
indicated that three of these four days had winds from directions other
than from Northwest Indiana.  

EPA Response:

On April 3, 2008, Indiana submitted a request that EPA redesignate Lake
and Porter Counties to attainment for the annual PM2.5 standard; this
submittal included culpability analysis for annual average PM2.5 for
various sites in the Chicago-Northwest Indiana area.  Indiana provided a
synopsis of this analysis as Appendix D of its comment letter to its
24-hour PM2.5 designations submittal of May 30, 2008.  EPA disagrees
with this comment for the following reasons.  First, this is not the
proper forum for addressing culpability for annual average PM2.5
concentrations.  This analysis of culpability for annual average PM2.5
concentrations does not clearly address contributions to 24-hour PM2.5
concentrations.  Second, based on 2005-2007 air quality data, Lake
County, Indiana, now has a violating monitor for the 24-hour standard. 
EPA’s analysis concludes that adjacent Porter County, Indiana (24,000
tons SO2; 30,000 tons NOx), is contributing to this violation.  Third,
the limited information presented in this analysis does not clearly
demonstrate that Lake and Porter Counties do not contribute to other
violating monitors in the Chicago nonattainment area.  In fact,
Indiana’s analysis is reasonably consistent with EPA’s findings,
concluding not that Lake and Porter counties contribute on every day
with high concentrations in Chicago, but that the winds carry
significant emissions from those counties toward violating monitors in
Illinois with sufficient frequency to warrant a determination that Lake
and Porter counties contribute to violations in Chicago.

EPA Response to Additional Modeling Results Referenced in Indiana
Comment Letter:

Indiana Department of Environmental Management (DEM) states that the
fine particle contribution from Indiana’s portion of the Chicago
nonattainment area (Lake and Porter counties) is less than 1.0 μg/m3
and that these results verify that Lake and Porter counties do not
contribute significantly to measured values in Illinois.  

μg/m3 contribution for the annual standard of 15 μg/m3 is clearly a
significant contribution and for a short term standard like the 24-hour
PM2.5 NAAQS.  This contribution could be much higher when monitors in
the southwest region of Lake Michigan indicate a violation of the
standard.  This contribution would also be larger if current year
emissions during 2005-2007 were used in the modeling analysis since
mobile source emissions are higher and other projected 2012 controls
would not be in place.  For these reasons, the modeling submitted by
Indiana is not appropriate for the purpose of designating the Chicago
24-hour PM2.5 nonattainment area, but clearly shows a significant
contribution to Chicago.

Chico, CA

CARB:

The only violating monitor in Butte County is located in the City of
Chico, which has a 2007 Design Value of 55 μg/m3.  Chico, the largest
urban area in Butte County, has a population three-to-five times other
areas in the county.  Based on the localized nature of the primary
emission contribution to winter PM2.5, CARB considers the urban area of
Chico an appropriate nonattainment boundary for PM2.5.

A diurnal analysis of concentrations at Chico and Gridley, during Chico
exceedance days, highlights the localized nature of the PM2.5 pollution
episodes.  The nighttime increases at Chico, the result of residential
wood burning, are not reflected at the monitoring site at Gridley.  The
majority of exceedance days occur during periods of stagnant or low
wind, keeping pollutants close to the emission sources.  

EPA Response:

CARB has recommended only the City of Chico as the nonattainment area of
Butte County.   There are a number of reasons why EPA disagreed with
this recommendation.  In developing the boundaries for Butte County, EPA
considered several factors contained in Agency guidance for PM2.5
designations, including air quality data, emissions data, population
data, commuting patterns, geography, topography, meteorology and other
information such as CES scores, chemical speciation data, and wintertime
inversion data.  EPA also considered information provided by the State
of California such as information on residential wood burning.

Exceedances of the PM2.5 standard most often occur in Chico during the
winter months and speciation data suggest that residential wood burning
and mobile source emissions are the most important sources.  Area source
data show that residential wood burning is the dominant source of PM2.5
and, therefore, could be linked to PM2.5 exceedances measured in Chico. 
This data also suggests that other population centers in Butte County
could experience high PM2.5 levels as a result of residential wood
burning.  With respect to mobile sources, the Butte County inventory
shows significant mobile source emissions which can also contribute to
regional PM2.5 pollution.  Since both sources are associated with the
population centers, EPA has determined that the nonattainment area in
Butte County should include Chico, Paradise and Oroville, and Gridley.  


The area recommended by California, namely the city limits of the City
of Chico, does not include all of Chico’s urban population nor does it
include other population centers in Butte County, including Paradise,
Gridley and Oroville.  The total population of Butte County is 214,000. 
According to the 2000 US Census, the population of Chico was 59,444, the
population of Paradise was 26,408 and the population of Oroville was
13,004.  Moreover, the City of Chico has been experiencing significant
growth, growing from 59,954 to 71,728 between 2000 and 2006, and we
would expect future growth to continue extending beyond the existing
city boundaries.  

Finally, topography is considered to be an important factor given that
inversion layers during the winter, when PM2.5 exceedances typically
occur, can contribute to higher pollution levels in the Sacramento
Valley.  In addition to affecting the City of Chico, these conditions
are expected to create similar pollution conditions in areas with large
populations and commuting patterns and, thereby, provide further reason
to expand the nonattainment boundary beyond the City of Chico.  

Cincinnati-Hamilton, OH-KY-IN

The State of Kentucky

Kentucky:

To have Boone County, Campbell County, and Kenton County designated
nonattainment would invoke additional, substantial, unnecessary
requirements on local government planning agencies with little or no
benefit to the area.  Substantial local emission reductions from Boone
County have already occurred, or will have occurred well before
attainment dates for this standard.  Drastic emission reductions are
scheduled to occur in the mobile sector throughout the next several
years that will greatly impact pollutant levels in the area.  Couple
these changes with those anticipated by the CAIR, or its replacements,
provisions which will further reduce SOX and NOX emissions within the
region, and the air monitoring data demonstrating attainment of the
PM2.5 standard, and the result should be that these counties be
designated attainment for the PM2.5 standard.

EPA Response:

EPA’s 9-factor analysis has determined that emissions from Boone
County, Campbell County, and Kenton County contribute to violations in 
violations in nearby Hamilton and Butler counties, Ohio, and that it is
appropriate to include these counties in the Cincinnati-Hamilton
nonattainment area.  EPA believes that consideration of all 9 factors
compels designating these counties nonattainment, consistent with the
existing nonattainment boundary for the annual PM2.5 standard.  As
stated in Section V, EPA determined the existence of violations of the
2006 24-hour fine particulate NAAQS based solely on the most recent
three complete years of certified monitoring data.  EPA is setting
boundaries based on available current data and analyses that best
represent present-day conditions, and not on projected emissions
reductions that may occur after designation areas are finalized.   This
is consistent with the explicit provisions of section 107(d), that
define “nonattainment” to be those areas that violate or contribute
to violations of the NAAQS.

While EPA acknowledges that local emission reductions have been and
continue to be achieved in the area, we are considering only the current
emission controls in place for designation purposes.  Any local emission
reductions in the area that occur before the required attainment date
will help further improve air quality and should be included in the
attainment plan.  However, the uncertainty of future emission
reductions, and also any future new or expanded sources, renders
assumptions about them unsuitable for use in establishing nonattainment
area boundaries.  

μg/m3 for the 2005-2007 period indicates that considerable emission
reductions are still needed for the area to attain.  Analysis of
chemical composition data for Cincinnati indicates that the same
components that make up most of the PM2.5 mass in the area on an annual
average basis (such as sulfate and direct PM2.5 carbon) also are key
contributors to the PM2.5 mass on days exceeding the 24-hour PM2.5
standard.  In addition, these data show that nitrate (formed from NOX
emissions) can also be an important component on high days in the cool
season.  For example, nitrate was 44 percent of PM2.5 mass on a cool
season high day in the 2005-2007 period.  These data indicate that the
same source categories that contribute to violations of the annual
standard also contribute to exceedances of the 24-hour standard.  EPA
concludes that counties that were designated as having emissions sources
contributing to fine particle concentrations which continue to exceed
the 1997 standards also contribute to fine particle concentrations on
the highest days.  This is another reason that EPA believes that Boone,
Kenton, and Campbell Counties should be included in the nonattainment
area for the 2006 24-hour standard as well.

The State of Kentucky makes the following points as additional rationale
for excluding Boone, Kenton, and Campbell Counties from the Cincinnati
nonattainment area.

Kentucky:

Kentucky believes that EPA’s use of the contributing emissions scoring
approach was skewed.  A review of actual percentages of emissions
contributions to the area shows that Boone, Kenton, and Campbell
Counties do not have the potential to contribute to PM2.5 levels within
the region.

EPA response: 

The CES is an initial screening tool used to determine potential
nonattainment area boundaries which can then be modified using
additional information provided by the State or other sources.  EPA has
recommended other types of information for this purpose through guidance
which could be considered in the final decision. 

The CES for Boone, Kenton, and Campbell are lower than those for other
contributing counties (Clermont, OH; Dearborn, IN) in the area. 
However, EPA does not find that emissions from these counties (Boone:
5383 tons of SO2, 10852 tons NOX, and 1600 tons direct PM2.5; Kenton: 
1300 tons SO2, 6316 ton NOX, 537 tons PM2.5; Campbell:  730 tons SO2,
4231 tons NOX, 412 tons PM2.5) are without the potential to contribute
to PM2.5 in the region.  One key source of emissions located in Boone
County is the Cincinnati-Northern Kentucky International Airport.

Kentucky:

Contributions from commuters and vehicle miles traveled in Boone,
Kenton, and Campbell Counties have been shown to have no potential to
impact PM2.5 levels within the region when compared to the levels from
other counties and therefore should not be used to determine
nonattainment status for this county.

EPA Response:

EPA believes that commuting and vehicle miles traveled data are an
important factor in the inclusion of these counties in the nonattainment
area.  Boone County has 1 million VMT and 43,000 commuters annually;
Kenton County has 1.6 million VMT and 74,000 commuters annually; and
Campbell County has 1 million VMT and 42,000 commuters annually.  EPA
has not suggested any threshold of significance for VMT and commuting,
but EPA believes these VMT and commuting data indicate that these
counties are an integral part of the Cincinnati-Hamilton urban area and
that vehicle emissions from these counties are likely contributing to
elevated fine particle concentrations on high days.   

Kentucky:

The populations of Boone, Kenton, and Campbell Counties are not
significant enough to have the potential to impact PM2.5 levels in the
region.  Each county’s population represents a small percentage of the
population of EPA’s proposed nonattainment boundaries.

EPA Response:

The 2005 population of Boone County is 106,000; Kenton County is
153,000; and Campbell County is 87,000.  All three have a moderate level
of population and population density when compared to other contributing
counties in the area and across the country.  Together, the population
of the northern Kentucky counties is substantial at around 350,000.  As
noted above, the residents of these counties are responsible for a
substantial amount of vehicle miles traveled and commuting in the
Cincinnati area, and EPA believes this contributes to PM2.5 NAAQS
exceedances in the area.

Kentucky:

There are no violating monitors in Boone, Kenton, or Campbell County.  

EPA Response:

μg/m3; Butler, OH with design value of 38 μg/m3).  

Kentucky:

Duke Energy East Bend Station Unit 2, which is located in Boone County,
Kentucky, is an electric power generating station consisting of one
pulverized coal-fired, dry bottom, wall-fired boiler.  The boiler (Unit
02) has an input capacity of 6313 mmBtu/hr.  Unit 2 is equipped with an
electrostatic precipitator (ESP), FGD unit, low nitrogen oxide burners
and a SCR unit.  As can be seen in the table on page 11 of Kentucky’s
comment letter, these controls provide for significant reductions in
precursor PM2.5 pollutants.  Although at this point in time it is
uncertain what reductions can be attributed to CAIR or its replacement,
it is certain that some manner of control similar to CAIR will be
implemented.  This will further reduce these emissions since the SCR
should be operated year-round instead of only during the ozone season. 
Kentucky’s NOX SIP Call regulations remain in effect and if CAIR
continues to be in limbo will cover the 2009 allocation timeframe.

EPA Response:

See the overview response to comments on power plant issues, located at
the beginning of Section IV.  

EPA acknowledges that the East Bend Station in Boone County has emission
controls in place for both SO2 and NOX.  Emissions in 2007 were 2452
tons of SO2 and 5563 tons of NOX.  However, EPA has not included Boone
County in the nonattainment area solely on the basis of the presence of
the East Bend plant.  Based on the amount of emissions from the plant,
the geographic location of the plant, and the meteorological data for
this area, and the speciated data for this area indicating that such
emissions play a role in the formation of the ambient PM2.5 in this
area, EPA has concluded that this source still contributes to violations
in the area, as that term is intended in section 107(d).  Therefore, it
is appropriate that the State and EPA will evaluate the source in the
nonattainment area SIP development process.   

EPA notes that reductions required by CAIR, or any replacement rule
under section 110(a)(2)(D) or section 126, to reduce interstate
transport would not be in lieu of the requirements under section 172 for
nonattainment areas.  Thus, additional controls may also be necessary
for local attainment needs.  Section 110(a)(2)(D) and section 126
address a different issue than section 107(d).

A number of other factors in EPA’s decision for inclusion of Boone
County are explained in the responses above.  

Kentucky:

The air monitoring data demonstrates attainment of the PM2.5 Standard.

EPA Response:

Although there are no ambient monitors in Boone and Campbell Counties
and the ambient monitor in Kenton County currently shows attainment,
there are several violating monitors in nearby Hamilton County, which
has a design value of 41 (g/m3, based on 2005-2007 data.  Analysis of
all additional data described in the TSD indicates these counties are
contributing to the violating monitor.  Section 107 requires designation
of areas that, although not they themselves violating, are contributing
to nearby nonattainment.

The State of Indiana

Indiana:

AEP-Tanners Creek (in Dearborn County) will be installing Selective
Non-Catalytics Reduction (SNCRs) on three of its four units, with
operation to begin in mid-2009.  This will achieve an additional 30
percent reduction in NOX.  Indiana believes that emissions from Dearborn
County do not affect the downwind area’s ability to attain the 24-hour
standard.  EPA must substantiate its assumption that this township is
contributing to upwind monitored violations with source apportionment
analysis and a model-based culpability analysis.

EPA Response:

See the overview response to comments on power plant issues, located at
the beginning of Section IV.

EPA disagrees with Indiana regarding its claim that emissions from the
Tanner’s Creek power plant located in the existing partial county
nonattainment area in Dearborn County do not contribute to PM2.5
concentrations in Cincinnati.  EPA’s analysis has evaluated emission
levels, speciation data, evidence from the pollution rose and other
meteorological data for this area that there is contribution from the
direction of this source to the rest of the nonattainment area, and the
level of emission controls on the source.  The Tanner’s Creek plant
has four units, none of which have a scrubber or SCR.  Emissions in
2005-2007 were substantial:  in the range of 33,000 to 46,000 tons of
SO2 and about 9000 tons NOX.  The plant is located close by, about 17
miles from the closest violating monitor.  EPA is assessing current
contributions to current violations, regardless of whether emissions may
decline in the future, consistent with the requirements of section
107(d).  Indiana does not address SO2 emission control at this plant;
SO2 emissions are currently relatively high, and EPA expects SO2
emissions from this facility to remain relatively high

Cincinnati is an existing PM2.5 nonattainment area which has not
attained the 1997 standards.  Its 2005-2007 annual design value is still
well above the standard at 17.3 μg/m3, indicating that considerable
emission reductions are still needed for the area to attain.  Analysis
of chemical composition data for Cincinnati indicates that the same
components that make up most of the PM2.5 mass in the area on an annual
average basis (such as sulfate and direct PM2.5 carbon) also are key
contributors to the PM2.5 mass on days exceeding the 24-hour PM2.5
standard.  In addition, these data show that nitrate (formed from NOX
emissions) can also be an important component on high days in the cool
season.  For example, nitrate was 44 percent of PM2.5 mass on a cool
season high day in the 2005-2007 period.  These data indicate that the
same source categories that contribute to violations of the annual
standard also contribute to exceedances of the 24-hour standard.  EPA
concludes that counties that were designated as having emissions sources
contributing to fine particle concentrations which continue to exceed
the 1997 standards also contribute to fine particle concentrations on
the highest days.  This is another reason that EPA believes that
Dearborn county should be included in the nonattainment area for the
2006 24-hour standard as well.  

EPA also disagrees with the commenter’s supposition that EPA must
“substantiate its assumption that this township is contributing to
upwind monitored violations with source apportionment analysis and a
model-based culpability analysis.”  Nothing in section 107(d) requires
that a modeling analysis be conducted to determine a “contribution”
and to make a nonattainment designation.  The best available evidence
indicates that Dearborn County contributes to violations in the
Cincinnati area.  

Clarksville, TN-KY

The State of Tennessee

Tennessee:

The State of Tennessee believes that Montgomery County should be
classified as “unclassifiable” by the December 18, 2008, deadline,
and that monitoring data support the State’s declaration that
Montgomery County will measure attainment.  The State of Tennessee
claims that ambient PM2.5 data for 2007 may be skewed by the impact of
Georgia and Florida wildfires during 2007.  Data for 2008 will be
certified before the February 2009 deadline to have relatively low
readings for 2006-2008 considered.

The State of Tennessee believes Stewart County and Humphreys County
should be classified as “attainment.”  However, if EPA designates
Humphreys and Stewart Counties as nonattainment, the State believes EPA
should designate only significantly contributing portions of Humphreys
and Stewart Counties.  The State notes that EPA has previously concluded
that a partial area of a county may be designated nonattainment if that
partial area includes the emissions point of concern.  Tennessee
believes this same situation exists for both counties of Humphreys and
Stewart, if both are non-attaining.  The State of Tennessee proposes as
boundaries for the partial areas census block groupings that include the
Tennessee Valley Authority (TVA) New Johnsonville Fossil Plant in
Humphreys County and TVA Cumberland Fossil Plant in Stewart County.

EPA Response:

EPA finds that ambient monitoring data for Montgomery County is complete
for 2005 – 2007 and therefore there is no basis to designate the
county “unclassifiable” at this time.  Therefore, EPA must designate
Montgomery nonattainment based on valid data for 2005 - 2007.

EPA acknowledges Tennessee has proposed as boundaries for the partial
areas census block groupings that include the TVA New Johnsonville
Fossil Plant in Humphreys County and TVA Cumberland Fossil Plant in
Stewart County.  EPA agrees with the proposed census block boundaries
and is designating only the contributing portions of Humphreys and
Stewart Counties.

EPA acknowledges that SO2 and NOX emissions have declined at TVA
Cumberland Fossil Plant and TVA Johnsonville Fossil Plant.  However, EPA
has determined that, even at current reduced emission levels, the
surrounding partial-county areas should be designated as nonattainment
based on their contribution to the violating monitor in nearby
Montgomery County.

Tennessee:

The primary sources of concern are two coal-fired power plants: the TVA
New Johnsonville Fossil Plant in Humphreys County and TVA Cumberland
Fossil Plant in Stewart County.  The proposed boundaries for the partial
areas that include these power plants are census block groupings as
listed in Table 1 and mapped in Attachment 1 (Humphreys County) and
Attachment 2 (Stewart County) of Tennessee’s comment letter.

If EPA determines that Humphreys and Stewart Counties must be designated
as nonattainment, the EPA should designate only the census block
groupings indicated in Table 1 and the attachments.  At each of the TVA
power plants, SO2 and NOX emissions continue to decline, and any further
control methodologies would be federally enforceable through operating
permit conditions and continued NOX SIP Call provisions within Tennessee
Air Pollution Control Regulations.  As stated in TVA's letter to EPA,
dated October 1, 2008, TVA Cumberland Fossil Plant in Stewart County has
two combustion units with each unit's emissions controlled with low-NOX
burners, SCRs, ESPs, and wet limestone scrubbers.  Also, the TVA
Johnsonville Fossil Plant in Humphreys County continues to utilize
electrostatic precipitators and selective non-catalytic reduction as
emission controls on all ten units.

EPA Response:

See the overview response to comments on power plant issues, located at
the beginning of Section IV.

The CAA requires that a nonattainment area must include not only the
area that is violating the standard, but also nearby areas that
contribute to the violation.  Thus, for each monitor or group of
monitors that indicate violations of a standard, EPA is establishing
nonattainment boundaries that cover a sufficiently large area to include
both the area that violates the standard and the areas that contribute
to the violations.  The Clarksville area includes the TVA New
Johnsonville Fossil Plant in Humphreys County (2007 emissions: 65,000
tons SO2 and 19,000 tons NOX) and TVA Cumberland Fossil Plant in Stewart
County (2007 emissions: 17,000 tons SO2 and 33,000 tons NOX).  The New
Johnsonville Plant currently has no scrubbers or SCR installed on its 10
units.  The Cumberland Plant, while relatively well-controlled, still
has high SO2 and NOX emissions and is located in close proximity to the
violating monitor.  EPA has concluded that these sources do factually
contribute to the violations in the nearby area, notwithstanding the
current level of control at the Cumberland plant.  Under section 107(d),
EPA believes that the level of control of emissions from a source is a
relevant inquiry in the designations process to assure that EPA and the
State are using up to date and accurate emissions information.  EPA does
not believe that the level of control alone should be used as a basis to
exclude from a nonattainment area a source that is otherwise deemed to
be contributing to the violations based upon information such as its
amount of emissions, geographic location, and meteorological connection
to the nearby violating area.  

The State of Kentucky

Kentucky:

To have Muhlenberg County designated nonattainment would invoke
additional, substantial, unnecessary requirements on local government
planning agencies with little or no benefit to the area.  Substantial
local emission reductions from Muhlenberg County have already occurred,
or will have occurred well before attainment dates for this standard. 
Drastic emission reductions are scheduled to occur in the mobile sector
throughout the next several years that will greatly impact pollutant
levels in the area.  Couple these changes with those anticipated by the
CAIR, or its replacements, provisions which will further reduce SOX and
NOX emissions within the region, and the air monitoring data
demonstrating attainment of the PM2.5 Standard, and the result should be
that these counties be designated attainment for the PM2.5 Standard.

EPA Response:

EPA commends Muhlenberg County for local emission reductions that have
been and continue to be achieved.  However, EPA’s 9-factor analysis
has determined that emissions from Muhlenberg County contribute to
nonattainment in Montgomery County, Tennessee, and that it is
appropriate to include this County in the Clarksville nonattainment
area.  EPA believes that consideration of all nine factors compels
designating this County nonattainment, in particular the boundaries
surrounding the TVA Paradise Power Plant and the Kentucky Utilities
Green River Power Plant.  As stated in Section V, EPA determined
violations of the 2006 24-hour fine particulate NAAQS based solely on
the most recent three complete years of certified monitoring data.  EPA
is setting boundaries based on available data and analysis that best
represents present-day conditions, and not on projected emissions
reductions that may occur after designation areas are finalized.  

The State of Kentucky makes the following points as additional rationale
for excluding Muhlenberg County from the Clarksville nonattainment area.

Kentucky:

Kentucky believes that EPA’s use of the contributing emissions scoring
approach was skewed.  A review of HYSPLITS on days of high monitor
readings shows that Muhlenberg County does not have the potential to
contribute to PM2.5 levels within the region.

EPA Response:

EPA disagrees with Kentucky that the HYSPLIT analysis is more refined
than the CES.  First, the HYSPLIT trajectories submitted by the State
show only days between 2004 and 2006.  Since Clarksville, TN violated
the 24-hour NAAQS with data collected from 2005 to 2007, data from that
three-year period were used to calculate the CES for the area. 
Furthermore, EPA trajectories were calculated for each “high” day at
three hour intervals across the day with each of the eight start times
starting at four heights at or below the mixing height.  They were
calculated at specific hours during which the air would have arrived at
the site during the high day.  

Second, the trajectories presented by Kentucky all arrive at the site
six hours after the end of the high day.  This is not as accurate in
determining whether air was passing over a contributing area on the
actual day of the exceedance.  Air could have been coming from other
directions and over other land areas during the high day in question
which would not be reflected in trajectories arriving at the site after
the high day.  EPA began trajectories at or below the mixing height in
all instances.  Only those points along the trajectory that remained
within the mixing layer were used in computing the CES.  It is not
certain whether the three starting heights the State used for their
backward trajectories were started at or below the mixing height.  Also,
since EPA started trajectories at four starting heights within the
mixing layer for eight different starting times, there were a total of
32 trajectories to examine to see if the air parcels followed consistent
paths throughout the day.  This level of robustness was not captured by
the State whose examples show several days where the three trajectories
that were run do not consistently follow the same path to Montgomery
County, TN.  When examining backward trajectories started at staggered
starting times across the “high” day and at multiple starting
heights, there are days when the trajectories consistently pass over
Muhlenberg County on their way to the Montgomery County, TN.  

Lastly, the results of the State run trajectories are plotted on maps
with only State boundaries given.  This provides no sense of spatial
resolution as to where Muhlenberg County is located in relation to the
trajectory paths.  

  

Kentucky:  

TVA Paradise, which is located in Muhlenberg County, is a facility that
consists of three cyclone-furnace coal-fired boilers.  All three
coal-fired boilers are equipped with staged overfire air and selective
catalytic reduction modules for nitrogen oxides emission control. 
Boiler Units 1 and 2 are equipped with venturi-type limestone slurry
flue gas desulfurization scrubbers for SO2 and particulate control. 
Boiler Unit 3 is equipped with an electrostatic precipitator and a wet
limestone FGD scrubber for particulate and SO2 control.  The Unit 3 FGD
scrubber that came online in 2006 has significantly reduced the TVA
Paradise SO2 emissions by 49,704 tons per year (tpy) given the Unit 3
2005 emissions (53,519 tpy) compared to the Unit 3 2007 emissions (3,815
tpy).  TVA Paradise’s total SO2 emissions in 2005 were 84,401 tpy
while in 2007 its SO2 emissions had decreased to 33,818 tpy due to the
addition of the Unit 3 scrubber.  In addition, although not considered
as BART since TVA had previously related to the Kentucky Division of Air
Quality (KYDAQ) its intent to install controls to mitigate SO2 emissions
at TVA Paradise, TVA plans to install hydrated lime injection controls
on TVA Paradise Units 1-3 at to address SO2 emissions and improve
visibility at Class I areas.  Given the TVA Paradise existing emission
controls that are in place that have significantly reduced the emissions
at TVA Paradise and additional controls planned, KYDAQ requests that EPA
reconsider its decision and not designate Muhlenburg County as
nonattainment for the 24-hour particulate matter standard (see table on
page 28 of Kentucky comment letter).

EPA Response:

See the overview response to comments on power plant issues, located at
the beginning of Section IV.  

In regard to the TVA Paradise plant in Muhlenberg County, an important
consideration for EPA is the level of emissions and contribution from
the plant even with emission controls in place.  As noted in the
State’s comment above, the TVA Paradise plant now has scrubbers and
SCR in place on all three units, yet 2007 emissions were still
substantial: 33,000 tons of SO2, and 43,000 tons of NOX.  In addition to
the TVA Paradise plant, the smaller Green River power plant is also
located nearby in Muhlenberg County.  It has two units without scrubbers
or SCR, and 2007 emissions totaling 22,000 tons of SO2 and 2000 tons of
NOX.  Thus, based on the most recent information available, the two main
sources of emissions in Muhlenberg County account for 55,000 tons of SO2
and 45,000 tons of NOX per year.  In addition, meteorological data
indicate that, on high PM2.5 days, emissions from Muhlenberg County can
be transported to the location of the violating monitor.  

Under section 107(d), EPA believes that the level of control of
emissions from the source is a relevant inquiry in the designations
process to assure that EPA and the State are using up to date and
accurate emissions information.  EPA does not believe that the level of
control alone should be used as a basis to exclude from a nonattainment
area a source that is otherwise deemed to be contributing to the
violations based upon information such as its amount of emissions,
geographic location, and meteorological connection to the nearby
violating area.  Based upon this evidence relevant to designations, in
accordance with section 107(d), EPA has concluded that emissions from
Muhlenberg County are contributing to violations in the Clarksville
area.  Thus, these sources will be evaluated by the State and EPA for
additional cortrols, as necessary, in the development of the
nonattainment area plan for this area.  It would be premature to
determine whether such sources are adequately controlled for local
nonattainment area needs at this time.

Kentucky:

Muhlenberg County, a county that is not in the Clarksville MSA and is
geographically separated from Montgomery County by other counties not
proposed as nonattainment, should not be identified as contributing to
such nonattainment status, and therefore should not be included within
the nonattainment boundary for the Clarksville nonattainment area.

EPA Response:

EPA agrees that, like Humphries County, TN, Muhlenberg County is not in
the Clarksville metropolitan statistical area and is geographically
separated from Montgomery County, TN, the site of the violating monitor
by other counties not proposed as nonattainment.  However, based on high
emissions from the power plants in these counties, coupled with
meteorological data indicating wind direction from these counties, EPA
believes sources in Humphries County, TN, and Muhlenberg County do
contribute to the violating monitor in Montgomery County, TN and thus
must be included within the designated nonattainment area.

Kentucky:

EPA’s proposal finds that only two of its nine delineated factors
weigh in favor of designating Muhlenberg County as contributing to
violations of the standard in Clarksville, Tennessee.  The results of
EPA’s assessment of the other seven factors should warrant designation
of Muhlenberg County as attainment.

EPA Response:

Section 107(d)(1)(A)(i) of the Act specifies that a nonattainment area
shall include “any area that does not meet (or that contributes to
ambient air quality in a nearby area that does not meet) the national
primary or secondary ambient air quality standard for the pollutant.” 
Based upon emission levels and emissions from electric generating units
in Muhlenberg County, EPA believes Muhlenberg County contributes to
ambient air quality that does not meet the 24-hour PM2.5 NAAQS in the
Clarksville area.  EPA considered all of the nine factors as well as
other data in each area in determining whether the area as a whole
contributed to nearby violating monitors.  If a few factors weighed
heavily in a conclusion that an area contributed, the area would be
included in the nonattainment area even if several other factors weighed
less significantly.  Further, the nine factors are not all equivalent
measures of potential nonattainment, and certain factors could carry
more weight in a given nonattainment area.  Based upon its analysis of
all the data, EPA continues to conclude that these areas should be
included in the Clarksville nonattainment area.

Kentucky:

The air monitoring data demonstrates attainment of the PM2.5 Standard.

EPA Response:

Although there are no ambient monitors in Muhlenberg County, the air
monitoring data demonstrates violating monitors in nearby Montgomery
County, TN, which has a design value of 38 (g/m3, based on 2005-2007
data.  Analysis of all additional data described in the TSD indicates
that Muhlenberg is contributing to the violating monitor.  Section 107
requires designation of areas that, although not violating, are
contributing to nearby nonattainment.

Cleveland-Akron-Lorain, OH

Ohio:

The State of Ohio objected to EPA’s inclusion of Ashtabula Township,
Ashtabula County within the Cleveland nonattainment area for the 2006
24-hour PM2.5 NAAQS.  Among other arguments, Ohio explained that a
significant change of circumstances at the power plant in Ashtabula
Township had drastically reduced emissions at that plant: 

“The actual emissions from Cleveland Electric Illuminating power
plant, located in Ashtabula Township (Ashtabula Co.), are less than a
three percent (3 percent) of its allowable emissions.  Besides having
electrostatic precipitators, the plant has only operated one unit since
2002.  In addition, Ashtabula County has insignificant emissions due to
low population and insignificant commuting traffic.  Ohio EPA believes
the above factors demonstrate that Ashtabula is not contributing to the
PM2.5 problems in this nonattainment area.  Moreover, a Cleveland area
windrose (enclosed) shows that, for most of the time, the wind is coming
from the southwest; consequently it is likely that high pollution levels
being detected in Ashtabula County are coming from counties south and
west of this County and not from local sources.”

By letter dated December 9, 2008, Ohio subsequently established that
three of the four boilers at this plant have permanently shut down, that
the permit for operating these boilers has expired, and that these
boilers are now prohibited from operating.  This letter confirmed that
the reductions of emissions from this plant are thus permanent.  

EPA Response:  

See the overview response to comments on power plant issues, located at
the beginning of Section IV.

EPA has determined that Ashtabula Township does not contribute to
violations of the 2006

24-hour PM2.5 NAAQS in the Cleveland area, and thus has concluded that
it is not necessary to include it within the designated nonattainment
area.  EPA notes that it is not the decrease in emissions in the
abstract that compels this conclusion.  EPA has considered the permanent
decrease in emissions at the source in light of other information
relevant to the designation.  With these shutdowns, the source has
decreased its emissions substantially.  These reductions have already
occurred, and are not future or contingent reductions.  Based on data
from the first six months of 2008, emissions from this plant were about
1500 tons SO2 and 500 tons NOX (equating to estimated annual emissions
of 3000 tons SO2 and 1000 tons NOX).  

These smaller amounts of emissions have rendered the source a relatively
small source in the context of the Cleveland area.  These reductions
should be viewed in terms of the meteorological data and other
information relevant to the Cleveland area for the 2006 24-hour PM2.5
NAAQS.  EPA examined back trajectory information for days with high
concentrations at the area’s violating monitors, and found that winds
blow occasionally but less frequently from Ashtabula County on these
critical days than from other area counties.  The pollution rose for a
critical monitor in Cuyahoga County (representative of conditions under
which high concentrations occur in the area) indicates that impacts from
the direction of Ashtabula Township to the violating monitors in
Cuyahoga and Summit Counties are relatively low, compared to
contribution from other areas.  The CES score provides an integrated
consideration of emission levels and frequency of being upwind; the CES
scores for the Cleveland area counties indicated that Ashtabula County
has among the smallest contributions to the Cleveland area.  As noted
elsewhere, EPA believes that the designations for the 2006 24-hour PM2.5
NAAQS must be based on the facts and circumstances of each geographic
area.  Unlike other partial counties within Ohio, EPA believes that the
source in Ashtabula Township is not contributing to violations of the
24-hour PM2.5 NAAQS in a nearby area.     

It should be noted that EPA does not agree with other arguments made by
Ohio to support exclusion of Ashtabula Township from the Cleveland area.
 The fact that actual emissions from the plant are only 3 percent of
“allowable emissions” does not eliminate the possibility for
contribution, and in fact would raise concerns that the source could
drastically increase emissions consistent with its allowable limits. 
The concept of “allowable emissions” is not germane to the
designations process under section 107(d), and the fact that a source
might be able to emit some amount of a pollutant without violating other
Federal or State statutory, regulatory, or permit limits does not
establish whether that source is contributing to violations of a NAAQS
in a nearby area.  Similarly, the fact that the source has certain
controls is relevant, but does not answer whether the source is
contributing notwithstanding the level of control.  The amount of
population in and commuting from Ashtabula County were not under
consideration as a basis for inclusion of Ashtabula Township; that
proposed inclusion was based upon concerns about the amount of emissions
from the electric generating plant.  While EPA agrees that the pollution
rose for the Cleveland area does support the exclusion of Ashtabula
Township, EPA would not agree that winds are always blowing from the
southwest in this area, or that all pollution transport in this area of
the country always comes from that direction.  To be clear, EPA believes
that the pollution rose, the CES score, and the back trajectories for
this area all reasonably support the conclusion that the source in
Ashtabula Township is not contributing to violations of the 2006 PM2.5
24-hour NAAQS in Cuyahoga and Summit Counties

It should also be noted that EPA’s decision not to include Ashtabula
Township within the Cleveland nonattainment area for the 2006 24-hour
PM2.5 NAAQS does not affect or negate the inclusion of Ashtabula
Township within the Cleveland nonattainment area for the 1997 annual
PM2.5 NAAQS.  To the extent that emissions of direct PM2.5, SO2, and NOX
from the source in Ashtabula Township contribute to violations of the
annual PM2.5 NAAQS in the Cleveland area, Ohio will be required to
evaluate those emissions, and control them as appropriate, to provide
for attainment of the annual PM2.5 NAAQS as expeditiously as practicable
in this area.  

Columbus, OH

Ohio:

The State of Ohio objected to EPA’s proposed inclusion of Franklin
Township, Coshocton County, within the nonattainment area for Columbus
for the 2006 24-hour PM2.5 NAAQS.  In its October 8, 2008, letter, Ohio
generally objected to this on the grounds that “the existence of an
electric generating unit in a township should not be the sole factor in
determining nonattainment.”  More specifically, Ohio argued in that
letter that Franklin Township should be excluded because the source
located there is adequately controlled, or will be in the future:

“The emissions from the Conesville power plant located in Franklin
Township (Coshocton Co.) have long been controlled with efficient
control equipment, and some of the units are expected to have additional
controls by the year 2009.  Currently there are four operating units and
two units that shut down permanently in 2006.  Besides electrostatic
precipitators, these operating units have installed or are planning to
install flue gas desulfurization systems.  Moreover, the largest unit
(800 MW) will have a SCR system in place and operating for the entire
ozone season, starting in June 2009.  In addition this unit (800 MW)
will have SCR equipment in continuous operation (the whole year)
starting in year 2011, increasing the nitrogen oxides emission
reductions.”  

EPA Response:

See the overview response to comments on power plant issues, located at
the beginning of Section IV.

EPA disagrees with the reasons provided by Ohio for exclusion of
Franklin Township from the Columbus nonattainment area.  First, Ohio is
incorrect that the “existence” of the source is the sole reason for
EPA’s inclusion of Franklin Township.  EPA’s analysis has evaluated
not just the existence of the source, but also other considerations
including:  the significant quantity of emissions of the pollutants
(including PM2.5 precursors and directly emitted PM2.5) that speciated
PM2.5 data in this nonattainment area show to be of concern; evidence
from the pollution rose for this area that there is contribution from
the direction of this source to the rest of the nonattainment area; and
the CES score that likewise indicates that emissions from Franklin
Township contribute to violations in the Columbus area.  

For example, emissions information available to EPA indicates that the
Conesville plant currently emits approximately 100,000 tons per year of
SO2 (over 106,000 tons in 2005, over 90,000 tons in 2006, and over
114,000 tons per year in 2007).  Similarly, the emissions of NOX from
this source are approximately 20,000 tons per year.  Emissions of this
magnitude place the Conesville plant among the largest sources of PM2.5
precursors in the country and make it the largest individual source in
the Columbus area.  Meteorological data derived in support of the CES
scores show that elevated levels of ambient PM2.5 in some cases occur on
days when winds are blowing generally from the direction of Franklin
Township towards the rest of the Columbus nonattainment area.  This
conclusion is further supported by the CES scores for the Columbus area,
which includes a score of 16 for Coshocton County (principally
attributable to Franklin Township), a score that is higher than for
other counties in the Columbus nonattainment area that Ohio agrees
should properly be within the nonattainment area.  Based upon this
evidence relevant to designations in accordance with section 107(d), EPA
concluded that emissions from Franklin Township are “contributing”
to violations in Columbus.  In the context of the nonattainment SIP for
this area, Ohio will have the opportunity and the obligation to perform
further analysis and attainment demonstration modeling that can further
explore the contribution of the source and the appropriate course of
action to reduce that contribution.

EPA also disagrees with Ohio’s conclusions relating to the level of
control of the Franklin Township source as a basis for excluding it from
the nonattainment area.  As noted above, EPA has concluded that the
source in Franklin Township does contribute to the violations,
notwithstanding its current level of control.  Under section 107(d), EPA
believes that the level of control of emissions from the source is a
relevant inquiry in the designations process to assure that EPA and the
State are using up to date and accurate emissions information.  EPA does
not believe that the level of control should be used as a basis to
exclude a source that is otherwise deemed to be contributing to the
violations, based upon information such as its amount of emissions,
geographic location, and meteorological connection to the nearby
violating area.  Whether sources in an area that is violating or is
contributing to violations of the NAAQS are adequately controlled is
more appropriately assessed during the development of the nonattainment
area SIP for an area.  Section 107(d) does not require EPA, in essence,
to construct a mock-up nonattainment area plan for an area and to
predetermine appropriate levels of pollution control on behalf of the
State, as a necessary step in promulgating the designation for the area.
 In short, the level of control of a source does not in isolation answer
whether the source is contributing to the violations.

For example, Ohio argues that this plant is currently well controlled
but will have additional controls, either in 2009 or in some unspecified
future plans.  As noted above, EPA has concluded that this source
currently has particularly high levels of PM2.5 and PM2.5 precursor
emissions and that other information such as meteorological data
confirms that these emissions are contributing to violations in the
Columbus area.  This contribution is occurring, notwithstanding that the
source in question has some significant pollution controls in place.  By
inclusion of Franklin Township within the nonattainment area, Ohio and
EPA will evaluate this source as part of the development of the
nonattainment area plan to assure that it is adequately controlled and
controlled in a timely fashion, to provide for expeditious attainment of
the 24-hour PM2.5 NAAQS in the Columbus area as required by the CAA. 
This inquiry may conclude that the source is appropriately controlled,
but that assessment must occur during the proper process contemplated by
the CAA for development of nonattainment area plans.

With respect to future or planned future emissions reductions at the
source, EPA has concluded that for evaluating contribution for
designations under section 107(d), it is appropriate to look at the
actual level of emissions in the relevant three year period, whether
2005-2007, or 2006-2008 if a State elects to submit additional data for
2008.  Given the present tense formulation of section 107(d), EPA does
not believe it is appropriate to base designations on projections of the
future.  Thus, Ohio has indicated that the source in Franklin Township
is installing a scrubber and SCR on its largest unit that will be
operational in the ozone season in 2009, and operational year round in
2011.  Although EPA supports these important steps by the source, they
nevertheless will result in emissions reductions that will happen beyond
the period that is appropriate for consideration in the designations
process.  However, these reductions will occur during the time that Ohio
is developing the nonattainment area plan for this area, and they may
properly be taken into account during that process.  Ohio may conclude,
and EPA may ultimately agree, that these reductions of SO2 and NOX are
all that is necessary to reduce NOX emissions from this specific source
within the PM2.5 nonattainment area plan for Columbus, but it would be
inappropriate to prejudge that outcome now.  

Davenport-Moline-Rock Island, IA-IL

The State of Illinois

Illinois:

The Illinois Environmental Protection Agency has worked closely with the
Iowa Department of Natural Resources (DNR) to perform the following
technical analyses, and both States agree that Rock Island County does
not contribute to violations of the 24-hour PM2.5 standard in Davenport.

EPA Response:

EPA reviewed the additional information provided by Illinois including
modeling analyses.  The modeling shows a minor contribution from the
Rock Island point sources.  Some additional contribution is expected
from area and mobile sources from the urbanized areas of Rock Island
County, Illinois and Scott County, Iowa.  So even with violating
monitoring being highly affected by a local source, sources from the
entire urban area also contribute to violation.  Contributing counties
are designated as nonattainment, even if the contribution is minor.  EPA
is designating partial county areas in Scott County, Iowa and Rock
Island County, Illinois to include the sources contributing to the
monitored violations in Davenport.

EPA Response to Additional Modeling Results Referenced in Illinois
Comment Letter:

The AERMOD modeling performed by Illinois EPA is not documented with any
supporting information.  No information is given about the methodology
used for the application of AERMOD.  For instance, the approach to
incorporate concentrations associated with calm wind speeds and the
selection of “urban/rural” deposition will make a noticeable
difference in model estimates.  

