UNITED STATES ENVIRONMENTAL PROTECTION AGENCY

REGIONS III, IV, and V

Technical Support Document  

EPA’s Notice of Proposed Rulemaking 

For 

Approval and Promulgation of Air Quality Plans; Determination of
Attainment of the 1997 Fine Particulate Standard

January 2011

                                         

  SEQ CHAPTER \h \r 1 Technical Support Document

INTRODUCTION……………………………………………………
………………...…........3

CLEAN AIR ACT REQUIREMENTS FOR FINE PARTICULATE (PM2.5) ATTAINMENT.3

PM2.5 Attainment Requirements
…………………………………………………….………3

Determination of Attainment
……………………………………………………….….....
....4

MONITORING NETWORK
………………………………………….………………..…
……5

LAWRENCE COUNTY HOSPITAL MONITORING SITE
.……………………..…………..6

EPA’S ANALYSIS
……………………………….………………………..……
……………..6

	A.  Analysis of Air Quality
Data……………………………………………..……………
……6 

B.  Addressing Missing
Data……………………………………………………..……
………..6

C.  Additional
Analysis………………………………………………………
………..……….14

D.  Air Quality in
2010…………………………………………………………
………...…….14

CONCLUSION……………………………………………………
………………..………….14

APPENDIX A Annual Means for the Huntington-Ashland Area PM2.5 FRM
monitors……...………..16 

I.	INTRODUCTION  TC \l1 "I.	INTRODUCTION 

	On April 1, 2010, Ohio requested that EPA determine that multiple
areas, including the Huntington-Ashland Area, are attaining the 1997
annual fine particulate (PM2.5) National Ambient Air Quality Standard
(NAAQS).  In addition, pursuant to 179(c)(1) of the Clean Air Act, EPA
is required to make a determination about the status of this area as of
its attainment date.  This technical support document specifically
addresses the Huntington-Ashland PM2.5 nonattainment area, including
portions in West Virginia, Kentucky, and Ohio (hereafter referred to as
“the Huntington-Ashland Area” or “the Area”).   In this
technical support document, EPA provides information supporting: the
proposed determination that the Huntington-Ashland Area has attained the
1997 annual PM2.5 NAAQS (hereafter referred to as “the annual PM2.5
NAAQS” or “the standard”) and that the area has attained the 1997
annual PM2.5 NAAQS by its applicable attainment date of April 5, 2010.  
         

	In this technical support document, EPA:

summarizes the statutory and policy requirements for PM2.5 nonattainment
area SIPs and EPA’s policy concerning attainment.

provides a description of the air quality monitoring network for the
Huntington-Ashland Area PM2.5 nonattainment area

provides a description of the air quality monitoring network for the
Huntington-Ashland Area 

describes the Lawrence County Hospital (LCH) monitoring site, and
provides an explanation for monitor shutdown;

describes our detailed analysis of our proposed determination of
attainment, including our analysis of the air quality data, and the
methodology used for substituting for missing data; 

describes air quality trends in the Huntington-Ashland Area 

summarizes our conclusions on the proposed determination of attainment
with respect to EPA requirements.       

II.	CLEAN AIR ACT REQUIREMENTS FOR FINE PARTICLE (PM 2.5) ATTAINMENT  
TC \l1 "II.	DESCRIPTION OF NOx RACT REQUIREMENTS AND NEW JERSEY'S NOx
RACT REGULATION 

A.	PM2.5 Attainment Requirements 

  SEQ CHAPTER \h \r 1 In 1997, EPA established the health-based NAAQS
for PM2.5, setting an annual standard at 15.0 micrograms per cubic meter
(ug/m3) based on a 3-year average of annual mean PM2.5 concentrations,
and a 24-hour standard of 65 ug/m3 based on a 3-year average of the 98th
percentile of 24-hour concentrations.  EPA established the standards
based on significant evidence and numerous health studies demonstrating
that serious health effects are associated with exposures to particulate
matter.  