Illinois EPA asserts that the maximum impact of all Rock Island County
sources combined is 1.17 μg/m3 based on their AERMOD simulations. 
These AERMOD simulations include eight stationary point sources, which
are unlikely to be inclusive of all point sources in the county. 
Additionally, this analysis does not include any other anthropogenic
emissions such as mobile sources and area sources.  If all anthropogenic
sources were included in addition to the eight stationary sources, the
impacts would be even larger than Illinois EPA’s estimated 1.17
μg/m3, which in itself could be considered a significant contribution.

Illinois EPA asserts that when using CAMx zero-out modeling, all
anthropogenic emissions sources in Rock Island County contribute a
maximum of 2 ug/m3 in Davenport.  This could be considered significant.

Illinois EPA asserts that Rock Island County contributes less than 1
percent of each sulfate, nitrate, and primary PM2.5 on the highest 2
percent of PM2.5 concentration days in 2002 using CAMx source
apportionment applied with a Central Regional Air Planning Association
(CENRAP) modeling platform.  EPA agrees that these results indicate that
Rock Island sources have a limited impact on secondarily formed
particulate impacting the grid cell containing the violating monitor. 
However, given the uncertainties in the modeling, as described further
in the TSD, EPA finds the results informative but not conclusive about
the impacts of Rock Island’s emissions on the violating monitor.  The
modeling results do not support the State’s position that emissions
sources in Rock Island do not contribute to the violating monitor.

The State of Iowa

Iowa:

The Davenport/Scott County monitor is located 150 meters from Blackhawk
Foundry.  The monitor records exceedances on days when other area
monitors record relatively lower values.  In addition, pollution roses
indicate that the greatest differences in readings between the Blackhawk
Foundry monitor and another local monitor (the Jefferson School monitor)
occur when winds blow from the foundry toward the monitor, suggesting
influences from foundry emissions.

EPA Response:

The comments relating to proximity of the Davenport (Blackhawk Foundry)
monitor to a source and the comparison of exceedance day readings to
readings at other monitors raise many of the same issues as described in
the Muscatine section of this document.  EPA incorporates its response
regarding the Muscatine monitor in its response regarding Davenport. 
EPA also notes that the Davenport monitor is a federal reference method,
source-oriented middle scale monitor, eligible for comparison to the
24-hour PM2.5 NAAQS, and that Iowa has not raised issues concerning the
siting of the monitor or validity of data collected at the monitor.

EPA acknowledges that the pollution roses provided by the state indicate
large differences in monitored readings between Blackhawk Foundry and
other monitors on exceedance days.  However, as described below, we are
unable to conclude on this basis that contributions from Blackhawk
Foundry alone are the sole basis for the monitored exceedances.

Iowa:

PSAT simulation is consistent with the zero-out results (discussed
above), showing very small contributions from sulfate and nitrate
emissions sources in Rock Island County, and similar contributions from
Scott and Muscatine Counties to particulate sulfate and nitrate
concentrations at the violating monitors.

EPA Response:

EPA agrees that the primary contributing pollutant from local sources in
the area to the violating monitors is direct PM2.5 rather than precursor
emissions.  EPA’s analysis of this information is discussed in more
detail in the TSD for Scott and Muscatine counties.  However, as
explained elsewhere, EPA believes that other nearby sources of PM2.5 in
each area nonetheless contribute, or have the potential to contribute. 
(See also, the discussion of contributing emissions in the TSD for Rock
Island.)

Iowa:

Iowa has re-reviewed the nine factors, in particular wind roses in the
two areas, as well as the growth pattern in Rock Island County (which
the commenter states indicates negative growth for the foreseeable
future), and believes that the data support the smaller boundary
originally recommended by the State.

EPA Response:

EPA has carefully considered all of the information submitted by Iowa
and Illinois regarding appropriate boundaries.  Wind roses, for example,
are instructive, but not dispositive of the appropriate boundary for an
area.  For example, as discussed above and in the TSD for Scott County,
point source modeling shows that other local point sources in the area
contribute to monitored exceedances.  EPA has reviewed this information,
and additional information made available to EPA after the 120-day
letter (and included in the docket), and has revised its 9-factor
analysis based on this review (see the TSDs for a detailed discussion of
the analyses).  Although this analysis has caused EPA to determine that
a partial county boundary is appropriate, EPA does not agree, for
reasons stated herein and in the TSDs, that the very small boundaries
suggested by the state are appropriate.	

Below are EPA’s responses to the State of Iowa’s contention that
modeling and other emissions impact analyses show that local point
sources contribute significantly to violating monitors even on days when
regional contributions are moderate to low, and that emissions from
other local sources, and sources outside the vicinity of the Blackhawk
Foundry monitor, do not significantly contribute to violations.

Iowa:

AERMOD dispersion modeling shows that on “most” exceedance days,
Blackhawk Foundry contributions were between 20 percent and 25 percent
of the local source contributions, and that the “vast majority” of
the remaining PM2.5 emissions are attributable to sources outside Scott
and Rock Island Counties.  This shows that the foundry “causes or
contributes” to the exceedances, thus justifying a small boundary
around the source’s area of impact.

EPA Response:

This comment was submitted in conjunction with Iowa’s October 20, 2008
response to EPA’s letter informing Iowa of its intent to modify the
state’s initial boundary designations for eastern Iowa (the 120-day
letter).  At that time, Iowa had not modeled impacts from other local
direct PM2.5 point sources on concentrations at the violating monitor. 
Iowa subsequently modeled (using AERMOD) additional local sources and
argued that they have a minimal impact on concentrations at the
violating monitor. In fact, the modeling of direct PM2.5 emissions shows
that other point sources in the area impact the monitor, and the
nonattainment boundary which EPA is promulgating will encompass all of
these PM2.5 sources.  In addition, EPA has determined that the
uncertainty regarding the AERMOD data and the lack of speciation data at
the violating monitor, as described in the TSDs, prevents us from
concluding that local sources of direct PM2.5 other than the Blackhawk
Foundry (i.e., sources outside the state’s recommended boundary) do
not contribute to the violations.  Whether these other sources should
ultimately be required to reduce emissions will be determined by the
state during the development of the implementation plan for the area. 

EPA has also performed a positive matrix factorization (PMF) analysis on
speciation data from the Jefferson School monitoring site in Davenport. 
This analysis is detailed in the TSD for Scott County.  The analysis
indicates that a unique source other than Blackhawk Foundry contributed
to the mass at the nearby Jefferson monitor.  Although this is not the
violating monitor, it indicates the influence of a source outside the
state’s recommended nonattainment area on a nearby monitor in the
area.  This finding supports EPA’s decision to promulgate a boundary
larger than that recommended by the state.  

Iowa:

The State’s recommended boundary includes the only sources of direct
PM2.5 emissions in the two-county area which have any significant
contributions to predicted concentrations of the violating monitor on
exceedance days.

EPA Response:

As stated in response to II)A), EPA does not believe the AERMOD modeling
is adequate to determine that the other point sources of direct PM2.5 in
the area are not contributing to violations.  In fact the modeling and
other analyses show that other sources do have some impact.  EPA
believes that other point sources in the area should be included in the
nonattainment boundary.

Iowa:

The results of CAMx model sensitivity analyses and AERMOD runs show that
Rock Island County should not be included in the nonattainment area due
to insignificant contributions to predicted concentrations at the
monitor, based on the following factors:  (1) the June 8, 2007 EPA
guidance does not provide for CBSA presumptive boundaries for 24-hour PM
violations;  (2) the mere presence of a broad mix of sources in Rock
Island does not indicate their potential contributions;  (3) a zero-out
run of direct PM emissions in Rock Island County results in an
insignificant reduction in base-case contributions at the violating
monitor and an even smaller reduction at other area monitors.  There are
similar negligible reductions when sulfate and nitrate precursors are
subject to zero-out runs; and (4) a sensitivity analysis of mobile
source commuting activity between Scott and Rock Island counties shows
that traffic patterns and motor vehicle emissions do not influence the
violating monitor on exceedance days.

EPA Response:

EPA has determined that the urbanized portion of Rock Island County will
be included in the nonattainment area, for reasons detailed in the
technical support document prepared by EPA for the State of Illinois and
included in the docket for the accompanying rulemaking.  EPA’s TSD for
Scott County also contains some discussion of the issues.  With respect
to Iowa’s specific comments, EPA provides the following responses:  

	(1)  EPA agrees that there are no presumptive boundaries for the 2006
24-hour PM2.5 standard designations.  The 2007 guidance does provide,
however, that one of the 9 factors in the boundary analysis is
jurisdictional boundaries, which could include consideration of CBSA,
MSA, or county boundaries.  However, EPA analyzes all of the factors,
and may conclude that smaller or larger boundaries are appropriate,
based on a case-by-case review.  In the case of Rock Island and Scott
counties, EPA has analyzed these factors and has concluded that the
urbanized portion of these counties should be included in the
nonattainment area.  This conclusion is not based on any presumptive
boundary for the area, but is the result of application of the 9-factors
to the specific information available for the Quad Cities area.

	(2)  EPA agrees that the “mere presence” of sources in an area does
not necessarily demonstrate their contribution to violations of the
NAAQS.  However, as described in the TSDs and the record for the
designations, including this response to comments, EPA has determined
that sources of emissions in Rock Island County may contribute to the
violations recorded in Davenport.  Therefore, EPA is including a portion
of the county in the nonattainment area.

	(3)   The zero-out modeling is instructive but not conclusive with
respect to the appropriate boundaries.  In addition, in establishing
boundaries for contributing sources or areas in the Quad Cities, we do
not believe it is appropriate to establish a bright line regarding the
amount of reduction in predicted concentration which an area must be
below to show that it is not contributing to violations of the NAAQS
(for example, that a 4% reduction in a zero-out run is insignificant,
but some higher amount is significant).  Instead, as explained in more
detail in the TSDs, in this response to comments, and in other materials
in the record, we are establishing the nonattainment boundary to
encompass most of the sources of PM2.5-related emissions in the area
which could, based on available information, contribute to PM2.5
concentrations at the violating monitor.  The states will subsequently
determine, through refined analyses during development of their
implementation plans, the emissions reductions needed, the portions of
the nonattainment area in which they must occur, and the sources which
must reduce emissions in order to achieve the NAAQS.  EPA believes that
the boundaries defined in this rulemaking will enable the states to
perform these analyses and develop strategies to achieve the standards.

	(4)  EPA agrees, as described in more detail in the TSDs and other
information supporting the decision, that direct PM2.5 emissions are a
critical component of the PM2.5 concentrations at the violating monitor
in Davenport. However, zero-out modeling results provided by the state
do show that emissions (including mobile) from Rock Island contribute to
the violating monitor on exceedance days and support EPA’s decision to
include a portion of Rock Island in the nonattainment area.  This is
discussed in more detail in the TSDs for the area, and most
significantly, in the Rock Island County TSD.

Iowa:

The modeling shows that sources in rural Scott County do not cause or
contribute to violations of the PM2.5 NAAQS.  Zero-out runs show that
eliminating emissions from rural portions of the county result in only
minimal reductions in PM2.5 concentrations at the violating monitor on
predicted exceedance days.

EPA Response:

As explained in the TSD, EPA believes that the rural portion of Scott
County, which does not contain point sources of PM2.5 emissions, likely
does not have a significant contribution to the PM2.5 concentrations at
the violating monitor, so as to require inclusion of that portion in the
nonattainment area.  However, EPA’s conclusion does not mean that
other portions of the county containing sources of PM2.5-related
emissions from point and area sources need not be included.  This is due
in part to the fact that, as stated in the TSD for Iowa, zero-out
techniques, although appropriate for resolving non-point sources, in
this instance, are not yet capable of reliably assessing the impacts of
a single source at the source receptor distances encountered in Scott
County.  Therefore, EPA is including the urbanized portion of the county
and those nearby rural portions containing PM2.5 emissions of concern in
the nonattainment boundary for the area.

Dayton-Springfield, OH

Ohio:

In its initial designation recommendations to EPA in December 2007, and
in its October 2008 letter responding to EPA’s proposed modifications
thereto, the State of Ohio has argued that Clark County should be
excluded from the Dayton nonattainment area for the 2006 24-hour PM2.5
NAAQS.  Ohio premised its initial recommendation on the lack of
monitored violations in Clark County, relatively low SO2 and NOX
emissions, population of 142,000 people, relatively low commuting, and
the geographic location of Clark County to the northeast of the
violations of the NAAQS in Montgomery County.  Subsequent data from
2005-2007 indicates that Clark is violating the NAAQS and EPA
accordingly notified Ohio of its intention to modify the State’s
recommendation to include Clark County in the Dayton nonattainment area.
 In its response to EPA’s proposed modification, Ohio acknowledged
“high pollution levels being detected in Clark County” but argued
that these concentrations are attributable to sources southwest of Clark
County and not due to local sources.

EPA Response:

Under section 107(d), EPA must designate as “nonattainment” both
those areas that are violating the NAAQS and those areas that are
contributing to violations in a nearby area.  Based upon 2005 to 2007
monitoring data, Clark County is itself violating the 2006 24-hour PM2.5
NAAQS.  Thus, Ohio’s arguments with respect to the level of
contribution of emissions from Clark County to Montgomery County do not
override the obligation of EPA to designate Clark County as
nonattainment.

EPA has concluded that it is appropriate to designate Clark County as a
part of the Dayton area because it is immediately adjacent to Montgomery
County (which is also violating the NAAQS) and Greene County (which is
contributing to violations in Montgomery).  EPA agrees that the
contribution of emissions in Clark County to violations in nearby
Montgomery County is relatively low.  For example, compared to
Montgomery County, Clark County has a lower total emissions, lower
population, and lower commuting levels of commuting, but the amount of
each of these is not insignificant in the context of this area.  In
particular, the number of commuters from Clark County to other Counties
within the area is approximately 60,000 and reflects a percentage of
commuting comparable to that of Montgomery and Greene Counties and a
number and amount of commuting comparable to that of Greene County. 
This information indicates some contribution from Clark County to
Montgomery County.  Moreover, the pollution rose for a critical
Montgomery County site supports the conclusion that there is
contribution from the direction of Clark County on some of the days with
relatively high ambient PM2.5 levels.  Finally, EPA considered that
Clark County is already a member of the Miami Valley Regional Planning
Commission and has typically been included with other Dayton area
counties in the nonattainment area for ozone.  Under these
circumstances, EPA is designating Clark County nonattainment and
including it within the boundaries of the Dayton area, rather than a
separate nonattainment area.   

Ohio argues, on the basis of wind data, that there is also contribution
from Montgomery County to Clark County.  As noted above, Clark County is
violating the NAAQS based upon 2005-2007 data.  The new information from
Ohio would perhaps support EPA designating Montgomery and Greene
Counties nonattainment based upon their contribution to Clark, but EPA
does not consider this necessary given that Montgomery County is itself
violating the NAAQS, and that both Ohio and EPA agree that Greene County
is contributing to violations in Montgomery County.  Thus, the wind data
could be support for designating the same three counties as part of the
Dayton area, but EPA does not consider this necessary.  

Detroit-Ann Arbor, MI

Michigan:

In our December 18, 2007, letter to EPA, we recommended that southeast
Michigan be divided into three separate nonattainment areas consisting
of St. Clair County, which appears to be influenced by Canadian
emissions; Wayne County, which has higher concentrations of PM2.5 than
the other counties due to local emissions sources; and the remaining
five counties as a third nonattainment area.

EPA Response:

EPA disagrees with Michigan’s recommendation to divide the Detroit
nonattainment area into three separate nonattainment areas.  EPA
generally will not divide areas because having one area allows for
better planning.  Instead of assessing the impact of a source on several
nonattainment areas, determining the impact is simplified when there is
just one area.  Having a single nonattainment area also eliminates the
need to integrate planning for multiple adjacent areas.

Specifically in the Detroit area, the metropolitan planning organization
addresses the seven counties of the existing nonattainment area and thus
is already designed to conduct planning for the prospective
nonattainment area as a whole.  Michigan wrote that the air quality in
Saint Clair County may be affected by international transport and that
local sources influence the air quality in Wayne County.  The air
quality in these counties is also affected by the same common sources as
all counties in the Detroit area.  The populations and commuting data
also indicates a connection between the metropolitan area counties. 
Therefore, it makes sense to have a single Detroit nonattainment area
instead of dividing the area into three parts.  Michigan can certainly
include a mix of controls implemented for area-wide air quality
improvement and specific control measures to address the unique sources
affecting the air quality in Saint Clair and Wayne Counties.  A single
seven county nonattainment area allows Michigan the greatest flexibility
in improving the air quality in the Detroit area.

Michigan:

In its October 14, 2008, letter to EPA, Michigan reiterates arguments
that “the industrialized area of Wayne County is unique to the rest of
the region and should be dealt with separately.”  To support this
argument, Michigan argued that only monitors in Wayne County have shown
violations of the annual PM2.5 NAAQS since 2003.  Michigan asserted that
because of this monitor, six other counties in the Detroit area have
“been kept in nonattainment.”   For the 24-hour PM2.5 NAAQS,
Michigan conceded that monitors in both Wayne and St. Clair show
violations of the NAAQS with 2006-2008 data (based on analysis using
data from the first half of 2008), but repeated arguments that
designating a large area is “unnecessary and heavy handed.”

EPA Response: 

EPA disagrees with Michigan’s approach to assessing what areas should
be designated nonattainment under section 107(d).  EPA reads the statute
to require designation of those areas that are violating and those areas
that are contributing to violations in nearby areas.  EPA does not
interpret the term “contributing” to be limited merely to the area
allegedly contributing the last increment of ambient PM2.5 that results
in a violation in the area; rather, EPA interprets the term to include
the full range of nearby sources of PM2.5 and PM2.5 precursors that are
cumulatively contributing to the violation.

5-2007 design values for Monroe (38 μg/m3), Oakland (40), St. Clair
(41), Washtenaw (39), and Wayne (43) Counties all are above 24-hour fine
particulate standard.  Michigan used data from only the first two
quarters of 2008 to make its projections of the 2006-2008 design values
in its October 2008 letter.  Design values must be calculated using
three complete years of monitoring data.  Considering that five of seven
counties in the Detroit nonattainment area have design value violating
the 24-hour fine particulate standard, the high particulate
concentrations are distributed through the area and not just affecting
two counties.  

 

Michigan:

Michigan argued that the PM2.5 problem in Southeast Michigan “has been
driven by the high values in the industrialized portion of Wayne
County.”  Michigan reiterated that six other counties are being
designated nonattainment for the 24-hour PM2.5 NAAQS

“because of one monitor that is in the most industrialized area of
Wayne County, directly downwind of a steel mill, auto manufacturing
plant and oil refinery.”  Michigan argued that this monitor is
“strongly influenced” by “local sources.”

Michigan argued that the “history” of PM10 and PM2.5 violations in
this industrial center supports “focused attainment plans” for a
separate nonattainment area in Wayne County.

EPA Response: 

EPA disagrees with Michigan’s implications that violations of the 2006
24-hour PM2.5 NAAQS are solely the result of emissions in the
“industrialized” portion of Wayne County.  EPA does not dispute that
the violating monitor is near a number of stationary sources, or that
the emissions from such sources do contribute to the violations at that
monitor.  However, EPA’s obligation is to include within the
nonattainment area not merely the nearest sources to the monitor, but
rather to include those areas that contain emissions sources that are
“contributing” to the violations at that monitor.  Michigan does not
contend that all of the ambient PM2.5 at the violating monitor in Wayne
is the result of emissions exclusively within the near vicinity of the
monitor, nor even exclusively within Wayne County.  Regardless of
whether a significant percentage of the ambient PM2.5 at the monitor was
the result of emissions from sources in the near vicinity of the
monitor, EPA must consider other sources that account for the remaining
ambient PM2.5 at the monitor, and EPA must promulgate boundaries that
correctly include those other sources of the remaining ambient PM2.5
within the nonattainment area.  

EPA’s view is that the information relevant to evaluating contribution
to violations in this area compels the conclusion that areas outside
Wayne County but nearby to Wayne County are contributing to the
violations, as contemplated by section 107(d).  Thus, for example,
evidence such as the emissions inventory of direct PM2.5 and PM2.5
precursors for other nearby counties, the pollution roses for this area,
the CES scores for other counties, and the amount of emissions reflected
by commuters and VMT for other counties, all tend to support that there
are emissions and emissions activities in the surrounding counties in
the Detroit area that are cumulatively contributing to the ambient
levels of PM2.5 at the violating monitors in both Wayne and St. Clair
Counties as well as at the other monitors in the area that are currently
violating the standard.  In accordance with the Clean Air Act, these
sources in these areas are thus appropriate for inclusion within the
nonattainment area boundaries, and thus for evaluation for control in
the nonattainment plan for this area.  

EPA also disagrees with Michigan’s suggestion that past history with
PM10, and implicitly with the designation for this area for PM10, is
supportive of the State’s view.  PM10 is a different NAAQS, with a
different size indicator, that forms differently and behaves differently
in the ambient area, typically emitted by different sources and
requiring different control strategies.  The relevance of past history
with PM10, especially given the facts of this area and the nature of the
PM2.5 particles in this area, is negligible.

Michigan:

To support its arguments with respect to Wayne County, Michigan
submitted a study commissioned by the Southeast Michigan Council of
Governments (SEMCOG) to analyze monitor data in the Detroit area. 
Michigan described the purpose of this study as to determine the
“drivers (i.e., local source, regional transport, or meteorology) for
high PM2.5 days in South East Michigan.”  The State argued that the
“drivers” of high PM2.5 at the “majority” of monitors with high
readings “are regional transport (background levels entering Southeast
Michigan) and/or poor atmospheric ventilation across the network.” 
Michigan also noted that the report indicates that ambient
concentrations are caused by “regional-scale contributions,
urban-scale contributions, and neighborhood-scale contributions.”   

EPA Response: 

EPA fully agrees with the SEMCOG contractor’s finding, as summarized
by Michigan, that ambient concentrations reflect contributions on
various scales ranging from regional scale to urban scale to
neighborhood scale.  Indeed, Michigan argues that the Dearborn monitor
records especially high values because it reflects impacts from both an
industrial zone plume and the urban scale plume.  EPA agrees with this
view as well.  Where Michigan appears to disagree with EPA is in
recommending a subdivision of the metropolitan area into three zones
that Michigan argues have separate sets of local contributors, without
regard for the resulting segregation of the common origins of the urban
scale contribution into three separate nonattainment areas.

EPA commends SEMCOG for sponsoring work to understand the origins of the
24-hour PM2.5 problem in Southeast Michigan.  However, this work is
likely to have more utility in the attainment planning phase than in the
designations phase.  The contractor finds that many monitors have high
concentrations if and only if other monitors also have high
concentrations.  This finding is consistent with the premise noted
above, that a significant element of high concentrations in the Detroit
area is an “urban plume.”  However, the majority of the
contractor’s work is designed not to define the origins of this urban
plume but rather to define the origins of concentrations in excess of
the combined impact of the urban plume and regional scale contributions.
 This work, by its very nature, is prone to focus on the contributions
of heterogeneously distributed sources that have disparate impacts at
different locations.  Conversely, this work, by its very nature, is
prone to disregard the impacts of sources that have relatively
homogeneous impacts, which can occur either because similar sources are
homogeneously distributed or because the emissions undergo a process
that disperses the impact (which would be expected to occur during
photochemical formation of secondary particles).  As a result, the study
is more useful for defining subsections of the area that may have unique
contributions than it is in defining what source area contributes to the
relatively homogeneous “urban plume.”

From this perspective, it is clear that the “drivers” identified by
Michigan represent evenly distributed sources of locally significant
sources that may cause some monitors to record concentrations that are
distinguishable from concentrations observed elsewhere.  It is equally
clear that the “drivers” identified by Michigan are almost by
definition not the sources distributed throughout the metropolitan area
that contribute to the urban plume described by the contractor.  Thus,
while EPA and Michigan agree that the “drivers” must be part of the
nonattainment area, EPA finds it equally important to include the range
of sources that contribute to the “urban plume.”

From another perspective, the concept of the “driver” of
concentrations on days with high ambient PM2.5 levels is merely using
different terminology to accomplish the same goal – to argue that only
the source or sources that add the last microgram of ambient PM2.5
levels above a standard is the only source or sources contributing to
the total ambient level.  EPA does not agree that this is the
appropriate method to evaluate the full range of sources that are
cumulatively contributing to the violation of the NAAQS.  

Second, it must be noted that it is likewise not appropriate to argue
that regional transport from greater distances, or even from other
States, is a grounds to exclude the nearby areas that are also
contributing to the violations in the area.  The Clean Air Act provides
other tools to address the influx of regional transport, especially
under section 110(a)(2)(D) and section 126.  In the context of
designations, it is the obligation of EPA to designate areas that are
violating the NAAQS and nearby areas that are contributing to those
violations.  EPA does not agree that all of the ambient PM2.5 at the
violating monitors in the Detroit area comes either from the sources
immediately adjacent to the monitors or from other States – there are
geographic areas between those extremes that contain sources with
emissions that are contributing to the violations.  

The report noted that the ambient concentrations are the result of
emissions from not just local and distant sources, but also from sources
on an urban-scale.  Having a single Detroit nonattainment area allows
Michigan to address sources through the seven county area that affect
the air quality and thus contribute to the violations.  While EPA is not
prejudging the mix of emission controls that Michigan should adopt to
provide for attainment, the Act mandates that EPA promulgate a
nonattainment area that includes the full area violating the standard
and nearby area contributing to the violations, in part to assure that
an integrated plan is developed based on consideration of control
options for all sources that contribute to the violations.

Michigan:

Michigan highlighted a conceptual diagram illustrating its view of the
progression of PM2.5 concentrations along a transect running from
southwest to northeast in support of its argument that the Dearborn
monitor in Wayne County is in a “unique” position “in the core of
the industrialized zone plume and urban scale plume, which results in
much higher values due to local sources than the other monitors in the
area.  The State further argued that data from the New Haven monitor
site, “downwind” from Detroit, did not show the same impacts of the
“urban scale contributions or the industrial zone plume.”

EPA Response: 

The diagram supplied by Michigan as Figure 2 in its letter illustrates
that the ambient concentration is affected by regional, urban, and
local-scale emissions.  Both EPA and the States have responsibilities
for programs to reduce regional-scale emissions.  The States must
develop control measures to reduce both the urban and local-scale
emissions to bring areas into attainment.  Specifically for Detroit,
Figure 2 suggests that the regional and urban-scale emission create
ambient concentrations near the PM2.5 standard (the Macomb County
monitor is close to the standard).  Thus Michigan would have to
eliminate almost all of the local-scale emissions from the “industrial
zone” to reach attainment unless there are also regional and
urban-scale reductions.  Therefore, Michigan’s diagram highlights the
importance of designating a broad Detroit nonattainment area that
includes a complete set of the sources in the Detroit metropolitan area
that contribute to these violations.

Michigan:

Michigan also highlighted Table 2 in its letter to support its arguments
regarding Wayne County.  According to Michigan, 13 of the 15 monitors in
the area, for 85 percent of the high days between 1999 and 2006,
occurred on days when “the entire network was exhibiting high PM.” 
By comparison, Michigan argued that the Dearborn and Wyandotte monitors
“frequently had exceedances on days when the entire network was not
exhibiting high PM.”  Michigan reasoned that this proves that these
monitors are “significantly influenced by nearby emissions sources.”

EPA Response: 

This illustrates that violations can be caused by local or broader scale
emissions.  Local sources will cause high concentrations at nearby
monitors, while emissions on a broader scale will cause high
concentrations over a wide area.  The fact that most or all of the
Detroit area monitors will record high concentrations on the same day
suggests there are broader scale sources that affect the entire area. 
That is, the existence of days when “the entire network was exhibiting
high PM” lends support to EPA’s decision to designate Southeast
Michigan as a single seven-county nonattainment area.

Michigan:

Michigan also pointed to Figure 3 in its letter as evidence that the
Dearborn monitor in Wayne County has “a much higher percentage of high
days than other monitors.”  As evidence, the State noted that the
Dearborn monitor has many more days above 30 μg/m3 compared to the
Allen Park monitor.  Again, Michigan argued that the higher readings at
the Dearborn monitor than at the Allen Park, which the State
characterizes as near but upwind from the Dearborn monitor, indicate
that the latter monitor has a “strong influence from local sources.”

EPA Response: 

As noted previously, the Act requires that EPA designate nonattainment
areas that include all nearby sources that contribute to nonattainment,
including both the local sources (i.e., sources nearby the monitor) and
sources elsewhere in the metropolitan area that contribute to the
violations.  To the extent that local sources cause a monitor to observe
more high concentration days, these local sources arguably make the
violation more difficult to address.  Nevertheless, while the Act gives
the State substantial discretion in selecting what combination of
controls of local sources, controls of metropolitan area sources, and
controls of sources elsewhere in the State to adopt to address the
nonattainment problem, the Act dictates that EPA define a nonattainment
area and planning area in which the full range of nearby contributing
sources are included.

Michigan:

μg/m3 excess of PM2.5, and that this quadrant is the direction of “a
steel mill, an auto manufacturing plant, an oil refinery and other
smaller industries.”  By comparison, Michigan notes that other
monitors show a lesser impact, less than 2 μg/m3.

EPA Response: 

EPA believes that Michigan is focusing on local deviations from broadly
high concentrations and fails to address the contributions from sources
throughout the metropolitan area to the “urban plume.”  The
“expected value” appears to be largely a reflection of the sum of
regional scale and metropolitan scale impacts, and the “excess”
appears to be a reflection of local (neighborhood scale) source impacts.
 Such a differentiation may be useful for attainment planning purposes,
but EPA’s obligation in promulgating designations is to define a
nonattainment area that includes nearby sources contributing to the
“expected value” as well as sources contributing to the
“excess.”

Again, Michigan evidently starts from the premise that the only areas
that should be designated as “nonattainment,” are the area in which
they claim there are sources that add the last increment of ambient
PM2.5 that results in violations at the Dearborn monitor, ignoring the
other contribution from other sources in nearby areas.  EPA does not
subscribe to this approach, as it is the cumulative impact of emissions
sources and activities throughout the Detroit area that are, in the
aggregate, contributing to the total amount of ambient PM2.5 at the
violating monitor.  The purpose of the designations process is not to
seek only the individual source next to the monitor that adds the last
increment, but rather to identify the range of sources that are
contributing.  This is logical, because these are the sources that will
need proper evaluation in the context of the development of the
nonattainment SIP, and control to assure expeditious attainment of the
NAAQS, as appropriate.   

Taken at face value, the analysis submitted by Michigan may well serve
to identify many of the specific sources that the State must control in
the nonattainment area SIP for Detroit to assure that this entire area
attains the 2006 24-hour PM2.5 NAAQS expeditiously.  However, this
analysis must look more broadly at all sources throughout the area
designated nonattainment and assure that all source categories are
evaluated for potential control, consistent with the requirements of the
CAA and consistent with attainment of the NAAQS as expeditiously as
practicable.  After all, the widespread violations being record by
monitors throughout the Detroit area are not being caused solely by a
few select sources in one portion of the area.  

In EPA’s view, it is meaningless to claim “more robust mathematical
support” for an analytical tool that pursues a different type of
information that is less relevant to the issue at hand.  The tool is
being used to identify locally significant sources, and there is no
dispute that locally significant sources should be included in the
nonattainment area.  Consequently, EPA’s tools, which better address
what areas have sources that contribute to the “urban plume,”
provide better information with which to determine nonattainment area
boundaries.

Michigan:

Michigan argued that the evidence it provided concerning the
“unique” aspects of Wayne County support designating it as a
separate nonattainment area within the Detroit area.

EPA Response: 

Michigan has recommended that seven counties in Southeast Michigan be
designated nonattainment, and EPA agrees with this recommendation.  The
disagreement between EPA and Michigan relates to Michigan’s
recommendation that these seven counties be divided into three separate
nonattainment areas rather than combined into a single nonattainment
area.  Since CAA section 107 requires each nonattainment area to include
the area that is violating and any nearby area contributing to that
violation, the recommended subdivision into multiple nonattainment areas
would require justification either that the subdivided areas are not
nearby or that the areas do not contribute to each other.  For example,
justification would need to be provided either that Wayne County is not
nearby to its neighboring counties or that there is no contribution from
Wayne County to the violations in various other Southeast Michigan
counties and no contribution from the other counties to the violations
in Wayne County.  Michigan provided no such justification, and EPA
believes that the evidence contradicts any such claim.

The information also suggests it would be difficult for the air quality
through the Detroit area to meet the standards relying just on localized
measures within Wayne County.  The fact that some portions are impacted
more by local sources is not unique.  This is common in cities around
the nation. This allows the State to determine how it will use area-wide
measures combined with controls focused on specific sources to bring the
area’s air quality into attainment.  EPA designated adjacent counties
in different nonattainment areas only when the counties were clearly in
separate areas.  Several of the factors suggest a link between the seven
counties in the Detroit area.  

Michigan:

In its October 14, 2008, letter, Michigan also argued that St Clair
County should be designated as a separate nonattainment area.  As
support for this proposal, Michigan again used the SEMCOG report.  The
State pointed to Figure 1 of its letter as evidence that the monitor in
St. Clair County is at the northern edge of the Detroit area and “may
experience different air masses than the other monitors” in the
Detroit area.  In addition, the State argued that the Port Huron monitor
in St. Clair County “appears to be strongly influenced by the
Canadian/Sarnia industrial core.

EPA Response: 

EPA does not share Michigan’s belief that St. Clair County should be a
separate nonattainment area from Detroit.  St. Clair County has the
second highest sulfur dioxide emissions in the Detroit area and
surrounding counties.  Its emissions of fine particulate and nitrogen
oxides are larger than several other Detroit area counties.  So,
emissions from within the county can contribute to the monitored
violations in other counties.  The commuting data shows a link to the
Detroit area.  St. Clair County is in the southeast Michigan Council of
Governments with other Detroit area counties, and is also in the Detroit
nonattainment areas for ozone and fine particulate under the 1997
standard.  These all point to a link to the Detroit area.  The pollution
rose for St. Clair County (Figure 11) shows the winds tend to come from
the south when the highest concentrations are monitored.  Some the fine
particulate concentration being monitored in St. Clair may come from
emissions generated at an industrial area to the southeast near Sarnia,
Ontario, Canada.  Some of the PM2.5 in St. Clair County may also come
from elsewhere in the Detroit area, which is to the southwest.  The
Belle River power plant in St. Clair is nearly straight south of the St.
Clair County monitor.  EPA concludes that the St. Clair County is being
impacted by emissions from local sources and sources throughout the
Detroit area, just as EPA is concluding that St. Clair County is
contributing to violations elsewhere in the Detroit area.  Therefore,
EPA determined it was appropriate to include St. Clair County in the
Detroit nonattainment area.  

Figure   SEQ Figure \* ARABIC  11 .  Pollution Rose for St. Clair
County, Michigan.

Michigan:

Citing Table 3 of its letter, Michigan noted that the three year annual
PM2.5 average at the Port Huron monitor is “one of the lowest” in
the Detroit area.  That State argued by contrast, that the 24-hour PM2.5
concentrations in this area are “some of the highest” and are
“distinctly different” than the averages in other areas.  Unlike
other areas, Michigan argued that the ambient air at the Port Huron
monitor are unchanged or worsening, compared to averages that appear to
be improving at other monitors in the Detroit area (except for 2005). 
Michigan asserted that this evidence suggests that the Port Huron
monitor “is being influenced by different sources than the other
monitors in Southeast Michigan.”

 

EPA Response: 

EPA does not believe that the annual PM2.5 average at the Port Huron
monitor is particularly germane to the designation for the 2006 24-hour
PM2.5 NAAQS.  As Michigan correctly stated, the Port Huron monitor is
showing high daily concentrations of PM2.5 well above the NAAQS, even
compared to other violating counties within the Detroit area.  The
design values for the area monitors for 2005-2007 are: Wayne 43, Oakland
40, Washtenaw 39, and Monroe 38, with St Clair at 41.  County design
values can show an occasional rise even when values have been declining.
 So the changes from the 2004-2006 design value to the 2005-2007 design
value should be not considered a trend.  Three of the Detroit area
counties, Oakland, St. Clair, and Washtenaw Counties, had 2005-2007
design values higher than its 2004-2006 values.  As Michigan noted, the
2005 monitored values were rather high throughout the nation.  If the
2006-2008 and beyond design values remain fairly steady for St. Clair
County while other Detroit area counties show a trend to lower values,
Michigan could consider additional local measures for the sources with
the greatest impact on the St. Clair County monitor.  The effect of
local-scale emissions causing the St. Clair County monitored values to
vary from the trends at other area monitors does not mean that it is a
different area.  The variations in the temporal patterns at different
monitors more likely is simply evidence that a few years of data do not
provide enough data to indicate a reliable trend.    

Michigan:

To support its contention that St. Clair should be a separate
nonattainment area, Michigan also argued that emissions from Canada are
contributing to the violations at the Port Huron monitor.  Michigan
utilized its “CPF” calculation in figure 9 of its letter to support
this point.  Michigan contends that the “excess mass” of PM2.5 at
the Port Huron monitor is coming from the south, southeast, and east,
and that “one of the most industrialized areas of Canada” lies in
that direction.  Michigan notes that the Sarnia area in Canada contains
significant sources such as oil refineries, petrochemical plants, and
chemical companies, and this area has well documented air quality
problems.

EPA Response:

Again, Michigan is focusing on “excess mass” and disregarding the
“urban plume” that is common to monitors throughout Southeast
Michigan.  EPA and Michigan agree that St. Clair County contributes to
violations in St. Clair County, but analyses that by their very nature
do not assess the contribution of sources broadly distributed throughout
Southeast Michigan and sources elsewhere with broadly distributed
impacts have limited utility in defining nonattainment areas.

As noted before, the violations are also due to the local emissions and
from contributions from other area counties.  International emissions
are not the lone source of fine particulate pollution in St. Clair
County.  Pollution from the power plant in St. Clair County, the rest of
the Detroit area, and from the nearby Canadian sources can all be
carried to the Port Huron monitor by southerly winds.  

In addition, EPA notes that even if there were contribution from
Canadian sources to the violations that the Port Huron monitor, that
does not negate EPA’s obligation to promulgate the designation for
this area to provide for attainment of the NAAQS in this area.  Section
107(d) does not provide for differential treatment in designations based
upon transport from Canada, and obligates EPA to make designations based
upon monitors and other relevant information to evaluate areas that are
violating and contributing to those violations.  

Although EPA obviously cannot define a nonattainment area that includes
emissions sources in Canada that may be contributing to violations in
the Detroit area, it is required to promulgate a designation that
includes the violating areas and nearby contributing areas in Michigan
within the nonattainment area.  As explained in the TSD for this area,
EPA has concluded that St. Clair is violating the NAAQS and must be
designated nonattainment.  St. Clair is also contributing to violations
in other nearby counties in the Detroit area.  St. Clair is not in a
separate airshed from the remainder of the Detroit area, it contains
sources and source activity that contributes to the remainder of the
Detroit area, and has similar sources that will require evaluation in
the nonattainment plan for the area as a whole.  Meteorological
information indicates that St. Clair County is not separate and distinct
from the rest of Detroit.  This information shows a slight tendency for
the winds to be coming from the direction of Wayne County on high
concentration days, but it also shows the winds often come from all
directions even on high concentration days.  So, it is fair to conclude
that St. Clair County contributes to PM2.5 levels in the Detroit
nonattainment area on high concentration days.

Michigan:

Michigan argued that because the high values at the Port Huron monitor
“are likely influenced by Ontario sources,” it “does not make
sense to include St Clair County within the same nonattainment area as
the remainder of Detroit.

EPA Response: 

EPA disagrees because Canadian emissions are not the lone source of fine
particulate pollution in St. Clair County.  The concentrations being
recorded at the Port Huron monitor are caused by pollution from local
sources, sources throughout the Detroit area, and from Canadian sources.
 St. Clair County sources may contribute to other Detroit area counties.
 EPA also determined that St. Clair County is linked to Detroit area. 
There may be some unique aspects to the violation being recorded in St.
Clair County.  Still, EPA determined St. Clair County is linked with the
other Detroit area counties and that all seven counties should be a
single nonattainment area.  

Evansville, IN

Indiana:

Concurrence from EPA concerning the exceptional events submission would
have altered the proposed designation of Dubois County as nonattainment.
 Publicly proposing counties as nonattainment in August 2008 without
addressing exceptional events leads to a false stigma in relation to air
quality in the area.  Furthermore, if the designations are to be based
on 2006-2008 monitored values, 2008 exceptional event submissions must
be reviewed and acted upon prior to the effective date.

EPA Response:

As discussed in the technical support document, EPA has now completed
its review of Indiana’s exceptional events claims.  The net effect of
the mix of concurrences and non-concurrences is that EPA finds Dubois
County to be attaining the standard for the 2005 to 2007 period.  EPA
also reviewed whether the county is contributing to nearby violations in
Evansville, and concluded it is not contributing and therefore Dubois
County is being designated as attainment for the 2006 24-hour NAAQS.

Indiana:

If EPA reviews and approves the 2005-2007 exceptional events, Dubois
County could be identified as attainment as well.  

EPA Response:

EPA has reviewed Indiana’s claims of exceptional events, with the
result that Dubois County is being treated as attaining the standards
for planning purposes.  Dubois County is being designated as attainment
for the 2006 standards because EPA has also judged that it does not
contribute to violations in the Evansville area.

Indiana:

Counties like Warrick County, that measure air quality below the
standard and are not proven to be significantly culpable for a downwind
violation, should be designated attainment.  

EPA Response:

In making determinations of contribution, EPA reviews all available data
including the nine factors and other relevant information as explained
in the TSD.  Based on the best available evidence, EPA has typically
concluded that nearby counties with high emission levels and other
supporting data such as meteorology do contribute to nearby violations. 
In this case EPA reviewed all the available data as outlined in the TSD
and EPA’s assessment of Warrick County indicates that it is
contributing to the monitored violations in Evansville, based on
emission levels (92,000 tons SO2 per year; 18,000 tons NOx per year),
combined with air quality and meteorological data.  EPA notes again that
it must designate as nonattainment counties that are contributing to
nearby nonattainment even if they are monitoring attainment.  Further,
section 107 does not provide for designation only of areas that have a
significant contribution to nearby violations.  In contrast, when
addressing long-range transport of pollutants under section 110(a)(2)(D)
the Act specifically looks only at significant contributions from one
state to another.  For designations section 107 requires inclusion of
all nearby areas that contribute to violations.  Of course EPA analyzes
all available data to be sure areas included are contributing, and in
this case has concluded based on analysis of all the factors and
analytic tools that Warrick County is contributing as outlined in the
TSD.

Fairbanks, AK

Alaska:

The Alaska Department of Environmental Conservation (ADEC) supplemented
its original boundary recommendation for Fairbanks, Alaska, stating that
the entirety of data supports a modification of both the original ADEC
recommended boundary and the EPA proposed boundary.  