	On January 5, 2005, (70 FR 944) EPA finalized its air quality
designations and classifications across the country with respect to the
1997 PM2.5 standard.  These designations became effective on April 5,
2005.  The Huntington-Ashland Area was designated nonattainment for the
1997 annual PM2.5 NAAQS (See 40 CFR 81.349 (West Virginia), 40 CFR
81.318 (Kentucky), and 40 CFR 81.336 (Ohio).   

These designations triggered the Clean Air Act’s requirements under
section 172(c), which requires the submission of an attainment plan for
each designated nonattainment area.  EPA’s PM2.5 implementation rule,
published on April 25, 2007 (72 Fed. Reg. 20586) specifies that states
must submit attainment plans for their nonattainment areas to the EPA by
no later than three years from the effective date of designation, that
is, by April 5, 2008.  EPA’s PM2.5 implementation rule outlines the
SIP requirements for the attainment plan, which includes among other
things, the submission of an attainment demonstration showing the
air-quality improvements expected to result from national and local
control measures, an analysis of reasonably available control measures
(RACM), including all reasonably available control technology (RACT),
reasonable further progress plan, and contingency measures.  EPA’s
implementation rule also sets the attainment date in these areas as
April 5, 2010, and the determination would be based on data from
2007-2009 (72 FR 20600).

The states of West Virginia, Kentucky, and Ohio, are subject to these
requirements since several counties in the states are located in a PM2.5
nonattainment area.  The Huntington-Ashland Area is composed of some or
all of the following counties in West Virginia, Kentucky, and Ohio: 
Cabell, Mason (part) and Wayne Counties in West Virginia; Boyd and
Lawrence (part) Counties in Kentucky; and Adams (part), Gallia (part),
Lawrence, and Scioto Counties in Ohio.  

B.	Determination of Attainment

Under EPA regulations at 40 CFR 50.7, the annual primary and secondary
PM2.5 standards are met when the annual arithmetic mean concentration,
as determined in accordance with 40 CFR part 50, Appendix N, is less
than or equal to 15.0 ug/m3.

	Data handling conventions and computations necessary for determining
whether areas have met the PM2.5 NAAQS, including requirements for data
completeness, are listed in Appendix N of 40 CFR Part 50.  A year meets
data completeness requirements when at least 75 percent of the scheduled
sampling days for each quarter have valid data.  The use of less than
complete data is subject to the approval of EPA, which may consider
factors such as monitoring site closures/moves, monitoring diligence,
and nearby concentrations in determining whether to use such data.  

	If this proposed determination of attainment is made final, EPA will
have met its statutory obligation pursuant to section 179(c)(1) of the
Clean Air Act (CAA) to make a determination based on the Area’s air
quality data as of the attainment date whether the area attained the
standard by that date.  Additionally, as a result of such determination,
the requirements for Kentucky, Ohio or West Virginia to submit an
attainment demonstration and associated RACM, a RFP plan, contingency
measures, and any other planning SIPs related to attainment of the 1997
annual PM2.5 NAAQS within the Huntington-Ashland Area would be suspended
until such time as: a) the area is redesignated to attainment, at which
time the requirements would no longer apply; or, b) EPA determines that
the area has violated the annual PM2.5 NAAQS, at which time, the area
would be required to submit such plans.  See 40 CFR 51.1004(c).

	A determination of attainment does not shield an area from other
required actions such as New Source Review (NSR), conformity and
emission inventory requirements, and provisions to address pollution
transport, which would require emission reductions at sources or other
types of emission activities contributing significantly to nonattainment
in other areas or states, or interfering with maintenance in those
areas.

Determinations as to whether individual areas have attained the PM2.5
standards and thus qualify for the application of a determination of
attainment are being made in the context of rulemakings for individual
areas.  This Technical Support Document (TSD) provides an explanation of
EPA’s analysis used to determine whether the Huntington-Ashland Area
has attained the annual PM2.5 standard by the applicable attainment
date.    

 

MONITORING NETWORK 

The Huntington-Ashland Area currently has three air monitoring
locations; one each in West Virginia, Kentucky, and Ohio.  West Virginia
also has a collocated monitor in place at the site for quality assurance
purposes.  The primary monitor, and not the collocated monitor, is used
to determine compliance with the PM2.5 NAAQS.  Since the collocated
quality assurance monitor takes many fewer readings than the primary
monitor, its average annual values may be unrepresentatively high even
if only a few high values are recorded. (See 40 CFR Part 50, Appendix N,
3(d)(1).  Two monitoring locations for Ohio are listed in Table 1 due to
the LCH monitor shut-down and Ironton DOT start-up during the 2007-2009
time period.  Table 1 shows the monitor design values (DVs) (i.e., the
3-year average of annual mean PM2.5 concentrations) for the years
2007-2009.  The area DV is the highest annual mean measured for a
federal reference method (FRM) monitor that had complete data in a
county.  

Table 1.— 2007-2009 Design Values for the 1997 Annual PM2.5 NAAQS for
the Huntington-Ashland Area Monitors in Micrograms per Cubic Meter
(ug/m3).

Site Name	County	Site Number	Annual Average Concentration (μg/m3)

Huntington	Cabell	54-011-0006	14.3

Ashland Primary (FIVCO)	Boyd	21-019-0017	12.4

Lawrence County Hospital	Lawrence	39-087-0010	13.3

Ironton DOT	Lawrence	39-087-0012	12.2



EPA meets annually with each state to discuss any problems or issues
concerning states’ air monitoring data and/or network.  In addition,
EPA and the states communicate many times during the year so that issues
can be addressed as they show up.  EPA regulations require states to
submit annual network plans to their respective Regions.  These plans
outline the current networks and any proposed changes in the upcoming 18
months.  Regions 3, 4, and 5 have been able to approve these plans due
to the quality of the West Virginia, Kentucky, and Ohio monitoring
networks.  

LAWRENCE COUNTY HOSPITAL MONITORING SITE 

The LCH monitor operated until February 11, 2008, when monitoring was
discontinued permanently at the site due to the building demolition.  
Because the monitor at this location could not be replaced, a new site
in this county was established at the Ironton DOT starting when
monitoring at the LCH ended.  As a result, the Ironton DOT monitor did
not obtain any data for 2007 and did not obtain complete data for the
first quarter of 2008.  Nevertheless, the available data for the
remainder of 2008 and all of 2009 indicate attainment at this monitor.  

	As discussed in greater detail below, the monitors in Lawrence County
were the only monitors in the Huntington-Ashland Area without valid
attaining DVs for 2007-2009 due to incomplete data.    This monitor
historically measured lower concentrations than other monitors in the
Area, including, for example, the monitor in Cabell County, West
Virginia. 

EPA’S ANALYSIS 

Analysis of Air Quality Data 

	EPA has reviewed the ambient air monitoring data for PM2.5, consistent
with the requirements contained in 40 CFR part 50 and recorded in the
EPA Air Quality System database for the Huntington-Ashland Area from
2007 through the present time.  Currently all monitors are measuring
concentrations averaging well below the annual standard of 15.0 ug/m3. 
The highest DV for the most recent 3 years, 2007-2009, for the
Huntington-Ashland Area is 14.3 ug/m3, based on monitoring data
collected in West Virginia at the Huntington monitor.  All 3-year
averages of the annual mean PM2.5 concentrations are below the NAAQS
from 2007-2009.  

	In order to determine whether the NAAQS has been attained for the LCH
monitor, EPA has applied a recently developed procedure to address the
missing data, and determine valid DVs for the monitor for the annual
standard for 2007-2009.  EPA’s procedure is described in the following
section.