Alternatively, ADEC requested that if EPA were unable to agree with the
modified boundary that ADEC either (1) be given up to an additional year
to collect additional monitoring data under CAA Section 107(d)(1)(B)(i),
or (2) EPA could consider and implement the proposal by ADEC to set a
smaller boundary now and then expand the boundary in the future, if
warranted, based on the data collected during the winter months of 2008.

EPA Response:

After EPA promulgated the latest PM2.5 NAAQS in December 2006, States
were required to submit recommendation for area designations by December
2007 based on PM2.5 monitoring data from 2004-2006.  The State submitted
its recommendation for the Fairbanks area in December 2007 as
nonattainment for the PM2.5 24-hour standard based on that monitoring
data.  The nonattainment area boundary recommended by the State followed
a 600 ft contour to the north and west of the City of Fairbanks, the
city boundary to the east and the Tanana River to the south.  In its
letter modifying the State’s recommendation, sent to the State of
Alaska on August 18 2008, EPA used data from 2005-2007 and available
data at its disposal to designate the area nonattainment and drew the
boundaries to include all potential sources within the Fairbanks North
Star Borough (FNSB) that could contribute to a violation of the PM2.5
24-hour standards at the Fairbanks Valley monitor.  This boundary
included major portions of the Borough to the south and east of the City
of Fairbanks, which captured the military training ranges of Tanana
Flats Training Area (TFTA) to the south of Fairbanks, and the Yukon
Training Area (YTA) to the east of the Fairbanks (GIS maps are included
in the Technical Support Document (TSD) for Alaska areas).  

Generally, the CAA requires EPA to designate as nonattainment areas
those which are violating the standard “or portions thereof.”  EPA
believes that county boundaries (or in some instances CSA or CBSA
boundaries for larger metropolitan areas) are an appropriate starting
point for analysis in designating nonattainment areas.  However, where
the facts and circumstances of a particular area and analyses of
relevant technical and factual information it is shown to EPA’s
satisfaction that only a portion of a county or jurisdictional area is
violating or contributing to a violation, the Agency will designate less
than a whole county or jurisdictional area.  This EPA believes the State
of Alaska to have done.  On the 20th of October 2008, the State
supplemented its response to EPA’s recommendation and submitted a
comprehensive technical analysis using the types of information
recommended by EPA in its guidance.  These analyses included, among
other analyses, information recommended by the nine factors approach, to
identify the sources that contribute to violations of the PM2.5
standards at the Fairbanks monitor.  After carefully reviewing the
relevant information, as explained in EPA’s TSD for the Fairbanks
area, EPA is in agreement with the State of Alaska regarding the
boundary of the nonattainment area.  

Because EPA has sufficient date to promulgate a designation, there is
not basis for the extention under CAA Section 107(d)(1)(B)(i).  While
the EPA encourages collection and analysis of more information that will
be helpful in solving the PM2.5 nonattainment issues in Fairbanks, it
does find any need to delay the designation to collect this information.

Alaska

Source-specific emission estimates show that area and non-road sources
are responsible for 99 percent of directly emitted PM2.5 and that point
sources are responsible for 79 percent of the SO2 and 56 percent of the
NOX emitted in Fairbanks.  A summary of major permitted facilities
showed that two facilities are not located within the FNSB or EPA’s
proposed nonattainment area.  Data presented for Eielson Air Force Base
showed that it is responsible for less than 5 percent of the NOX and SO2
emitted within the Borough.  Data provided for military training ranges
located to the south of Fairbanks showed very limited activity during
winter months.

EPA Response:

EPA appreciates the updated emission inventory information provided by
the State of Alaska.  Along with the data submitted by Fort Wainwright
Army Base and Eielson AFB, EPA has concluded that activity in TFTA,
including the Blair Lakes Training Facility (BLTF), and the YTA do not
contribute to violations of the PM2.5 NAAQS in Fairbanks.  However,
emissions data alone do not provide conclusive evidence to exclude EAFB
from the NAA boundary.  See EPA’s TSD for Fairbanks, Alaska for
further detail.

Alaska:

Prior to last winter, the only source of PM2.5 monitoring data was from
the SLAMS monitor at the State office building in downtown Fairbanks. 
New monitoring data from other locations paint an inconsistent picture. 
The Eielson Air Force Base concentrations from an earlier winter
remained well below the 24-hour PM2.5 standard for an entire winter
season and comparisons showed there were large differences between
values recorded on base and those recorded at the downtown monitor.  The
Fort Wainwright values from an earlier winter show that, despite its
proximity to the downtown area, the values recorded over an entire
winter season never exceeded the standard.  The differences between the
values recorded on base and those recorded at the downtown monitor,
however, were much smaller.  Data collected during an episode this past
winter showed high concentrations at multiple locations.  The military
values suggest that concentrations throughout the region are not uniform
and the data collected last winter during one episode show there may be
additional areas with higher concentrations.  Clearly, the data do not
support a conclusion and suggest the need for an intensive monitoring
program, which is what ADEC and Borough are planning for the coming
winter.

EPA Response:

EPA considered this additional air quality data from Alaska in
conjunction with other technical data relevant to the nine factors, and
developed a scientifically-based boundary, which includes Fort
Wainwright, but does not include Eielson Air Force Base.

Alaska:

The annual VMT estimate reported by EPA for Fairbanks is significantly
lower than values reported by the Northern Region of the Alaska
Department of Transportation and Public Facilities (ADOT&PF).  EPA
reports a Borough wide value of 321 million miles in 2005; discussions
with ADOT&PF reported 723 million miles of travel in 2006.  Roughly 58
percent of the travel (i.e., 418.7 million miles) occurred within the
Fairbanks Metropolitan Area Transportation System (FMATS) area. 
According to comments submitted by the ADOT&PF, EPA only reported VMT
for a single category of roads (i.e., collectors) and failed to report
travel for the rest of the road system.  

Despite the error in EPA’s estimate of travel within the Borough, the
conclusion with regard to potential impacts of commuters is correct. 
The long distances to the Borough borders and low overall population
density of the region ensures that external commutes are not
contributing to elevated PM2.5 concentrations in Fairbanks.

EPA Response:

EPA obtained the data from a submission from the State of Alaska in
December 2007.  We appreciate the State correcting and providing us
accurate data (see the TSD for FNSB area).

Alaska:

The data presented above demonstrate that the long-term growth rate
throughout the populated areas of the Borough has been stable on a
long-term basis, roughly 1 percent per year, but erratic on a
year-to-year basis.  This insight confirms there is no need to expand
the nonattainment boundaries to ensure that emissions from projected
growth within the Borough are captured and controlled.

EPA Response:

EPA appreciates this information and has used this factor in combination
with all other factors in its final decision (see the TSD for FNSB area.

Alaska:

High PM2.5 days in Fairbanks are the result of very cold surface
temperatures and shallow temperature inversions, calm winds creating
stagnant conditions and inhibiting the transport and/or dispersion of
pollutants, and local emissions in each community simultaneously
producing localized air pollution increases and PM2.5 concentrations
high enough to exceed the standard in some areas.  These factors
indicate that the emission sources contributing to high pollution
concentrations in Fairbanks are fairly localized and that the
nonattainment boundary should be constrained to the populated areas
where elevated concentrations occur.  The large distances between the
military ranges and the populated areas of Fairbanks, combined with an
absence of southerly winds during PM2.5 episodes, demonstrate that the
limited emissions from these facilities do not contribute to exceedances
recorded in Fairbanks.  Similarly, data collected at Eielson show there
is no transport of its emissions into Fairbanks prior to or during
episodes except for brief periods of southeasterly flow that is shown to
be part of drainage flow along the Tanana.  Data collected at Fairbanks
International Airport demonstrate that the dominant flow prior to and
during episodes is from the northeast and there is little evidence of
any flow from the west.  These findings demonstrate that EPA’s
expansive boundaries are overly conservative and unwarranted and provide
a basis for redefining the boundaries to the south, east, and west.

The State also submitted further analysis of data for three episodes in
2004 and 2005.  Based on this analysis for Eielson, the data show a
consistent pattern of low concentrations and calm conditions at the
surface that prevent any transport to populated areas.  Data for the
upper air, into which the bulk of Eielson’s pollutants are emitted, is
coarse (i.e., soundings are taken twice per day), offering limited
insight into these short, one day episodes.  Despite the aloft data
showing some evidence transport from the southeast, analysis of the
conditions during those times shows that the transport was unlikely to
affect recorded concentrations in Fairbanks due to decoupling of the
aloft and surface layers by the temperature inversions examined.

EPA Response:

EPA generally concurs with this analysis.  In combination with other
factors, this supports the exclusion of and the military areas of TFTA,
YFTA, the BLTF, and the Eielson AFB.  In Fairbanks, filter analysis
submitted by the State of Alaska indicates that a large proportion
(40-55 percent) of PM2.5 is secondarily formed.  This indicates that the
potential sources that contribute could do so by emitting direct PM2.5
or by emitting gaseous precursors that react in the atmosphere to create
fine particles and PM2.5.  The analysis by the State also shows that the
SO4 mass shows a strong correlation to PM2.5 concentrations, which in
turn indicates that sources of SO2 are of importance when contributing
sources are considered.  SO2 is a fairly stable gaseous compound and can
transport long distances over a long periods of time.  Although the
sources at Eielson AFB emit SO2, the amount is less than 5 percent of
the total source contribution in the region, and there is weak
meteorological evidence to show a transport of such emissions and
available secondary conversion mechanism for such SO2 into SO4.  Because
of these reasons, EPA concludes that sources in EAFB do not contribute
to the violations of the PM2.5 standard in Fairbanks and should
therefore be excluded from the Fairbanks PM2.5 NAA see the TSD for FNSB.

Alaska:

In light of the information, presented above, the State in concert with
the Borough developed a recommended nonattainment boundary.  The
starting point for these recommendations was the FMATS area.  Revisions
to that boundary are primarily based on consideration of population
density, meteorology, terrain, emissions and the lack of growth. The
Final TSD for the State of Alaska includes the final PM2.5 nonattainment
boundary for Fairbanks North Star Borough. In addition to the factors
noted above, care was taken to ensure the boundary is consistent with
ownership (i.e., lots were not split) and that entire neighborhoods were
included within the proposed nonattainment area unless they were divided
by geographical features (e.g., ridgeline) that distinguished their
potential to impact Fairbanks.  

Starting with the south, the proposed boundary is consistent with the
FMATS boundary, which is located just to the north of the Tanana River. 
The eastern edge follows the FMATS boundary, which excludes Eielson, but
is expanded to include populated areas adjacent to Chena Lakes, east of
Nordale Road and north of Badger Road.  The areas excluded to the east
include undeveloped areas and swamp land.  Some of the excluded areas
also appear to include populated areas; however, a discussion with the
Borough demographer indicated that these were artifacts of arbitrary
census boundaries and in fact no one lived in those locations (the
density reflects the average of the area represented, not the location
of where people lived).  The northern end of the eastern boundary is
selected to incorporate the higher density valley to the west of Gilmore
Dome but to exclude communities farther to the east.  The low population
density of these communities and distance from the higher density areas
of Fairbanks and North Pole is seen to limit their potential impacts
despite the predominant northeast wind flow.

Recognizing the potential of Goldstream Valley to impact Fairbanks, the
FMATS boundary was expanded well to the north to include all areas with
the potential to contribute to the drainage flow.  The northern boundary
is not located at the top of the ridge separating the Chatinika Valley
from the Goldstream Valley as recommended by EPA.  Instead the northern
edge of the populated areas was selected, hence the jog in the middle of
the northern boundary.  To the west, the FMATS boundary was expanded to
include the higher population density areas with the potential to
contribute drainage to Goldstream Valley.  This includes the area to the
east of Ester Dome.  The areas along Murphy Dome Road further to the
west were excluded because of the combination of low population density,
distance from the higher density populated areas and prevailing
meteorology.  The southwestern FMATS boundary was expanded to include
Ester Valley.  This area, located to the south of Ester Dome and East of
Chena Ridge is seen as having the potential to contribute to drainage
into Fairbanks.

EPA Response:

Because of the collective evidence presented by the technical
information provided by the State of Alaska, EPA agrees with the State
of Alaska and the recommended boundary for the FNSB PM2.5 nonattainment
area.  For more information that the EPA used in arriving at this
boundary, please refer to the Technical Support Document for Alaska
PM2.5 Designations.

Fort Belknap Indian Community, MT

Fort Belknap Indian Community:

The Fort Belknap Indian Community sent a letter to EPA’s Region 8
office on September 16, 2008, asking that the lands within the exterior
boundaries of the reservation be designated as unclassifiable for the
annual and 24-hour standards for PM2.5.

EPA Response:

EPA’s Region 8 office responded to the Fort Belknap Indian Community
on October 31, 2008, informing them of EPA’s intended designations
schedule for the 2006 24-hour PM2.5 NAAQS, and offering to discuss Clean
Air Act programs the Tribe may want to consider to address air quality
on the Reservation.

Grand Rapids, MI

Michigan:

The EPA believes that Ottawa County should be included with Kent County,
one primary reason being that the emissions from the J.  H.  Campbell
plant EGUs in Ottawa County are impacting air quality in Kent County. 
However, SCR controls have been installed on one of three EGUs at the
Campbell facility, which will result in significant NOX emission
decreases.  Plans are also in place to install scrubbers to reduce SO2
emissions.  Both of these controls significantly reduce PM2.5 and are
likely to result in continued improvements in the PM2.5 levels in Kent
County.

EPA Response:

See the overview response to comments on power plant issues, located at
the beginning of Section IV.  

EPA bases area designations on current air quality data and current
emissions information.  The comment appears to indicate that the only
reason for the inclusion of Ottawa County in the Grand Rapids area is
the presence of the Campbell power plant.  While this plant is the
largest source in the county, it is not the only reason for the
inclusion of Ottawa County.  Ottawa County has a large population
(255,000), significant commuting (119,000) and VMT, and high emissions
(46,000 tons SO2, 28,000 tons NOx, and 3100 tons direct PM2.5)  The
Campbell plant has three units without scrubbers and emissions of 30,000
tons SO2 and 9000 tons NOx.  Planned emission control projects at the
J.H.  Campbell plant in Ottawa County may significantly reduce emissions
in the future.  Under section 107(d), however, EPA is assessing
violations and contribution to those violations during the relevant
three year period, i.e., 2005-2007, or 2006-2008 if the State elects to
submit 2008 data.  Even with the reductions from one unit at the
Campbell facility, the emissions from Ottawa County remain high.  The
emissions data and meteorological data indicate a contribution to the
monitored violations, and the commuting data provide a link to Kent
County.  Kent and Ottawa Counties form the Grand Rapids ozone
maintenance area, a well established connection between the counties. 
For these reasons, EPA has determined that Ottawa County should be
included as a contributing county in the Grand Rapids nonattainment
area.  

Green Bay, WI

The Oneida Tribe of Indians of Wisconsin

Oneida Tribe:

“A unique governmental and administrative burden is created due to our
political jurisdiction as an Indian Nation in which we have portions of
two of Wisconsin’s 72 counties located in our reservation boundaries. 
Unique to our situation is that Brown County would be in nonattainment
status and Outagamie County would be in an attainment designation.  We
find this situation presents a significant burden affecting our
responsibilities as a government in addressing the implementation of the
Clean Air Act’s two types of national air quality standards.  …The
72 counties are one level of government in the state of Wisconsin as
well as the 11 tribal units of government.”

EPA Response:

The Oneida Tribe’s situation is not unique.  Every state within which
EPA is designating a nonattainment area also has area that will be
designated attainment.  We believe it is not a significant burden either
to states or to tribes (or for EPA on behalf of the tribes) to
administer some jurisdictions as nonattainment and some jurisdictions as
attainment.  EPA further believes that the Clean Air Act does not
authorize EPA to designate an area that does not contribute to
violations as nonattainment simply because the governing authority has
jurisdiction over land elsewhere that warrants a nonattainment
designation, or conversely to designate an area that does contribute to
violations as attainment due to an attainment designation elsewhere. 
EPA is prepared to work closely with the Tribe to facilitate air quality
planning for both attainment and nonattainment areas.

Oneida Tribe:

The Oneida Tribe of Indians of Wisconsin recommends that the air
management area within the jurisdiction of the Oneida Tribe of Indians
of Wisconsin exterior boundaries not be determined at this time due to
the fact that parts of two counties (Brown and Outagamie Counties) are
located in our jurisdiction and that one is designated nonattainment and
the other attainment by the EPA.  

EPA Response:

With respect to state lands, EPA is required by the Clean Air Act to
promulgate designations within two years of the promulgation of an air
quality standard, in this case in December 2008.  An extension may be
granted in cases where information is insufficient to promulgate
designations, but EPA does not believe those criteria for an extension
are met here.  EPA understands that, pursuant to the Tribal Authority
Rule, EPA may grant tribes flexibility with respect to the schedules by
which tribes meet relevant requirements.  However, designations are an
EPA action, not a tribal action, and EPA believes that promulgating
tribal designations on a different timetable than state designations
would foster needless confusion.  EPA has sufficient data to make
designations for tribal areas and thus must do so at this time
consistent with the statute.  Therefore, EPA is promulgating
designations for the Oneida Tribe and other tribal lands at the same
time as it promulgates designations for state lands.

Oneida Tribe:

We would like to have you consider the following in future determination
of the impact of pollution activities and development on the Oneida
Tribe of Indians of Wisconsin membership and jurisdictional areas;

The concept of cumulative impact analysis involves environmental, social
and cultural impacts not always adequately covered in the environmental
impact assessment and statement in the normal National Environmental
Policy Act processes.

The concept of cultural catchment areas has a direct impact on our
cultural and spiritual well being.   We view that the status of our
health and community well being is based on the interconnectedness of
the elements of science, social, economic and cultural considerations to
address our holistic health approach to our world view.

EPA Response:

These comments are more germane to the attainment planning process than
to designations.  EPA intends to work with the Oneida Tribe of Indians
of Wisconsin towards development of air quality plans in a culturally
sensitive manner that considers the interconnections that are important
to the tribe.

Huntington-Ashland, WV-KY-OH

The State of West Virginia

West Virginia:

West Virginia agrees with the data and analysis presented in EPA’s
analysis relating to Factor 1 (Emissions data), Factor 2 (Air quality
data), and Factor 3 (Population density and Degree of urbanization). 
West Virginia also agrees that since there are no topographic or
geographic features in the air shed, factor 7 (Geography/topography) is
not a significant factor within the overall analysis, including our
consideration of whether to include the Graham Tax District within the
nonattainment area.  West Virginia contends that the data presented in
EPA’s analysis underlying each of these cited factors supports
exclusion of Mason County, including the Graham Tax District, from the
Huntington-Ashland nonattainment area.  

EPA Response:

EPA’s recommendation to include the Graham Tax District within the
Huntington-Ashland nonattainment area was based upon our review of
available data and implementation of EPA policy and guidance as further
and more fully explained in the technical support document (TSD)
supporting the designation of this area for the 2006 PM2.5 NAAQS.  EPA
reviewed available technical data relevant to each of the factors cited
by West Virginia.  EPA’s inclusion of the Graham Tax District as part
of the nonattainment area is supported by EPA’s analysis in its
entirety and not solely upon individual analysis of select factors
therein.  EPA believes that the overall analysis supports a finding that
the Graham Tax District within Mason County contains significant
emissions sources that contribute to PM2.5 ambient concentrations in the
nonattainment area, and that these sources are overwhelmingly large in
comparison to emissions from the remainder of the county.  Recognition
that the county in whole, or that the portion containing the Graham Tax
District is a lower ranking candidate based upon certain factors does
not preclude EPA from determining on the basis of the analysis in its
entirety that the Graham Tax District “contributes to” nonattainment
of the 24-hour PM2.5 NAAQS in the Huntington-Ashland area.  Indeed, this
is among the reasons that EPA concluded that it was appropriate to
designate only a portion of the county instead of the entire county
nonattainment.

West Virginia’s comment with regard to the exclusion of Mason County
from the 1997 annual PM2.5 nonattainment area is outside the scope of
this action to designate PM2.5 nonattainment areas under the 24-hour
PM2.5 standard.  Justification for the use of the tax boundary is
described in section 6.3.6.2 of the TSD prepared for designations under
the 1997 annual PM2.5 standard.

West Virginia:

West Virginia states that EPA’s analysis set forth in Factor 4
(Traffic and commuting patterns) includes VMT and commuter data for all
of Mason County, rather than apportioning the VMT and commuter
contribution from the Graham Tax District portion of the county, thus
overstating the relative contribution of the tax district.  The State
argues that EPA’s assertion that Mason County has more commuters than
Adams and Gallia Counties in Ohio is not relevant because Adams County
is not adjacent to the MSA.  Finally, the State argues that Mason
County’s contribution of commuter traffic to the violating county, and
also to and within the statistical area, is small.   

EPA Response:

As was the case with EPA’s technical analyses for all areas, including
the Huntington-Ashland area, the Agency relied upon 2005 VMT for our
analysis set forth in Factor 4 (Traffic and commuting patterns).  In
general, EPA’s analysis used the most recently available actual data,
rather than long-term VMT projections (as in the case of the
supplemental information for 2020 conformity projections provided by
West Virginia as part of its supplemental information) in our analysis
of traffic and commuting patterns.  For the partial county areas, EPA
assessed the factors based upon county level information (with the
exception of the contribution of emissions from large power plants).  

EPA believes that its analysis of the available transportation data is
accurate and relevant.  EPA appropriately compared Mason County’s
contribution of commuter traffic to the violating county along with
other counties subject to analysis, including Adams County, OH. 
Although Adams County is not adjacent to the MSA, a portion of Adams
County is included in the nonattainment area for the 1997 PM2.5 NAAQS. 
EPA’s analysis accurately reflects that the counties of Mason, WV;
Adams, OH; and Gallia, OH all have commuters traveling between 200 and
300 million miles based upon available data.  EPA recognizes that there
are counties within the area analysis which have greater number of
commuters who drive to and within the Huntington-Ashland area.  EPA did
not represent that Mason County (or the Graham Tax District) contributes
a regionally significant portion of total area VMT or vehicle-related
emissions.  We reiterate that our assessment of this individual factor
does not preclude EPA from determining that on the basis of all relevant
facts and scientific data, including the nine factors, that the Graham
Tax District contributes to nonattainment of the 24-hour PM2.5 NAAQS in
the Huntington-Ashland area.  Again, this is among the reasons that EPA
concluded that it was appropriate to designate only a portion of the
county instead of the entire county nonattainment.

         

West Virginia:

West Virginia agrees with the population and VMT growth estimates for
Mason County presented by EPA in Factor 5 (Growth rates and patterns) of
its technical analysis of the Huntington-Ashland area.  However, West
Virginia argues that use of predictive, travel demand modeling is more
appropriate than use of historical VMT growth data in assessing growth
rates for this analysis.  West Virginia submitted supplemental
information in the form of transportation conformity analysis for Cabell
and Wayne Counties and for the Graham Tax District portion of Mason
County.  West Virginia concluded that this data supports exclusion of
Mason County from the nonattainment area on the basis of this factor.  

EPA Response:

EPA consistently used historical county-level population and VMT growth
estimates for all areas of the country.  EPA used the same methodology
for estimating historic VMT growth in all nonattainment areas analyzed. 
Consistent with the CAA, EPA’s designations reflect EPA’s analysis
of data reflecting current conditions, as opposed to predictions of
future conditions.  EPA does not believe that the use of predictive
travel demand modeling would be consistent with the CAA because such
modeling does not reflect current conditions and the Act requires
designation of areas based on current conditions.  Rather, predictive
travel demand modeling is a tool for predicting conditions that may
occur after designation determinations are finalized.  EPA agrees that
population and VMT growth in Mason County are both relatively low,
compared to the other counties in the area of analysis.  West
Virginia’s data supports EPA’s assessment that VMT growth in Mason
County is growing at a faster rate, by percentage, than in most counties
in the Huntington-Ashland area.  EPA’s assessment of this single
factor is not out-come determinative and does not preclude it from
determining that the substantial emissions from the Graham Tax District
contributes to increased ambient PM2.5 levels in the Huntington-Ashland
area and should thus be included in the nonattainment area.  Again, this
is among the reasons that EPA concluded that it was appropriate to
designate only a portion of the county instead of the entire county
nonattainment.

West Virginia:

With respect to Factor 6 (Meteorology – weather transport patterns) of
EPA’s technical analysis, West Virginia provided supplemental
information in order to support its contention that emissions from Mason
County, including the Graham Tax District, do not contribute to the
Huntington-Ashland nonattainment area and should not be included in the
nonattainment designation.  West Virginia used NOAA Air Resource
Laboratory’s HYSPLIT trajectory model (using 100, 250, and 500 meter
trajectory heights) to provide a limited analysis of air patterns during
each of the 26 days during which monitoring data within the
Huntington-Ashland nonattainment area exceeded 30 μg/m3.  Using this
model, WVDEP calculated 26 backward trajectories to track pollutants
from the ambient monitor site to determine which trajectories “passed
over” one of the two coal-fired power plants located in the Graham Tax
District.  Based upon the data provided, West Virginia concluded that,
in most instances, backward trajectories originating from the
nonattainment monitor did not pass through the Graham Tax District. 
West Virginia therefore concludes that emissions from Mason County do
not contribute to the Huntington-Ashland nonattainment area and should
not be included in the nonattainment designation.

 

West Virginia also commented on EPA’s pollution rose data analysis (or
pollution trajectory plot) for Cabell County.  West Virginia suggests
that based upon such data,

“one would conclude that (the Graham Tax District) .  .  .  has no
significant influence on the Huntington-Ashland nonattainment area, and
should not be included in the Huntington-Ashland nonattainment area.”

West Virginia also argues that EPA Regions 3, 4, and 5 drew different
conclusions from the same Pollution Trajectory Plot for Cabell County,
WV.  

EPA Response:

EPA does not believe that West Virginia’s analysis supports a finding
that emissions from Mason County do not contribute to the
Huntington-Ashland nonattainment area and should not be included in the
nonattainment designation.

 

West Virginia’s limited backward trajectory analysis of air patterns
represents a limited “direct causal relationship” test that, if
adopted by EPA, would contravene the goal of the statute by narrowing
its application.  EPA believes that the State’s proposed alternative
analysis is insufficient for purposes of determining whether emissions
from the Graham Tax District “contribute to” ambient air quality in
the Huntington-Ashland nonattainment area.  

Section 107(d)(1)(A)(i) provides that nonattainment areas must include
“any

area that does not meet (or that contributes to ambient air quality in a
nearby area that does not meet)” the applicable NAAQS.  42 U.S.C.  §
7407(d)(1)(A)(i).  The definition of “nonattainment” specifically
includes any area that “contributes to” violations of NAAQS in a
nearby area, not only those that literally “cause” such violations
at specific monitoring sites.  Ambient PM2.5 at every monitor reflects
the cumulative impacts of many types of emissions from many sources,
near and far, that result in primary and secondary formation of
particles.  Pursuant to section 107(d), EPA must evaluate which
“nearby” areas contribute to the violations and include them within
the nonattainment area.

Additionally, EPA believes that West Virginia’s application of
NOAA’s HYSPLIT backward trajectory analysis is incomplete.  For
example, West Virginia’s analysis used only one trajectory for each
day where measurements exceeded 30 μg/m3.  If back trajectories are to
serve as the sole source for determining meteorological contribution, a
more thorough analysis using multiple trajectories for each day in
question should have been completed to account for the entire duration
of the 24-hour PM2.5 standard.  Furthermore, West Virginia’s analysis
focuses on back trajectories from the monitoring site, and, therefore,
does not adequately demonstrate that emissions from the Graham Tax
District do not “contribute to” ambient air quality in the
Huntington-Ashland nonattainment area.  West Virginia’s limited
analysis fails to take into account that both of the coal-fired power
plants located in the Graham Tax District are significant sources of
direct PM2.5 and PM2.5  precursor (NOx and SO2) emissions.  A more
complete and relevant analysis would compute multiple forward HYSPLIT
trajectories from the coal-fired power plants in question to determine
if emissions from those facilities could have “passed through” any
portion of the nonattainment area at issue.  Recognizing that NOAA’s
HYSPLIT Model is a multifaceted tool, EPA believes a more complete
analysis would support its contention that emissions from the Graham Tax
District “contribute to” ambient air quality in the
Huntington-Ashland nonattainment area.

 EPA developed pollution trajectory plots (or “pollution roses”) to
understand the prevailing wind direction and wind speed on the days with
highest fine particle concentrations.   EPA believes that such analysis
is one of several factors in evaluating emission contributions to the
Huntington-Ashland nonattainment area.  With regard to West Virginia’s
comments regarding EPA’s pollution trajectory plot for Cabell County,
EPA believes that the data presented in the August TSD was accurate. 
However, as was noted in the August TSD, EPA noted at that time that
eight high days were not plotted in that Cabell County pollution rose
due to missing or variable wind data.  However, since then EPA has
updated its pollution rose data to reflect more current information, so
that the pollution rose (Figure 6) presented in the updated, December
2008 TSD now covers the period from 2005-2007.  It shows that four days
with monitored 24-hour PM2.5 values greater than 35 and 40 µg/m3 in
Cabell County, WV in 2005-2007 occurred when the wind came from the
north-northeast.  This data suggests that emissions from the
north-northeast, i.e., from Mason County, WV as well as Lawrence and
Boyd Counties in Ohio, likely contribute to 2006 24-hour PM2.5 NAAQS
violations.  Eight other days with monitored PM2.5 values greater than
35 and 40 µg/m3 show winds from the southwest, east, and south.  This
indicates likely PM2.5 contributions from Wayne County, WV, and Boyd and
Lawrence Counties in Kentucky.

West Virginia:

West Virginia contends that, based upon the presence of emission
controls and the prospect of future emission controls for the two
coal-fired power plants in the Graham Tax District, and also that EPA
has provided States flexibility to address SO2  and NOx emissions from
nearby sources (within 200 km) via its PM2.5 Implementation Rule, it is
unnecessary to include the Graham Tax District in the Huntington-Ashland
area.

EPA Response:

Section 107(d) defines nonattainment areas as those that are violating
the NAAQS, and those that are contributing to such violations in nearby
areas currently.  Thus, EPA determined violations of the 2006 24-hour
fine particulate NAAQS based solely on the most recent three complete
years of certified monitoring data.  EPA set boundaries based on
available data and analysis that best represents present-day conditions,
and not on projected emissions reductions that may occur after
designation areas are finalized.  EPA recognized that due to new
controls at large EGUs, there may be emission reductions of SO2 and NOX
subsequent to 2005 that are not accounted for in its August 2008
analysis and EPA gave States the opportunity to submit supplemental
information on emissions controls to aid in making final designation
decisions.  

West Virginia agrees with EPA’s summary data of 2002-2007 NOx and SO2
annual emissions and heat input for the two large, coal-fired power
plants (Mountaineer and Philip Sporn) in Mason County.  West Virginia
also agrees that, while Mountaineer’s SO2 emissions were greatly
reduced beginning in 2007 due to the installation of a scrubber, Philip
Sporn lacks scrubbers to control SO2 emissions.  EPA agrees that NOx
emission rates have decreased for both facilities, however, based upon
the most recent available data (2007), heat input levels have increased
enough to keep facility-wide NOx and SO2 emission levels either flat or
higher than 2005 levels.  

EPA reviewed West Virginia’s comments and determined that they
presented no new additional information with regard to emissions
controls for these power plants.  EPA acknowledged in our analysis that
SO2 emissions will likely be reduced by the 2007 installation of wet
scrubber units at Mountaineer.  However, EPA maintains that, based upon
currently available data, the combined NOx and SO2 emissions from both
facilities, along with other data, support a finding that emissions from
the Graham Tax District contribute to PM2.5 ambient air quality in the
Huntington-Ashland nonattainment area.   

The State’s reference to the requirements of the PM2.5 implementation
rule are at most an analogy, as that rule applies to the 1997 PM2.5
NAAQS.  Even by analogy, the opportunity to take emissions reductions at
sources up to 200km from a nonattainment area is premised upon the state
being able to demonstrate that those emission reductions would have an
impact in the nonattainment area.  Moreover, the fact that emission
reductions from such sources could be shown to make an improvement in
the nonattainment area is not the sole test for designations under
section 107(d).  Under this provision, EPA must ascertain which sources
are “nearby” for purposes of the designation for the nonattainment
area, given the facts and circumstances of each area.  Other tools under
the CAA, such as section 110(a)(2)(D) and section 126 are designed to
deal with interstate emission impacts, especially regional impacts.

West Virginia:  

The portion of Mason County, the Graham Tax District, proposed to be
included in the

Huntington nonattainment area encompasses the Mountaineer and Philip
Sporn power plants, both owned by American Electric Power (AEP). 
Mountaineer has one 1,300 MW unit which is equipped with an ESP for
particulate control, an SCR for NOx control and a wet limestone scrubber
for SO2 control.  These controls are federally enforceable through
regulations, permits and a federal Consent Decree (AEP Consent Decree). 
Philip Sporn has five units, Units 1-4 are rated at 150 MW each and Unit
5 is rated at 450 MW.  All five units are equipped with an ESP for
particulate control, and low NOx burners for NOx control.  These
controls are federally enforceable through regulations and permits.  In
addition, AEP has been issued a no permit needed letter (dated September
25, 2008) for the installation of SNCR on two of the five units at
Philip Sporn (Units 3 and 4).  AEP has indicated that the installation
of SNCR on these units is to comply with the provisions of the AEP
Consent Decree which requires that “A total of at least 600 MW from
the following list of units: Sporn Units 1-4, Clinch River Units 1-3,
Tanners Creek Units 1-3, and/or Kammer Units 1-3” be retired,
retrofit, or re-powered.  The expected NOx reduction is 25 percent in
addition to the 60 percent already achieved by the LNBs.

Historical data shows that between 2002 and 2007 the SO2 and NOx
emission rates decreased at Mountaineer and Philip Sporn while heat
input increased.  The Mountaineer SO2 emission rate decreased by 95.3
percent, and the NOx emission rate decreased by 17.6 percent.  The
Philip Sporn SO2 emission rate decreased by 18.2 percent, and the NOx
emission rate decreased by 24.4 percent.  In addition, the West Virginia
Department of Environmental Protection (DEP) has conducted a RACT
analysis (which was included in the Parkersburg Annual PM2.5 SIP
revision submitted to EPA on September 9, 2008) to determine the
appropriate level of controls for EGUs.  It was determined that SCR and
wet scrubbers are not economically feasible for units the size of those
at Philip Sporn.  DEP believes that these units are equipped with
controls that meet the definition of RACT and there is no air quality
benefit to be gained by designating the Graham Tax District as
nonattainment.  Furthermore, EPA in the final PM2.5 Implementation Rule
stated that for SO2 and NOx “EPA believes that States could justify
considering not only all emissions in the nonattainment area but also
emissions within a distance that may be up to 200 kilometers from the
nonattainment area” [72 FR 20636, 25APR2007].  Since EPA has provided
States the flexibility to consider emissions from sources within 200
kilometers of the nonattainment area, it is not necessary for EPA to
include the partial counties within the nonattainment areas (see table
on page 14 of WV comment letter).

EPA Response:

See the overview response to comments on power plant issues, located at
the beginning of Section IV.  

EPA disagrees with the reasons provided by West Virginia for exclusion
of the Mason County Mountaineer and Philip Sporn power plants from the
Huntington-Ashland nonattainment area.  EPA’s analysis has evaluated
pollutant emissions; evidence from the pollution rose for this area that
there is contribution from the direction of these sources to the rest of
the nonattainment area; and the CES score that likewise indicates that
emissions from the pertinent sources contribute to violations in the
Huntington-Ashland area.  

The 2007 emissions from the Mason County power plants were about 42,000
tons SO2 and 20,000 tons of NOx.  The plants are located about 30 miles
from the closest violating monitor in the Huntington-Ashland area, a
distance at which significant formation of PM2.5 from precursors emitted
by these plants can occur.  The technical support document provides
pollution roses for both the Cabell County site and the Scioto County
site, providing evidence that the emissions from these plants can be
expected on some high concentration days to have impacts on the
violating monitors.  EPA finds that current emissions data and
meteorological information indicate that in accordance with section
107(d), Mason County is contributing to violations in the
Huntington-Ashland area.  It should be emphasized that the designations
process is intended only to determine those areas which are contributing
to a violation of the standard.  Section 107(d) was not intended to also
require EPA and the States to determine whether emission controls are or
are not economically feasible for a particular source.  The evaluation
of technically and economically feasible emission controls is required
as part of the SIP development process.

The State of Kentucky

Kentucky:

To have Boyd and Lawrence Counties designated nonattainment would invoke
additional, substantial, unnecessary requirements on local government
planning agencies.  Substantial local emission reductions from Boyd and
Lawrence Counties have already occurred, or will have occurred well
before attainment dates for this standard.  Drastic emission reductions
are scheduled to occur in the mobile sector throughout the next several
years that will greatly impact pollutant levels in the area.  Couple
these changes with those anticipated by the CAIR, or its replacements,
provisions which will further reduce SOx and NOx emissions within the
region, and the air monitoring data demonstrating attainment of the
PM2.5 standard, and the result should be that these counties be
designated attainment for the PM2.5 standard.

EPA Response:

EPA commends Boyd and Lawrence Counties for local emission reductions
that have been and continue to be achieved.  However, EPA’s 9-factor
analysis has determined that emissions from Boyd and Lawrence (partial)
Counties contribute to nonattainment in nearby Cabell County, West
Virginia, and that it is appropriate to include these counties in the
Huntington-Ashland nonattainment area.  EPA believes that consideration
of all nine factors compels designating these Counties nonattainment,
consistent with the existing nonattainment boundary for the annual PM2.5
standard.  As stated in Section V, EPA determined violations of the 2006
24-hour fine particulate NAAQS based solely on the most recent three
complete years of certified monitoring data.  EPA is setting boundaries
based on available data and analysis that best represents present-day
conditions, and not on projected emissions reductions that may occur
after designation areas are finalized.  

Kentucky:

American Electric Power (AEP), which is located in Lawrence County,
Kentucky, consists of two pulverized coal-fired boilers: Unit 01 (BSU1:
2512 mmBTU/hour 260 MW) is a pulverized coal-fired, dry bottom,
wall-fired unit constructed on or before January 1963, equipped with
overfire air, low NOX burners and an electrostatic precipitator.  Unit
02 (BSU2: 7914 mmBTU/hour 800 MW) is a pulverized coal-fired, dry
bottom, wall-fired unit constructed on or before October 1969, equipped
with an electrostatic precipitator, ammonia flue gas conditioning, low
NOX burners, and selective catalytic reduction.  Pursuant to a 10/9/2007
consent decree and for BART, AEP must install a SO2 FGD scrubber on the
larger Big Sandy Unit 2 by December 31, 2015. Assuming a conservative 90
percent control on the 2007 SO2 emissions from Unit 2, the required
scrubber would significantly reduce SO2 emissions from AEP Big Sandy by
an estimated 36,971 tons per year when the scrubber is operational on
Unit 2. Also, pursuant to the consent decree AEP Big Sandy Unit 1 now is
required to continue to burn coal with a sulfur content of no greater
than 1.75 lb/mmBTU on an annual basis. In addition to the FGD scrubber
on Unit 2 for BART, AEP is also required for BART to install ammonia
injection on Unit 1 to address inorganic condensable emissions to
improve visibility in Class I areas.  Given the existing and future new
controls required for AEP Big Sandy by consent decree and BART, KYDAQ
requests that EPA consider this information regarding the reduction in
emissions at AEP for the attainment/nonattainment designations.

(see table on page 54 of Kentucky comment letter)

EPA Response:

See the overview response to comments on power plant issues, located at
the beginning of Section IV.

The CAA requires that a nonattainment area must include not only the
area that is violating the standard, but also nearby areas that
contribute to the violation. Thus, for each monitor or group of monitors
that indicate violations of a standard, EPA is establishing
nonattainment boundaries that cover a sufficiently large area to include
both the area that violates the standard and the areas that contribute
to the violations.  The current 2007 emissions of AEP Big Sandy in
Lawrence County are 46,000 tons SO2 and 15,000 tons NOx.  As described
in the State’s comment above, a scrubber is required on one unit by
2015 and other emission reduction activities are planned for the future.
 Meteorological data indicate potential impacts from Lawrence county on
high PM2.5 days.  While Kentucky anticipates installation and operation
of additional controls, EPA must judge whether current emission levels,
supported by current regulations that provide assurance of continued
operation of existing controls, contribute to existing violations.  If
controls are required either by the Clean Air Interstate Rule or by an
equivalent rule, and SIP quality modeling based on enforceable emission
levels in Kentucky and elsewhere (with or without the Clean Air
Interstate Rule, as the case may be) demonstrates timely attainment,
then the attainment planning requirement resulting from a nonattainment
designation will not be  as burdensome to the Commonwealth.  However,
EPA cannot agree to designate areas attainment where the best evidence
indicates current contribution to nonattainment.  Accordingly, EPA has
included Lawrence County, Kentucky in the Huntington-Ashland
nonattainment area.  

The State of Kentucky makes the following points as additional rationale
for excluding Boyd County from the Huntington-Ashland nonattainment
area.

Kentucky:

Kentucky believes that EPA’s use of the contributing emissions scoring
approach was skewed.  A review of actual percentages of emissions
contributions to an area shows that Boyd County does not have the
potential to contribute to PM2.5 levels within the region.

EPA Response:

The CES is an initial screening tool used to determine potential
nonattainment area boundaries which can then be modified using
additional information provided by the State or other sources.  EPA has
recommended other types of information for this purpose through guidance
which could be considered in the final decision.  

The CES for Boyd County is lower than that of some contributing counties
(Cabell, WV; Gallia, OH; Lawrence, OH; Scioto, OH; Mason, WV; and Adams,
OH) in the area, but higher than that of others (Wayne, WV and Lawrence,
KY).  Additionally, EPA does not find that emissions from Boyd County
(10,501 tons of SO2; 10,123 tons NOx; and 3,458 tons direct PM2.5) are
without the potential to contribute to PM2.5 in the region.  One key
source of emissions located in Boyd County is the traffic and commuting
patterns.

Kentucky:

The population of Boyd County is not significant enough to have the
potential to impact PM2.5 levels in the region.  Population in this area
has shown a continuing decline over the last several years and that
decline is anticipated to continue.  Boyd County’s population actually
represents only 12 percent of the population within EPA’s proposed
nonattainment boundaries.

EPA Response:

Although the population of Boyd County may have declined in recent
years, the County remains moderate in population size and density.  Of
the Huntington-Ashland CSA counties that EPA stated intentions to
designate as nonattainment in our August 2008 letters to State
Governors, about 23 percent reside in Cabell County, WV, compared to
about 12 percent living in Boyd County.  The population densities of the
two counties are similar at 327 persons per square mile and 305 persons
per square mile Cabell and Boyd County, respectively.  Boyd County has
the highest population of the Kentucky counties in the area and it is
one of the most densely populated counties in the area.  EPA’s finds
this factor adds support to EPA’s decision to include Boyd County in
the Huntington-Ashland nonattainment area.

Kentucky:

The air monitoring data demonstrates attainment of the PM2.5 standard.

EPA Response:

Although the ambient monitor in Boyd County has a design value of 33
µg/m3, based on 2005 – 2007 data, EPA believes sources in Boyd County
contribute to the violating monitor in nearby Cabell County, WV, with a
design value of 37 µg/m3, and also to Scioto County, with a design
value of 36 µg/m3.  In.  In addition, EPA notes the design value of the
Boyd County monitor is up from the 2004 – 2006 design value of 32
µg/m3.  Specifically, Boyd County has substantial levels of emissions
and meteorological data that indicate a contribution to PM2.5
concentrations at the violating monitors on the high days. 
Additionally, Boyd County, which includes the city of Ashland, Kentucky,
has a high degree of population density, urbanization, and commuting
within the area.

The State of Kentucky makes the following points as rationale for
excluding Lawrence County, Kentucky, from the Huntington-Ashland
nonattainment area.

Kentucky:

Kentucky believes that EPA’s use of the CES approach was skewed.  A
review of actual percentages of emissions contributions to an area shows
that Lawrence County does not have the potential to contribute to PM2.5
levels within the region.

EPA Response:

The CES is an initial screening tool used to determine potential
nonattainment area boundaries which can then be modified using
additional information provided by the State or other sources.  EPA has
recommended other types of information for this purpose through guidance
which could be considered in the final decision.  

The CES for Lawrence County is lower than those for the other
contributing counties (Cabell, WV; Gallia, OH; Lawrence, OH; Scioto, OH;
Mason, WV; Adams, OH; Wayne, WV; and Boyd, KY) in the area.  However,
EPA does not find that emissions from Lawrence County (50,239 tons of
SO2; 13,761 tons NOx; and 5,134 tons direct PM2.5) are without the
potential to contribute to PM2.5 in the region.  One key source of
emissions located in Lawrence County is the Big Sandy Power Plant.

Kentucky:

The air monitoring data demonstrates attainment of the PM2.5 standard.

EPA Response:

Although there is no ambient monitor in Lawrence County, EPA believes
the resident power plant contributes to the violating monitor in nearby
Cabell County, WV, with a design value of 37 µg/m3.  Thus, the
appropriateness of additional controls for this plant will be evaluated
by the State and EPA during the development of the nonattainment area
plan for this area, as necessary,  in accordance with section 172.
Analysis of all additional data described in the TSD indicates these
counties are contributing to the violating monitor.  Section 107
requires designation of areas that, although not they themselves
violating, are contributing to nearby nonattainment.

The State of Ohio

Ohio:

In its October 2008 letter to EPA, Ohio objected to EPA’s proposed
inclusion of Lawrence and Scioto Counties in the Huntington-Ashland
area.  Specifically, Ohio argued that:

“Lawrence and Scioto Counties have low population and insignificant
commuting traffic.  Moreover, due to the location of the monitor in
Lawrence County, it is likely that the PM2.5 levels being detected are
coming from sources across the Ohio River in West Virginia and
Kentucky.”

In addition, Ohio questioned why EPA would include Lawrence and Scioto
Counties in Ohio, but exclude Greenup County Kentucky from this area. 
Ohio argued that this Kentucky County is located across the river from
Lawrence and Scioto, and based upon CES scores and other analytical
tools, Greenup, Boyd, and Lawrence Counties in Kentucky should be
included in the Huntington-Ashland area.

EPA Response:

EPA disagrees with Ohio’s objections to the Huntington-Ashland area. 
First, Scioto County is violating the 2006 24-hour PM2.5 NAAQS based
upon 2005- 2007 monitoring data.  Thus, in accordance with section
107(d), EPA is required to designate that area nonattainment based on
violations not contribution.  Even if EPA were to agree that Scioto has
modest population and commuting, relative to other counties within this
area, the violation of the NAAQS in Scioto overrides those
considerations.   Further, Scioto County’s geographic location and
meteorological relationship to violations in the Huntington-Ashland area
could still render its emissions, population, and commuting sufficient
to conclude that it is contributing to violations in the area as well.  


Second, EPA has concluded that Lawrence County, Ohio is contributing to
violations in the Huntington-Ashland area based upon consideration of
various forms of information that indicate it contributes to violations
in the area.  Lawrence County is immediately adjacent to both Scioto
County, Ohio, and Cabell County, West Virginia, each of which is
violating the NAAQS based upon 2005-2007 data.  Emissions inventory
information indicates that emissions from sources in Lawrence are low
relative to other nearby counties with large stationary sources, but
nevertheless are adding to the mix of emissions in the area – over
1,000 tons per year of direct PM2.5 and over 500 tons per year of SO2
and over 3,700 tons per year of NOx.  The nature and sources of this
contribution are reflected by the relatively large population of
Lawrence relative to all other counties in this area (nearly 63,000 –
placing it only behind Cabell and Scioto in population) and the
relatively large amount of commuting to violating counties throughout
the area and large amount of overall VMT indicating contribution from
mobile sources.  The CES scores for this area indicate that Lawrence
County contributes to violations, and contributes nearly as much as
Cabell and Gallia Counties and more than other nearby counties.  EPA has
concluded that the geographic proximity of Lawrence County to violations
in both Cabell and Scioto Counties, the CES score (confirming
contribution), and the relatively high degree of commuting supports
inclusion of Lawrence County, Ohio, within the nonattainment area.  This
will assure that the emissions activities within Lawrence County will be
evaluated and controlled, as appropriate, in the nonattainment area plan
developed for this area as a whole.

Finally, EPA disagrees with Ohio’s contention that inclusion of Scioto
and Lawrence Counties in Ohio necessarily compels the inclusion of
Greenup, Lawrence, and Boyd Counties Kentucky in the Huntington-Ashland
area.  As noted above, Scioto is violating the standard so a comparison
of its relative degree of contribution is not germane.  EPA is including
Boyd County, Kentucky within the area because of its contribution, so
that comparison is likewise does not support Ohio’s argument. 
Similarly, EPA is including most of the emissions of Lawrence County,
Kentucky, in the nonattainment area, so that comparison as well does not
support Ohio’s argument.  Only Greenup County is excluded from the
nonattainment area, because on balance its emissions are lower,
particularly for directly emitted particulate matter, than those of
other counties in the area such as Lawrence County.

Ohio:

The State of Ohio objected to EPA’s proposed inclusion of Monroe and
Sprigg Townships, Adams County, and Cheshire Township, Gallia County,
within the nonattainment area for Huntington-Ashland for the 2006
24-hour PM2.5 NAAQS.  In its October 8, 2008, letter, Ohio generally
objected to this on the grounds that “the existence of an electric
generating unit in a township should not be the sole factor in
determining nonattainment.”  More specifically, Ohio argued in that
letter that each of these townships should be excluded because the
source located in each are adequately controlled, or will be in the
future:

”Some emissions from the electric generating units located in Monroe
and Sprigg Townships (Adams Co.) and Cheshire Township (Gallia Co.) are
already controlled with highly efficient control equipment, and other
emissions are expected to be controlled within year 2009.  Besides
electrostatic precipitators, all these units have installed flue gas
desulfurization systems, which are currently operational or will be
operational by July 2009.  Moreover, all the units have SCR systems,
which will be operating for the entire ozone season, controlling and
reducing nitrogen oxides emissions.  In addition, some units, in both
counties, will have SCR equipment in continuous operation (the whole
year) starting in year 2009, increasing the reduction of nitrogen oxides
emissions.”

EPA Response:

See the overview response to comments on power plant issues, located at
the beginning of Section IV.

EPA disagrees with the reasons provided by Ohio for exclusion of each of
these areas from the Huntington-Ashland nonattainment area.  First, Ohio
is incorrect that the “existence” of the source is the sole reason
for EPA’s inclusion of these townships.  EPA’s analysis has
evaluated not just the existence of the source, but also other
considerations including:  the significant quantity of emissions of the
pollutants (including PM2.5 precursors and directly emitted PM2.5) that
speciated PM2.5 data in this nonattainment area show to be of concern;
evidence from the pollution rose for this area that there is
contribution from the direction of this source to the rest of the
nonattainment area; and the CES score that likewise indicates that
emissions from the pertinent townships contribute to violations in the
Huntington-Ashland area.  

In the case of Monroe and Sprigg Townships, 2005 emissions inventories
indicated that sources in Adams County emit nearly 6,000 tons per year
of direct PM2.5 and over 126,000 tons of SO2 and over 33,000 tons of NOx
per year.  The vast majority of these emissions came from the J.M. 
Stuart and Killen Station coal fired power plants in Monroe and Sprigg
Townships.  These Adams County plants are only about 40 miles from the
violating monitor in Scioto County and about 70 miles from the violating
monitor in Cabell County, West Virginia, distances at which significant
formation of PM2.5 from precursors emitted by these plants would have
occurred.  The technical support document provides pollution roses for
both the Cabell site and the Scioto County site, providing evidence that
the emissions from these plants can be expected on some high
concentration days to have impacts on the violating monitors.  

As part of the public comment period, EPA received information from
DP&L, the owner of both plants, indicating that a scrubber had been
installed at the Killen plant in 2007 and scrubbers had been installed
on all four units at the Stuart plant in mid-2008.  EPA is not aware of
any action that has been taken to ensure that the controls on the Killen
unit are federally enforceable at this time, while the controls at the
Stuart plant were installed to satisfy the provisions of a consent
decree.  The consent decree also requires the controls to be required in
a federally-enforceable permit once the consent decree is satisfied. 
Based on this information, the emissions in 2009 from these Adams county
plants are estimated to be about 7200 tons SO2 and 20,000 tons NOx. 
Taking all of this information into account, EPA finds that the Adams
county plants have taken important steps to reduce emissions.  However,
not all of the controls are federally enforceable and the remaining
emissions are not so small as to be considered to have no contribution. 
Based upon this evidence relevant to designations in accordance with
section 107(d), EPA concluded that the emissions from Monroe and Sprigg
Townships are currently still “contributing” to violations in the
Huntington-Ashland area.  

In the case of Gallia County, the two largest sources are located in
Cheshire Township:  the Kyger Creek and Gavin coal fired electric
generating plants.  The two units at the Gavin plant have scrubbers and
SCR in place, but 2007 emissions are still substantial:  29,000 tons SO2
and 33,000 tons of NOx.  The five units at the Kyger Creek plant do not
have scrubbers – scrubbers are planned for 2010 – but they do have
SCR installed.  Like the Gavin plant, current 2007 emissions are also
substantial:  57,000 tons SO2, and 13,000 tons NOx.  The plants are
located about 40 miles from the violating monitors in the
Huntington-Ashland area, a distance at which significant formation of
PM2.5 from precursors emitted by these plants can occur.  The pollution
rose and other meteorological data indicate that on high PM2.5 days,
impacts at the violating monitors can come from the direction of these
plants.  Moreover, the CES scores for the Huntington-Ashland area
indicate that there is contribution from emissions in Gallia County to
elevated levels of ambient PM2.5 on days on when winds blow generally
from the direction of those sources to the rest of the
Huntington-Ashland area.  The CES score for Gallia County is 100, the
highest score of any contributing county in the area.  Based upon this
evidence relevant to designations, in accordance with section 107(d),
EPA concluded that emissions from Gallia County are “contributing”
to violations in the Huntington-Ashland area.  

In the context of the nonattainment SIP for this area, Ohio will have
the opportunity and the obligation to perform further analysis and
attainment demonstration modeling that can further explore the
contribution of the sources in these townships in Adams and Gallia
Counties, and the appropriate course of action to reduce that amount of
contribution.

EPA also disagrees with Ohio’s conclusions relating to the level of
control of the sources in Monroe, Sprigg, and Cheshire Townships as a
basis for excluding them from the nonattainment area.  As noted above,
EPA has concluded that these sources do factually contribute to the
violations in the nearby area, notwithstanding the current level of
control at these sources.  Under section 107(d), EPA believes that the
level of control of emissions from the source is a relevant inquiry in
the designations process to assure that EPA and the State are using up
to date and accurate emissions information.  EPA does not believe
however that the level of control alone should be used as a basis to
exclude a source that is otherwise deemed to be contributing to the
violations, based upon information such as its amount of emissions,
geographic location, and meteorological connection to the nearby
violating area.  For a source in an area that is violating, or is
contributing to violations of, the NAAQS, the adequacy of control is
more appropriately assessed during the development of the nonattainment
area SIP for an area.  Section 107(d) does not require EPA, in essence,
to construct a mock-up nonattainment area plan and to predetermine
appropriate levels of pollution control on behalf of the State, as a
necessary step in promulgating the designation for the area.  In short,
the level of control of a source does not in isolation answer whether
the source is contributing to the violations.

Ohio provided information on the status of emission controls at four
plants in the Huntington-Ashland area.  Some of the emissions have long
been controlled with effective control equipment, some of the emissions
have become well controlled more recently, and some of these emissions
are expected to be controlled within a few years.  In Gallia County,
Gavin Station has long had control equipment reducing SO2 emissions and
more recently has installed controls to reduce NOx emissions, but
emissions remain high.  Also in Gallia County, Kyger Creek Station is
expected to install SO2 emission controls by 2010, but current emissions
remain high.  In Adams County, JM Stuart is subject to a consent decree
requiring control of both SO2 and NOx.  Also in Adams County, Killen
Station has installed control equipment both for SO2 and for NOx,
although EPA is aware of no enforceable requirement for the company to
operate this equipment, particularly if the D.C. Circuit Court of
Appeals vacates CAIR.  More importantly, even with these controls, total
emissions from these plants remain somewhat high.  The company, in
comments submitted in response to EPA solicitation of public comments,
stated that emissions with control would be approximately 7200 tons per
year of SO2 and 20,000 tons per year of NOx.  The emissions from these
facilities even with controls are sufficient enough that EPA judges that
these facilities contribute to violations in Scioto and Cabell Counties.
 Also of concern is that no permit or other enforceable document
requires operation of the control equipment at Killen Station. 
Therefore, EPA is including Monroe and Sprigg Townships in Adams County
and Cheshire Township in Gallia County, along with Scioto and Lawrence
Counties, in the Huntington-Ashland nonattainment area.

By inclusion of these townships within the nonattainment area, Ohio and
EPA will evaluate these sources as part of the development of the
nonattainment area plan to assure that they are adequately controlled,
and controlled in a timely fashion, to provide for expeditious
attainment of the 24-hour PM2.5 NAAQS in the area, as required by the
CAA.  This inquiry may conclude that the sources are appropriately
controlled, but that assessment should occur during the proper process
contemplated by the CAA for development of nonattainment area plans.

With respect to future or planned future emissions reductions at these
sources, EPA has concluded that for evaluating contribution for
designations under section 107(d), it is appropriate to look at the
actual level of emissions in the relevant 3 year period, whether 2005
– 2007, or 2006 – 2008 if a State elects to submit additional data
for 2008.  Given the present tense formulation of section 107(d), EPA
does not believe it is appropriate to base designations on projections
of future levels of emissions or controls.  Thus, although EPA supports
the plans for future controls, EPA believes it would be inappropriate to
consider the controls expected in 2010 in Cheshire Township, since these
emissions reductions will happen beyond the period that is appropriate
for consideration in the designations process.  However, these
reductions will occur during the time that Ohio is developing the
nonattainment area plan for this area, and they may properly be taken
into account during that process.  Ohio may conclude, and EPA may
ultimately agree, that these reductions are all that is necessary to
reduce emissions from these specific sources within the PM2.5
nonattainment area plan for Huntington-Ashland, but it would be
inappropriate to prejudge that outcome now.  

Imperial County, CA

CARB:

CARB requests that EPA modify the proposed nonattainment area boundaries
to be consistent with California recommendation, that only the City of
Calexico and not the entire Imperial County be designated as
nonattainment for the revised PM2.5 standard.  

Population:  The majority of Imperial County is largely unpopulated. 
Calexico has 24 percent of the population of Imperial County within its
boundaries with the second largest population and the highest population
density.  The largest population area, El Centro, only nine miles north
of Calexico, is in attainment of the standard.  

Calexico accounts for only 5 percent of the population and 4 percent of
the land area of the combined Calexico/Mexicali urban area.  The
population density of Imperial County is less than a fifth of the
Municipality of Mexicali, in an area of roughly the same size.  A
similar situation is faced at the border area of Nogales, Arizona.  This
population disparity was noted by EPA in considering the Nogales area as
a focused nonattainment area for PM2.5, retaining the rest of Santa Cruz
County in attainment.  CARB believes that air quality in the City of
Calexico is similarly overwhelmed by the much larger City of Mexicali
across the border and requests similar consideration.

Transport from the northern portion of Imperial County:  In its
response, EPA showed HYSPLIT model results that implied a contribution
from emissions throughout Imperial County to elevated levels at the
Calexico-Ethel site.  However, other sites in the county showed much
lower concentrations during Calexico exceedance days, indicating that
the high concentrations at Calexico were unlikely to be due to a
northern influence.

Transport from Mexico:  With regard to transport of emissions from
Mexico, in reply to a request from EPA, CARB has supplied a table
showing 2005 Imperial County and Mexicali emissions.  The table shows
the great disparity between Imperial County and Mexicali emissions
including stationary, area and mobile sources.  

As noted in the EPA response table, the emissions inventory for Imperial
County shows a 24 percent contribution from carbon.  Chemical
composition data for Calexico specifically from exceedance days at
Calexico show an organic carbon contribution of over 50 percent.  The
seasonal pattern shows the strong wintertime increase in organic carbon.
 We believe the majority of these carbon emissions are the result of
transport from the City and municipality of Mexicali, Mexico where
residential trash and wood burning are largely unregulated.  In
addition, the majority of the exceedance days occurred during the
December/January time period when there are increased volumes of smoke
across the border.  

EPA Response:

As requested by EPA, CARB provided additional information showing that
the PM2.5 emissions in the Calexico area are different from those in the
rest of Imperial County.  The 2007 emission inventory for the entire
Imperial County indicates the major sources are fugitive windblown dust,
unpaved road dust and farming operations.  The violating monitor in
Calexico indicates that PM2.5 is 48 percent organic carbon and 22
percent ammonium nitrate.  This speciation data indicates that primary
sources in Calexico are residential and/or open burning and mobile
sources.  These sources correlate with population centers and location
of major highways.  

California has noted that emissions from activities in Mexicali and at
the border crossing contribute significantly to PM2.5 levels in
Calexico.  The Calexico population is 30,000, compared to 1 million in
Mexicali, Mexico.   Heavy truck traffic at the border crossing
contributes to mobile source emissions.  The data also show that the
high levels of PM2.5 emissions occur primarily in the winter months when
residents south of the border heavily depend on wood as a fuel source.

EPA agrees that the PM2.5 designation for Imperial County should be
limited to the areas where local sources account for the contribution to
the fine particle levels on exceedance days.  Therefore, EPA redefined
the PM2.5 nonattainment area in Imperial County to encompass those areas
within Imperial County that contain the bulk of the sources of PM2.5 and
PM2.5 precursor emissions, which EPA has concluded are located primarily
in the the cities of Calexico, El Centro and Brawley and environs.  
This nonattainment area covers 690 square miles, or 15 percent of the
entire County, and it captures 86 percent of the population, as well as
all major highways and stationary sources.  

The Torres Martinez Desert Cahuilla Indians Reservation

Torres Martinez:

The Torres Martinez Tribe asks EPA to consider a possible designation
for the Torres Martinez Reservation as “unclassifiable” rather than
nonattainment for the 2006 PM2.5 air quality standard because the
following factors that EPA used in its analysis do not accurately
reflect the true nature of the Torres Martinez Reservation.   

The Torres Martinez Air monitoring station does not have a PM2.5
analyzer and has not monitored for PM2.5.  This should be enough for a
non-classifiable determination for a PM2.5 designation for the Torres
Martinez Tribe.

The emissions data that EPA used to determine a nonattainment status for
the Reservation was based mainly from the Calexico Ethel Street air
monitoring site in Imperial County.  That air monitoring site is
approximately 85 miles from the Headquarters of the Torres Martinez
Tribe.  The Torres Martinez air program staff believes that the South
Coast Air Quality Management District (SCAQMD) site in Indio in
Riverside County, which is located about 13 miles from Tribal
Headquarters, would be a more appropriate PM2.5 database to draw
conclusions from.  

   The Torres Martinez Reservation is located in the Eastern Valley, not
the Imperial

   Valley, as EPA’s designation proposal factors illustrate.

EPA Response:

The Torres-Martinez Reservation spans both Imperial and Riverside
Counties.  The nearest PM2.5 monitor to the tribal area is the Indio
monitoring station operated by South Coast AQMD (SCAQMD) in Riverside
County.  The design value for the Indio site (for 2005-07) is 24 ug/m3,
which is below the standard of 35 ug/m3.  The Tribe states that tribal
lands are located in the eastern valley and that the SCAQMD monitoring
data are more representative for their circumstances.  EPA appreciates
the clarification.  The Tribe also states that the Torres Martinez
Reservation does not have a PM2.5 monitor.  If a tribe has no data of
its own, EPA looks to the nearest monitoring sites for an indication of
the air quality in the vicinity of the tribe.  In this case, as the
Tribe has pointed out, the nearest monitors are operated by SCAQMD and
the monitor in Indio is not violating the PM2.5 standard.  Consequently,
EPA believes that a designation of attainment/unclassifiable is
appropriate for the Torrez-Martinez Tribe.

Torres Martinez:

While jurisdictional boundaries facilitate planning and organizational
structure for air quality mitigation, it is the opinion of the Torres
Martinez Environmental Department that NAAQS pollutants be considered on
a one-by-one basis and not grouped with other pollutants for designation
proposals.

EPA Response:

EPA considers pollutants on a one-by-one basis.  The pollutants are
those specifically mentioned in the CAA, namely, carbon monoxide, ozone,
sulfur dioxide, nitrogen dioxides, lead and particulates.  EPA sets the
standards for each of these pollutants and makes designations based on
whether areas are meeting or violating the separate standards.  However,
jurisdictional boundaries are a factor in the decision-making process,
including existing boundaries for other pollutants.

Indianapolis, IN

Indiana:

Indiana strongly believes that a number of Indiana counties were
improperly designated as nonattainment under the annual PM2.5 standard. 
Indiana cites the outer counties of the Indianapolis area as an example
in which the counties are attaining the standard and area and mobile
sources are not culpable for the urban excess once measured in the core
county.

EPA Response:

This is not the appropriate forum for challenging designations for the
annual PM2.5 standard.  At issue is whether counties that are monitoring
attainment are nevertheless contributing to violations nearby.  As noted
above, evidence with respect to Indianapolis indicates that the area
continues to monitor a significant urban increment, and, contrary to
Indiana’s assertion, sources in the counties designated nonattainment
contribute to the violations recorded in the violating county.

Indiana:

The monitors within Marion County that the surrounding counties would
influence the most measure air quality below the standard.  EPA should
review the 2006-2008 monitoring data prior to designating any counties
to nonattainment.  Not only will the designation of non-contributing
counties in Central Indiana be unnecessary and serve no air
quality-related purpose, it will result in a tremendous amount of
unnecessary work for the Indiana Department of Environmental Management
(IDEM) and EPA in redesignating the area shortly after designations are
effective.

EPA Response:

Proximity to monitors measuring attainment suggests that the counties
are not by themselves causing violations measured at more distant
monitors, but these facts do not refute EPA’s findings that these
counties have an impact at the violating monitors that warrants
concluding that the counties contribute to the violation.  EPA plans to
review 2006 to 2008 data (if available from the State) and apply an
attainment designation if all monitors in the area are found to be
measuring attainment.  Also, EPA is not designating any noncontributing
counties as nonattainment.  If the Indianapolis area is in fact
designated nonattainment, the designation as nonattainment of Marion
County and the four additional counties that contribute to violations in
Marion County will facilitate the important air quality planning that
the CAA requires for such areas.

Juneau, AK

Alaska:

After reviewing the available data and EPA’s analyses, the Alaska
Department of Environmental Conservation (ADEC) believes the available
scientific evidence does not support EPA’s boundary recommendations
which substantially expand upon those recommended by us.  ADEC believes
public health will be protected and the applicable legal requirements
met by taking the actions described in Alaska’s letter.  For Juneau,
we are requesting EPA revisit certain assumptions and include data from
2008 before making a final decision on whether a nonattainment
designation is warranted, and if so, the appropriate boundaries of the
nonattainment area.

EPA has presented no data demonstrating a need to expand the State’s
proposed PM2.5 nonattainment boundary.  A combination of population
density and topographical data shows there are several distinct airsheds
within Juneau.  The meteorology data show there was no transport from
any of those airsheds into the Mendenhall Valley prior to or during high
concentration episodes.  Vehicle travel into Juneau from outside areas
during the winter is essentially nonexistent and traffic count data
demonstrate that travel within Juneau declined over the past decade. 
The available monitoring data show that concentrations within the
Mendenhall Valley are largely the result of combustion emissions during
the winter and that design values for Lemon Creek did not exceed the
24-hour PM2.5 standard.  To address the concern about elevated PM2.5
concentrations recorded within the Mendenhall Valley, a new ordinance
was implemented to call for burn bans when concentrations approached the
standard.  Finally, an analysis of monitoring data recorded in the
Mendenhall Valley suggests that the 2005–2007 design value does not
exceed the 24-hour PM2.5 standard.  

EPA Response:

After EPA promulgated the latest PM2.5 NAAQS in December 2006, States
were required to submit recommendation for area designations by
December, 2007 based on PM2.5 monitoring data from 2004-2006.  Based on
this data Juneau was in attainment of the 24-hour PM2.5 NAAQS.  However,
the monitor in Mendenhall Valley measured several exceedances in
December 2007.  A preliminary review of the 2005-2007 PM2.5 monitoring
data in April 2008, indicated that the area may potentially be in
violation of the 24-hour PM2.5 NAAQS.  Based on this review, EPA invited
the State to provide a recommendation for the Juneau area based on a
through technical analysis of the data.  The State submitted its
recommendation for the Juneau area in June 2008, as nonattainment for
the PM2.5 24-hour standard based on the monitoring data and recommended
the old PM10 nonattainment area boundary as the PM2.5 boundary without
any further technical justification.  In its letter modifying States
recommendation sent to the State of Alaska on August 18, 2008, EPA used
data from 2005-2007 and available data at its disposal to designate the
area nonattainment and drew the boundaries to include all potential
sources that could contribute to a violation of the PM2.5 24-hour
standards at the Mendenhall Valley monitor.  

Generally, the CAA requires EPA to designate as nonattainment areas
those which are violating the standard “or portions thereof”.  And
while commonly, and only as a point of departure, EPA relies on county
boundaries (or in some instances CSA or CBSA boundaries) in designating
nonattainment areas, where, based upon the unique facts and
circumstances of a particular area and analyses of relevant technical
and factual information it is shown to EPA’s satisfaction based upon
those analyses that only a portion of a county or jurisdictional area is
violating or contributing to a violation, we will designate less than a
whole county or jurisdictional area.  This we believe the State of
Alaska to have done.  On the 20th of October 2008, the State
supplemented its response to EPA’s recommendation and submitted a
comprehensive technical analysis using the types of information
recommended by EPA in its guidance.  These analyses included, among
other analyses, information recommended by the nine factor approach, to
identify the sources that contribute to violations of the PM2.5
standards at the Mendenhall Valley FRM for the years 2005-2007.  After
analyzing all available monitoring data form the Mendenhall Valley FRM
for the years 2005-2007, EPA has determined that area is in violation of
the PM2.5 NAAQS.  After carefully reviewing the relevant information, as
explained in EPA’s TSD for the Juneau area, EPA concurs with the State
of Alaska that the existing PM10 boundary in Juneau constitutes an
appropriate boundary for the PM2.5 nonattainment area also.  This means
that sources that contribute to the violation of the 24-hour PM2.5 NAAQS
are within this boundary and controlling those sources will help the
area attain the 24-hour PM2.5 standards expeditiously see the TSD for
Juneau area.  EPA notes that it has not chosen these boundaries because
they applied for the PM10 NAAQS, but rather because EPA has determined
that the same area happens to be sufficient to include the violating and
contributing areas for Juneau, given the very unusual topographical and
meteorological considerations of this area.

Alaska:

The State requests that EPA revisit the design value calculation to
determine if, in fact, there is a PM2.5 nonattainment area in Juneau. 
Second, the State requests EPA consider all 2008 data and recalculate
the design value for the 2006-2008 period to insure that the area is
clearly nonattainment.  Finally, if it is determined that there is
justification for a nonattainment area, the State recommends that EPA
adopt the existing PM10 nonattainment area for the Mendenhall Valley.  

In a conversation with Neil Frank (Senior Advisor EPA/OAR/OAQPS/AQAD)
during the recent AQS conference in Milwaukee (August 2008), Barbara
Trost (Acting Air Monitoring Program Manager, Alaska Department of
Environmental Conservation/Air Quality) explained the State’s concern
and objection to how the design values for the Floyd Dryden site in
Juneau had been calculated.  Mr. Frank indicated that consecutive
sampling days should not be included in the design value calculation. 
Given that Mr. Frank believed that the calculation of the Juneau design
value may be flawed and considering the information provided above, the
State respectfully requests that EPA review and revisit the Juneau
design value to ensure that it has been properly calculated with respect
to the national ambient air quality standard.

EPA Response:

40 CFR 58.12 "Operating schedules" of the monitoring regulations
provides opportunity for States to request approval for seasonal
sampling from the EPA Regional Administrator and Appendix N to 40 CFR
Part 50 has a provision for computing "the annual 98th percentile values
when sampling frequencies are seasonal."  In light of the extra
collected samples, in addition to those required with the routine once
in three day schedule, EPA reconsidered the 2005-07 Juneau design value
to verify if the Juneau data would be eligible for this alternative
computational approach.

Upon review, EPA determined that the data do not conform to a seasonal
sampling schedule.  The Juneau sampling approach was not described in
the State monitoring plan and it did not receive prior approval by the
EPA Regional Office.  Furthermore, the reported sampling days do not
appear to follow a systematic plan and the reasons for sampling on some
extra days and not other extra days were not provided with the State
comments.  In particular, the relationship between extra sampling days
and potential for high concentrations is not known.  Given the lack of
specific documentation and prior approval for the realized sampling
plan, EPA does not have a basis to use 40 CFR Part 50, Appendix N,
Section 4.5(a)(2) .  Consequently, EPA used the "Regular procedure for
identifying annual 98th percentile values," as described in Appendix N,
Section 4.5(a)(1).  Based on this finding, the Juneau Design Value does
indicate a violation of the 24-hour PM2.5 NAAQS.  Accordingly, as
explained above, EPA has determined that the existing PM10 boundary in
Juneau constitutes an appropriate boundary for the PM2.5 nonattainment
area also.

Klamath Falls, OR

Oregon:

Oregon recommends the Klamath Falls Air Quality Zone (AQZ) as the
nonattainment boundary and provided significant additional technical
information justifying this boundary.

EPA Response:

EPA agrees with the State’s recommendation of using the AQZ as the
boundary of the Klamath Falls nonattainment area (NAA).  Oregon provided
additional technical information on population and woodstove density,
land use, area photographs, PM2.5 filter analysis, and meteorology.  As
discussed in the TSD to this action, EPA agrees with the State and now
believes that only the Klamath Falls AQZ should be designated
nonattainment based on EPA’s conclusion that areas outside of the
Klamath Falls AQZ are not contributing to the violating monitor.  EPA
agrees that the bulk of potential contributing residential and
industrial sources are within the AQZ.

EPA considered a partial county boundary in situations where a unique
situation is shown to exist in a county in which the sources captured by
the partial county boundary account for nearly all of the total fine
particle mass on exceedance days.  In such cases, given compelling
evidence, EPA considered designations of partial counties to be
appropriate.  EPA selected the boundaries for these partial county
nonattainment areas by relying on legally recognized governmental
boundaries (e.g., townships) in which the sources are located.  For the
Klamath Falls nonattainment area, EPA used township-range to closely
match the Air Quality Zone.

Knoxville-Sevierville-LaFollette, TN

Tennessee:

The State of Tennessee accepts EPA’s proposed Greater Knoxville area
boundaries under one condition:  EPA should not determine final
designations of the Knoxville area boundaries until all data flag
requests have been resolved.  In their response letter, dated October
20, 2008, the State of Tennessee claims that a number of flags have been
requested yet no replies have been issued.  The State of Tennessee
believes that, even if no data flags are approved, Knox, Anderson,
Blount, London, and Roane counties are very likely to achieve attainment
based on 2006 through 2008 design values.  

EPA Response:

EPA has reviewed flagged data that was collected in 2007 and does not
concur with the flagged data for Knoxville because a public review and
comment period was not conducted as required by the Exceptional Events
rule.  For an event to qualify as an exceptional event, the
demonstration submitted to EPA must meet all the procedural and
substantive requirements of the Exceptional Events Rule at 40 CFR 50.1
and 50.14.  EPA prepared a response to the exceptional event
demonstration and mailed it on November 21, 2008, to the Knox County
Department of Health and to the State of Tennessee.

Tennessee:

The Tennessee Valley Authority (TVA) will have scrubbers online at Bull
Run Fossil Plant by the end of 2008 and, in two phases, at the Kingston
Fossil Plant in 2009 and 2010.  

EPA Response:

See the overview response to comments on power plant issues, located at
the beginning of Section IV.

The CAA requires that a nonattainment area must include not only the
area that is violating the standard, but also nearby areas that
contribute to the violation.  Thus, for each monitor or group of
monitors that indicate violations of a standard, EPA is establishing
nonattainment boundaries that cover a sufficiently large area to include
both the area that violates the standard and the areas that contribute
to the violations.  

The Greater Knoxville area includes the Bull Run plant in Anderson
county (2007 emissions:  40,000 tons SO2 and 12,000 tons NOx) and the
Kingston plant in Roane county (2007 emissions:  51,000 tons SO2 and
12,000 tons NOx).  All of Anderson county and a partial county area
around the Kingston plant were designated as part of the Knoxville PM2.5
nonattainment area in 2005.  While the State anticipates installation
and operation of scrubber units at the Bull Run Fossil Plant by the end
of 2008, similar controls at the Kingston Fossil Plant will not be
operational until 2010.  EPA must judge whether current emission levels,
supported by current regulations that provide assurance of continued
operation of existing controls, contribute to existing violations. 
Furthermore, Tennessee has indicated agreement with EPA’s
nonattainment designation for the Greater Knoxville area upon resolution
of all monitoring data flags.  As stated above, EPA has reviewed flagged
data that was collected in 2007 and does not concur with the flagged
data for Knoxville.

Logan, UT-ID

Utah:

Creation of a multi-State nonattainment area for Cache County and
Franklin County will unnecessarily complicate the planning process by
the duplication of agencies on either side of the State border.

EPA Response:

We note that the State initially made their comment as a suggestion in
the October comment letter but subsequently expressed further opposition
during a meeting with EPA personnnel to the combination of portions of
Cache County and Franklin County as one nonattainment area.  Rather than
complicating air quality planning, EPA believes that designating these
two counties as a single nonattainment area will help ensure the
coordination of SIP planning.  There are numerous examples nationwide of
multi-State and multi-jurisdictional nonattainment areas that have been
successfully working together on attainment plans for meeting the
standard for many years.  Where areas in two different jurisdictions
contribute to nonattainment, EPA believes the most effective way to
achieve timely attainment is for the areas to work together to produce
one air quality plan that addresses pollution activities in both
jurisdictions.

Utah:

EPA produced pollution rose diagrams for the Cache Valley using
non-local weather stations.  Wind data from ASOS stations located at the
Salt Lake City International Airport and Pocatello Regional Airport
(both greater than 100 km away from the Cache Valley) were used as
surrogates for wind in the Cache Valley.  EPA did not discuss why local
wind data from the AWOS station at Logan Airport (KLGU) or data from the
DAQ Logan monitor was not used.  The topography that controls the
diurnal variations in wind speed and direction in the Cache Valley is
completely different than that of the topography surrounding Salt Lake
City and Pocatello.  Further, using data from meteorological stations
separated by over 200 km to represent wind conditions for locations
separated by 35 km within the Cache Valley is not technically defendable
(e.g., see the 90 degree difference in wind direction for pollution
roses in Attachment 2, Appendix 1.B).  For these reasons, the pollution
rose diagrams produced by EPA must be removed as a technical
justification for any argument put forth.

EPA Response:

Nationally, EPA used the nearest ASOS station to violating monitors to
represent pollution roses during exceedance episodes at those monitors. 
These pollution roses were created in a national analysis to provide a
consistent approach and data set for use in all areas.  These wind roses
were only one factor EPA considered in making designations, and were
more significant in analyses where they were located close to the
violating monitor.  Specifically in the Cache Valley, the pollution rose
played a relatively small role in decisions relative to the Cache Valley
nonattainment area boundaries.  The Cache Valley is the prototypical
trapped valley: a flat basin nearly surrounded by high mountain ridges. 
Thus, the boundary would be as EPA and Utah recommended independent of
the choice of wind rose.  The only modification to boundaries occurred
in Idaho in EPA’s response to the Idaho boundary proposal presenting
data showing that a smaller area than the entire county contributed to
the violating monitor.

Utah:

The mixing height data for Cache Valley that shows average mixing
heights of about 1000 feet during PM2.5 episodes may not be accurate. 
The State cites alternative data from mountain side temperature sensors
indicating that the mixing height is more complex than EPA described and
may be as low as 300 feet.

EPA Response:

We acknowledge that vertical temperature profiles may vary by
time-of-day and episode and that the mixing depth during episodes in
Cache Valley may typically be lower than those that form in the Wasatch
Front.  We believe the vertical temperature information from Utah State
will be helpful in PM2.5 attainment planning for the Cache Valley. 
However, horizontal transport of emissions between the Cache Valley
portions of Franklin County, Idaho and Cache County, UT will occur
regardless of the mixing depth since there is no topographic feature
that will inhibit air movement between the two areas within the valley. 
Thus mixing height is not a major factor in determining whether a
portion of Franklin County, ID should be included within a single
non-attainment area.

Louisville, KY-IN

The State of Kentucky

Kentucky:

To have Bullitt and Jefferson Counties designated nonattainment would
invoke additional, substantial, unnecessary requirements on local
government planning agencies, especially in Bullitt County where a
thorough review of information shows the County is being impacted by
emissions coming from outside the County.  Drastic emission reductions
are scheduled to occur in the mobile sector throughout the next several
years that will greatly impact pollutant levels in the area.  Couple
these changes with those anticipated by the CAIR, or its replacements,
provisions which will further reduce SOx and NOx emissions within the
region, and the air monitoring data demonstrating attainment of the
PM2.5 standard, and the result should be that these counties be
designated attainment for the PM2.5 standard.

EPA Response:

EPA’s 9-factor analysis has determined that emissions from Bullitt
County contribute to nonattainment in adjacent Jefferson County, and
monitoring data shows Bullitt and Jefferson Counties to be violating the
24-hour PM2.5 NAAQS.  Therefore, it is appropriate to include these
Counties in the Louisville nonattainment area.  It may be true that
emissions from outside Bullitt County are also contributing to the
violations monitored in the county, but section 107 requires designation
of all areas that are currently violating the standard regardless of
where the emissions are coming from.  EPA believes that consideration of
all nine factors compels designating these counties nonattainment,
consistent with the existing nonattainment boundary for the annual PM2.5
standard.  As stated in Section V, EPA determined violations of the 2006
24-hour fine particulate NAAQS based solely on the most recent three
complete years of certified monitoring data.  As required by the Act,
EPA is setting boundaries based on available data and analysis that best
represents present-day conditions, and not on projected emissions
reductions that may occur after designation areas are finalized.  

The State of Kentucky makes the following points as additional rationale
for excluding Bullitt County from the Louisville nonattainment area.

Kentucky:

Kentucky believes that EPA’s use of the CES approach was skewed.  A
review of actual percentages of emissions contributions to an area shows
that Bullitt County does not have the potential to contribute to PM2.5
levels within the region.

EPA Response:

The CES is an initial screening tool used to determine potential
nonattainment area boundaries which can then be modified using
additional information provided by the State or other sources.  EPA has
recommended other types of information for this purpose through guidance
which could be considered in the final decision.  

The CES for Bullitt County is lower than those for other contributing
counties (Jefferson, KY; Floyd, IN; and Clark, IN) in the area, but
higher than that of one other county (Jefferson, IN).  However, EPA does
not find that emissions from this County (857 tons of SO2, 3,140 tons
NOx, and 1,318 tons direct PM2.5) are without the potential to
contribute to PM2.5 in the region.  In addition, the ambient monitor in
Bullitt County shows a violation of the 24-hour PM2.5 standard and thus
Bullitt County must be included within the nonattainment area. 

Kentucky:

Emissions data, population, and commuter data show that the actual
percentage of contribution from Bullitt County itself is exceptionally
low compared to other counties within the region.  This analysis
actually points to Bullitt County monitor being impacted by emissions
from somewhere else within the region.

EPA Response:

EPA agrees that that Bullitt County is likely being impacted by
emissions coming from outside the County as well as those from within
the County.  However, Section 107(d)(1)(A)(i) of the Act specifies that
a nonattainment area shall include “any area that does not meet (or
that contributes to ambient air quality in a nearby area that does not
meet) the national primary or secondary ambient air quality standard for
the pollutant.”  The design value for the Bullitt County monitor,
based on 2005 – 2007 data, is 36 µg/m3, which is in violation of the
24-hour PM2.5 NAAQS.  Thus, Bullitt County must be designated
nonattainment as violating the NAAQS even if the violation is caused in
large part by emissions from outside the county.

Kentucky:

The air monitoring data demonstrates attainment of the PM2.5 Standard
and a downward trend in monitored values.

EPA Response:

After review of all flagged data and exceptional events claims submitted
by the Commonwealth of Kentucky, EPA has determined the 2005-2007
24-hour PM2.5 design value for Bullitt County to be 36 µg/m3, which is
in violation of the NAAQS.  EPA also notes this design value is actually
an upward trend from Bullitt County’s 2004-2006 design value of 34
µg/m3.

The State of Kentucky makes the following points as additional rationale
for excluding Jefferson County from the Louisville nonattainment area.

Kentucky:

EPA should have determined that the correct data were being used for the
VMT assessment.  Also the VMT data should include locally derived
current data which would give a more accurate picture.

EPA Response:

For 2005, a full VMT database at the county, roadway type, and vehicle
type level of detail was developed from Federal Highway Administration
(FHWA) information.  For States and local areas that submitted VMT data
that were incorporated in the 2002 NEI, the 2002 NEI VMT data were grown
to 2005 using growth factors developed from the FHWA data, and these
grown VMT data replaced the baseline FHWA-based VMT data.  The resulting
VMT database prepared for 2005 include data for all 50 States, the
District of Columbia, Puerto Rico, and the Virgin Islands for each of
the 12 Highway Performance Modeling System (HPMS) functional roadway
types and the 28 MOBILE6 vehicle classes, for a total of 336 records per
year per county.  The data were prepared in the NMIM National County
Database BaseYearVMT table format. At this point, States also had
opportunity to submit 2005 estimates that replaced EPA's estimates for
2005.  EPA believes that it used the most accurate data available for
characterizing VMT in the area.

 

Kentucky:

Each monitor in the Louisville area would meet the daily standard if EPA
approves Kentucky’s submitted exceptional events flags.