Addressing Missing Data

Additional analysis was necessary for the LCH monitor (site number
39-087-0010) because this monitor is no longer operating (due to
building demolition) but recorded a violation in the most recent three
year period in which it collected complete data.  An objective
statistical method was used to determine if this monitor would likely
have met the NAAQS in subsequent time periods.  

The general concept behind the statistical method is to establish a
linear regression relationship between the monitor of interest and
another “comparison monitor” in the nonattainment area based on data
obtained when the monitor of interest was still operating.  The method
then estimates concentrations at the monitor of interest for the period
after monitor shutdown by applying a linear regression to the more
recently observed concentrations at the other monitor.  The results are
checked with an additional statistical technique, known as
bootstrapping, to verify if the conclusion of attainment is correct.  

The monitor of interest in this analysis is the LCH monitor.  This site
ceased monitoring in February 2008 and had an incomplete DV of 12.9
µg/m3 for 2007-2008.  The goal of this methodology is to appropriately
fill in the missing data from the quarters 1-4 of 2008, and quarters 1-4
of 2009 to judge whether this location would likely have monitored
attainment of the NAAQS had the monitor continued to operate.  The
Huntington-Ashland Area has two monitoring locations with complete data
for 2007-2009, listed in Table 1 above, both of which are similar enough
to warrant comparison to the LCH monitor.

Criteria to be met

EPA established criteria to assure that the DV for the LCH monitor in
the Huntington-Ashland Area, as estimated by the method, was robust
enough to allow consideration of the results in a weight of evidence
decision.  EPA established criteria both for the operating period of the
comparison monitor and for the number of days of samples that the
comparison monitor and the LCH monitor have in common.  EPA has not
established bright lines or specific requirements for determining DVs
for monitors with less than complete data.  

Operating period

In order to provide satisfactory concentration estimates for the LCH
monitor for the period after the monitor shut down, comparison monitors
in the Huntington-Ashland Area would need to have operated from well
before from February 2008 to the end of 2009.  Operation for a
substantial period before February 2008 is necessary to support a robust
regression relationship with concentrations at the LCH monitor, and
operation from February 2008 to the end of 2009 is necessary to support
concentration estimates for this period.  

Paired samples

Pairs of sample data for the LCH monitor and comparison monitors from
the most recent 20-quarter period of operation of the LCH monitor were
used.  Comparison monitors were required to have at least 100 historical
sample pairs in common with the target monitor during this period.  When
there were more than 100 pairs in the 20-quarter period, all available
pairs were used.  In addition to the 100 pairs over the 20-quarters,
comparison monitors were required to have at least 20 paired days of
data in common with the LCH monitor for each quarter (adding together
the number of days in a given calendar quarter [Q1, Q2, Q3, or Q4]
across the various years in the 20-quarter period).

Both comparison monitors, Ashland Primary and Huntington, met the
eligibility criteria of 100 pairs over the 20-quarters, and at least 20
paired days of data in common with the LCH monitor for each calendar
quarter. 

Method for determining DVs

Step 1 – Organize data and perform linear regression(s) between the
LCH monitor and the comparison monitors.

For both the Ashland Primary and the Huntington monitors, all the
available paired data within the 5-year period described above were used
to develop a linear regression expressing concentrations at the LCH
monitor in relation to concentrations at the comparison monitor.  

Step 2 – Select candidate monitor

The Huntington-Ashland Area has 2 monitors which could have been used as
comparison monitors, Ashland Primary and the Huntington.  The following
steps were used to select the comparison monitor.  

1) Comparison monitors were placed into order of highest correlation
(highest r-squared in the linear regression) of concentrations with
concentrations at the LCH monitor.  The comparison monitor showing the
highest correlation of concentrations with concentrations at the LCH
monitor was used first.

2) Using the regression equation relating concentrations at the LCH
monitor and the comparison monitor, estimates of the 24-hour average
concentrations at the LCH monitor were made for each day after this
monitor was shut down that the monitor otherwise would have been
scheduled to measure concentrations, except that no estimate was made if
the comparison monitor had no valid measurement.  The estimates were
combined with any reported data at the LCH monitor in the recent 3-year
period and the data capture rates were computed for each originally
incomplete quarter.  