EPA Response:

EPA’s response to Kentucky’s comments on our review of the
Commonwealth’s Exceptional Events submittal for Louisville is provided
below.

Kentucky:

The air monitoring data demonstrates attainment of the PM2.5 Standard
and a downward trend in monitored values.

EPA Response:

After review of all flagged data and exceptional events claims submitted
by the Commonwealth of Kentucky, EPA has determined the 2005-2007
24-hour PM2.5 design value for Jefferson County to be 39 µg/m3, which
is in violation of the NAAQS.  EPA also notes this design value is
actually an upward trend from Jefferson County’s 2004-2006 design
value of 36 µg/m3.

Review of Kentucky’s Comments on EPA’s Exceptional Events Review for
Louisville

Comments on Step 1 Monthly Average Comparison:

Kentucky:

Annual variations in measured monthly means especially in urban settings
not appropriate in every event.

EPA Response:

EPA believes that typical monthly averages over several years give a
good overall initial picture of the normal concentrations during the
specified time period.  Another analysis method is to center the day in
question over 30 days and then average that across several similar time
periods to get a 'normal' concentration base for comparison.  Other
approaches we used considered the variability within each month. 
Accordingly, we used the upper 84th percentile against which we judged
air quality for the event day under consideration.  The 84th percentile
is approximately one standard deviation from the mean and was chosen as
a convenient way to initially identify concentrations which are higher
than typical values.

Kentucky:

Kentucky believes EPA’s use of 24-hour NAAQS as the upper benchmark
appears to be in conflict with the guidance provided in 72 FR 13570, Use
of a “But For” Test, 2.  Final Rule.

EPA Response:

EPA considered all values which exceeded or violated the annual or
24-hour NAAQS.  In the two-step analysis values that did not exceed the
24-hour NAAQS and for which the site monthly average was at or above the
annual NAAQS did not meet the screening criteria of the analysis. 
However, in our response, it is stated that the two-step analysis is
used only as a screening tool, not as a determinative tool, to identify
which values are more likely to meet the requirements of the Exceptional
Events Rule.  EPA analyzed each claimed exceptional event to determine
whether it caused an exceedance or a violation consistent with the
requirements of the exceptional events rule.

Kentucky:

EPA calculated 95th percentile values that do not agree with
Kentucky’s calculated 95th percentile values.  Explanation of how
EPA’ values were calculated is requested.

EPA Response:

EPA does not specify a method for daily PM2.5 NAAQS calculations, but
for our purposes of evaluating exceptional events submittals the Excel
spread sheet formulas was used.  There are many common approaches for
estimating the percentiles. For the purposes of this analysis, EPA finds
the Excel formula to be appropriate.  Excel uses the following formula:

	vp = vk + d(vk + 1 - vk)

	

	Where:	p is the percentile of the dataset (i.e.  95 or 84)

		vp is the pth percentile of an ordered dataset

		vk is the kth value in the ordered dataset

		N is the number of values in the dataset

		k is the integer component of n

		d is the decimal component of n

		(k+d=n)

Comments on EPA denial on July 4, 2004 Submittal:

Kentucky:

EPA states that the Louisville Metropolitan Air Control District (LMACD)
did not clearly demonstrate a clear causal relationship between the
measured concentration and the event.  

EPA Response:

EPA notes that the impacts from the fireworks did not cause the
exceedance, as demonstrated in the Sonoma Technology analysis (attached
in the TSD), therefore, §50.1(j)(B) and (D) of the exceptional events
rule requiring a clear causal relationship to qualify as an exceptional
event were not met.

Kentucky:

Kentucky believes EPA’s two-step analysis that uses monthly means to
determine if the event meets requirement C under the definition of an
exceptional event is inappropriate for fireworks.

EPA Response:

Although EPA screened the event using the two-step analysis as stated,
each of the firework events were analyzed independently by both EPA
Regional staff, OAQPS and by Sonoma Technology, Inc. In the exceptional
event attachment to the TSD, we explain that the two-step analysis and
the inclusion of the  monthly means  is used only as a screening tool,
and not as a determinative tool, to identify which values are more
likely to meet the requirements of the Exceptional Event Rule.  EPA’s
finding considered its own analyses together with the evidence provided
by the State.

Comments on EPA’s denial of July 3-4, 2005 submittal:

Kentucky:

There is an error in Table 2 for AQS ID 21-111-0051.  The observed
concentration of 28.9 occurred on 7/03/05 not 7/04/05.

EPA Response:

EPA concurs and has corrected this error.

Kentucky:

Kentucky contends that continuous instruments meet requirement B and
provide evidence of “But For” which meets requirement D of the
exceptional events rule.

EPA Response:

The empirical data collected by the continuous instruments without
analysis of the data does not meet the requirements of §50.1(j)(B) and
(D) or of the procedural and substantive requirements of 50.14 of the
Exceptional Events Rule.  EPA concludes that the fireworks did not cause
an exceedance or a violation, as demonstrated in the Sonoma Technology
analysis entitled Exceptional Events Analysis Louisville, Kentucky,
September 30, 2008 (attached to the TSD).

Comments on EPA’s denial of July 3-4, 2006 submittal:

Kentucky:

EPA’s denial appears to be based on monthly means and concentrations
below 35 μg/m3.  EPA concurred with a value of 31.4 μg/m3 for the
Jeffersonville, IN monitor for the same event.

EPA Response:

Although EPA screened the event using the two-step analysis as stated,
each of the firework events were analyzed independently by both EPA
Regional and OAQPS staff and by Sonoma Technology, Inc. For the review
of these analyses, EPA concludes that the fireworks did not cause an
exceedance or violation of the NAAQS. This is explained in the
attachment to the TSD.  

Comments on November 11-12, 2005 Fort Knox Range Fire:

Kentucky:

The observed concentrations for 21-111-0044 on 11/11/2005 should be
28.8.  There was no sample collected for site 21-111-0048 on 11/11/2005.

EPA Response:

EPA concurs and has corrected this error.

Comments on July 21, 2004 Events:

Kentucky:

The data in Table 5 is incorrect.  Incorrect data was used in EPA’s
analysis therefore, EPA should reanalyze the event.

EPA Response:

EPA used the Kentucky Technical Demonstration submittal for conducting
all data analysis for Kentucky’s request for exclusion of flagged data
due to exceptional events.  A review of the evidence provided by
Kentucky and the research done by Regional staff confirms that the
correct data was used during the analysis of the event on July 21, 2004.
(See attachment to TSD)  The error occurred inputting the data into
tables for the TSD.  The tables were copied from one event to the next
and modified to enter the data specific for each event.  As such, the
error was not material and did not affect the analysis of the event. 
The error was a typographical error, has been corrected and does not
affect the outcome of the analysis of the event.   

Kentucky:

The Statement “The NOAA satellite smoke maps show no smoke plume
coverage over the Louisville, KY- IN MSA from the 20th through the 25th
of July 2004” is misleading.  The maps were used to show that a smoke
plume had passed through the area during the event.

EPA Response:

EPA is correcting its TSD to indicate that the maps do show a smoke
plume. Although the smoke map shows that the plume passed over the area,
however, there is no evidence that the smoke left the higher elevations
and reached ground level where it then impacted the air quality.  The
Sonoma analysis supports EPA’s conclusion.  The summary for this event
is in  Sonoma’s Exceptional Event Analysis Louisville, Kentucky, page
13, which states that the “trajectory does not support long-range
transport of PM2.5 from Canada/Alaska to Louisville.”

Comments on Sep 8-13, 2005 Event:

Kentucky:

The event was influenced by wildfires in two separate regions and
therefore the demonstration for smoke impact from Arkansas, Mississippi,
Texas and the Northeastern United States.

EPA Response:

EPA understands that Kentucky meant that the event was influenced by
wildfires from Arkansas, Mississippi, Texas and the North Western United
States as indicated by HYSPLIT Model for September 7-10, 2005. 
Satellite images depict smoke plumes at very high elevations; Kentucky
did not provide any evidence to indicate that the smoke reached ground
level where it would then impact air quality.  The Sonoma analysis
further supports EPA’s conclusion.  The summary for this event is in
Sonoma’s Exceptional Event Analysis Louisville, Kentucky, page 40,
which states that the “trajectories do not show a clear impact from
the fire area on the Louisville areas.”

Kentucky:

The assumption that the contribution of organic carbon has to be high
enough to cause the exceedance ignores the contribution of other
chemical species that may be attributed to the wildfire smoke.

EPA Response:

EPA calculated organic mass increment using the following formula: 

OMI = 2x(OCobserved - OCaverage) as explained on page 14 of the
Louisville TSD.  This calculation uses a multiplier of 2.0 to
approximate the total PM2.5 mass associated with smoke from wildfires
(Turpin and Lim 2001).  Kentucky did not provide any evidence that
non-carbonaceous components resulted from the fire or contributed to an
exceedance.  

Kentucky:

The Heading September 13, 2006 for the chart is incorrect.  It should be
September 13, 2005.

EPA Response:

EPA concurs and has corrected this error.

Comments on July 18-20, 2006 Event: 

Kentucky:

EPA states the HYSPLIT backward trajectory does not indicate the air
mass traveling from the Kansas Wildfires.  Kentucky contends that the
trajectory begins in the area portion of the plume on July 16th.   

EPA Response:

Satellite images depict smoke plumes at very high elevations; Kentucky
did not provide any evidence to indicate that the smoke reached ground
level where it would then impact air quality.  The Sonoma analysis
supports EPA’s conclusion.  The summary for this event indicate that
back- trajectories indicate stagnant conditions, and source impact
trajectories do not show a clear impact from the fire area sites,
including July 16th. 

Kentucky:

The assumption that the contribution of organic carbon has to be high
enough to cause the exceedance ignores the contribution of other
chemical species that may be attributed to the wildfire smoke.

EPA Response:

EPA calculated organic mass increment using the following formula: OMI=
2x(OCobserved - OCaverage) as found on page 14 of the Louisville TSD. 
This calculation uses a multiplier of 2.0 to approximate the total PM2.5
mass associated with smoke from wildfires (Turpin and Lim 2001). 
Kentucky did not provide any evidence that non-carbonaceous components
resulted from the fire or contributed to an exceedance.

Comments on August 25-26, 2006 Event:

Kentucky:

The statement that the NOAA satellite smoke maps show no smoke plume
coverage over Louisville, KY-IN on August 25th and 26th is misleading. 
The maps indicate that the smoke passed through the area.

EPA Response:

Satellite images may depict smoke plumes at very high elevations, but
Kentucky did not provide any evidence to indicate that the smoke reached
ground level where it would then impact air quality.  The Sonoma
analysis supports EPA’s conclusion.   The summary for this event is in
Sonoma’s Exceptional Event Analysis Louisville, Kentucky, page 80,
which states that the “the analysis does not indicate any impact from
the area of the fires on 8/25/06 or 8/26/06.”

Kentucky:

The assumption that the contribution of organic carbon has to be high
enough to cause the exceedance ignores the contribution of other
chemical species that may be attributed to the wildfire smoke.

EPA Response:

EPA calculated organic mass increment using the following formula:

OMI = 2 x (OCobserved - OCaverage) as found on page 14 of the Louisville
TSD.  This calculation uses a multiplier of 2.0 to approximate the total
PM2.5 mass associated with smoke from wildfires (Turpin and Lim 2001). 
Kentucky did not provide any evidence that non-carbonaceous components
resulted from the fire or contributed to an exceedance.  

Comments on June 2, 2007 Event:

Kentucky:

EPA states that this was a stagnation event.  The mild winds from the
direction of the smoke plume are also consistent with the source impact
tool analysis performed by EPA.  Therefore denying concurrence to the
event due to air stagnation event appears contradictory.

EPA Response:

Louisville was experiencing an air stagnation event from May 30 through
June 2, 2007.  It was originally believed that the exceedance was caused
by the lack of air movement and the build up of emissions from local
sources and not by the Georgia/Florida fires.  The existence of the
stagnation event was not the basis for the initial denial of
Kentucky’s request.  The request was initially denied because it was
believed that the exceedance was, in fact, caused by anthropogenic
emissions which were exacerbated by the stagnation event.  

The analysis found in Sonoma’s Exceptional Event Analysis Louisville,
Kentucky, page 92  uses the Kenton and Bullitt County sites as
surrogates for the Louisville sites based on the close physical
proximity to the Louisville monitor sites.  The finding of this analysis
as presented in the TSD is that “there is likely impact from the
Georgia/Florida fires throughout Kentucky on 6/2/07.”  After reviewing
the Sonoma analysis and the potential effect of cloud coverage on the
initial evaluation of the NOAA analyzed smoke maps and the NAAPS models;
EPA now concurs with the request for exclusion of flagged data at the
following sites:

AQS ID	Date	Concentration	Concurrence

21-111-0044	6/2/07	36.8	Yes

21-111-0048	6/2/07	37.2	Yes

21-111-0051	6/2/07	36.3	Yes



Kentucky:

The sections appear to contradict each other.  Section C states Sulfates
are below seasonal averages and Section D states the maps indicate a
regional event of elevated sulfate concentrations and the increased
levels of sulfate negates that “But For” the event there would not
have been an exceedance.

EPA Response:

EPA agrees that a contradiction exist between statements made in Section
C and D.  EPA has revised this analysis and corrected the contradiction.
 Maps in Figures 22 and 23 show low modeled levels of sulfate and
moderate smoke.  The moderate level of smoke, the strong trajectory
analysis as depicted in Figure 21, and the Sonoma Exceptional Event
Analysis, Louisville, KY were all considered and resulted in EPA
revising its analysis.  The revision grants concurrence on three of the
values for this date.  

Comments on August 2-4, 2007 Event:

Kentucky:

NOAA Satellite Fire Detection Maps do not sufficiently establish a
causal.

EPA Response:

Satellite images depict smoke plumes at very high elevations; Kentucky
did not provide any evidence to indicate that the smoke reached ground
level where it would then impact air quality.  The Sonoma analysis
supports EPA’s conclusion.  The summary for this event is in
Sonoma’s Exceptional Event Analysis Louisville, Kentucky, page 105,
which states that the “the analysis does not indicate any impact from
the area of the fires on 8/2/07 - 8/4/07.”

Kentucky:

Reference to wind speeds and calm conditions made by EPA do not match
the local meteorological data provided for August 2, 2007

EPA Response:

Wind speeds were at 8 mph for approximately 45 percent on August 2,
2007.  However, for the majority of the rest of the day speeds were at
or below 4mph.  These data are the basis for EPA’s conclusion that
calm conditions existed.

Kentucky:

The flagged values are clearly above the monthly mean, the 84th
percentiles and the 95th percentiles and therefore are above historical
fluctuations, including background.

EPA Response:

Kentucky did not provide sufficient evidence to demonstrate both a
causal relationship and that “but for” the event there would have
not been an exceedance of the NAAQS.  For an event to qualify as an
exceptional event it must show, among others, “a clear causal
relationship” between the measured exceedance or violation and the
event.   The summary for this event is in Sonoma’s Exceptional Event
Analysis Louisville, Kentucky, page 105 which states that “Trajectory
analysis does not indicate any impact from the area of the fires on
8/2/07 – 8/4007.”  Maps supporting this statement can be found on
pages 98 and 99.  

Kentucky:

The assumption that the contribution of organic carbon has to be high
enough to cause the exceedance ignores the contribution of other
chemical species that may be attributed to the wildfire smoke.

EPA Response:

EPA calculated organic mass increment using the following formula: OMI=
2x(OCobserved - OCaverage) as found on page 14 of the Louisville TSD. 
This calculation uses a multiplier of 2.0 to approximate the total PM2.5
mass associated with smoke from wildfires (Turpin and Lim 2001). 
Kentucky did not provide any evidence that non-carbonaceous components
resulted from the fire or contributed to an exceedance.

Comments on September 6, 2007 Event:

Kentucky:

Absence of a smoke plume over the area is misleading.  Kentucky asked
for additional clarification as to why EPA believes the connection was
not established.

EPA Response:

Satellite images depict smoke plumes at high elevations on September 5,
2007.  Kentucky did not provide any evidence to indicate that the smoke
reached ground level where it would then impact air quality.  The Sonoma
analysis supports EPA’s conclusion.  The summary for this event is in
Sonoma’s Exceptional Event Analysis Louisville, Kentucky, page 118
(attached in the TSD) which further states that the “Trajectories do
not show a clear impact from the fire area on the Louisville area.” 
EPA made its finding on the collection of many factors and the lack of
clear evidence that the smoke plume impacted the ground level monitor.

Kentucky:

Data for September 1, 2007 was below annual standard and annual average.
 In addition data during the event indicated five values over the 95th
percentile.  Therefore, Kentucky contends that data submitted was
sufficient to demonstrate values were in excess of normal historical
fluctuations, including background.

EPA Response:

Kentucky did not provide sufficient evidence to demonstrate both a
causal relationship and that “but for” the event there would have
not been an exceedance of the NAAQS.  The Sonoma analysis supports
EPA’s conclusion.  The summary for this event is in Sonoma’s
Exceptional Event analysis Louisville, Kentucky, page 118 which states
that “Trajectories do not show a clear impact from the fire area on
the Louisville area.” Maps supporting this conclusion can be found on
pages 110 and 111.

The State of Indiana

Indiana:

Clark, Floyd, and Jefferson Counties in Indiana are downwind of
Louisville.  

EPA Response:

Indiana presented wind roses for “elevated PM2.5” days in 2005,
2006, and 2007.  Indiana concluded from this that the winds generally
are southerly on the analyzed days.  The wind roses show more 20 percent
calm days for each year.  EPA has a pollution rose in its Technical
Support Document which displays information on the winds and pollution
levels in the 2005-2007 period.  The pollution rose for Clark County,
Indiana shows the winds coming from all directions, even when higher
PM2.5 concentrations occur.  The pollution rose also shows many of the
high concentrations occurring when the winds are light.  This is
consistent with Indiana’s findings of many calm days.  The
meteorological data EPA used does not show a tendency for the winds to
come from any direction, so EPA feels it is inaccurate to say the
Indiana portion of the Louisville area is downwind from the Kentucky
portion.

Indiana:

EPA needs to reconsider its determination that Townships with
significant power plants should be designated as nonattainment based
upon the cause or contribute rationale.  At a minimum, EPA should
consider the size of the power plant and the emission controls it has
implemented.  For example, the only significant source in Madison
Township in Jefferson County is IKEC’s Clifty Creek power plant.  Five
of the six 217 MW units at this plant have SCRs and scrubbers are
currently under construction for all six units with various star up
dates in 2009.  Unless there is a documented significant contribution
from this plant (or Township) to some nonattainment area, the entire
designation and SIP process will accomplish nothing but employ EPA and
Indiana staff in a paperwork process that will not improve the
environment at all, and in fact, will divert resources from our fully
addressing our actual air quality problem in Marion County.

EPA Response:

See the overview response to comments on power plant issues, located at
the beginning of Section IV.  

EPA has determined that while future emission controls are planned for
the Clifty Creek facility in 2009, the current emissions of the Clifty
Creek plant in Madison Township (more than 60,000 tons SO2 and 20,000
tons NOx annually) contribute to PM2.5 NAAQS violations in the
Louisville area.  Louisville is also an existing PM2.5 nonattainment
area which has not attained the 1997 standards.  EPA finds that it is
appropriate to include the same counties, including Madison Township in
Jefferson county, in the nonattainment area for the 24-hour standard
since air quality data indicate that the same PM2.5 components that are
the biggest contributors to PM2.5 mass on an annual average basis (e.g.,
sulfate, carbon, nitrate) in this area are also key components of PM2.5
mass on the highest days.  The Clifty Creek plant, located about 35
miles from violating monitors in Louisville, is a large emitter of PM2.5
precursors which can readily convert to particle form over this distance
and contribute to nearby violations.   

In this analysis, EPA has considered the size of the power plants,
including any emission controls that have been implemented and are
subject to requirements that will assure that those emission controls
remain in place.  We commend the future installation of control
equipment at Clifty Creek, but, as Indiana notes, much of this control
is not currently in place and cannot be considered in an assessment of
current air quality and current contributions to that air quality.  In
addition, to our knowledge, this facility is subject to no enforceable
requirement that would mandate that these controls or controls at other
power plants in Indiana be operated, particularly if CAIR is vacated.

The Act mandates that any area with a violation must be designated
nonattainment, and the designated area must include all nearby areas
that contribute to the violation.  The designated nonattainment area
serves as a primary planning area, wherein lie the primary contributors
to nonattainment and thus the primary candidates for control in order to
achieve attainment.  Indiana implies that the planning process for a
larger nonattainment area including townships with large power plants
requires more resources than for a smaller area, but Indiana does not
further justify this view.  Indiana states that a SIP-level analysis of
the impact of power plants should be conducted before determining
whether to require the State to expend the resources of a SIP-level
analysis on a nonattainment area that includes the power plants.  EPA
believes that the analytical approach used in the designations process
is adequate for determining that large sources of emissions are
important contributors to nearby violations.  Indiana also considers the
CAA mandate for States to develop plans assuring expeditious attainment
of the PM2.5 standards to require useless effort “in a paperwork
process that will not improve the environment at all.”  EPA disagrees,
and believes that contributing power plants should be included in
nonattainment areas.  Proper evaluation of such sources by the State, in
accordance with the requirements of the CAA as part of the nonattainment
area SIP developed for each area may result in additional controls for
such sources that will help to achieve attainment in the area.  As such
sources are frequently by far the largest individual sources of direct
PM2.5 and PM2.5 precursors in an area, and such emissions play a large
role in the ambient PM2.5 in the area, this evaluation is appropriate
and consistent with the intent behind the designations process in
section 107(d).

Finally, section 107(d) does not provide for designation only of sources
that “significantly” contribute to violations; it specifically
refers to areas that contribute, which is thus presumably less than the
contribution intended by section 110(a)(2)(D). EPA concludes that the
sizeable emissions from the Clifty Creek facility clearly contribute
along with various other emissions in the area to the violating
monitors.

Indiana:

The partial counties with power plants that EPA stated intent to
designate nonattainment are more downwind than upwind of the violating
monitors within the areas U.S. EPA proposed as nonattainment.  The
stationary sources in those townships are heavily controlled and will
continue to be.  Should a mandate for CAIR be issued, Indiana will have
an equivalent control program in place prior to the mandate being
effective.  Since all areas are projected to attain based on SIP-quality
modeling, and Indiana’s control program for EGUs will be equivalent to
RACT, not only are reductions from facilities within the townships that
U.S. EPA proposed as nonattainment not necessary, their being designated
nonattainment will not result in further emission reductions or serve
any purpose.  Counties and townships for which monitoring data does not
exist, and are not proven to be significantly culpable for a downwind
violation, should be considered unclassifiable and designated
attainment.  

EPA Response:

See the overview response to comments on power plant issues, located at
the beginning of Section IV.  

Contribution of a source reflects a combination of emissions level and
frequency of winds blowing from the source to the violating monitor. 
Current emissions are quite high, and the winds blow sufficiently
frequently from the source to the violating monitor for a significant
impact to occur.  The power plants at issue have numerous opportunities
for further control beyond current control levels, and emissions are
expected to remain relatively high even with anticipated controls. 
Indiana anticipates adopting regulations that will require installation
and operation of substantial controls, but EPA must judge whether
current emission levels, supported by current regulations that provide
assurance of continued operation of existing controls, contribute to
existing violations.  If the scenario that Indiana anticipates in fact
occurs, i.e.  if controls are required by an equivalent Indiana rule in
the event the Clean Air Interstate Rule is vacated, and SIP quality
modeling based on enforceable emission levels in Indiana and elsewhere
(with or without the Clean Air Interstate Rule, as the case may be)
demonstrates timely attainment, then the attainment planning requirement
resulting from a nonattainment designation will not be burdensome to
Indiana.  Again, EPA cannot agree to designate areas as attainment based
on potential future controls where the best evidence indicates current
contribution to nonattainment.  As noted above, the Act does not contain
a significance threshold and EPA analysis concluded that these partial
counties are contributing to the violating monitors.

Madison-Baraboo, WI

Wisconsin:

Your August 2008 correspondence relied on a series of assumptions that
concluded Columbia, County contributes to the nonattainment violations
in Dane County, and therefore should be considered as nonattainment.

EPA Response:

The Clean Air Act instructs EPA to include contributing counties in
nonattainment areas.  If only counties monitoring a violation were
designated as nonattainment, that would place all the burden of reducing
emissions to meet the air quality standards on sources in that county or
counties.  Including contributing counties as required by the CAA allows
the emission reductions to come from both sources near the violating
monitor and from nearby sources throughout the area that are
contributing to the violation.  The State thus has the flexibility to
determine the best way to reduce emissions in the area to improve air
quality.  In Columbia County, the Columbia Generating Station may be
subject to future emission controls due to the Mercury Multipollutant
Rule, but its current emissions contribute to the violation in Dane
County and EPA must make designations based on current air quality data.
 EPA is designating Pacific Township in Columbia County as nonattainment
because the power plant generates a significant fraction of Columbia
County’s emissions which contribute to the violating monitor.

Wisconsin:

Dane County, in partnership with the State, City of Madison, local
businesses, utilities, and citizens has also taken numerous actions to
reduce emissions on a voluntary basis.

EPA Response:

EPA commends the improved air quality results from the voluntary
emission reduction programs in the Madison area.  The air quality
improvements may be enough for the Madison area 2006-2008 monitoring
data to show the area meets the standard.  If that is the case,
Wisconsin can early certify and submit this data to EPA following the
2008 data policy.  If appropriate, EPA can designate the counties in the
Madison area as attainment based on this new data demonstrating that the
area is attaining the standard.  The planned emission reduction will
help further improve air quality, but they cannot be used in making the
designation which must be based on current conditions.  The uncertainty
of future emission reductions and new or expanded sources, as well as
the voluntary nature of the reductions, does not allow EPA to make an
accurate determination of the nonattainment area if these assumptions
were used.  Thus, EPA considers the current emission controls in place
for designations as required by the Act.

Milwaukee-Racine, WI

Wisconsin:

EPA’s August 2008 correspondence relied on a series of assumptions
that concluded Racine and Waukesha Counties contribute to the
nonattainment violations in Milwaukee County, and therefore should be
considered as nonattainment.

EPA Response:

The Clean Air Act instructs EPA to include contributing counties in
nonattainment areas.  If only counties monitoring a violation were
designated as nonattainment, that would place all the burden of reducing
emissions to meet the air quality standards on sources in that county or
counties.  Including contributing counties as required by the CAA allows
the emission reductions to come from both sources near the violating
monitor and from sources throughout the area that are contributing to
the violation.  The State thus has the flexibility to determine the best
way to reduce emissions in the area to improve air quality.

Wisconsin requests Racine and Waukesha Counties in the Milwaukee area be
designated attainment.  Waukesha County is included in the nonattainment
area because it has high population, high commuting, and moderate levels
of emissions (12,000 tons NOx per year; 1,000 tons SO2 per year; 2,000
tons direct PM2.5 per year).  Racine County is included in the
nonattainment area because it has high population, high commuting, and
moderate levels of emissions (5,800 tons NOx per year; 700 tons SO2 per
year; 1,200 tons direct PM2.5 per year).  Meteorological data indicates
contributions predominantly from the south and west on high
concentration days, which further supports the inclusion of Waukesha and
Racine Counties in the Milwaukee nonattainment area.

Muscatine, IA

The State of Iowa

Iowa:

Measurements from local and regional monitoring networks indicate that
the violating monitor in Muscatine County is influenced significantly by
local (City of Muscatine) sources and by regional events outside the
State.  Exceedances at the Muscatine monitor occur on days when other
monitors in the area record considerably lower values, suggesting that
the exceedances are due to local sources.  The monitor is located
approximately 500 meters from the Grain Processing facility, suggesting
substantial influence by emissions from the local sources on monitored
values.

EPA Response:

EPA’s detailed analysis of air quality data for Muscatine County is
included in the Technical Support Document (TSD) for the area
designation for Muscatine County.  (The Muscatine County TSD was
prepared by Region 7, and the Scott County, Iowa and Rock Island County,
Illinois TSDs, also referenced in this document, are a joint effort by
Regions 5 and 7. References herein to “TSDs” are to relevant
portions of two or more TSDs as indicated by the context.)  EPA agrees
that on a number of exceedance days the Muscatine monitor is influenced
significantly by “local” events and by regional events outside the
state.  EPA also acknowledges that the monitor is located near the Grain
Processing facility.  However, the Clean Air Act (section
107(d)(1)(A)(i)) instructs the state and EPA to designate, as
nonattainment, areas which violate a National Ambient Air Quality
Standard (NAAQS), and nearby areas which contribute to violations.  Iowa
uses the variations in area monitored values to support its argument
that only a very few sources in the Muscatine area contribute to
violations in any meaningful way.  EPA does not believe this provides
sufficient information (without filter analyses or similar studies)
alone, or in conjunction with other factors in Iowa’s “cumulative
weight-of-evidence” approach, to conclude that some local PM2.5
sources contribute to violations but other nearby sources do not.

We note that the Muscatine (Garfield School) monitor near the Grain
Processing facility is a federal reference method monitor (classified as
population-oriented, neighborhood scale), and is designated under 40 CFR
Part 58 as eligible for comparison with the 24-hour PM2.5 NAAQS.  The
monitor meets EPA siting criteria, and the state has not claimed that
any specific data are invalid or otherwise unrepresentative of air
quality in the area.  Iowa has not provided sufficient information
concerning values recorded by the monitor for EPA to conclude that local
source influence does not include contributions from sources outside the
narrow boundary recommended by the state.  Also, as explained below and
in the Muscatine TSD, there is no speciation data available for the
monitor which would provide further evidence of source contribution.

Iowa:

On exceedance days, an average of 32 percent of local source emissions
is from the Grain Processing and Muscatine Power and Water facilities.

EPA Response:

EPA acknowledges that the AERMOD modeling results predict significant
impacts from these facilities.  However, the state did not attempt to
model impacts from other sources in the area to determine whether other
sources might also contribute.  In addition there is no speciation data
for the Muscatine monitor to indicate source contributions.  Therefore,
EPA does not consider this information sufficient to support the small
boundary recommended by the state.

Iowa:

A zero-out analysis of all Muscatine County emissions outside the City
of Muscatine result in only minimal reductions in PM2.5 concentrations
at the violating monitor on predicted exceedance days.

EPA Response:

The zero-out modeling is instructive but not conclusive with respect to
the appropriate boundaries.  In establishing boundaries for contributing
sources or areas in the Muscatine area, we do not believe it is
appropriate to establish a bright line regarding the amount of reduction
in predicted concentration which an area must be below to show that it
is not contributing to violations of the NAAQS (for example, that a 4%
reduction in a zero-out run is insignificant, but some higher amount is
significant).  Instead, as explained in more detail in the TSDs, in this
response to comments, and in other materials in the record, we are
establishing the nonattainment boundary to encompass sources of almost
all of the nearby PM2.5-related emissions in the area which could, based
on available information, contribute to PM2.5 concentrations at the
violating monitor.  The states will subsequently determine during
development of their implementation plans the specific reductions
needed, and which sources in the nonattainment area they will come from
in order to achieve the NAAQS.  EPA believes that the boundaries defined
in this rulemaking will enable the states to perform these analyses and
develop strategies to achieve the standards.

Iowa:

PSAT simulation is consistent with the zero-out results (discussed
above), showing very small contributions from sulfate and nitrate
emissions sources in Rock Island County, and similar contributions from
Scott and Muscatine counties to particulate sulfate and nitrate
concentrations at the violating monitors.

EPA Response:

EPA agrees that the primary contributing pollutant from local sources in
the area to the violating monitors is direct PM2.5 rather than precursor
emissions.  EPA’s analysis of this information is discussed in more
detail in the TSD for Scott and Muscatine counties.  However, as
explained elsewhere, EPA believes that other nearby sources of PM2.5 in
each area nonetheless contribute, or have the potential to contribute. 
(See also, the discussion of contributing emissions in the TSD for Rock
Island.)

Iowa:

Iowa has re-reviewed the nine factors, in particular wind roses in the
two areas, as well as the growth pattern in Rock Island County (which
the commenter states indicates negative growth for the foreseeable
future), and believes that the data support the smaller boundary
originally recommended by the State.

EPA Response:

EPA has carefully considered all of the information submitted by Iowa
regarding appropriate boundaries.  Wind roses, for example, are
instructive, but not dispositive of the appropriate boundary for an
area.  (For example, as discussed above and in the TSD, for Scott
County, other local point sources in the area contribute to monitored
exceedances.)  EPA’s review of the information, and additional
information made available to EPA after the 120-day letter (and included
in the docket) and has revised its 9-factor analysis based on this
review (see the TSDs for a detailed discussion of the analyses). 
Although this analysis has caused EPA to determine that a different
(partial county) boundary is appropriate, EPA does not agree, for
reasons stated herein and in the TSDs, that the very small boundaries
for each area suggested by the State is appropriate for these areas.	

Nez Perce Tribe

Nez Perce Tribe

The Nez Perce Tribes recommended that EPA classify the Nez Perce
Reservation as in an attainment/unclassifiable area for particulate
matter PM2.5 standards based on three years of FRM monitoring data
collected from 2005-2007.  The tribe provided information related to the
monitoring program, a quick look report from the EPA’s Air Quality
System for the monitor and the computed annual and 24-hour design values
for the monitor.  

EPA Response:

After review of the details of the monitoring program and the computed
design values, EPA is pleased to announce that the Nez Perce Reservation
will be designated as in an attainment/unclassifiable area.  

Oakridge, OR

Oregon:

The Lane Regional Air Protection Agency (LRAPA) submitted comments to
the EPA Docket on the recommended PM2.5 nonattainment boundary for
Oakridge.  ODEQ fully supports LRAPA's recommendation and analysis that
the Oakridge Urban Growth Boundary (UGB), as recommended by Governor
Kulongoski in December 2007, is the most appropriate nonattainment
boundary.  ODEQ requests that LRAPA's comments and technical analysis be
considered in the same manner as a State comment.  ODEQ has worked
closely with LRAPA to address PM2.5 pollution in Oakridge, and we value
this partnership to maintain clean air in Oregon.

EPA Response:

The Lane Regional Air Protection Agency (LRAPA) submitted their comments
to the EPA Docket as public comments, and EPA has responded in full to
LRAPA’s comments in EPA’s Response-to-Public-Comments document.

Paducah-Mayfield, KY-IL

The State of Kentucky

Kentucky:

To have McCracken County designated nonattainment would invoke
additional, substantial, unnecessary requirements on local government
planning agencies with little or no benefit to the area.  Substantial
local emission reductions from McCracken County have already occurred,
or will have occurred well before attainment dates for this standard. 
Drastic emission reductions are scheduled to occur in the mobile sector
throughout the next several years that will greatly impact pollutant
levels in the area.  Couple these changes with those anticipated by the
CAIR, or its replacements, provisions which will further reduce SOx and
NOx emissions within the region, and the air monitoring data
demonstrating attainment of the PM2.5 Standard, and the result should be
that this County be designated attainment for the PM2.5 Standard.

EPA Response:

EPA commends McCracken County for local emission reductions that have
been and continue to be achieved.  However, monitoring data shows
McCracken County to be violating the 24-hour PM2.5 NAAQS.  EPA therefore
finds that it is appropriate to include this County in the
Paducah-Mayfield nonattainment area.  EPA believes that consideration of
all nine factors compels designating this County nonattainment.  As
stated in Section V above, EPA determined violations of the 2006 24-hour
fine particulate NAAQS based solely on the most recent three complete
years of certified monitoring data.  EPA is setting boundaries based on
available data and analysis that best represents present-day conditions,
and not on projected emissions reductions that may occur after
designation areas are finalized.  

The State of Kentucky makes the following points as rationale for
excluding McCracken County from the Paducah-Mayfield nonattainment area.

Kentucky:

The monitor located in McCracken County would meet the daily standard if
EPA approves Kentucky’s submitted exceptional events dates.

EPA Response:

EPA’s response to Kentucky’s comments on our review of the
Commonwealth’s Exceptional Events submittal for Paducah-Mayfield is
provided below.

Kentucky:

The population of McCracken County has shown a continuing decline over
the past several years and that decline is anticipated to continue.

EPA Response:

EPA understands the population of McCracken County has shown a
continuing decline over the past several years, however, the County
currently makes up 47 percent of the population in the Paducah-Mayfield
CSA counties that EPA stated intentions to designate as nonattainment in
our August 2008 letters to State Governors.  McCracken County also has a
significantly higher population density then the other four counties in
this area.  EPA’s finds this factor adds support to EPA’s decision
to include McCracken County in the Paducah-Mayfield nonattainment area.

Kentucky:

The air monitoring data demonstrates attainment of the PM2.5 Standard
and a downward trend in monitored values.

EPA Response:

After review of all flagged data and exceptional events claims submitted
by the Commonwealth of Kentucky, EPA has determined the 2005-2007
24-hour PM2.5 design value for McCracken County to be 36 µg/m3, which
is in violation of the NAAQS.  EPA also notes this design value is
actually an upward trend from McCracken County’s 2004-2006 design
value of 33 µg/m3.

Review of Kentucky’s Comments on EPA’s Exceptional Events Review for
Paducah

Comments June 21-24, 2005, Events:

Kentucky:

The flagged values are clearly above the monthly mean, the 84th
percentiles and the 95th percentiles and therefore are above historical
fluctuations, including background.  In support of an exceptional event
determination EPA Region 4 declares that “the requested values passed
both steps” of the two-step analysis.

EPA Response:

The two-step analysis was used as an aid, not a determinative tool, to
help decide which values were more likely to meet the requirements of
the Exceptional Event Rule.  In the demonstrations submitted to EPA,
Kentucky did not provide sufficient evidence to demonstrate both a
causal relationship and that but for the event there would not have been
an exceedance of the NAAQS.  The Sonoma analysis supports EPA’s
conclusion.  The summary for this event is in Sonoma’s Exceptional
Event Analysis, Kentucky, page 10, which states that the Trajectories
show impact from the NE but not from fire area and indicates that the
exceedance was more likely impacted by a regional sulfate event as
opposed to a smoke event.”  This is also discussed in the TSD.

Kentucky:

Kentucky contends that NOAA “Hot spot” maps, wind roses and
historical data is sufficient evidence to establish a causal
relationship.  

EPA Response:

The Sonoma analysis supports EPA’s conclusion.  The summary for this
event found in Sonoma’s Exceptional Event Analysis Kentucky, page 9
states that “trajectory analysis shows impact from the NE, not from
fires.” This is attached to the TSD.  Furthermore, Kentucky did not
provide evidence expected from local events such as news media reports
or reports from the local Fire Marshal.

Kentucky:

Kentucky contends that organic carbon is 1.3 times higher than
background level measurements, thereby confirming Kentucky’s
justification for exclusion under the Exceptional Event Rule.

EPA Response:

Kentucky has requested an exclusion of data under the Exceptional Event
Rule for a wildfire event.  Sulfate matter accounted for greater than
fifty percent of the mass of the speciated fine particulate measured on
June 21, 2005.  Therefore, the request to exclude data impacted by the
wildfire event identified on June 21st and June 24th in Kentucky’s
exceptional event submittal was denied.  The Sonoma analysis supports
EPA’s conclusion.  The summary for this event, found in Sonoma’s
Exceptional Event Analysis Kentucky, page 9, states that “High sulfate
and low OC values (with respect to sulfates) indicate a sulfate event
rather than a fire event.”  The average three-year standard deviation
organic carbon in the month of June at all of the monitoring sites in
Kentucky is 1.6 and the average concentration is approximately 4 μg/m3,
therefore, the organic carbon values 1.3 times higher than background
measurements are not especially high and are well within expected
values.

Kentucky:

Kentucky contends the following:

“the understanding of atmospheric chemistry does not allow for
generalizations to estimate contributions based o species of PM2.5; that
precursor gases react with air pollutants to form a variety of
compounds; and also that the burning of biomass results in elevated
sulfate concentrations.”

EPA Response:

EPA disagrees with Kentucky’s first statement.  There are known major
components of smoke from biomass burning.  Through the well established
speciation network, there are obvious and evident markers that show
exceptional contributions of wildfire or biomass smoke in an area. 
While the statement regarding precursor gases reactions is an obvious
statement, EPA is not sure it is completely relevant to PM2.5 speciated
compounds.  Kentucky did not provide supportive discussion or include
evidence from available speciation measurements where there were
atypical concentrations of non-carbonaceous constituents.

EPA is not aware of research that definitively links significant amounts
of sulfate production 

from biomass burning.  We referenced Turpin and Lim (2001) paper, which
is the basis for our OMI calculations used in the TSD and provide the
research paper “Gaseous and Particulate Emissions from Prescribed
Burning in Georgia" (Lee et al.  2005) that identifies the PM2.5 mass
from fires as composed of 60 percent organic carbon and 0.2 percent
sulfate which further supports our use of the OMI method.  Kentucky did
not include any evidence that shows atypical levels of sulfates
resulting from the event, particularly distinguishing areas impacted by
the smoke compared to other locations in the region.

Comments September 10, 2005, Event:

Kentucky:

The flagged values are clearly above the monthly mean, the 84th
percentiles and the 95th percentiles and therefore are above historical
fluctuations, including background.  In support of an exceptional event
determination EPA Region 4 declares that “the requested values passed
both steps” of the two-step analysis

EPA Response:

The two-step analysis is used as an aid, not as a determinative tool, 
to help decide which values were more likely to meet the requirements of
the Exceptional Event Rule.  Kentucky did not provide sufficient
evidence to demonstrate both a causal relationship and that but for the
event there would have not been an exceedance of the NAAQS.   

 

Kentucky:

Kentucky does not concur with EPA’s conclusion that request for
concurrence “solely based on wind speed and wind direction”

EPA Response:

EPA notes that Kentucky provided NOAA satellite images and historical
data.  However, the Commonwealth's discussion of the September 10, 2005
event consists solely of the statement that “The wind direction and
wind speed indicate that an air mass held the smoke plume over Kentucky
for 3 days.” This statement alone does not meet the requirements for
an event to qualify as an exceptional event.  Kentucky did not provide
any evidence that indicated that the smoke reached ground level where it
would then impact air quality or other evidence such as newspaper
articles documenting local haze or discussions in support of the impact
from this event.  

The summary for the Sonoma analysis for this event is  in the
Exceptional Event Analysis Louisville, Kentucky, page 47 which states
that “Sulfate concentrations are very high throughout the region
during 9/8-9/13/05, indicating a regional sulfate event.”  It also
states that “meteorological analysis shows a surface high centered
over Kentucky that is likely causing the sulfate event.”  The
conclusion is that the high PM2.5 is not impacted from forest fires.

Kentucky:

EPA provides the seasonal averages for sulfates and organic carbon on
September 10, 2005 when no speciation samples were collected on that
day.

EPA Response:

The VIEWS maps, both for sulfate and organic carbon, are created using
the measured concentrations collected from the speciation monitors and
entered into the Air Quality Systems (AQS) data base.  The concentration
gradients shown on the maps are spatially averaged between the monitors
that collected samples on any given day.  Empirical data points were not
available for September 10th as speciation samples were not collected by
Kentucky.  However, as stated, the concentration is calculated from the
spatial averaging of speciation data collected by other samplers
throughout the country.  If you refer back to the TSD the spatially
averaged sulfate concentration for September 10th, 2005 was between 14
and 16 μg/m3 and organic carbon was between 2 and 4 μg/m3,
respectively (see Exceptional Event Analysis Louisville , Kentucky,
Exceptional Event, 9/8/05-9/13/05, page 30, maps 9/10/05)

Kentucky:

EPA provided a map with the same particulate matter trace but identifies
the trace as SO4.  

EPA Response:

As noted the sulfate trace for the VIEWS SO4f map generated is the same
as that of the smoke plume.  The VIEWS maps, both for sulfate and
organic carbon, are created using the measured concentrations collected
from the speciation monitors and entered into the AQS database.  The
concentration gradients shown on the maps are spatially averaged between
the monitors that collected samples on any given day.  The difference
between the smoke analyzed map and the views map as discussed in the TSD
is that sulfate is based on measured concentrations and is therefore at
ground level and the estimated smoke is at unknown elevations.

Kentucky:

Kentucky contends the following: the understanding of atmospheric
chemistry  does not allow for generalizations to estimate contributions
based on species of PM2.5; that precursor gases react with air
pollutants to form a variety of compounds; and also that the burning of
biomass results in elevated sulfate concentrations.

EPA Response:

EPA disagrees with Kentucky’s first statement.  There are known major
components of smoke from biomass burning.  Through the well established
speciation network, there are obvious and evident markers that show
exceptional contributions of wildfire or biomass smoke in an area. 
While the statement regarding precursor gases reactions is an obvious
statement, we're not sure it is completely relevant to PM2.5 speciated
compounds.  Kentucky did not provide supportive discussion or include
evidence from available speciation measurements showing atypical
concentrations of non-carbonaceous constituents.

EPA is not aware of research that definitively links significant amounts
of sulfate production from biomass burning.  We referenced Turpin and
Lim (2001) paper, which is the basis for our OMI calculations used in
the TSD and provide the research paper “Gaseous and Particulate
Emissions from Prescribed Burning in Georgia" (Lee et al.  2005) that
identifies the PM2.5 mass from fires as composed of 60 percent organic
carbon and 0.2 percent sulfate which further supports our use of the OMI
method.  Kentucky did not include any available empirical evidence that
shows atypical levels of sulfates resulting from the event, particularly
distinguishing areas impacted by the smoke compared to other locations
in the region.

Comments July 19, 2005, Event:

Kentucky:

The measured values flagged are clearly above the monthly mean, the 84th
percentiles and the 95th percentiles and therefore are above historical
fluctuations, including background.  In support of an exceptional event
determination EPA declares that “the requested values passed both
steps” of the two-step analysis

EPA Response:

The two-step analysis is used as an aid, not as a determinative tool, to
help decide which values were more likely to meet the requirements of
the Exceptional Event Rule.  Kentucky did not provide sufficient
evidence to demonstrate both a causal relationship and that “but
for” the event there would have not been an exceedance of the NAAQS. 
In the Sonoma Analysis the summary for this event is in the Exceptional
Event Analysis, Kentucky, page 20 which states that “PM2.5
concentration of 36.7 μg/m3 is within the 95th percentile of summer
concentrations for 2006 and just above the 95th percentile of summer
concentrations for 2000-2006.”  EPA agrees that the concentrations
were higher than the typical concentrations including background.
However, this is only one of the requirements specified by the
exceptional events rule.  As explained in the TSD, the State did not
demonstrate that the exceedance or violation would not have occurred but
for the event.

Kentucky:

The data analysis along with evidence from an independent organization,
NOAA, and wind speed/wind direction graphs provide evidence of the
causal relationship between the measured exceedance and the exceptional
event.

EPA Response:

Satellite images depicting the location of local fires and wind roses
providing wind speed and winds directions traveling predominantly from
the direction of local wildfires may show that there was air quality
impact, but not by itself demonstrate that the event caused the
exceedance or NAAQS violations. In addition, such local events should
have news media reports or reports from the local Fire Marshal but
Kentucky failed to submit any such reports.  Satellite images depict
smoke plumes at high elevations that may not impact the ambient air
monitoring network samples.  Kentucky did not provide sufficient
evidence that the smoke from forest fires caused the exceedance or
violation  The Sonoma analysis (attached in the TSD) cites a lack of
strong trajectories; high regional sulfate concentrations, and
meteorological conditions that were conducive to high secondary PM
formation.

Kentucky:

The Paducah site was discontinued in March of 2006.  EPA has not
demonstrated that a sulfate event occurred in the Southeast on July 19,
2006.  The NOAA analyzed smoke map trace is the same as EPA VIEWS OCf
map with the same particulates matter trace but identifies the trace as
organic carbon .

EPA Response:

As noted the organic carbon trace for the VIEWS OCf map generated is the
same as that of the smoke plume.  The VIEWS maps, both for sulfate and
organic carbon, are created using the measured concentrations collected
from the speciation monitors and entered into the Air Quality Systems
(AQS) data base.  The concentration gradients shown on the maps for July
19 are spatially averaged between the monitors that collected samples on
any given day.  These maps indicate a sulfate event rather than a smoke
event. 

Kentucky:

Kentucky contends that the high sulfate concentrations do not preclude
defining the event as an exceptional event.  TVA Shawnee electrical
utility company reported zero hours of noncompliance during the third
quarter.  Therefore, the exceptional event meets the definition as
defined in 40 CFR 50.1(j).

EPA Response:

First, sulfate concentrations can result from routine emissions of SO2
from nearby and regional sources.  EPA is not aware of research that
definitively links significant amounts of sulfate production from
biomass burning.  We referenced Turpin and Lim (2001) paper, and the
research paper “Gaseous and Particulate Emissions from Prescribed
Burning in Georgia" (Lee et al. 2005) that identifies the PM2.5 mass
from fires as composed of 60% organic carbon and 0.2% sulfate which
further supports our stand.  Kentucky did not provide supportive
discussion or include evidence from available speciation measurements
showing atypical concentrations of non-carbonaceous constituents.  As
stated previously, organic carbon was low on that day indicating a
non-smoke event.

The State of Illinois

Illinois:

The largest emission source in Massac County is the Joppa Steam Plant. 
The Joppa

Steam Plant Consists of six boilers, each rated at 183 MW.  Currently,
NOx control is achieved by the use of low NOx burners, with overfire air
also being applied at boiler 6.  Currently, sulfur emissions at the
Joppa plant are reduced by use of low sulfur coal in all boilers. 
Particulate matter emissions are controlled by cold-side ESPs.  Illinois
EPA urges EPA to consider that Ameren, the owner of the Joppa Steam
Plant, has committed to significantly reduce both its NOx and SO2
emissions through the installation of new control equipment, which will
be operational before the likely attainment date of the 24-hour PM2.5
standard.  All the low NOx burners will be supplemented with separated
overfire air (SOFA) in 2010 to increase combustion efficiency.  In
addition, boilers 1 and 2 and boilers 5 and 6, respectively, will get
new scrubbers and baghouses in 2013 and 2014.  A copy of Ameren's
commitment to install controls at the Joppa plant is included as
Attachment 6 of the Illinois comment letter.

EPA Response:

See the overview response to comments on power plant issues, located at
the beginning of Section IV. 

EPA has determined that while future scrubbers and baghouses are planned
for the Joppa Steam Plant in 2013 and 2014, the current emissions of the
plant (more than 25,000 tons SO2 and 5000 tons NOx annually) and
meteorological data indicate that it currently contributes to PM2.5
NAAQS violations in the nearby Paducah area.

The emissions from the Joppa Steam Plant are significant. The planned
emission controls for the Joppa plant have the potential to greatly
reduce the plant’s emission. The air quality in the Paducah area will
benefit from reductions at the Joppa plant. However, EPA designates
areas based on current information regarding emissions and air quality
concentrations. Projections of future emissions reductions or increases
are not reliable because plans can change. It would be misleading to
designate area as attainment due to anticipated emission reductions or
to designate an area as nonattainment based on projected new or
expanding sources. The Joppa plant accounts for a majority of Massac
County emissions, so in Illinois, EPA is designating just Hillerman
Precinct of Massac County as part of the nonattainment area. This
partial county area along with McCracken County, Kentucky, will make up
the Paducah nonattainment area.

Parkersburg-Marietta, WV-OH

West Virginia:

West Virginia agrees with the data and analysis presented in Factor 1
(Emissions data), Factor 2 (Air quality data), and Factor 3 (Population
density and degree of urbanization) of EPA’s technical analysis.  West
Virginia also agrees that, since there are no topographic or geographic
features in the air shed, Factor 7 (Geography/topography) is not a
significant factor within the overall analysis.  However, West Virginia
contends that the data presented in EPA’s analysis supports exclusion
of Pleasants County, including Grant Tax District, from the
Parkersburg-Marietta nonattainment area.   

EPA Response:

EPA’s recommendation to include the Grant Tax District within the
Parkersburg-Marietta nonattainment area was based upon our review of
available data and implementation of EPA policy and guidance.  EPA
reviewed all available technical data relevant to each of these factors
cited by West Virginia.  EPA’s recommendation, to include the Grant
Tax District as part of the nonattainment area, is supported by EPA’s
analysis in its entirety and not solely upon individual analysis of
select factors therein.  Pleasants County contains two large coal-fired
power plants (Pleasants and Willow Island) and EPA’s analysis supports
a finding that emissions from these plants contribute to PM2.5 formation
in the nonattainment area.  Recognition that Grant Tax District is a
lower ranking candidate based upon certain factors does not preclude EPA
from determining on the basis of all factors that the Grant Tax District
“contributes to” nonattainment in the Parkersburg-Marietta
nonattainment area.           

West Virginia:

Relating to Factor 4 (Traffic and commuting patterns) of EPA’s
technical analysis, West Virginia agrees with EPA’s conclusion that,
in comparison to Washington and Wood Counties, Pleasants and Wirth
Counties are lower ranking candidates for nonattainment designation on
the basis of traffic and commuting data.  West Virginia submitted
supplemental VMT growth projections (2002-2025) from the
Wood-Washington-Wirt Interstate Planning Commission for Wood and
Pleasants Counties to demonstrate the that the Grant Tax District of
Pleasants County has a smaller overall VMT contribution to the overall
nonattainment area than that of the entire county that EPA analyzed.   

EPA Response:

Consistent with the CAA, EPA’s designations reflect EPA’s analysis
of data reflecting current conditions, as opposed to predictions of
future conditions.  For purpose of consistency, EPA relied upon 2005 VMT
for our analysis of commuting patterns within all proposed nonattainment
areas.  In general, EPA’s analysis used the most recently available
historic data, rather than long-term VMT projections.  EPA does not
believe that the use of predictive travel demand modeling would be
consistent with the CAA because such modeling does not reflect current
conditions.  Rather, predictive travel demand modeling is a tool for
predicting conditions that may occur after designation determinations
are finalized.  For the partial county areas, we assessed the factors
based upon county level information (with the exception of the
contribution of large power plants).   

West Virginia’s supplemental 2002 VMT information compares Grant Tax
District VMT data to Wood County VMT data.  EPA’s analysis indicates
that Pleasants County has low VMT relative to the other counties in the
area.  West Virginia’s supplemental VMT information does not conflict
with EPA’s analysis relating to traffic and commuting patterns.  Nor
does this VMT information support an overall conclusion that “Grant
Tax District  .  .  .  emissions should not only be insignificant in
2002, but also will be even lower in 2005.”  Recognition that Grant
Tax District is a lower ranking candidate based upon traffic and
commuting patterns does not preclude EPA from determining on the basis
of all factors that the Grant Tax District “contributes to”
nonattainment in the Parkersburg-Marietta nonattainment area.

 

West Virginia:

West Virginia states that EPA reached no definitive conclusion for
Factor 5 (Growth rates and patterns) of its technical analysis for the
Parkersburg-Marietta area.  West Virginia argues that future VMT
increases (based on its transportation conformity projections between
2002 and 2025) for the area are not expected to result in growth in
highway motor vehicle emissions.  West Virginia contends that conformity
projections provided by the Wood-Washington-Wirt (WWW) Planning
Commission demonstrate decreased NOx and direct PM2.5 emissions in
future conformity analysis years extend to 2025, and as a result, that
highway emissions growth (in the WWW study area counties) is an
insignificant factor for the Parkersburg area.       

EPA Response:

EPA used historical county-level population and VMT growth estimates for
all areas of the country.  EPA did not take into consideration projected
long-term trends in VMT growth beyond the PM2.5 designation timeframes,
such as future year transportation conformity analyses.  EPA used the
same methodology for estimating historic VMT growth in all nonattainment
areas analyzed.

West Virginia:

With respect to Factor 6 (Meteorology-weather transport patterns) of
EPA’s technical analysis, West Virginia provided supplemental
information in order to support its contentions that emissions from
Grant Tax District do not contribute to violations monitored at the Wood
County Ambient Monitor and that Pleasants County, including Graham Tax
District, should not be included in the nonattainment designation.  West
Virginia used NOAA Air Resource Laboratory’s HYSPLIT trajectory model
(using 100, 250, and 500 meter trajectory heights) to provide a limited
analysis of air patterns during each of the 19 days during which
monitoring data within the Wood County Ambient Monitor exceeded 30
μg/m3.  Using this model, WVDEP calculated backward trajectories to
track pollutants from the site of the monitor.  West Virginia analyzed
these 19 backward trajectories to determine which trajectories “passed
over” Grant Tax District.  Based upon the data provided, West Virginia
observed that “none of the 19 trajectories on high PM2.5 days pass
through the Grant Tax District (Pleasants County partial).” 
Therefore, West Virginia concludes that emissions from Pleasants County
do not contribute to the Parkersburg nonattainment area and should not
be included in the nonattainment designation.  

West Virginia also commented upon the fact that EPA’s Technical
Analysis did not provide a pollution trajectory plot depicting weather
patterns in the Parkersburg-Marietta area.  West Virginia concludes
that, due to the absence of such plot, EPA has inadequate justification
based on weather patterns to conclude that Pleasants County (partial)
should be included in the nonattainment designation for the
Parkersburg-Marietta area.

EPA Response: 

EPA does not believe that West Virginia’s analysis supports a finding
that “emissions from Pleasants County (partial) do not contribute to
the Parkersburg-Marietta nonattainment area and should not be included
in the nonattainment designation.”  

As noted above in EPA’s response to West Virginia’s comments
pertaining to the Huntington-Ashland nonattainment area, West
Virginia’s backward trajectory analysis presents a “direct causal
relationship” test that inappropriately limits the geographic scope of
“ambient air quality in a nearby area” to the location of an air
quality monitor.    The analysis presented is limited in that it seeks
to show the impact of emissions from Grant Tax District only on the Wood
County Ambient Monitor.  Section 107(d)(1)(A)(i) provides that
nonattainment areas must include “any area that does not meet (or that
contributes to ambient air quality in a nearby area that does not
meet)” the applicable NAAQS.  42 U.S.C.  § 7407(d)(1)(A)(i).   The
definition of “nonattainment” specifically includes any area that
“contributes to” violations of NAAQS in a nearby area, not only
those that literally “cause” such violations at specific monitoring
locations.  Ambient PM2.5 at every monitor reflects the cumulative
impacts of many types of emissions from many sources, near and far, that
result in primary and secondary formation of particles.  

μg/m3.  West Virginia’s analysis focuses on back trajectories from
the monitoring site, and, therefore, does not adequately demonstrate
that emissions from the Grant Tax District does not “contribute to”
ambient air quality in the Parkersburg nonattainment area.  West
Virginia’s limited analysis fails to take into account that both of
the coal-fired power plants located in the Grant Tax District are
significant sources of direct PM2.5 and precursor (NOx and SO2)
emissions.  A more complete and relevant analysis would compute multiple
forward HYSPLIT trajectories from the coal-fired power plants in
question to determine if emissions from those facilities could have
“passed through” any portion of the nonattainment area at issue. 
Recognizing that NOAA’s HYSPLIT Model is a multifaceted tool, EPA
believes a more complete analysis would support its contention that
emissions from the Grant Tax District “contribute to” ambient air
quality in the Parkersburg nonattainment area.     

 

Although EPA’s Technical Analysis for the Parkersburg-Marietta area
did not provide a pollution trajectory plot depicting weather patterns,
EPA reasonably developed an analytical approach to evaluate the
meteorological influence that counties within in the
Parkersburg-Marietta area have on violations of the 24-hour PM2.5 NAAQS
in Wood County.  EPA provided an analysis of NWS wind direction and wind
speed data for 2004-2006, including warm season trajectory factor values
which represent the meteorological influence that a particular
county’s emissions have on the violating county.  EPA explained that
for the Parkersburg-Marietta area it calculated warm season trajectory
factor values because the data indicates that all of the high PM2.5 days
occur in the warm season.  EPA explained that a higher trajectory factor
value means that air masses -frequently traversed the county on their
way to the violating county.  EPA believes that the analysis presented
weighs in favor of a finding that emissions from Pleasants County have a
meteorological influence on Wood County.  The absence of a pollution
trajectory plot depicting the data does not negate the meteorological
analysis presented by EPA.

West Virginia:

West Virginia contends that, based upon presence of emission controls
and the prospects for future emission controls for the two coal-fired
power plants in the Grant Tax District, it is unnecessary to include
Grant Tax District in the Parkersburg-Marietta nonattainment area.  West
Virginia contends that because these units are equipped with controls
that meet the definition of RACT, there is no air quality benefit to be
gained by designating the Grant Tax District as nonattainment.  The
State also contends that EPA has provided States flexibility to address
SOx and NOx emissions from nearby sources (within 200 km) via its PM2.5
Implementation Rule.

EPA Response:

EPA determined violations of the 2006 24-hour fine particulate NAAQS
based solely on the most recent three complete years of certified
monitoring data.  EPA set boundaries based on available data and
analysis that best represents present-day conditions, and not on
projected emissions reductions that may occur after designation areas
are finalized.  EPA recognized that due to new controls at large
electric generating units (EGUs), there may be emission reductions of
SO2 and NOx subsequent to 2005 that are not accounted for in its August
2008 analysis and EPA gave States the opportunity to submit supplemental
information on emissions controls to aid in making final designation
decisions.  

Washington County, OH and Pleasants County, WV are the only counties in
the Parkersburg-Marietta area with EGUs which, during 2006, had SO2 plus
NOx emissions greater than 5000 tons.  For the two coal-fired power
plants in Pleasants County (the Pleasants and Willow Island power
stations), West Virginia agrees with EPA’s summary data of 2002-2007
NOx and SOx annual emissions and heat input for these two power plants,
as well as the level of controls expected to be in place on these
facilities in the near term.  Furthermore, for each of these power
stations, West Virginia agrees with EPA that no new controls to reduce
NOx or SO2 have been (or will soon be) installed since 2005.  

The Pleasants Power Station’s SO2 emission rate decreased slightly
between 2005 and 2007, while the NOx emission rate increased slightly
over the same period.  Heat input levels have decreased since 2005;
these decreases have resulted in slightly lower SO2 and NOx emissions
from this facility.  The Willow Island facility’s SO2 and NOx emission
rates decreased between 2005 and 2007, while heat input levels increased
since 2005, resulting in slightly lower SO2 and NOx emissions from the
facility.  However, EPA reiterates that these reductions did not stem
from emissions controls added since the 2005 emissions inventory that
was used by EPA in considering the emissions contribution of Pleasants
County to the nonattainment area under Factor 1 of our analysis.

EPA reviewed West Virginia’s comments and determined that the State
has presented no new additional information with respect to emissions
controls for these power plants.  EPA maintains that, in light of
current emissions controls, consideration of the total emissions of NOx
and SOx from both facilities, along with other data, support a finding
that emissions from the Grant Tax District contribute to ambient air
quality in the Parkersburg-Marietta nonattainment area.  

West Virginia:

The portion of Pleasants County, the Grant Tax District, proposed to be
included in the

Parkersburg nonattainment area encompasses the Pleasants and Willow
Island power plants, both owned by Allegheny Energy.  Pleasants has two
units rated at 650 MW each, and both are equipped with an ESP for
particulate control, an SCR for NOx control and a wet limestone scrubber
for SO2 control.  The particulate and SO2 controls are federally
enforceable through regulations and permits.  The year round operation
of the SCR is required by State Consent Order, which was included in the
Parkersburg PM2.5 SIP revision (submitted to EPA September 9, 2008). 
Willow Island has two units, Unit 1 rated at 54 MW and Unit 2 rated at
181 MW.  Both units are equipped with an ESP for particulate control and
overfire air (OFA) for NOx control, which are federally enforceable
through regulation and permits.

In addition, in December of 2007 Pleasants began scrubbing 100 percent
of the flue gas, permanently eliminating the 15 percent bypass.  This
results in a 95 percent SO2 removal rate for the additional 15 percent
of the flue gas that is now scrubbed, for a total removal efficiency of
95 percent.  Historical data shows that between 2002 and 2007 the SO2
and NOx emission rates at Pleasants and Willow Island decreased, while
heat input remained steady at Pleasants and decreased slightly at Willow
Island.  The Pleasants SO2 emission rate decrease by 5.4 percent, and
the NOx emission rate decreased by 34 percent.  The Willow Island SO2
emission rate decreased by 48.9 percent, and the NOx emission rate
decreased by 29.7 percent.  In addition, DEP has conducted a RACT
analysis (which was included in the Parkersburg PM2.5 SIP revision
submitted to EPA on September 9, 2008) to determine the appropriate
level of controls for EGUs.  It was determined that SCR and wet
scrubbers are not economically feasible for units the size of those at
Willow Island.

DEP believes that these units are equipped with controls that meet the
definition of

RACT and there is no air quality benefit to be gained by designating the
Grant Tax District as nonattainment.  Furthermore, EPA in the PM2.5
Implementation Rule stated that for SO2 and NOx “EPA believes that
States could justify considering not only all emissions in the
nonattainment area but also emissions within a distance that may be up
to 200 kilometers from the nonattainment area” [72 FR 20636,
25APR2007].  Since EPA has provided States the flexibility to consider
emissions from sources within 200 kilometers of the nonattainment area,
it is not necessary for EPA to include the partial counties within the
nonattainment areas. (see table on page 25 of WV comment letter)

EPA Response:

See the overview response to comments on power plant issues, located at
the beginning of Section IV.  

EPA disagrees with the reasons provided by West Virginia for exclusion
of the Pleasants County Pleasants and Willow Island power plants from
the Parkersburg nonattainment area.  EPA’s analysis has evaluated
pollutant emissions; evidence from the pollution rose for this area that
there is contribution from the direction of these sources to the rest of
the nonattainment area; and the CES score that likewise indicates that
emissions from the pertinent sources contribute to violations in the
Parkersburg area.  

The 2007 emissions from the Pleasants County power plants were about
42,000 tons SO2 and 11,000 tons of NOx.  The plants are located 14-23
miles from the closest violating monitor in the Parkersburg area, a
distance at which significant formation of PM2.5 from precursors emitted
by these plants can occur.  The technical support document provides
pollution roses and other meteorological data for the Parkersburg area,
providing evidence that the emissions from these plants can be expected
on some high concentration days to have impacts on the violating
monitors.  EPA finds that current emissions data and meteorological
information indicate that in accordance with section 107(d), the partial
county area of Pleasants County is contributing to violations in the
Parkersburg area.  It should be emphasized that the designations process
is intended only to determine those areas which are contributing to a
violation of the standard.  Section 107(d) was not intended to also
require EPA and the States to determine whether emission controls are or
are not economically feasible for a particular source.  The evaluation
of technically and economically feasible emission controls is required
as part of the SIP development process.

The State’s reference to the requirements of the PM2.5 implementation
rule are at most an analogy, as that rule applies to the 1997 PM2.5
NAAQS.  Even by analogy, the opportunity to take emissions reductions at
sources up to 200km from the nonattainment area is premised upon the
state being able to demonstrate that those emission reductions would
have an impact in the nonattainment area.  Moreover, the fact that
emission reductions from such sources could be shown to make an
improvement in the nonattainment area is not the sole test for
designations under section 107(d).  Under this provision, EPA must
ascertain which sources are “nearby” for purposes of the designation
for the nonattainment area, given the facts and circumstances of each
area.  Other tools under the CAA, such as section 110(a)(2)(D) and
section 126 are designed to deal with interstate emission impacts,
especially regional impacts.

Pascua Yaqui Tribe

Pascua Yaqui:

The Pascua Yaqui Tribe cannot accept the decision regarding the Section
107(d)(1) Clean Air Act Designation of attainment/unclassifiable for the
2006 24-hour PM 2.5 health standard.  This decision represents an
arbitrary measure that has no basis in science.  Pima County air testing
stations northeast of the Pascua Yaqui Reservation would be incapable of
presenting a representative air sample.  The Pima County air testing
stations are all placed east and north of the Pascua Yaqui Reservation. 
They are placed strategically to sample the air for the population
center of Tucson, Arizona, but they are ineffectively placed for the
Pascua Yaqui Reservation and the San Xavier District of the Tohono
O’odham Nation.  

According to the Western Regional Climate Center, the prevailing winds
in Arizona from 1992- 2002 indicate a prevailing southeast wind.  Any
Pima County air testing stations northeast and on the other side of the
Tucson Mountains would be incapable of providing an air sample
representative of the Pascua Yaqui Reservation.  The southwest side of
Tucson, Arizona occupies a bowl formed by the Tucson Mountains on the
north and the high rock mountains with passes for major road on the
east.  The populations of the communities of Drexel Heights, Tucson
Estates, Valencia West, and Three Point have all increased substantially
in recent years.  Increased populations equate to substantially more air
pollution opportunities.

EPA Response:

EPA requires State and local agencies to have a representative
monitoring network for the criteria pollutants of concern.  These
monitoring networks measure pollution in the air and that data is
recorded in the EPA Air Quality System, or AQS.  A monitoring network
usually has several monitors placed throughout an area and EPA accepts
the monitoring data from the network as being representative of the
entire area.  The Pima Department of Environmental Quality (PDEQ) has an
approved monitoring network and EPA views the data from that network as
a reliable indicator of the air quality in the Tucson area.  The data
indicates that the Tucson area is attaining the PM 2.5 annual standard. 
Almost the entire State of Arizona, excluding Nogales, meets the PM 2.5
standard.  Since there are no monitors on the reservation and there are
no monitors within the area surrounding the reservation that are showing
violations, there is no scientific evidence that there is a PM 2.5
problem on the Pasqua Yaqui Reservation.  Absent monitoring data that
shows a violation, we cannot make a designation other than
“attainment/unclassifiable”.

Philadelphia-Wilmington, PA-NJ-DE

By letter dated October 15, 2008, Delaware submitted additional
information to EPA to support its recommendation that EPA establish New
Castle County as a separate single-county nonattainment area and not
include New Castle County as part of the Philadelphia-Wilmington,
PA-NJ-DE nonattainment area for the 2006 PM2.5 NAAQS.  This letter also
requested a technical meeting with EPA, Region 3 representatives.  On
November 20, 2008, EPA, Region 3 met with representatives of
Delaware’s Department of Natural Resources and Environmental Control
(DNREC).  During this meeting, DNREC representatives submitted
supplemental information, which were in the form of two power point
presentations.  The first presentation, entitled “New Castle County,
Delaware PM2.5 Nonattainment Area Boundaries,” was presented by Ali
Mirzakhalili, Administrator, DNREC Air Quality Management Section.  The
second presentation, entitled “CALGRID Modeling for Assessing
Delaware’s Emissions Impact on Downwind Counties in the PM2.5
Nonattainment Area,” was presented by Mohammed A.  Majeed of DNREC. 
The following reflects EPA’s response to each of the above
submissions.

Delaware:

Delaware contends that “1) The boundaries of New Castle County
encompass the full area that is violating the standard, based on actual
monitoring data, 2) Delaware emissions do not significantly impact any
part of the Philadelphia combined statistical area (CSA), based on EPA
modeling, 3) emissions from a broad area encompassing the States of
Ohio, Indiana, Pennsylvania, Michigan, West Virginia, New York,
Maryland, District of Columbia, New Jersey, North Carolina, and Illinois
impact Delaware (i.e., an area much broader than the Philadelphia CSA
boundaries), based on EPA modeling, and 4) while New Castle County, and
the Philadelphia CSA, and all other areas in the region share a
transport problem, there is no evidence that supports the linking of
transport mitigation to small CSA scale areas.”

EPA Response:

While the Philadelphia area and the northeastern United States are
impacted by regional transport, based on the types of information
recommended by EPA in its guidance including nine factors, and any other
relevant information, as explained in detail below, EPA has determined
that emissions from New Castle County contribute to nonattainment
downwind in the Philadelphia area, and it is appropriate to include New
Castle County in the Philadelphia-Wilmington nonattainment area.

Factor 1:  Emissions data 

Delaware:

Delaware stated that while New Castle County emissions were high
relative to other counties in the CSA, federally enforceable
requirements will reduce emissions between 2009 and 2012.  Delaware
believes that EPA’s use of 2005 National Emissions Inventory (NEI) and
Contributing Emissions Score (CES) are not reflective of its impact on
the Philadelphia CSA.  Delaware has already required significant
reductions in PM2.5 and PM2.5 precursor emissions such that New Castle
County emissions will be reduced 1) to a level lower than other areas of
the CSA and 2) by a percentage that is significantly more than any other
county in the CSA.  Delaware also indicated that these reductions are
required at least two years before the attainment date.

EPA Response:

EPA determined violations of the 2006 24-hour fine particulate NAAQS
based solely on the most recent three complete years of certified
monitoring data.  EPA set boundaries based on available data and
analysis that best represents present-day conditions, and not on
projected emissions reductions that may occur after designation areas
are finalized.

States may account for other projected emissions reductions in their
State Implementation Plans for achieving attainment, where appropriate. 
States will also be able to take advantage of any air quality benefits
that may have already occurred from recently enacted rules by submitting
the 2008 monitoring data to EPA by February 20, 2009.  EPA will review
2008 monitoring data submitted by this date, and will modify designation
areas if appropriate.

Furthermore, Delaware’s own data indicates that even with these
emission reductions in place in 2012, New Castle County still has the
highest SO2 emissions in the Philadelphia CSA, and the second highest
total emissions.   It is notable that sulfate emissions dominate on days
with the highest fine particle concentrations in the Philadelphia area. 


Delaware:

Delaware objects to EPA’s use of the 2005 NEI as the basis for Factor
1 (Emissions data) and believes EPA should have used Delaware’s 2009
and 2012 projection inventories.

EPA Response:

In this designation process, nationwide EPA used the 2005 NEI Version 1
as the starting point for all its technical evaluations for factor 1
(Emissions Data).  The NEI is the national database of air pollutant
emissions.  The NEI is a comprehensive inventory that is created from
many sources, including State/Local and tribal emissions data.  The NEI
can and has been used by EPA for modeling analyses, projecting future
year control strategies, tracking progress to meet requirements of the
CAA, calculating risk, and responding to public comment.  In accordance
with the CAA, EPA believes that the 2005 NEI provides emissions data
relevant to its determination as to which areas are currently violating
the PM2.5 24-hour NAAQS or are currently contributing to violations of
the PM2.5 24-hour NAAQS.  With respect to emission controls, as
explained and for the reasons stated above, EPA is only considering
emission controls in place and federally enforceable at the by December
2008.  Therefore, in accordance with the CAA, EPA has determined that it
is not appropriate to consider projections of future emissions such as
Delaware’s 2009 and 2012 projection inventories.

Delaware:

Delaware commented that EPA Clean Air Interstate Rule (CAIR) modeling
has already demonstrated that the PM2.5 transport problem is a regional
problem, and explicitly demonstrates that Delaware does not
significantly contribute to the Philadelphia CSA problem.  The CES
indicates that New Castle County impacts the Philadelphia area more than
any other county, and the EPA CAIR modeling indicates that the entire
State of Delaware does not contribute significantly to any part of the
Philadelphia CSA (i.e., they reach opposite conclusions).

EPA Response:

EPA does not agree with Delaware’s assertion that the CAIR modeling
has demonstrated that Delaware does not contribute to exceedances of the
PM 2.5 24-hour NAAQS in the Philadelphia CSA.  CAIR modeling was
performed to assist States in determining which areas significantly
contributed to exceedances of the Ozone 8-hour and PM 2.5 annual NAAQS. 
The PM 2.5 concentrations resulting from the CAIR modeling are annual
average concentrations.  In addition, section 107(d) requires EPA to
include areas that are “contributing” to violations in a nearby
area, and EPA interprets this to be a different test than that of
section 110(a)(2)(D) which refers to “significant” contribution. 
Moreover, section 107(d) refers to current facts, unlike the analysis
for CAIR which examined contribution in 2010. Therefore, it is not
appropriate to use CAIR modeling results to determine whether Delaware
is contributing to exceedances of the PM 2.5 24-hour NAAQS in the
Philadelphia-Wilmington Area.  

Factor 2:  Air quality data 

Delaware:

Delaware stated that in its December 12, 2007 recommendation letter, an
analysis of data from the four New Castle County air monitors showed
that the City of Wilmington’s PM2.5 concentrations at the Martin
Luther King Boulevard (MLK) monitor range from 4 to 5 micrograms per
cubic meter (µg/m3) higher than the other monitors located northeast,
west and south of the MLK monitor, and that the relatively high downtown
concentrations drop off quickly to below the NAAQS within the boundaries
of New Castle County.  Delaware stated that these other monitors are
placed in areas which represent most of the compass, thereby
“encircling” the MLK monitor with “clean” ones.  Delaware added
that the Bellefonte monitor, which is four miles downwind from the MLK
monitor, recorded 11 percent lower concentrations than the MLK monitor. 
In addition, Delaware stated that it obtained preliminary 2008 air
quality monitoring data from the Pennsylvania Department of
Environmental Protection (PADEP) for the downwind Chester monitor in
Delaware County, and that this monitor will be back into attainment
using the preliminary 2006-2008 data.

Delaware commented that the fact that the design values for the
Pennsylvania monitors near New Castle County increased from 2004-2006 to
2005-2007 while the New Castle County monitors’ design values remained
the same is evidence that Delaware emission contributions and emission
transport in general, are not responsible for those higher Pennsylvania
values.  Delaware added that, otherwise, emissions from Delaware or
other upwind States would have caused Delaware monitored levels to rise
as well.  Delaware concluded that this confirms its assertion that the
PM2.5 nonattainment problem is highly localized in the CSA and that
contributions from Delaware are not significant.

EPA Response:

EPA does not agree with Delaware’s assertion that evaluation of the
24-hour PM2.5 design values for air quality monitors in the
Philadelphia-Wilmington area, based upon 2005-2007 data, shows that the
“nonattainment problem” in New Castle County is limited to the
boundaries of New Castle County.  EPA agrees that there is a local
component to the PM2.5 nonattainment problem in downtown Wilmington. 
However, as explained in more detail below and in the TSD for the
Philadelphia-Wilmington area, EPA has determined that emissions from New
Castle County contribute to violations of the 2006 24-hour PM2.5
standard in the Philadelphia-Wilmington area.

Considering the most recent three years of data, 2005-2007, monitors in
Chester and Delaware Counties in Pennsylvania, which border New Castle
County, are violating the standard.  Furthermore, other monitors
downwind of New Castle County in Camden and Philadelphia Counties also
violate the standard.  As explained above, with respect to 2008 air
quality monitoring data at the Chester monitor in Delaware County, EPA
determined violations of the 2006 24-hour fine particulate NAAQS based
solely on the most recent three complete years of certified monitoring
data.  EPA set boundaries based on available data and analysis that best
represents present-day conditions, and not on projected emissions
reductions that may occur after designation areas are finalized.  

The Delaware County design value did go up slightly, from 35 µg/m3
using 2004-2006 data to 36 µg/m3 using 2005-2007 data.  The Chester
County 2005-2007 design value is 37 µg/m3.  No 2004-2006 design value
could be calculated for the Chester County monitor due to incomplete
data.  Therefore, it is not possible to say whether monitored values
have risen there.  However, the New Castle County design value has been
consistently high, at 37 µg/m3.

Delaware:

In its October 15, 2008 comment letter, Delaware states that,
“Delaware has shown that the air quality is bad in downtown
Wilmington, the air quality is bad in downtown Philadelphia, and the
monitors between the two demonstrate that these two areas are separate
and distinct.  EPA’s analysis did not once reference the
continually-attainment Bellefonte monitoring data in its evaluation,
which was clearly discussed in Delaware’s recommendation.  The
Bellefonte monitor is downwind of the Delaware’s only non-attaining
monitor, and is upwind of Philadelphia counties.  It lies close to major
point sources (Refinery and Conectiv EGU5), and also is located between
those sources and Philadelphia Counties.  Its design value is 33; and
this holds true for 2004-2006 and 2005-2007 data.  Based on an analysis
of this factor, the area of non-attainment is clearly limited to the
city of Wilmington, which is completely contained within New Castle
County.”

EPA Response:

EPA recognizes that the monitors between downtown Philadelphia and
downtown Wilmington are not as impacted by high levels of localized
emissions as those monitors located in the downtown Wilmington and
downtown Philadelphia areas.  That does not mean that downtown
Philadelphia and downtown Wilmington do not impact the air quality in
downwind areas on a daily basis.  Evaluation of the 24-hour PM2.5 design
values for air quality monitors in the Philadelphia-Wilmington area
alone cannot support a finding that New Castle County does not
contribute to exceedances of the PM2.5 24-hour NAAQS in the Pennsylvania
and New Jersey portions of the Philadelphia-Wilmington area.  Based on
other relevant information and data and EPA’s best scientific analysis
thereof as explained in further detail below and in the TSD for the
Philadelphia-Wilmington area, New Castle County should be included in
the Philadelphia-Wilmington area 2006 24-hour PM2.5 nonattainment area.

 

Delaware:

In 2004, Dr.  Phil Hopke performed receptor modeling on the MLK monitor.
 Based on the results of this modeling, Delaware concludes that
long-range transport of sulfate and local mobile sources of PM2.5
dominate the mix of PM at the MLK monitor.  Delaware further concludes
that local mobile source emissions near the MLK monitor, including
traffic on MLK Boulevard and Interstate I-95, a large bus depot, and the
CSX/Norfolk Southern Railroad, are one of the primary causes of the
nonattainment problem.  

EPA Response:

The modeling described by Delaware is limited in scope and does not
evaluate how sources of PM2.5 in New Castle County may be impacting
other portions of the Philadelphia CSA.  Receptor models are
mathematical or statistical procedures for identifying and quantifying
the potential sources of air pollutants at a receptor location.  Unlike
photochemical and dispersion air quality models, receptor models do not
use pollutant emissions, meteorological data and chemical transformation
mechanisms to estimate the contribution of sources to receptor
concentrations.  Instead, receptor models use the chemical and physical
characteristics of gases and particles measured at source and receptor
to both identify the presence of and to quantify potential source
contributions to receptor concentrations.  As stated above, receptor
models are not designed to quantify the impacts of specific sources on a
given location.  Therefore, the results of the Hopke study give very
little insight to the extent of the impact PM2.5 emissions from New
Castle County may be having in other areas.  

Delaware:

On November 20, 2008, DNREC presented maps to EPA depicting 2004-2006
and 2005-2007 24-hour PM2.5 design value gradients in northern New
Castle County, southeast Pennsylvania and portions of southwest New
Jersey.  These maps appear to show concentration declines between the
Martin Luther King Boulevard monitor in downtown Wilmington (the MLK
monitor) in New Castle County and other nearby monitors in the
Philadelphia area.  DNREC concludes that air quality problems in New
Castle County are limited to Wilmington, are completely contained within
the boundaries of New Castle County, and are separate from the problems
of the rest of the Philadelphia area.  

EPA Response:

PM2.5 design values presented by DNREC are “unpaired in time.” 
These design values represent the three year average of the 98th
percentile 24-hour PM2.5 concentration, which could occur on different
days at different monitors.  An analysis of actual PM2.5 concentrations
on days when concentrations are relatively high may show little or no
difference in concentrations between New Castle County monitors and
monitors in southeastern Pennsylvania and portions of southwestern New
Jersey.

Factor 3:  Population density and degree of urbanization (including
commercial development) 

Delaware:.  

Delaware asserts that population and population density are not
indicators of air quality;  many areas with high population densities
are not exceeding the PM2.5 24-hour NAAQS.  Delaware objects to EPA’s
using population data as an indicator of population-based emissions
(i.e.  area sources) that might contribute to nonattainment, including
downwind nonattainment, because emissions are evaluated in Factor 1, and
it is not appropriate to evaluate emissions under both factors.  Also,
Delaware stated that area sources make up only six percent of overall
PM2.5 and the primary precursor emissions (SO2 and NOx) in New Castle
County.  However, Delaware comments that if the EPA is to consider
population density in drawing nonattainment boundaries, then the
population density of New Castle County is in line with the lower
density counties in the CSA, and is 8 ½ times less than the highest
county.  Therefore, analysis of population data does not provide
justification to include New Castle County in a Philadelphia based
nonattainment area.  Based upon these observations, Delaware believes
this criterion for boundary considerations should be of low priority.  

EPA Response:

Population and population density were one of nine factors that EPA used
nation-wide to determine the appropriate boundaries for PM nonattainment
area.  EPA continues to believe that population, population density, and
degree of urbanization are appropriate surrogate measures of the
relative level of contribution between counties under consideration for
a nonattainment designation.  Population data gives an indication of
whether it is likely that population-based emissions might contribute to
violations of the 24-hour PM2.5 NAAQS.

Of the nine counties in the Philadelphia-Wilmington nonattainment area
for the 1997 PM2.5 NAAQS, New Castle County has the fifth highest
population and the fifth highest population density.  EPA believes that
population density data does not distinguish New Castle County from any
other county in the Philadelphia-Wilmington nonattainment area for the
1997 PM2.5 NAAQS, each of which EPA has determined should be included in
the Philadelphia-Wilmington nonattainment area for the 2006 PM2.5 NAAQS.
 Rather, EPA continues to believe that New Castle County population data
supports a finding that New Castle County, as a densely populated area
located immediately south of Philadelphia, PA, likely contributes to
violations of the 2006 PM2.5 NAAQS within the Pennsylvania and Delaware
areas of the Philadelphia-Wilmington area.  EPA continues to believe
that an area with a large and dense population, such as New Castle
County, more likely contributes to violations than a rural area with a
very low and widely dispersed population.  Population density is but one
factor considered in determining whether to include an area in a
designated nonattainment boundary, and is not usually outcome
determinative.  Instead, the entire suite of technical and factual
considerations and our analysis thereof informs our decisions whether an
area should be included.

Factor 4: Traffic and commuting patterns

Delaware:

Delaware commented that in its December 12, 2007 recommendation letter,
Delaware demonstrated that less than one percent of the commuters in the
Philadelphia-Wilmington area are from New Castle County.  Delaware
reiterated that many of these commuters are likely to use public
transportation, that the Southeastern Pennsylvania Transportation
Authority (SEPTA) serves commuters from the Bucks, Chester, Delaware,
Montgomery, Philadelphia, and New Castle Counties, and that SEPTA brings
many of them into Center City Philadelphia.  