3) If performing the process in #2 results in all 12 quarters at the LCH
monitor having at least 75 percent of scheduled values, the comparison
monitor was selected for use in the remaining steps. If 75 percent of
the scheduled values for the LCH monitor were not obtained, the steps in
#2 were repeated for the next best correlated comparison monitor.  This
process was repeated until the LCH monitor had at least 75 percent of
the number of values expected under its original operating schedule.

Step 3 – DV Computed

The 3-year DV was calculated using the actual observations at the LCH
monitor and the regression-based estimates of daily concentrations.  The
resulting DV for the LCH monitor was 13.3 µg/m3, which is below the
annual PM2.5 NAAQS.

Step 4 –Statistical confidence of a DV below the NAAQS checked

In order to show that there is a minimal probability that the actual DV
(had the LCH monitor operated more days) was above the annual PM2.5
NAAQS, further analysis was performed as follows:

(a) 	Residuals were calculated from the regression.

(b) 	A bootstrap analysis was performed, repeating the regression-based
substitution in Step 2 but adding to each estimated 24-hour value that
substitutes for a missing scheduled monitored value a plus or minus
residual that has been randomly drawn (with replacement) from the pool
of regression residuals.  This step randomly applies real residuals from
the linear regression to the imputed current-period target values.

(c)  	Repeat (b) for a total of 1000 runs.  

(d)  	For each of the 1000 bootstrap trials, the 3-year DV was
calculated.  This provides a pseudo confidence interval for the
partially imputed current-period DV for the target.

(e) 	Since none of the 1000 runs resulted in a DV above the annual PM2.5
NAAQS, the LCH monitor was considered to have met the annual PM2.5
NAAQS.   

 

Determining the DV for Lawrence County Hospital (AQS ID: 39-087-0010)

The analysis described above was performed to estimate concentrations at
the LCH monitor.  The results of the regressions can be found in Table
2.  The LCH monitor is identified in this table as Site A.  The
comparison monitors in this table are all identified in this table as
Site B.  The Ashland Primary monitor (AQS ID: 21-019-0017) had the
highest correlation with the LCH monitor and met the operating period
and paired data eligibility criteria for being the comparison monitor. 
The results of the regression between the LCH monitor and the Ashland
Primary monitor are shown in Figure 1.  Estimates for the missing data
for quarters 1 thru 4 of 2008 and for quarters 1 thru 4 of 2009 for the
LCH monitor were made.  These quarters were the quarters with less than
75 percent data capture.  As a result of filling in the missing data,
the LCH monitor now had 32, 30, 31, 30 values for quarters 1 thru 4 for
2008, respectively, and 30, 30, 27, 31 values for quarters 1 thru 4 of
2009, respectively.  A DV of 13.3 (g/m3 was calculated for the LCH
monitor when the missing data was filled in, as shown in Figure 2.

The statistical confidence of the DV was checked using the method
outlined in Step 4 above.  The results of the 1000 bootstrap trials
provided the following results:  243 DVs) of 13.2 µg/m3, 223 DVs of
13.3 µg/m3, 177 DVs of 13.1 µg/m3, 157 DVs of 13.4 µg/m3, 81 DVs of
13.0 µg/m3, 61 DVs of 13.5 µg/m3, 21 DVs of 13.6 µg/m3, 19 DVs of
12.9 µg/m3, 8 DVs of 13.7 µg/m3, 6 DVs of 12.8 µg/m3, 2 DVs of 12.7
µg/m3, and 2 DVs of 13.8 µg/m3.  None of the 1000 bootstrapping trials
resulted in a violating DV, which provides statistical confidence that
the air quality at the LCH monitor is below the annual PM2.5 NAAQS.