EPA Response:

Over 237,000 New Castle County commuters work within the Philadelphia
statistical area, including New Castle County.  Of the nine counties in
the Philadelphia-Wilmington nonattainment area for the 1997 PM2.5 NAAQS,
New Castle County has the fifth highest number of commuters into and
within the statistical area.  As stated in EPA’s August 18, 2008
letter, as a general matter, it is likely that commuters from most
counties in the Philadelphia-Wilmington nonattainment area rely heavily
on public transportation.  However, EPA also recognized that currently
available data does not clearly indicate the percentage of commuters
from New Castle County to Pennsylvania which commute via SEPTA or other
public transportation versus non-public transportation (such as private
automobiles).  The percentage of commuters from New Castle County that
rely on public transportation for purposes of commuting within the
Philadelphia statistical area is not clearly documented. Therefore,
there is currently insufficient data to quantify the percentage of
commuters from New Castle County using public transportation to commute
within the Philadelphia statistical area.  EPA considered available VMT
data as one indicator of emissions that might contribute to
nonattainment, including downwind nonattainment.  The currently
available VMT data does not support a finding that New Castle County, DE
should be a separate single-county nonattainment area; rather, EPA
believes that such data continues to support a finding that it is
appropriate for New Castle County, DE to be included as part of the
Philadelphia-Wilmington, PA-NJ-DE nonattainment area for the 2006 PM2.5
NAAQS.

Delaware:

Delaware stated that while the number of New Castle County commuters to
any violating county is 228,630, this figure includes New Castle
residents going to work within New Castle County.  This large number of
New Castle County commuters within New Castle County, and the small
number of New Castle County commuters to Philadelphia, indicate this
factor supports New Castle County as being separate and not part of a
Philadelphia CSA based nonattainment area.  Additionally, Delaware
stated that it was not appropriate for EPA to present information on a
relationship between emissions and vehicle miles traveled (VMT) as part
of its analysis of this factor, because all emissions, including mobile
source emissions, are included under factor 1 above.

EPA Response:

Mobile source emissions constitute a large portion of the emissions
inventory in urban areas.  Thus, mobile source use in an area, as
reflected in information like the number of drivers and the VMT in an
area, is relevant to determining whether an area contributes to
violations in nearby areas.  Of the nine counties in the
Philadelphia-Wilmington nonattainment area for the 1997 PM2.5 NAAQS, New
Castle County has the second highest number of residents traveling into
any violating county, with Philadelphia County having the highest.  As
with most counties in the Philadelphia area, the majority of New Castle
County commuters commute within New Castle County, and the majority of
Philadelphia County commuters commute within Philadelphia County.  As
explained above, one important indicator of contribution of a county is
the emissions within that county.  Commonly, these data strongly
indicated contribution of relevant counties, and information regarding
the location of mobile sources and their emissions had little influence
on EPA’s judgment.  Nevertheless, since some mobile source control
programs are applied based on the home station of the vehicle, one
factor in judging the contribution of a county is the number of vehicles
that are housed in the county.  Mobile sources are important
contributors to PM2.5 concentrations, and population data can be an
important indicator of the home station of these sources.

Furthermore, a county with numerous commuters and high VMT is generally
an integral part of an urban area and is likely contributing to fine
particle concentrations in the area.  Of the nine counties in the
Philadelphia-Wilmington nonattainment area for the 1997 PM2.5 NAAQS, New
Castle County has the third highest VMT.  EPA believes that the data
does not distinguish New Castle County from any other county in the
Philadelphia-Wilmington nonattainment area for the 1997 PM2.5 NAAQS, all
of which EPA has determined should be included in the
Philadelphia-Wilmington nonattainment area for the 2006 PM2.5 NAAQS.

Factor 5:  Growth rates and patterns

Delaware:

Delaware commented that in its December 12, 2007 recommendation letter,
it compared New Castle County’s population and VMT growth rates to
counties such as Gloucester, which are monitoring attainment, but
EPA’s August 2008 response does not appear to evaluate the “growth
rate and pattern” factor at all, but rather concentrates on total
population.  Delaware stated that this factor, as evaluated in its
December 12, 2007 recommendation letter, supports New Castle County as
being different from the other counties in the Philadelphia area.

EPA Response:

Of the nine counties in the Philadelphia-Wilmington nonattainment area
for the 1997 PM2.5 NAAQS, New Castle County has the fourth highest
population growth rate and the sixth highest VMT growth rate.  EPA
believes that the data does not distinguish New Castle County from any
other county in the Philadelphia-Wilmington nonattainment area for the
1997 PM2.5 NAAQS, all of which EPA has determined should be included in
the Philadelphia-Wilmington nonattainment area for the 2006 PM2.5 NAAQS.

Factor 6:  Meteorology (weather/transport patterns).  

Delaware:

Delaware commented that the “wind direction from New Castle County on
high PM days does not mean sources in New Castle County are the only
contributors (or even minor contributors) to downwind Counties.” 
Delaware provided National Oceanic and Atmospheric Administration (NOAA)
HYSPLIT Model 24-hour backward trajectories for the Philadelphia County,
Broad Street and Lycoming monitors for the ten-highest PM days for 2004,
2005 and 2006.  Each of these ten highest-PM days was also high
concentration days in New Castle County.  Delaware stated that its
trajectories do not correspond well with EPA’s pollution roses. 
Delaware commented that on these high-PM days, its HYSPLIT back
trajectories do not point to Delaware, and therefore, Delaware does not
“understand how EPA can conclude that on high PM days New Castle
County is a major contributor.”

EPA Response:

In accordance with the CAA, EPA need not make a determination as to
whether, or not, “sources in New Castle County are the only
contributors (or even minor contributors) to downwind Counties.”  EPA
recognizes that ambient PM2.5 at every monitor reflects the cumulative
impacts of many types of emissions from many sources, near and far, that
result in primary and secondary formation of particles.  EPA believes
the backward trajectory information provided by Delaware is not
dispositive as to whether or not New Castle County “contributes to”
violations of NAAQS in a nearby area EPA’s proposed determination. 
Additionally, EPA believes that the proposed analysis is limited in
scope and therefore, is not complete.  Delaware used HYSPLIT to
calculate backward trajectories for the ten highest PM days in the years
2004-2006 at the Philadelphia County, Broad Street and Lycoming
monitors.  Delaware’s analysis is not directly relevant because it
seeks to determine whether emission sources outside New Castle County
directly caused violations of the 2006 PM2.5 NAAQS at certain monitors
on certain days.  The CAA definition of “nonattainment” specifically
includes any area that “contributes to” violations of NAAQS in a
nearby area, not only those that literally “cause” such violations
at specific monitoring locations.  Delaware’s analyses are not
complete in part because the analysis produced only one trajectory for
each day at each monitor.  A more robust analysis should have included
multiple trajectories for each day to account for the entire duration of
the twenty-four hour PM2.5 standard.  Furthermore, a significant number
(seven out of 16) of the back trajectories included in the report pass
trough New Castle County and directly impact the Philadelphia area. 
Therefore, EPA does not believe that Delaware’s analysis supports a
determination that New Castle County should be separated from the
Philadelphia nonattainment area for the 2006 PM2.5 NAAQS.

Delaware:

DNREC used the California Photochemical Grid (CALGRID) model to assess
the effects of New Castle County’s emissions on other monitors in the
Philadelphia PM2.5 nonattainment area.  The CALGRID modeling presented
by Delaware on November 20, 2008 included zero-out and worse-case
analyses.  The zero-out analysis is meant to show the difference between
the base case (with no anthropogenic emissions) and emission control
strategy runs.  The worst-case impact analysis is meant to show New
Castle County’s emissions impacts on downwind nonattainment areas
counties.  DNREC contends that the modeling results indicated minimal
impacts from direct PM2.5 emissions and several other PM2.5 precursors
from New Castle County to the rest of the Philadelphia area.

EPA Response:

In general, it is difficult to consider this modeling relevant for
making an assessment about the contribution from any county in Delaware
to the Philadelphia nonattainment monitor(s).  There is no supporting
evidence that the July 6 to 23, 2002 and July 30 to August 16, 2002
episodes that were modeled fully represent the highest observed PM2.5 in
Philadelphia during the 2005-2008 time period.  It is not clear that
July and August 2002 even represent high PM2.5 in Philadelphia.  No
model performance for CALGRID at the Philadelphia monitor (or any
monitor locations) is presented, making it impossible to determine when
the model is performing adequately for PM2.5 sulfate, nitrate, and other
primary species.  The CALGRID model is not typically used for PM2.5 SIP
submittals.  Therefore, documentation regarding model chemistry and
physics is needed to assess its comparability to other commonly used
modeling systems.  In general, a modeling system applied for less than
two months in the same season is not considered an adequate
representation of assessing the multitude of different meteorological
conditions that can lead to high 24-hour PM2.5 in the Northeastern
United States.

The CALGRID modeling presented by DNREC is not appropriate for
determining 2006 24-hour PM2.5 nonattainment area boundaries for the
following reasons:  

CALGRID is designed as a tool to “Screen emission control strategies
for urban and regional attainment” (See Jeff Underhill’s
presentation for the OTC/MANE-VU Annual Meeting Philadelphia, PA, July
21-23, 2003, CALGRID Modeling Overview, A First Look, at
http://www.4cleanair.org/OTCClearSkiesmodelling.pdf.)  For absolute
PM2.5 predictions and response to emissions changes, results from
CALGRID were not shown to be comparable to models such as the Community
Multiscale Air Quality (CMAQ) model and the CAMx, which are commonly
used for regulatory purposes.

There is not any supporting information for the CALGRIDv2.45 model that
shows that model chemistry and physics related to PM2.5 (like aqueous
phase chemistry, inorganic chemistry, deposition processes, etc) are
comparable to state of the science one atmosphere regulatory models like
CMAQ and CAMx.  

There is no model performance given for monitor(s) in Philadelphia so it
is impossible to determine if the modeling system is performing
adequately for PM2.5 mass.

The two episodes modeled with CALGRID represent warm-season simulations
where regional sulfate loading is probably responsible for most of the
elevated PM2.5 concentrations in the Philadelphia PM2.5 nonattainment
area.  “Local” transport may be more important for cool-season
episodes and thus any impact of Delaware/New Castle County sources may
be underestimated using Delaware DNREC’s CALGRID runs.  Consideration
should be given to high winter time PM2.5 episodes in Philadelphia.  If
there are none then that should be stated and supported.

The modeling episode selected for this analysis does not have any
supporting information that indicates this is a typical high PM2.5
episode in the Philadelphia area.  Multiple episodes that capture the
variety of high PM2.5 formation regimes are necessary when making a
determination about contribution.  

Delaware:

DNREC contends that long-range pollutant transport to the Philadelphia
area overwhelms the contribution from New Castle County.  To illustrate
this point, DNREC provided a figure which showed PM2.5 measurements at
22 FRM monitors in the Mid-Atlantic Regional Air Management Association
(MARAMA) region every day from April through June 2003.  DNREC states
that PM2.5 concentrations often rise and fall sympathetically across
much of the MARAMA Region.  At the same time, DNREC contends that its
monitors do not correlate well with other monitors within the current
Philadelphia PM2.5 nonattainment area.  This contention was used to
support DNREC’s argument that New Castle County is generally
“isolated” from the rest of the Philadelphia PM2.5 nonattainment
area.

EPA Response:

DNREC did not provide the statistical analysis supporting this
contention.  It would seem unlikely that there would be significant
differences between monitors that are relatively close to one another. 
The monitoring trends from across the MARAMA region presented by DNREC
appear to indicate a good correlation across a large area of the
Mid-Atlantic region, which is counter to their assertion that New Castle
County is “isolated” from the rest of the region.

Delaware:

DNREC contends that CAIR modeling and its CALGRID modeling show that the
entire State of Delaware’s emissions do not contribute significantly
to any violation of the PM2.5 NAAQS in any part of the CSA.

EPA Response:

As stated above, the CALGRID modeling presented by DNREC is not
appropriate for determining 2006 24-hour PM2.5 nonattainment area
boundaries.  While EPA appreciates Delaware’s efforts to complete
modeling to support its PM2.5 boundary recommendation, CALGRID is not
shown to be a state-of-the-science one-atmosphere model.  In addition,
no model performance is presented, and the episodes selected do not
appear to reflect high PM2.5 in Philadelphia.  These issues make it
difficult to consider this modeling as being representative of
Delaware's (New Castle County) contribution to high PM2.5 in
Philadelphia.  Furthermore, EPA disagrees with Delaware’s assertion
that the CAIR modeling has demonstrated that Delaware does not
significantly contribute to exceedances of the PM 2.5 24-hour NAAQS in
the Philadelphia CSA.

Factor 8:  Jurisdictional boundaries (e.g., existing PM areas)

Delaware:

While Delaware, Pennsylvania, and New Jersey do work cooperatively
together, no cooperative air planning effort to date has occurred as a
result of being in a common nonattainment area.  These States have
historically worked together only as part of larger efforts, like
Mid-Atlantic/Northeast Visibility Union (MANE-VU) and the Ozone
Transport Commission (OTC).  No control measures have been developed as
a cooperative effort amongst these States outside of MANE-VU and OTC
context.

EPA Response:

Delaware has conceded that it has worked with the other States in the
Philadelphia area as part of MANE-VU and the OTC.  Therefore, as stated
in its August 18, 2008 letter, EPA does not anticipate that including
New Castle County as part of the Philadelphia nonattainment area for the
2006 PM2.5 NAAQS will be an undue burden on Delaware.

Delaware:

Delaware commented that no CSA scale efforts are necessary for 2006
PM2.5 standard planning because the New Castle County nonattainment
problem is separate and distinct from the Philadelphia problem.

EPA Response:

As explained above and in responses below, as well as in the TSD for the
Philadelphia-Wilmington nonattainment area, EPA has determined that
emissions from New Castle County contribute to downwind fine particulate
concentrations in the Philadelphia CSA.  Therefore, CSA-scale planning
is necessary.

Factor 9:  Level of control of emission sources

Delaware:

Delaware stated that in its analysis, EPA is only considering controls
in place and federally enforceable at the time of designation, i.e., by
2008.  Delaware does not agree with this for the reasons discussed under
factor 1, above.  Delaware commented that EPA’s position does not
makes sense in the context of factor 9, because the purpose of this
factor is to not evaluate the level of emissions (i.e., not to
re-evaluate Factor 1), but instead to evaluate how well sources are
controlled in the area.  Delaware argues that significant emission
control measures have been adopted, and approved into Delaware’s SIP
by the EPA, and that these measures are being implemented; permits have
been issued and construction has commenced, and in some cases
construction has been completed.  Delaware concludes that this factor
cannot be evaluated without considering SIP approved control measures.

EPA Response:

In accordance with the CAA, this factor evaluates control measures that
are in place and federally enforceable by December 2008 and were not
reflected in the 2005 NEI.  This factor is intended to consider emission
reductions that were not incorporated into factor 1, but which are
effective before EPA makes its final designations.  In accordance with
the CAA, EPA is tasked to make determinations as to which areas are
currently in violation of the 2006 PM2.5 NAAQS or contributing to
violations of the 2006 PM2.5 NAAQS in nearby areas.  Therefore, and as
explained above, EPA has determined that analysis of emission controls
should not include projections of future emission reductions.

Delaware:

Delaware commented that the Premcor refinery has installed scrubbers on
its largest SO2 emitting sources, a fluid cracking unit and a fluid
coker.  Delaware believes proper evaluation of this factor demonstrates
that these two large units at the Premcor refinery are very well
controlled relative to SO2, i.e., this is Best Available Control
Technology (BACT).

EPA Response:

EPA recognizes that the Premcor (formerly Motiva) refinery has
significantly reduced its SO2 emissions.  However, the Premcor refinery
is not the only large source of emissions in New Castle County.  All
such sources have not yet achieved such significant emission reductions,
and even considering the Premcor reductions, New Castle County’s SO2
emissions are highest in the CSA.

Delaware:

Delaware stated that under its SIP approved Reg. 1146, EGU
Multi-Pollutant regulation, Edge Moor Units 3, 4, and 5 are subject to
stringent NOX and SO2 emission limits in 2009 and 2012.  Delaware
believes that proper evaluation of this factor demonstrates that these
units are well controlled relative to NOX and SO2, i.e., this is BACT.  

Delaware submitted the following additional information for the EGUs in
New Castle County, DE with controls in place after 2005 (i.e., controls
not reflected in the 2005 NEI).

Table   SEQ Table \* ARABIC  1 .  Delaware EGUs.

Unit	Edge Moor 

Unit 3	Edge Moor 

Unit 4	Edge Moor 

Unit 5

Plant Name, City, and County, 	Conectiv Edge Moor, Wilmington Delaware
Conectiv Edge Moor, Wilmington Delaware	Conectiv Edge Moor, Wilmington
Delaware

Emission Unit, fuel use, and megawatt capacity	Coal	Coal	Oil

Controls Installed/ Controls not installed	Yes	Yes	Yes

Type of emission control that has been or will be installed, date on
which the control device will become operational, and the emission
reduction efficiency of the control device	SO2: limited to 0.26 lb/MMBtu

NOX: limited to 0.12 lb/MMBtu	SO2: limited to 0.26 lb/MMBtu

NOX: limited to 0.12 lb/MMBtu	SO2:  0.5% Sulfur oil (max)

NOX:  limited to 0.12 lb/MMBtu

The estimated pollutant emissions for each unit before and after
implementation of emission controls

2002	2012	2002	2012	2002	2012

	SO2	3,344	560	5,051	970	2,133	977

	NOx	922	314	1,096	544	1,289	548

Control device operation federal enforceable date, and instrument by
which federal enforceability will be ensured.	Approved in Delaware SIP
on 09/29/2008	Approved in Delaware SIP on 09/29/2008	Approved in
Delaware SIP on 09/29/2008



EPA Response:

EPA commends Delaware on imposing these emission limits on the Edge Moor
facility.  The above information is informative regarding estimated
reductions in emissions that are anticipated by 2012.  Delaware’s
Multi-Pollutant regulation phases in emissions limits, starting in 2009,
with the bulk of the reductions occurring in 2012.  The estimated
emissions limits given in DNREC’s table, above, reflect the 2012
emissions limits.  

Effective on January 1, 2009, the Edge Moor facility is subject to the
following annual emissions limits:  

Table   SEQ Table \* ARABIC  2 .  Moor Facility Annual Emissions Limits.

Unit	Annual NOX Mass Emissions Limits (tons)	Annual SO2 Mass Emissions
Limits (tons)

Edge Moor 3	773	1391

Edge Moor 4	1339	2410

Edge Moor 5	1348	2427

		

These limits are significantly higher (about 2.5 times higher) than the
emission limits given in Delaware’s October 15, 2008 letter.

Starting in May 2009, the Edge Moor facility will be subject to the
following emissions limits:  0.37 lb/MMBTU heat input for SO2 and 0.125
lb/MMBTU for NOX.  Again, these emission limits are significantly higher
than the emission limits given in Delaware’s October 15, 2008 letter.

The above data does not provide additional information regarding
controls currently in place at the Edge Moor Facility.  With respect to
emission controls, as explained and for the reasons stated above, EPA is
only considering controls in place and federally enforceable at the time
of designation, i.e., by December 2008.  Therefore, in accordance with
the CAA, EPA has determined that analysis of emission controls should
not include projections of future emission reductions.

Delaware:

Delaware states that in evaluating the “level of emissions control”
factor, EPA notes that Premcor has installed scrubbers on its largest
SO2 emitting sources; a fluid cracking unit and a fluid coker.  However,
EPA then seems to give much credence to the fact that even with these
controls New Castle County emissions are highest in the CSA.  This
indicates the EPA has not evaluated this factor at all (i.e., level of
control), but rather they have again evaluated factor 1 (i.e.,
emissions).

Delaware believes proper evaluation of this factor demonstrates that
these two large units at the Premcor refinery are very well controlled
relative to SO2 – BACT.

In evaluating EGUs, the EPA collected data that shows emissions and
controls (current and projected) for EGUs with SO2 plus NOX emissions
greater than 5000 tons.  They obtained this data from the 2006 National
Electric Energy Data System (NEEDS) database.  EPA notes that with the
exception of the Brunner Island facility in York County, which has a
projected date of 2008 for a scrubber on one of its three units, none of
the EGUs in the counties in the Philadelphia-Wilmington nonattainment
area for the 1997 PM2.5 NAAQS put controls in place between 2005 and
2008.  Therefore, the level of control of EGUs is not a major factor in
this analysis.  Delaware believes this analysis makes no sense.  EPA
should not be looking at the NEEDS database to determine future
controls; they should look to the State and SIP approved regulations!
The Delaware units identified by EPA are Edge Moor Unit 3, 4, and 5. 
Under Delaware’s SIP approved Reg.  1146, EGU Multi-Pollutant
regulation, each of these units is subject to stringent NOX and SO2
emission limits in 2009 and 2012.  Delaware believes that proper
evaluation of this factor demonstrates that these units are well
controlled relative to NOX and SO2 – BACT (see table on page 16 of
Delaware comment letter).

In summary, by 2012, New Castle County will achieve a 75 percent
reduction in SO2 emissions, a 47 percent reduction in NOX emissions, and
a 62 percent reduction in overall PM2.5 and PM2.5 precursor emissions,
from a 2002 baseline.  Other counties in the Philadelphia CSA have made
only a fraction of these emission reductions.  This, plus the individual
unit discussion above demonstrates that the level of control of emission
sources in New Castle County is greater than other areas in the CSA, and
evaluation of this factor does not support including New Castle County
within CSA non-attainment boundaries.

EPA Response:

See the overview response to comments on power plant issues, located at
the beginning of Section IV.  

EPA has evaluated each county in the area on the weight of evidence of
all factors as well as any additional information provided by the State.
 For example, EPA considered the information on emission controls
currently in place in light of other factors such as total current
emissions in making a final decision on the designation of New Castle
County.  The fact that current emissions from New Castle county are
quite high in comparison to other counties in the existing Philadelphia
nonattainment area (despite facilities like Premcor having emission
controls in place), along with other factors such as commuting and
meteorology, were important considerations in this decision.  

Delaware notes that some State rules have been adopted to achieve
reductions beginning in 2009 and significant reductions are planned to
be achieved by 2012.  EPA commends Delaware for taking these actions. 
However, as noted in the overview response at the beginning of Section
IV, EPA interprets section 107(d) as requiring the designation of areas
which currently (i.e., through 2008) emit pollutants contributing to
violations of the 24-hour standard.  Therefore, in making decisions
about contributing areas, EPA did not take into consideration emission
controls that would be implemented after the December 2008 date for the
Agency’s final action on designations.  EPA recognizes that
implementation of future emission controls will provide important
improvements in air quality and they should be included in the State
implementation plans and attainment demonstrations for these designated
nonattainment areas.  These emission reductions should be an important
element of Delaware’s plan for attaining the PM2.5 standards.

Pinehurst, ID

Idaho:

Since December 14, 2007, the Idaho Department of Environmental Quality
(DEQ) conducted further discussion with land managers to discuss at
length smoke behavior in the area surrounding the Silver Valley.  The
DEQ feels that it will be advantageous in the implementation process to
include input from the land managers.  These land managers have
significant experience with prescribed fire and slash burning in
Shoshone County, Idaho and the Pinehurst area.  Based on input from the
land managers, DEQ decided to reconsider the recommended NAA boundary. 
The boundary included in DEQ’s December submittal was drawn along
ridgelines so as to include the two main drainages into the Pinehurst
valley (Pine Creek/Amy Gulch and Deer Lick Gulch), the French Gulch
drainage into Kingston, and the South Fork Ridge north of the entrance
to the Pine Creek drainage (see Figure 1, Letter dated October 17, 2008
from Toni Hardesty to the Docket and Elin Miller, RA, Region 10,
hereafter called the letter).  The concept was to include areas where
the possible occurrence of slash burning could impact Pinehurst within
the NAA areas.  The consultation with land managers familiar with the
area have indicated that for any fires west of Pinehurst, when there is
sufficient loft for smoke to clear the ridge between French Gulch and
Pinehurst, the smoke stays aloft.  As a result of these discussions and
refined modeling, DEQ recommends that the nonattainment area boundary be
modified as explained in Figure 1 of Comment 1 attached to the letter.

EPA Response:

EPA has reviewed the State of Idaho’s submission.  EPA commends the
State of Idaho for consulting with land managers in the affected area,
and getting their input to clarify smoke behavior and its interactions
with terrain and meteorology in the area.  With regards to the comment
that “for any fires west of Pinehurst, when there is sufficient loft
for smoke to clear the ridge between French Gulch and Pinehurst, the
smoke stays aloft” the State has not provided any substantiating
information other than this statement.  Data to support this would
include analysis of plume behavior in the area under different
meteorological and seasonal conditions.  Without supporting information
and data to substantiate this statement, it is not possible for an
objective reviewer to verify this information.

Idaho:

Information from the land managers prompted DEQ to reassess potential
air impacts on Pinehurst.  In this analysis, DEQ remodeled theoretical
slash burn sources using actual terrain and meteorological data.  The
predicted impacts on Pinehurst were categorized to show the impact
relative to a source located immediately next to the monitor.  The “10
percent line” from this analysis depicts the location of all sources
that could produce an impact in Pinehurst equivalent to 10 percent of a
centrally located source.  Another way to think of this contour is that
any source outside of the line will contribute less than 10 percent, and
any source inside will contribute more than 10 percent.  DEQ staff
believes this is a good tool to delineate the shape of an appropriate
particulate matter impact boundary or area.

When compared with the original NAA boundary proposal (shown in red on
Figure 1 in the comment letter), the 10 percent equal impact contour
suggests that the areas of concern for impact on Pinehurst are shifted
to the north and east.  The 10 percent contour line also generally
agreed with the boundary proposed by the land managers.  In addition,
overlaying the annual and seasonal wind roses on the monitor location in
Pinehurst (see Figures 2 and 3 in the comment letter) confirms that the
two drainages south of Pinehurst are of greatest concern for smoke
impacts on Pinehurst.  

Therefore, DEQ recommends the nonattainment boundary be changed to that
shown in black in Figure 1.

EPA Response:

After review of this information, EPA has determined that there is not
enough supporting information to verify the validity of the modeling
that supports the analysis.  Specific information that may have helped
clarify the analysis could include specifics such as:

Treatment of plume near ridgetops by the model;

Meteorological data used and gridding information;

Emissions data used and gridding information;

Sensitivity of results to magnitude of emissions; or

Analysis of flow from all channels that could potentially interact with
the drainages that could impact the Pinehurst monitor.

These would have provided specificity to assess the validity of the
information shared by the land managers and to consider whether a change
in boundaries would be warranted.  However, under the present
circumstances EPA feels that the boundary recommended by the State in
its December 14, 2007 letter and agreed to by EPA is well supported
technically and represents a technically sound boundary for the
Pinehurst Area PM2.5 nonattainment area.

Sacramento, CA

CARB:

CARB believes that only Sacramento County should be included in the
PM2.5 nonattainment area.  The only violating monitors in the Sacramento
area (Sacramento T Street, Sacramento-Del Paso and Sacramento-Health
Department) are located in the city of Sacramento.  Data from air
quality monitors in surrounding counties are well below the new standard
and far outside the zone of influence established by Sacramento-Del Paso
monitoring site.  According to the California Regional PM10/PM2.5 Air
Quality Study (CRPAQS) study by Chow (Chow, et al., 2006), a zone of
influence is defined in which a concentration varies by 20 percent. 
Only the monitoring sites at Sacramento-T Street and Sacramento-Health
Department fall within this zone.  CARB also pointed out that the 2006
Design Value for the monitoring site in Placer County was 31 μg/m3, and
therefore, was not a violating monitor as stated by EPA.

The use of county-wide emissions for areas such as Placer and El Dorado
Counties, mountainous regions with large rural populations, does not
adequately reflect the reality of emissions within these areas. 
Although the majority of the population of El Dorado County resides in
the western portion of the county, the population of the eastern
portion, South Lake Tahoe and the surrounding mountainous area is over
25,000.  The majority of the urban population of Placer County resides
in the western part of the county, but almost a third resides in
unincorporated areas.  

Sacramento County, which encompasses the majority of the population in
the region, is the only area that violates the new PM2.5 standard.  CARB
analysis continues to support that violations in Sacramento are due to
localized wood smoke emissions.  CARB notes that Table 2 of EPA’s
Response is incorrect with respect to residential wood combustion
emissions in Eldorado County.

EPA Response:

EPA does not agree that only Sacramento County should be designated as
nonattainment for PM2.5.  EPA’s boundary designation for the
Sacramento PM2.5 nonattainment area includes all of Sacramento County
and parts of Yolo, Solano, Placer, and El Dorado Counties.  This area
includes a majority of the population, and all major point sources and
highways in this area.  Locations west of this area contain small,
dispersed populations and do not include major travel corridors, so they
were not included in the nonattainment area.  

In the 2006 Chow study cited by CARB, the "zone of representation" for
carbon ranged from 3 km to 21 km (2 mi to 13 mi), and averaged 13 km (8
mi).  While not intended to set boundaries for planning purposes, these
results do suggest the size of the area that is represented by a PM2.5
air monitor in California.  And conversely, they suggest a range of
influence of urban areas, which have a relatively dense collection of
organic carbon sources, i.e. wood combustion.  EPA believes this study
supports setting nonattainment area boundaries at least 5-10 miles
beyond the edge of urbanized areas.  Additional consideration of the
ammonium nitrate PM2.5 fraction from mobile source NOx emissions led EPA
to designate larger areas, to capture the contribution of such sources
to PM2.5 NAAQS violations.

EPA’s boundary for Sacramento PM2.5 nonattainment area includes the
areas with violating monitors, and the nearby contributing areas.  When
EPA proposed nonattainment boundaries in August 2008, the 2004-2006 and
2005-2007 design values for Placer County, based on air quality data in
EPA’s AQS database, were 38 μg/m3 and 30 μg/m3, respectively.  Since
then, data from July 9, 2005, which affected the 2004-2006 design value,
were removed from AQS by ARB based on lab error and thereby reducing the
2004-2006 design value to 31μg/m3.  Consistent with comments provided
CARB, these design values do not represent a violation of the PM2.5
standard.  However, the revised design values do not affect EPA’s
premise for including the western portion of Placer County, or the other
surrounding counties, in the Sacramento nonattainment area.   

Based on speciation data provided by CARB, organic carbon and nitrates
were identified as the major components of PM2.5 where violations
occurred, which were attributed to residential woodburning and mobile
sources, respectively.  As both sources are associated with urban areas,
the Sacramento nonattainment areas is intended to capture the full
extent of the urban areas associated with the City of Sacramento so that
sources potentially contributing to the violating monitors are included.
 Even though violations are not recorded in the surrounding counties,
such as Placer County, the Sacramento urban area clearly extends into
the surrounding counties.  

With respect to the most recent residential wood burning emissions data
for El Dorado County provided in CARB’s October 15 letter, EPA agrees
that this data suggests significantly less emissions than given in
ARB’s Almanac (updated August 8,2007).  However, until the values in
ARB’s Almanac are revised to reflect these modifications, EPA will
continue to cite the data presented in ARB’s Almanac.  

Finally, while EPA agrees that woodsmoke emissions are more localized,
we do not agree that the contribution of mobile sources can be
discounted.  EPA views mobile source emissions as a significant
component of regional PM2.5 levels in the Sacramento Valley, and it
appears that the combination of this regional pollution and local
woodsmoke emissions in the Sacramento area lead to violations of the
PM2.5 standard, particularly during stagnant conditions.  Considering
this, EPA looked at the location of the violating monitors and the
sources of both woodsmoke and nitrates to determine the appropriate
nonattainment boundary.  This includes Sacramento and the surrounding
urban areas where there could be significant emissions from residential
wood burning.  It also includes the major transportation and commuting
corridors associated with Sacramento and the surrounding urban areas
which can be a significant source of the nitrate component.

Saint Louis, MO-IL

Missouri:

Information previously submitted by the State and information contained
in the 120-day response continue to support the State’s recommendation
that all counties in Missouri should be designated as attainment or
unclassifiable.

EPA Response:

Pursuant to section 107(d) of the Clean Air Act, EPA must designate as
nonattainment those areas that violate the NAAQS and those nearby areas
that contribute to violations.  EPA’s review of the nine factors for
the four counties and the City of St. Louis (including five counties in
Illinois) shows that the Missouri portion of the bi-state area
contributes to the violating monitors.  

Missouri counties contribute approximately 50% of the total tons per
year (tpy) of direct PM2.5, approximately 57% of the total tpy of NOx,
and approximately 74% of the total tpy of SO2 emitted in the bi-state
area.

70% or greater of the exceedance days occur during the warm season and
are influenced by SO2.  As Missouri counties contribute 74% of the SO2
emissions in the bi-state area, and are upwind during certain episode
days, it is probable that Missouri counties contribute to the
violations.  

78% of the total population in the bi-state area lives on the Missouri
side of the St. Louis bi-state area.   Since these populated areas are
upwind of the violating monitors during certain episode days; it is
probable that those counties are candidates for consideration in the
nonattainment area designation.  

Surface wind direction for high PM2.5 days in Madison County, Illinois
(which is the county with the violating monitors) is from the
south-southwest. A majority of the Missouri part of the bi-state area is
located to the south and southwest of the violating monitors.

It is important to consider that the area in Missouri is already part of
a nonattainment area for existing Ozone and PM violations. EPA’s
boundary recommendation for this standard is generally consistent with
the previous approaches to establish boundaries in that it is inclusive
of the majority of probable emissions contributors in the metropolitan
area and will facilitate air quality planning. 

Missouri:

Analysis of data for days with concentrations over the standard at the
two Granite City sites shows, in general, two types of episode days;
summer days with high concentrations (and high regional sulfate)
throughout the area but higher concentrations at the Granite City sites,
and other days with high concentrations only at one or both of the
Granite City sites, suggesting the strong influence of local sources on
violations of the standard.

EPA Response:

There is evidence to support the claim that there is a local
contribution to the Granite City violating monitors from a limited
number of local sources in Granite City.   However, Missouri also
indicates in its comment, that there are likely monitored episodes which
are regional in nature.  Missouri has not provided information that
confirms that that the preponderance of the mass captured on the PM2.5
filter (at the violating monitors) for all exceedance days is from local
sources alone; nor has Missouri provided conclusive evidence that nearby
Missouri counties are not contributing to these violations.  In
addition, Missouri’s submittal does not address the 2005-2007
violations at the Alton monitor, which is located approximately 20 miles
away.  Therefore, the analysis is not only inconclusive, but is also
incomplete, and does not provide a basis for concluding that emissions
from the Missouri portion of the metropolitan area do not contribute to
the violations.

Missouri:

Chemical mass balance source apportionment based on chemical analysis of
filters from the Granite City VFW site shows a clear difference between
days when the site was upwind and downwind of the local point source. 
Upwind days at the site, when wind vectors from rural Illinois
predominated, consistently showed little impact from the local Granite
City point sources, and were similar in total mass concentration to the
rest of the sites in the metro area.  Downwind day analysis supports
attribution of a significant fraction of the PM2.5 mass measured at the
site to the local source in Granite City.  

	

EPA Response:

EPA agrees that there is evidence to support the claim that there is a
local contribution to the Granite City violating monitors from a limited
number of local sources.   However, Missouri has not provided
information which confirms that that the preponderance of the mass
captured on the PM2.5 filter (at the violating monitors) for all
exceedance days is from local sources.  In fact, as Missouri indicates
in an earlier comment, there are likely monitored episodes which are
regional in nature.  Also, as explained previously, the state’s
analysis does not consider contributions to the Alton monitor. 
Consequently, insufficient evidence has been provided to support the
claim that that nearby Missouri counties are not contributing to these
violations, and EPA’s technical analysis indicated that the Missouri
counties did contribute to the violations.  

Missouri:

Air quality modeling analysis conducted for the annual PM2.5 SIP using
2002 met data show that additional sulfur dioxide controls at St. Louis
area facilities will have minimal effect on reducing 24-hour PM2.5
concentrations on days over the standard.  In contrast, the model shows
direct PM2.5 emissions controls at sources near the violating monitors
will provide the necessary reductions.

EPA Response:

The State utilized CMAQ to estimate the effectiveness of control
strategies on future PM2.5 concentrations in the area of the violating
monitors using strategies being considered for the 2012 annual PM2.5
attainment demonstrations.  Four monitored PM2.5 exceedance days (at one
of the Granite City sites) during 2002 were selected for evaluation.  In
addition, three future (2012) emission scenarios were evaluated: a
baseline scenario (NOX and SO2 reductions at several area facilities); a
scenario that included a 90 percent reduction in SO2 emissions at
Herculaneum (a lead smelter in the area); and a scenario that included
both of the SO2 reductions previously described plus a 16 percent
reduction in primary PM2.5 emissions for US Steel Granite City.  The
results are described in Table 3.  

Table   SEQ Table \* ARABIC  3 .  PM2.5 Modeling Results.  From the MDNR
10-18-08 response.

The state modeling study relies on Missouri emission reductions of PM2.5
precursors that are not yet federally enforceable and for the most part,
not yet state enforceable.  It is therefore, not appropriate to rely on
these potential future reductions to eliminate sources or areas from
consideration in establishing current nonattainment boundaries. 
However, if one could assume that these reductions were permanent,
enforceable and quantifiable, the modeling indicates that SO2 and NOx
reductions (including reductions from sources on the Missouri side of
the metropolitan area) in combination with reductions of direct PM2.5
emissions from local sources in Granite City could result in a reduction
of on average of 1 ug/m3 at the violating Granite City monitors.  

Consequently, the model actually confirms that both localized emissions
reductions and regional emissions reductions (some of which may
originate from the Missouri side of the metropolitan area) could be
effective in addressing PM2.5 violations for the 24-hour standard.  The
analysis therefore provides further support for inclusion of the
Missouri portion of the area in the nonattainment area.

Missouri:

Because there is very limited speciation data available for the Granite
City sites, the CES analysis done by the EPA underestimated the effect
of local sources adjacent to monitors and overestimated the contribution
of urban-wide area sulfur oxides.  Contrary to the assertion by EPA,
Missouri analysis shows that the large portion of the PM2.5 sulfate in
the St. Louis area is from multi-State regional transport.

EPA Response:

EPA did not rely solely on the CES scores in determining what boundaries
are appropriate for this non-attainment area.  CES scores were just one
small component of the emissions and meteorology criteria (two of the
nine criteria) that EPA considered when making its determination. 
Despite the fact that these speciation data were not available at the
time that CES scores were calculated, a review of these data indicates
that they likely would have resulted in relatively small changes to the
CES.  Furthermore, these data do not confirm overwhelming impact from
local sources in Granite City to the exclusion of contribution from
sources throughout the metropolitan area, and do not provide definitive
conclusions about contributions from Missouri sources.  

Missouri:

It would be instructive to examine PM2.5 speciation measurement results
on the days of interest (high days), although speciation data is not
available for the two violating monitors in Illinois.  It is the
State’s position that by evaluating this data, along with correlation
data (given as R2 values in the response) the State could easily show
that the two violating monitors in Illinois are impacted greatly by a
local source (Granite Steel) and that the two monitors have little
similarity to Missouri monitors.  

EPA Response:

EPA agrees that speciation data from the violating monitors on episode
days would be of use in establishing nonattainment boundaries.  However,
we also agree that such speciation data are not available.  Using
speciation data from other monitor sites (e.g. Blair Street and Arnold)
has analytical limitations.  The state indicated that on the two days
for which speciation data for both non-violating Missouri sites are
available (2/18/2004 and 6/24/ 2005), the composition of the two sites
are similar. On these days, neither Blair St. nor Arnold is exceeding
but at least one of the violating monitor sites in Illinois is (Table
4).

EPA agrees that there is likely a local component (in Granite City)
which is contributing emissions to the violating monitors.  However,
Missouri has not provided information which would confirm that this
local contribution is overwhelming when compared to other sources; nor
has Missouri demonstrated that emissions from nearby Missouri counties
(and the City of St. Louis) are insignificant in terms of their
contribution to the violating monitors.   EPA’s technical analysis of
all factors and analytic tools shows contribution from the Missouri
Counties which is not overcome by the Missouri submittal.

Table 4.  Days with High Concentrations at Granite City or VFW.

From the MDNR October 18, 2008 response.

Salt Lake City-Provo, UT

Utah:

Utah Valley is a basin that traps its own air during stagnant
winter-time temperature inversions.  This difference is most pronounced
between Utah and Salt Lake Counties where the two air basins are
separated by a high promontory called Point of the Mountain.

EPA Response:

The two counties are not separate air basins based on several factors as
described in our August 18, 2008 9-factor technical analysis.  The key
factors were:

Both EPA and the State evaluated a 6,000 ft inversion height which
exceeds the height of the lowest point in the gap (Point of the
Mountain) between Utah and Salt Lake Counties by at least 1000 ft.  

The presence of strong temperature inversions being sustained for up to
three weeks

Similar 2005-2007 design values for Salt Lake County at 48 micrograms
and Utah County at 45 micrograms

Cold dense air will move from the higher elevation of Utah Valley at
4,700 ft  mean sea level (MSL) to the lowest point in the area which is
Salt Lake City at 4250 ft MSL.  

	

Furthermore, as described in the 9-factor analysis (specifically Factor
6) EPA also identified that at approximately 6,000 ft MSL there is a
4.75 mile wide opening between the two counties at Point of the Mountain
that allows air masses to move uninhibited.  The State did not
contradict this information and only provided data from the Herriman
monitor to support their comment that Utah and Salt Lake Counties are
two separate air basins.  	

EPA used the best available meteorological data from the Salt Lake
Airport and has recently evaluated additional meteorological data
showing that the air mass is interconnected, moves back and forth
between these two counties and is not separated by any topographical
barrier.

Utah:

Based on a variety of arguments, Utah asserts that Utah County should be
a separate nonattainment area from the remainder of the Wasatch Front. 
Among other things, Utah argues that Utah County is geographically
separated from Salt Lake County, that other areas around the Country
have been separated into more than one nonattainment area even where a
single nonattainment area could have been created, and that planning
efforts will be simplified and expedited if Utah County is designated as
a separate nonattainment area.  

EPA Response:

EPA has amended our August 18, 2008 proposal, relative to separating or
combining Utah County and Salt Lake County.  EPA has given further
consideration to the State recommendation to separate the two counties
and agrees to Utah County (partial) as being a separate nonattainment
area.  This decision is based on a broader evaluation of jurisdictional
issues that the State claims will facilitate SIP planning.  Regardless
of whether they are a single or separate nonattainment area, both Utah
County and Salt Lake County are violating the 2006 24-hour PM2.5
standards and have similar design values.  EPA anticipates that the
State of Utah will have to insure that these two areas work closely
together to develop a consistent strategy for attaining the NAAQS,
including a combined modeling demonstration and consistent control
measures.  However, EPA’s analysis of speciated data, pollution roses,
and other analytical tools indicates that these two areas are
interconnected and thus we expect that SIPs for both areas will fully
account for this influence.

Utah:

Both Utah and Salt Lake County PM10 nonattainment areas were able to
develop implementation plans that successfully met the goals and
requirements of the Clean Air Act.

EPA Response:

PM10 nonattainment area boundaries do not give an indication of what
boundaries are appropriate for PM2.5.  PM10 and PM2.5 are different size
particles, behave differently in the atmosphere, are often the result of
emissions from different types of sources, and will probably require
different controls strategies.  Thus, the existing boundaries for PM10
nonattainment areas are of limited relevance in this area.  

EPA did approve the original PM10 plans for these two separate areas in
1994 but has not been able to approve Utah’s 2005 redesignation
request for Salt Lake County for several reasons including recent
violations of the PM10 standard.  The State acknowledges problems with
separating the two Counties due to the modeling demonstration used to
support the attainment of the PM10 standard. While Utah County has not
had recent violations of the PM10 standard prior to 2008, there have
been significant exceedances.  Most recently, preliminary 2008 data (not
certified) show a potential PM10 violation in Utah County in 2008.  Our
experience with the PM10 attainment plans has shown that it is more
difficult to have two separate nonattainment areas and for the same
reason mentioned by the State which is the need to use one modeling
domain.  

Utah:

Other EPA Regions have retained a separation of nonattainment areas with
examples for Ohio and Pennsylvania.  

 

EPA Response:

EPA’s designations guidance for the 2006 PM2.5 standard states that
the criteria for determining appropriate boundaries is to be made on a
case-by-case basis considering the nine factors.  Not every factor was
relevant in each area (e.g., in many eastern locations topography is not
an issue), nor was every factor equally important to each area.  Due to
the complex and variable nature of PM2.5, the important factors varied
from one area to another based on local circumstances.  Moreover, EPA
consciously did not impose any mandatory “brightline” tests for any
of the nine factors.  

	

Additionally, the guidance called for retaining existing boundaries and
to the extent possible making them identical to existing ozone
boundaries.  The PM2.5 nonattainment areas for Ohio and Pennsylvania are
examples of maintaining existing nonattainment boundaries.  Utah is
being designated nonattainment for the 24-hour PM2.5 standard for the
first time.  And thus the guidance on existing boundaries does not apply
and the comparison to Ohio and Pennsylvania is not relevant.  

	

Nationwide, there are numerous examples of nonattainment boundaries in
place from prior NAAQS designations or currently being proposed based on
a case-by-case analysis of site specific conditions encompassing very
large multi-county, multi-state and multi-jurisdiction nonattainment
areas. Historically, the complexity of jurisdictions within a
nonattainment area has not been an impediment to air quality
improvements. 

	

For instance, the Denver 8-hour ozone area is comprised of seven whole
counties and two partial counties encompassing over 8,287 square miles
including national forest land, national parks and three wilderness
areas: Indian Peaks (Boulder County),  Comanche Peak (Larimer County)
and Lost Creek (in Jefferson County).

	

At the time of the Denver 8-hour ozone designation only two of the nine
counties had violating monitors.  In addition, there are two separate
Metropolitan Planning Organizations (MPOs) which have been successfully
working together to identify emission budgets for SIP planning purposes.
 This area also encompasses four separate Core Based Statistical Areas
(CBSAs) with a population of over 3.1 million people.  In terms of
jurisdiction, Colorado and Utah both have State level air quality
departments responsible for SIP development.  The area being proposed
for the Utah Wasatch Front is 150 miles in length and approximately
3,800 square miles and is less than half the area of the Denver 8-hour
ozone nonattainment area.

	

Utah:

As technical justification for this portion of its proposal, EPA cites a
4-mile opening beneath a 6,000 ft inversion level operating as a
geophysical connection between the two valleys.  While the actual
connectivity and degree of air transport under cold-pool conditions are
difficult to quantify, we would ask you to consider the ambient air
monitor at Herriman, a location near the divide.  It has a design value
of only 24 μg/m3 (for 2005-2007) which suggests that the high values
observed to the North and South are in fact the result of two distinct
air-sheds dominated by two separate urban centers.

EPA Response:

While the Herriman monitor is the southernmost monitor in Salt Lake
County (see figure 12), it is not in a flow path between the Salt Lake
and Utah valleys; rather it is located 7.3 miles to the west of the
north/south gap called “Point of the Mountain.”  It is approximately
two miles north of the 6800 ft South Mountain, and in line with the
mouth of Rose Canyon, which extends upward into the Oquirrh mountains
which form the western border of the Salt Lake Basin.  In this location,
the monitor may be subject to inputs of clean air from the west.  The
Magna monitor, also on the west side of the Salt Lake City Basin, but
far from Point of the Mountain, is the second lowest monitor in the
County, with similar siting, and also attains the NAAQS for 2005-2007.

	

μg/m3  level in 2002-2004), it is still subject to very high levels of
PM2.5 (1st maxima of 69.1  μg/m3in 2001, 62.1 μg/m3 in 2004 and 57.4
μg/m3 in 2007), indicating that very elevated levels of PM2.5 occur in
this part of Salt Lake County under some meteorological conditions.  

	

Even closer to Point of the Mountain is the Highland monitor in Utah
County, only 5.8 miles to the southeast.  Highland has similar siting
considerations to Herriman, being in line with the mouth of Dry Creek
Canyon, and lying south of the ridge separating the eastern Salt Lake
County communities from the northeast Utah County communities.  As is
the case for Herriman in Salt Lake County, Highland often shows the
lowest PM2.5 concentration for Utah County.  Despite being nearer Point
of the Mountain, Highland violates the NAAQS with 2005-2007 data.

 

Figure   SEQ Figure \* ARABIC  12 .  Position of Herriman and Highland
relative to Point of the Mountain.

Utah claims that, since Herriman is between the maximum concentration
Salt Lake City monitors and the maximum concentration Provo monitors,
and shows a significantly lower design value, the conclusion is that
there are two independent PM2.5 airsheds with a minimum point in
between.  A different picture is presented by examining actual high
concentration days.  On January 27, 2007 Herriman monitored its maximum
concentration.  Concentrations for area monitors on January 27, 2007 are
shown in Figure 13, along with winds at Point of the Mountain.  Winds
were consistently from south to north, around 5 mph, for the first
twelve hours of the day, and then changed to 10 to 15 mph winds from the
north from noon to sunset, showing a diurnal down valley/up valley
pattern with significant transport velocities.  Rather than a
north/south gradient in concentration with a minimum at Point of the
Mountain, Figure 13 shows more strongly an east/west gradient, with
highest concentrations along the east edge of the Salt Lake and Utah
Lake basins in the high population density areas, and lower
concentrations at Herriman and Magna on the west side of the Salt Lake
City Basin.

	

Figure 14 shows winds at Point of the Mountain collected by the Utah
Department of Transportation (UDOT) monitor along I-15.  The monitor
shows moderate southerly winds in the morning and stronger northerly
winds in the afternoon through the gap.

 

Figure   SEQ Figure \* ARABIC  13 .  Concentration and Point of the
Mountain Surface Winds on the Peak PM2.5 Day of 2007.

 

Figure   SEQ Figure \* ARABIC  14 .  Point of the Mountain Meteorology,
January 27, 2007.

Given the presence of significant transport winds through Point of the
Mountain, the significance of low values at the Herriman monitor has a
different implication.  If the Herriman monitor were actually in the gap
between Salt Lake County and Utah County, and if it consistently saw
concentrations at 24 μg/m3 at that location on high pollution days,
this would indicate that air at 69 percent of the NAAQS is moving from
one basin into the other.  On an exceedance day, this level of PM2.5 in
an “outside” air injection shows that the upwind basin is
contributing significantly to high PM2.5 readings in the downwind basin.
 The input air leaves very little room for local emissions above this
incoming “background.”  In fact, because of the siting of the
Herriman monitor, EPA expects that the air passing Point of the Mountain
has considerably higher concentrations than those measured at Herriman. 
Data from the UDOT meteorology station at Point of the Mountain
indicates that surprisingly high transport winds pass through this gap
consistently during inversion periods.

Utah:

Under EPA’s proposal, where the entire area would be a single
nonattainment area for PM2.5 , an inadequate conformity finding due to
the planning of either MPO would result in a lapse of federal funding
that would encompass the areas administered by both.  This creates a
scenario under which one area could be held responsible for actions
taken by an organization in which it had no representation.

 

EPA Response:

A collaborative process is outlined in Section XII “Transportation
Conformity Consultation” of Utah’s SIP and addresses issues such as;
transportation conformity consultation, specific roles and
responsibilities, interagency collaboration and consultation process,
and the dispute resolution process.  

There are options available to administer the transportation conformity
requirements in the entire nonattainment  PM2.5  boundary area for Salt
Lake County and Utah County (and for Cache County, UT and Franklin
County, ID).  To address these transportation conformity issues, and
many other situations, EPA’s Office of Transportation and Air Quality
(OTAQ) has produced two documents; “Companion Guidance for the July 1,
2004, Final Transportation Conformity Rule; Conformity Implementation in
Multi-Jurisdictional Nonattainment and Maintenance Areas for Existing
and New Air Quality Standards” (EPA420-B-04-012, July, 2004) and
“Interim Guidance for Implementing the Transportation Conformity
Provisions in the Safe, Accountable, Flexible, Efficient Transportation
Equity Act:  A Legacy for Users (SAFETEA-LU)” (EPA420-B-06-90,
February, 2006.)

	

The “Companion Guidance” document (EPA420-B-04-012) outlines options
for establishing the PM2.5 motor vehicle emissions budgets (MVEB) on a
nonattainment area-wide basis or sub-area basis.  In addition, the MPOs
may revert from using nonattainment area-wide budgets to demonstrate
conformity by meeting their selected respective sub-area emission
budgets if the MPOs make concurrent conformity determinations that
demonstrate consistency of their respective plans and programs with
their individual sub-area budgets.  Likewise, at any time in the future,
the MPOs may switch from using sub-area budgets to using nonattainment
area-wide budgets.  This is allowed as long as they once again perform a
joint conformity determination and the sum of their sub-area motor
vehicle emissions are equal to or less than the established
nonattainment area-wide PM2.5  MVEBs.  This process is allowed provided
that it is established in the SIP.

	

As noted above the EPA “Companion Guidance” document allows great
flexibility for the MPOs (both Wasatch Front Regional Council and
Mountainland Association of Governments) to jointly meet the necessary
SIP budgets and no one MPO is penalized.  This represents a unified
process for addressing transportation conformity, and to work towards
attainment of the PM2.5 24-hour NAAQS.

	

For example, EPA notes this process is currently being employed by the
two MPOs in the Denver North Front Range 8-hour ozone nonattainment area
which are; the Denver Regional Council of Governments (DRCOG), and the
North Front Range Transportation and Air Quality Planning Council
(NFRT&AQPC).  These two MPOs, and the State of Colorado, which elected
to represent areas in the 8-hour ozone nonattainment area not covered by
an MPO, have worked cooperatively to address transportation conformity
issues associated with the development of the Denver North Front Range
8-hour ozone attainment SIP.  Through the development of this ozone
attainment SIP, Colorado has collectively considered a unified approach
for addressing its regional ozone issue for an area that includes seven
whole counties, two partial counties, contains approximately 3.1 million
people, and a land area of approximately 8,287 square miles.

Utah:

There is also concern with respect to general conformity.  EPA’s
proposal to annex a portion of Tooele County includes the Deseret Peak
Wilderness area.  Inclusion of this area would force the Forest Service
to compile an abundance of documentation.

EPA Response:

By definition a “Wilderness area” is roadless; therefore, the Forest
Service should not have to provide much documentation for transportation
budgets.  In terms of general conformity, the Forest Service would have
to prepare NEPA documents to assess the emissions from a project
regardless of whether the area is attainment or nonattainment.

Furthermore it is not unusual for nonattainment areas to encompass
national forest, parks and wilderness areas.  For example, the Denver
North Front Range 8-hour ozone nonattainment area includes national
forest land, Rocky Mountain national park and three wilderness areas:
Indian Peaks (Boulder County),  Comanche Peak (Larimer County) and Lost
Creek (in Jefferson County).  

Utah:

The “satellite” areas of Tooele and Box Elder Counties should not be
included in the Salt Lake City-Ogden-Clearfield CSA nonattainment area. 
Concerning the proposal to include portions of Tooele and Box Elder
Counties in the single nonattainment area for the Wasatch Front, Utah
recognizes that EPA, in its evaluation of the nine factors probably gave
less weight to the actual monitored data than the State did and arrived
at a different conclusion.

 

EPA Response:

With regard to including portions of Box Elder and Tooele Counties in
the Salt Lake City nonattainment area, EPA is required under section
107(d) to designate not only violating areas, but nearby areas that
contribute to those violations.  Thus, a key objective of the
designation process was to ascertain those nearby areas with emissions
sources or emissions activities that contribute to violations.   EPA
based its evaluation of these two areas on the types of information
recommended in EPA’s guidance and additional relevant information,
including traffic and commuting, growth, meteorology, topography, and
emissions.  Taken together, this information supports the conclusion
that both Box Elder and Tooele contribute to the PM2.5 violations in
adjacent counties.  The fact that neither area is currently monitoring
nonattainment does not address whether they contribute to violations in
nearby areas.  

We note there are no topographical barriers between Brigham City and
Salt Lake City; the two areas are part of a single very large air basin.
 Based on emission transport during long periods of stagnation under
persistent temperature inversions, sufficient mixing occurs allowing
both Box Elder and Tooele Valley emissions to reach the maximum
concentration monitors in Salt Lake City and Ogden and contribute to
NAAQS violations.

	

μg/m3 and the 2005-2007 design value was 29 μg/m3 ; in Tooele the
2005-2007 design value was 31 μg/m3 .  While not violations, these
values demonstrate that these areas are subject to poor air quality at
times, and it is likely that these high concentrations contribute to
violations in adjacent counties on days when winds blow from this
direction towards the rest of this area, and contribute to area wide
ambient levels during inversions.  If exceedances are caused by local
emissions, it indicates emission levels high enough to leave very little
margin to the NAAQS and that area emissions are high enough to
contribute to the overall air quality issues of the larger basin.  If
exceedances are from transport from Salt Lake City then it demonstrates
that Box Elder and Tooele County are in the same air shed, with
sufficient mixing for the central and outlying areas to interact.  

Utah:

There is really nothing to be gained, in terms of air quality
mitigation, by making a nonattainment designation in these areas.  As
pointed out in Utah’s recommendation to EPA, industrial sources are
not excused from emission controls simply because they choose to locate
outside a nonattainment area.  Will these permitting tools remain
effective when trading would be allowed between sources located 150
miles apart?

EPA Response:

We acknowledge that Utah requires PSD BACT in all major attainment area
permits Statewide; therefore, we agree that industrial sources locating
outside the nonattainment area are not excused from emission controls. 
Also, EPA recognizes that Utah requires even minor sources apply
“State” defined BACT prior to obtaining a permit to construct. 
However, a nonattainment designation includes additional requirements
that are designed to bring significant improvements to air quality.  For
example, Nonattainment Area New Source Review (NA NSR) includes
requirements for major stationary sources, such as Lowest Achievable
Emission Rate (LAER), offsets, and net air quality benefit that go
beyond PSD permitting requirements.  

	

Regarding the claim that offsets would be allowed between sources
located 150 miles apart, it is our view that such offsets would only be
allowed if they produce a positive net air quality benefit in the
affected area.  In accordance with this requirement, Utah’s
SIP-approved rule R307-403-3(3)(d) requires that emission offsets “…
provide a positive net Air Quality benefit in the affected area of
nonattainment.”  Unfortunately, Utah has merely assumed that
offsetting reductions in emissions will provide a positive net air
quality benefit regardless of the location of the emissions and the
offsetting emission reductions.  EPA commented on this issue in letters
dated December 9, 2004, regarding the Summit Vineyard NSR permit action,
and June 30, 1999, regarding the Kennecott Bingham Canyon NSR permit
action.  At the very least, there would have to be reasonable progress
toward attainment of the applicable NAAQS.  In addition, it may be
appropriate for the State to identify geographical restrictions within a
single nonattainment area on obtaining and using offsets.

Utah:

Also concerning the Box Elder appendage, EPA had indicated that there
will likely be an effort to “harmonize” areas of nonattainment for
both PM2.5 and ozone.  This area has measured ozone concentrations that
are very close to the 2008 ozone standard, but only under meteorological
conditions that include a steady wind from the south.  This supports the
notion that Brigham City is in fact being adversely impacted by the core
area of ozone nonattainment.  This is likely also the case with PM2.5. 
EPA’s proposal, however, presumes the opposite – that Brigham City
is adversely impacting on the core nonattainment area and should
therefore become part of the nonattainment area.

EPA Response:

EPA’s 9-factor analysis did not include any discussion of the ozone
exceedances that have occurred in both Box Elder and Tooele Counties. 
While maximum PM2.5 and ozone concentrations in Brigham City may occur
with southerly flows, when Brigham City is downwind from the largest
concentration of emissions in Salt Lake County (as would be expected),
this does not mean that Box Elder emissions play no role in Salt Lake
City PM2.5 and ozone concentrations.  No topographical barrier exists
between Brigham City and Salt Lake City; the two areas are part of a
single very large air basin.  Based on emission transport during long
periods of stagnation under persistent temperature inversions,
sufficient mixing occurs allowing Box Elder and Tooele Valley emissions
to reach the maximum concentration monitors in Salt Lake City and Ogden
and contribute to NAAQS violations.

	

The monitor in Box Elder County has shown significant exceedances of the
PM2.5 standard as well as three year design values near the 2006 NAAQS
level (and above that level for some periods historically).  In Box
Elder, the 2004-2006 design value was 35 μg/m3 and the 2005-2007 design
value was 29 μg/m3 ; Tooele’s 2005-2007 design value was 31 μg/m3. 
While not violations, these values demonstrate that these areas are
subject to poor air quality at times.  If exceedances are caused by
local emissions, it indicates emission levels high enough to leave very
little margin to the NAAQS and that area emissions are high enough to
contribute to the overall air quality issues of the larger basin.  If
exceedances are due primarily to transport from Salt Lake City then it
demonstrates that Box Elder and Tooele County are included in the same
air shed, with sufficient mixing for the central and outlying areas to
interact.  Taken together, the nine factors indicate that both Box Elder
and Tooele contribute to the PM2.5 violations in adjacent counties and
that the adjacent counties contribute to NAAQS exceedances in Box Elder
and Tooele Counties.  Based on 2008 preliminary data, the design value
for Box Elder appears to be moving upward from the 2005-2007 design
value, further demonstrating this interconnection.

Utah:

As we have already noted, the application of backward wind trajectories
used to justify the inclusion of not only these areas, but the appendage
of Utah County as well, demonstrates a flawed understanding of
meteorological processes at work in Utah.  These trajectories do not
recognize terrain effects or the trapping of the critical boundary
layer, and thus are not representative of actual air flow.  Furthermore,
EPA’s proposal did not include any technical details in support of the
trajectory model runs, making it difficult to understand some of the
assumptions that were made.

EPA Response:

EPA agrees that attempts to model air motion in the Wasatch Front during
wintertime inversions have not yielded completely satisfactory results. 
Whether considering HYSPLIT results using archived national coarse grid
meteorological data sets, or Utah’s more detailed approaches using
UAMAERO/MM5, model performance in matching measured low level inversion
period winds has been relatively poor.  This difficulty in modeling cold
pool winds during wintertime stagnation events has long been an issue in
other western States as well.  While uncertain, and providing limited
usefulness in quantifying transport between different parts of the
Wasatch Front, the methods suggest levels of transport that are
consistent with measured surface winds and wind patterns, particularly
over extended multi-day, or multi-week episodes with persistent strong
wintertime inversions.  For example, EPA used three sets of
backtrajectories to illustrate transport from Utah County, Tooele County
and Box Elder County toward Salt Lake County.  Examining PM2.5 data
collected by the Utah Division of Air Quality (Utah DAQ) along the
Wasatch front, PM2.5 measurements at Brigham City in Box Elder County
are consistent with the transport winds indicated by the HYSPLIT
backtrajectories.  On January 13, 2004 and January 26, 2006,
backtrajectories indicate general flow along the Wasatch front from
south to north.  On both days, the Brigham City monitor recorded PM2.5
exceedances, consistent with transport of pollutants northward from the
center of the nonattainment area.  On January 22, 2004, the
backtrajectory indicated general flow from north to south; on that day,
Brigham City was the only Wasatch Front monitor to record PM2.5 levels
below the PM2.5 NAAQS, consistent with local Box Elder emissions being
blown south into the rest of the nonattainment area.  The three
backtrajectories used by EPA align with measured PM2.5 data; the
response to the last comments from the Utah DAQ on the CES, in Section
II, of this document shows that the backtrajectories used are also
consistent with surface wind data collected by the Utah DAQ.

	

Under such wintertime conditions, even light winds can transport
emissions throughout the PM2.5 nonattainment area proposed by EPA.  EPA
has developed data sets of surface wind conditions from the Utah DAQ
meteorological monitoring network that show some degree of organized
transport winds during all of the inversion days analyzed (that is,
periods of time lasting several hours in which most or all of the Utah
DAQ met sites simultaneously show northerly wind components or, at other
times, southerly wind components).  While HYSPLIT may not accurately
represent actual local wind conditions, the EPA assessment shows that
winds measured on the ground by Utah DAQ during wintertime inversions
are consistent with the archived wind fields utilized by HYSPLIT.  The
HYSPLIT trajectories would not be accurate enough for some applications
(quantitative source attribution, for example), but are accurate enough
to provide a demonstration of gross air movement within the Wasatch
Front, and result in conclusions consistent with both PM2.5 measurements
and surface wind measurements.

 

Utah:

Furthermore, we believe that the technical basis underlying the EPA’s
proposal is incorrect in its use of back-trajectory analyses that fail
to represent cold–pool conditions, meteorological data that was
collected at the wrong locations, arbitrary inventories that were not
even seasonally adjusted (and therefore overstate carbon emission) and
commuting patterns that were misrepresented by a factor of 10.

EPA Response:

EPA utilized the best available tools and information in a nationally
consistent manner.  EPA has examined actual Utah DEQ surface wind data
to evaluate the accuracy of the HYSPLIT trajectories, as well as the
representativeness of the pollution roses collected at Salt Lake
International Airport relative to data from other areas in the Wasatch
Front.  

	

While the accuracy of these methods is important, it must be noted that
Great Salt Lake Valley and adjacent Utah Valley are closed air basins. 
There are no low elevation outlets.  The presence of a strong, multi-day
temperature inversion under wintertime conditions traps all surface
level emissions, and for longer events ample time is provided for mixing
along the length of the Wasatch Front given the observed non-zero wind
velocities and patterns.  We note that Utah has not provided any
countervailing evidence, and EPA’s methods and conclusions were the
best available at the time of our August 18, 2008, 9-factor technical
analysis.  Subsequently, we have included updated pollution roses in our
final Technical Support Document, which includes wind data from an
additional meteorological station located at Hill Air Force Base in
Ogden, Utah.  The issues related to emission inventories and commuting
patterns are addressed under separate comment/response.

Utah:

Along the Wasatch Front, EPA’s pollution rose diagrams were created
using inappropriate wind data.  Wind data from the Salt Lake City
International Airport to create pollution rose diagrams for Utah County,
Weber County, Davis County, and Toole (sic) County while wind data from
Pocatello, Idaho was used to create a pollution rose diagram for Box
Elder County, Utah.  These pollution rose diagrams were then used as
evidence to support the exchange of precursor emissions between outlying
areas and the Salt Lake City-Ogden-Clearfield CSA and Provo-Orem
CBSA’s.  The DAQ would again like to point out that because topography
controls the wind, a wind rose from Salt Lake City is not representative
of the wind rose in Ogden or Provo and a wind rose from Pocatello, Idaho
(mountain ranges away) is completely inappropriate to be used for Box
Elder County, Utah.

EPA Response:

Utah did not provide wind roses for Box Elder, Weber, Davis Tooele or
Utah Counties to demonstrate that the wind rose used was inappropriate. 
EPA does not disagree that local pollution roses would be more
representative of local conditions.  Given the siting at Salt Lake
International Airport, however, the wind data there is likely to be
representative of much of the southern Great Salt Lake area, and of
overall flow within the greater basin.  EPA has done an analysis of the
Utah DAQ meteorological data to assess regional wind patterns during
inversion episodes.  Results show periods of uniformity in air flow. 
The Utah DAQ site between the Airport and the Lake shows more lake
effect forcing than do other sites, but is not an outlier within the
network as a whole in terms of directional patterns.  Overall the
analysis supports the conclusion reached independently using the Salt
Lake International Airport wind rose.  Throughout much of the valley,
there is often widely distributed simultaneous northerly or southerly
motion.

Utah:

The two-sentence statement on page 36, following Table A.3-5 of the
comment letter states a conclusion that is simply not supported by the
data in the table:

“Many of the counties that are candidates for nonattainment show a
higher percentage of commuters going to Salt Lake County than are
commuting from Salt Lake to other counties.  The counties of Box Elder
at 24.1 percent, Tooele at 43.8 percent and Utah at 12.9 percent are all
higher than Salt Lake at 5.4 percent which shows that emissions related
to traffic and commuting from those areas are contributing to violations
of the PM2.5 standard.” 

Table 2 (located at the end of the comment letter) clarifies the actual
contribution of VMT, and thus emissions, to the proposed nonattainment
area from Box Elder and Tooele Counties.  The fact that outlying
counties have a higher percentage of drivers commuting outside the
county than does Salt Lake County does not mean that those counties are
significant contributors to the mobile emissions inventory along the
Wasatch Front.  Salt Lake County had an estimated 8.9 billion VMT per
year in 2005.  This is roughly 10 times the VMT reported for either Box
Elder or Tooele.  Therefore, any comparison of these relative
percentages is misleading.  It is more appropriate to examine in terms
of actual VMT, the impact on the core nonattainment area that is due to
these outlying regions.  Table 1 in the comment letter shows that there
is only about a 1.5 percent increase in VMT to each of the proposed
nonattainment counties from commuters in Box Elder and Tooele Counties,
hardly a compelling reason to conclude that this impact contributes to a
violation in the nonattainment area.  This is based on the assumption
that approximately 27 percent of the annual VMT comes from driving to
and from work.

EPA Response:

EPA presented commuting data and VMT data as supporting information with
regard to the contribution of emissions to the proposed nonattainment
area.  This is only data for one factor (Factor 4) of the EPA’s
9-factor analysis.  The State only considered information from Factor 4
for 2005 and did not elaborate on the data in Factor 5, which shows
significant predicted growth in both population and VMT for Box Elder
and Tooele Counties.  The population growth can be a surrogate measure
of emissions activities, and a gauge of whether the area is integrated
economically into the larger area, both of which indicate a greater
degree of contribution to ambient PM2.5.

	

EPA also notes that commuting VMT into the nonattainment counties does
contribute to PM2.5 violations in those counties.  The populated area of
both Box Elder and Tooele Counties are in the eastern portion of these
Counties.  The overall total VMT and the associated emissions from each
of these Counties contributes to the PM2.5 and precursor emissions to
PM2.5 concentrations in the airshed.  These emissions and their future
growth for the overall VMT should not be discounted merely because these
emissions are not specifically related to commuting.

	

The State uses a default factor of 27 percent for VMT from commuting to
and from work from a 2004 FHWA document entitled “Summary of Travel
Trends” to significantly discount the VMT.  There is no indication
that the State conferred with the MPO for the Wasatch Front (Wasatch
Front Regional Council - WFRC) to see if they agree with this 27 percent
figure and if it is relevant for Utah.  This input from WFRC would
provide a localized interpretation, rather than a national default, for
consideration for commuters coming from Box Elder and Tooele Counties to
Davis, Salt Lake, and Weber Counties.  

Utah:

Other than exceptional wind and smoke events, elevated PM2.5
concentrations in Utah occur during wintertime quiescent ridge dominated
synoptic conditions.  The lack of synoptic scale forcing allows strong
cold pools to form in basin or valley bottoms.  Within and around these
cold pools, weak large scale forcing allows terrain driven flow to
develop and vertical mixing is weak to non-existent.  Topography
controls the wind during these conditions.

EPA Response:

EPA acknowledges the States comment as mostly true.  However, the last
sentence that “Topography controls the wind during these conditions”
implies that there are no organized wind patterns within the cold pool. 
This may not be accurate.  EPA has done an assessment of wind speed and
direction during inversion episodes using the surface wind data
collected by Utah DAQ across its network.  EPA’s conclusions are that
while some sites experience winds dominated by local topography (down
valley flow at canyon mouths, etc.) the network as a whole shows a
significant amount of uniform winds.  There are recurring, multi-hour
periods when all or most of the network records similar wind directions
at the same time; rather than being controlled by local topography, the
basin as a whole shows recurring, relatively uniform responses to
changes in conditions, to the extent that the regional analysis diagrams
are termed “Slosh Diagrams.”

San Carlos Apache Indian Reservation, CA

San Carlos Apache:

With regard to EPA’s letter stating that EPA intends to designate the
San Carlos Apache Reservation as attainment/unclassifiable for the 2006
24-hour PM2.5 health standard, the Tribe finds this designation
“challenging.”  The Reservation, which consists of about 1.9 million
acres, is largely pristine and undeveloped.  Much of the particulate
matter in the air on the Reservation originates from mining related
activities to the east, west, and south of the Reservation.

EPA Response:

There are no monitors on the reservation and there are no monitors in
the  area  surrounding the reservation that are showing violations. 
Absent monitoring data that reflects a violation, we cannot make a
designation other than “attainment/unclassifiable.  Also, since the
San Carlos Apache Reservation is not included within a nonattainment
area and since they are not contributing to nonattainment elsewhere, EPA
considers the appropriate PM2.5 designation for the Reservation as
attainment/unclassifiable.  

San Carlos Apache:

The Tribe encourages the EPA to take enforcement actions which will
contain and reduce fugitive emissions and particulate point sources on
the Federal, State and private lands in the region which surrounds the
Reservation.  

The Tribe asks the EPA to take aggressive measures necessary to control
particulate emissions, such as requiring the U.S.  Forest Service,
Bureau of Land management, State Land Department, and Gila, Graham and
Pinal Counties to control industrial pollution, off road vehicle use,
and the dust from unpaved roads in their areas of responsibility.

EPA Response:

EPA wants to ensure that air quality for tribal residents is healthy and
will take action to address sources that are not complying with
applicable laws.  EPA hopes the San Carlos Apache Tribe will bring those
sources to our attention.  However, the designation process does not
address individual sources.  The designation process identifies those
areas of the country that meet or do not meet the PM2.5 standard for
small particulates.  

EPA requires each State to have a representative monitoring network for
the criteria pollutants of concern.  These monitoring networks measure
pollution in the air and that data is recorded in EPA’s AQS.  A
network usually has many monitors placed throughout a State and EPA
accepts the monitoring data from the network as being representative of
the State’s air quality, provided the network meets EPA requirements. 
The Arizona Department of Environmental Quality (ADEQ) has an
EPA-approved monitoring network and EPA views the data from that network
as a reliable indicator of the air quality in the State of Arizona.

EPA also considers air quality data from monitors that the tribes
operate on their reservations.  It is EPA’s understanding that are no
such air monitors on the San Carlos Apache Reservation.  Consequently,
EPA relied on air quality data from the nearest State monitors to
determine whether there was a PM2.5 problem in the area of the San
Carlos Apache Reservation.  There is no data indicating that a violation
of the PM2.5 standard is occurring anywhere in the State of Arizona
except in Nogales, Arizona.  As a result, EPA made a designation of
attainment/unclassifiable for the San Carlos Apache and all the other
Tribes in Arizona because there is no scientific evidence that there is
a PM2.5 problem on tribal lands.  

However, it is possible that activities are causing PM10 issues on the
San Carlos Apache Reservation.  Several of the activities mentioned in
the letter are associated with dust issues throughout the State, and we
can certainly work with the tribe on those issues.

San Carlos Apache:

The Tribe further requests that the EPA cause Federal agencies and the
ADEQ to refrain from issuing air quality permits and other permits which
will further degrade the air quality of the Reservation.

EPA Response:

The San Carlos Apache comment letter raised the issue of permits being
issued by ADEQ that can further degrade air quality.  The permits issued
by ADEQ are designed to impose controls on the sources that hold the
permit, which ensures that the source cannot degrade air quality to
levels that would harm public health.  The Tribe has complained that
they were not informed about permits that were pending prior to
issuance.  EPA contacted the ADEQ Air Quality Division Director and the
ADEQ Tribal Liaison to discuss the complaint and to request that the San
Carlos Apache be notified of any potential permit actions that might
affect the Tribe.  ADEQ committed to doing so.

San Carlos Apache:

As part of the consultation process, the Tribe requests that EPA make
available to the Tribe all data which was reviewed in the process of
reviewing sources of air pollution in the area surrounding the
reservation and all documentation of EPA’s decision making process.  

The Tribe also requests consultation with EPA and for EPA to develop a
comprehensive plan to bring activities which in surrounding areas
contribute to poor air quality within the Reservation, and all
documentation of your decision making process.   

EPA Response:

EPA can provide the analysis submitted by the State of Arizona which
shows the results of the State monitoring network.  The only area in
Arizona that is being designated as nonattainment for the PM2.5 standard
is the city of Nogales.  EPA has prepared a TSD which provides technical
information on the reasons for the designation of Nogales as
nonattainment, which EPA can be found in the docket for the PM2.5
designations at www.regulations.gov; the EPA docket number for 24-hour
PM2.5 designations is:  EPA-HQ-OAR-2007-0562

Youngstown, OH

Ohio:

Ohio EPA’s initial recommendation to EPA in December 2007 recommended
Mahoning and Trumbull Counties as nonattainment for the new 24-hour
PM2.5 standard.  However, since that time air quality has improved in
the area.”

EPA Response:

hˆ

|

ß

Ù

 

 

 

6

7

8

9

Y

Z

[

u

u

v

w

y

z

{

|

}

~

š

›

œ

¼

½

¾

Ø

Ù

Ú

Ü

Ý

Þ

ß

à

á

ý

ý

þ

ÿ

 

 

 

8

8

9

:

;

^

_

`

z

{

|

~



€

‚

ƒ

Ÿ

 

¡

¢

j¸

j;

 ¢

¶

·

¸

Ò

Ó

Ô

Ö

×

Ø

Ù

Ú

Û

÷

ø

ù

ú

j¬

j/

j

摧琈

t

t

h

h

h

hƒ

hƒ

hƒ

hƒ

ç

h

h

h¾O

h¾O

h¾O

«

¬

í

ð

ª

«

h¾O

摧ࢩ

hï

hƒ

hï

hï

hï

h-

hï

h

h

hˆ

J

m

n

h—

옍)

㜀$␸䠀$摧ӭr

摧ӭr

㜀$␸䠀$摧ӭr

h

㜀$␸䠀$摧ᅒ»	㜀$␸䠀$摧ӭr

hs

h—

h—

h

h

h

h»

h—

h»

h»

h»

摧噞©

h—

h—

hÒ

h—

h—

hÒ

h—

hÒ

h—

h

h

h

h

	



m

p

£

¦

¯

¯

Á

Â

2

5

r

t

¹

º

ß

â

ð

ó

ù

û

h—

h—

h—

h—

h—

옍)

h—

ༀ梄ᄁ预㟾$␸䠀$葞Ũ葠ﺘ摧䈇/	㜀$␸䠀$摧ᅒ»
㜀$␸䠀$摧஗Á

h—

S

T

0

1

‚

ƒ

摧氓ª

O

R

T

©

ñ

1

:

r

t

w

x

y

‚

ƒ

h²

 ƒ

ï

ð

h—

h

h²

h—

h²

h²

h²

h—

h—

h—

h—

h—

h—

摧㔵£	㜀$␸䠀$摧஗Á	㜀$␸䠀$摧、I

h—

h—

h—

h

h—

h—

h—

h—

h—

h—

h—

h—

h—

ó

þ

h—

	ÿ

h—

h—

h—

h—

h—

h—

h—

h—

h—

h—

摧灅c

h—

h—

h—

摧䋬a	㜀$␸䠀$摧஗Á

h—

h—

옍)

«

¬

옍)

Ò

Ó

ÿ

h—

㜀$␸䠀$摧糛0	㜀$␸䠀$摧஗Áᘀ

@

@

@

@

@

@

@

@

옍#

옍#

옍#

옍#

}

‰

h

h

h

h

h

h

h

h

옍)

摧ڠMጀth of the Youngstown area counties had 2004-2006 design values
above the standard.  Ohio pointed out the air quality is improving in
the Youngstown area and throughout the State.  The most recent design
values for 2005-2007 show that Trumbull County now meets the standards
and that Mahoning County is now just above the standard.  Based on the
most recent information, EPA is designating Mahoning and Trumbull
Counties as nonattainment.  

However, EPA is inviting States to submit complete, quality assured,
certified 2008 air quality data to EPA earlier than the usual June 30,
2009 deadline.  This early submittal deadline will be approximately 45
days prior to the 90 days after publication effective date of the final
24-hour PM2.5 designations.  If EPA agrees that a change of designation
status is appropriate based on 2006-2008 air quality data, then EPA will
withdraw the nonattainment designation prior to the effective date of
final 24-hour PM2.5 designations and the area would be designated as in
attainment.  EPA will make such a change only when the 2008 establishes
that there is no violation in the area; EPA will not revisit the
designated boundaries of nonattainment areas that are still violating
the NAAQS.

Yuba City-Marysville, CA

CARB:

The only violating monitor in the Feather River Air Quality Management
District is located in Yuba City, the largest urban area in Sutter
County, home to over 65 percent of the County’s population.  18
percent of Yuba County’s residents live in Marysville, located in Yuba
County but sharing a border with Yuba City.  Combined, the two cities
account for 44 percent of the population of the two counties.  Based on
the localized nature of the primary emission contribution to winter
PM2.5, CARB considers the combined urban area of Yuba City/Marysville an
appropriate nonattainment boundary for PM2.5. 

Almost 55 percent of the PM2.5 on exceedance days in Yuba City is
composed of carbon, primarily from residential wood combustion.  The low
wind speeds exhibited during times of PM2.5 exceedances, as noted in the
pollution wind rose on page 16 of the EPA response, only reinforces the
exceedances as resulting from localized sources such as residential wood
burning.

 

	EPA Response:

EPA designation of the PM2.5 nonattainment area for Yuba City and
Marysville area includes all of Sutter County and part of Yuba County
with a boundary drawn to the eastern ridge line (based on inversion
layer and topography) that fully captures the City of Marysville and the
surrounding populations.

Violation of the PM2.5 standard for the 2005-2007 period occurred at the
monitor located in Yuba City (design value of 40 ug/m3).  Data provided
in CARB’s October 15 letter indicates that 54 percent of the PM2.5 is
carbon and 38 percent is nitrates, with crustal and sulfates making up
the remaining 8 percent.  Again, this data suggests that residential
wood burning and mobile sources are the most important sources
contributing to PM2.5 exceedances, particularly during the winter
season.  Consistent with the rationale used in other areas, the PM2.5
nonattainment area for Yuba and Sutter is intended to capture the full
extent of urbanization associated with Yuba City and Marysville, as well
as the major transportation corridors.  This boundary represents 100
percent of Sutter County and approximately 75 percent of Yuba County. 
As a partial county designation, the nonattainment area in Yuba County
includes approximately 93 percent of the population, approximately 99
percent of the non-truck traffic and approximately 97 percent of the
truck traffic.  

Topography was also an important consideration in developing the
boundaries for Sutter and Yuba County.  That is, given Sutter County is
entirely within the Sacramento Valley below the approximate boundary of
the winter inversion layer, there is the potential for pollution levels
measured in Yuba City to represent a larger area.  Given the mostly flat
terrain, there was not a reason to limit the designation to a partial
county.  This rationale was also applied to Yuba County, where the
Sierra Nevada Mountains on the eastern side of the County, combined with
the winter inversion layer, was seen as physical boundary.  Considering
this physical boundary, which also resembles the California Air Basin
boundary for the Sacramento valley, EPA defined the eastern boundary
within Yuba County with township and range lines to capture this
topographical boundary, as well as the major population centers and
transportation networks within the County.

 See the regulations on the revised standard at 40 CFR 50.13(c); 71 FR
61224, October 17, 2006.  See also monitoring regulations at 40 CFR Part
58, as revised on October 17, 2006 (see 71 FR 61236); and procedures for
using these data to determine whether a violation has occurred in 40 CFR
Part 50 Appendix N, as revised on October 17, 2006 (see 71 FR 61144).

  Draxler, R.R.  and Rolph, G.D., 2003.  HYSPLIT (HYbrid Single-Particle
Lagrangian Integrated Trajectory) Model access via NOAA ARL READY
Website (http://www.arl.noaa.gov/ready/hysplit4.html).  NOAA Air
Resources Laboratory, Silver Spring, MD.

 Exceptional Event Analysis, Sonoma Technology Inc., Sep 30, 2008, 

 Draxler, R.R.  and Rolph, G.D., 2003.  HYSPLIT (HYbrid Single-Particle
Lagrangian Integrated Trajectory) Model access via NOAA ARL READY
Website (http://www.arl.noaa.gov/ready/hysplit4.html).  NOAA Air
Resources Laboratory, Silver Spring, MD.

Page   PAGE  3  of   NUMPAGES  194 

g/m3          

Afternoon Winds,  10-15 mph             

Lindon 74.5 g/m3          

West Valley 80.6 g/m3          

71.4 g/m3          

57.4 g/m3          

65.1 g/m3          

Hawthorne 86.6 g/m3          

Cottonwood 85.0 g/m3          

Alpine Met Station AM Winds, 5 mph               

Point of the Mountain

Dry Creek Canyon

Rose Canyon

5.8 miles

7.3 miles

Alpine

Bluffdale

HYSPLIT for the Period:  January 26 Salt Lake City air mass generally
originates to the south, with very low velocities and variable
directions mid-day

Met Data: January wind for the day are mostly southerly from noon the
previous day, but with some local southerly flow in the morning of Jan. 
22.

Conclusion:  HYSPLIT hypothesis (Utah County emissions were transported
to SLC) is well supported by surface meteorology measurements

Salmon Box indicates winds with a southerly component (from SW, S or SE)

Blue Box indicates winds with a northerly component (from NW, N or NE)

DEQ Met Stations

S

g/m3			49.3 g/m3		40.2 g/m3

Northerly PM

Winds

Local S winds

Before noon

HYSPLIT for the Period:  January 22 Salt Lake City air mass originates
to the north

Met Data: January wind for the day are mostly northerly from noon the
previous day, but with some local southerly flow in the morning of Jan.
22.

Conclusion:  HYSPLIT hypothesis (Utah County emissions were transported
to SLC) is well supported by surface meteorology measurements

 háQ

⼀g/m3			56.8 g/m3		46.3 g/m3

Morning southerly

 periods

HYSPLIT for the Period:  January 13 Salt Lake City air mass originates
to the south

Met Data: January 13 winds for the day are southerly from before
midnight to approximately noon

Conclusion:  HYSPLIT hypothesis (Utah County emissions were transported
to SLC) is supported by surface meteorology measurements

Salmon Box indicates winds with a southerly component (from SW, S or SE)

Blue Box indicates winds with a northerly component (from NW, N or NE)

DEQ Met Stations

S

N		1/11/04			1/12/04		1/13/04

SLC PM2.5g/m3		48.2 g/m3			63.9 g/m3