 

 

	

 C. 	Additional Analysis

	Additional statistical analysis was performed for the LCH monitor using
a different comparison monitor.  EPA re-ran the analysis using data from
the Huntington monitor (site number 54-011-0006), a monitor that is
similarly correlated with the LCH monitor as the Ashland Primary
monitor.  The results of the analysis using the Huntington monitor as
the comparison monitor produced an estimated 2007-2009 DV at the LCH of
13.1 µg/m3, which is also below the NAAQS of 15.0 ug/m3.  That is, the
concentration estimates provided by the method described herein indicate
a DV at LCH that is below the annual PM2.5 NAAQS, regardless of which
area monitor is used as the basis for the estimates. 

D.         Air Quality in 2010

	The states have submitted monitoring data for 2010.  EPA examined 2008
to 2010 data as an indication of whether the area continues to attain
the standard.  Table 3 shows the average concentration at each of the
sites in the Huntington-Ashland Area.  All sites have average
concentrations well below the standard, with average values ranging from
12.2 to 13.1 µg/m3.  Therefore, EPA believes that 2008 to 2010 data,
interpreted in conjunction with data from 2007 to 2009, supports a
determination that the Huntington-Ashland Area is attaining the annual
PM2.5 NAAQS.

Table 3.— 20008- 2010 annual average PM2.5 concentrations for the
Huntington-Ashland Area monitors in micrograms per cubic meter (ug/m3).

Site Name	County	Site Number	Annual Average Concentration (μg/m3)

Huntington	Cabell	54-011-0006	13.1

Ashland Primary (FIVCO)	Boyd	21-019-0017	11.4

Ironton DOT	Lawrence	39-087-0012	12.2



CONCLUSION

	EPA is proposing to determine that the Huntington-Ashland Area has
attained the annual 1997 PM2.5 NAAQS by its attainment date of April 5,
2010.

	EPA has reviewed the ambient air monitoring data for PM2.5, consistent
with the requirements contained in 40 CFR part 50 and recorded in the
EPA Air Quality System database for the Huntington-Ashland Area.  The
3-year averages of the annual mean PM2.5 concentrations continue to be
below the NAAQS of 15.0 ug/m3.  

	EPA has concluded that use of less than complete data from the LCH
monitoring site is warranted for determining attainment of the NAAQS.
The States of West Virginia, Kentucky, and Ohio have been diligent in
the number and placement of PM2.5 monitors in the nonattainment area. 
The monitor at LCH was shut down due to a planned building demolition
and could not be replaced.  In light of such facts, EPA employed a
procedure using a statistical method to estimate concentrations at the
LCH monitor for the remainder of the time period at issue; and, the
resulting analysis supports a finding of attainment of the annual PM2.5
NAAQS at this location.

	As provided in 40 CFR 51.1004(c), if EPA finalizes this determination,
it would suspend the requirements for this area to submit attainment
demonstrations, reasonably available control measures, reasonable
further progress plans, and contingency measures related to attainment
of the 1997 annual PM2.5 NAAQS so long as the Area continues to attain
the 1997 annual PM2.5 NAAQS.

Appendix A

Annual Means for the Huntington-Ashland Area PM2.5 FRM monitors

 

 West Virginia has a collocated monitor in place at the same site for
quality assurance purposes.  The primary monitor, and not the collocated
monitor, is used to determine compliance with the PM2.5 NAAQS.  Since
the collocated monitor takes fewer readings than the primary monitor,
its average annual values may be unrepresentatively high. See 40 CFR
Part 50, Appendix N, 3(d)(1).

 Data from the Lawrence County Hospital and Ironton DOT sites were
incomplete due to shut-down of the Lawrence County Hospital site and
startup of the Ironton DOT site in February 2008.

 The Lawrence County Hospital Site was shut-down in February 2008.  The
Ironton DOT site began operation in 2008 after the Lawrence County
Hospital Site ceased monitoring.

 West Virginia has a collocated monitor in place at the same site for
quality assurance purposes.  The primary monitor, and not the collocated
monitor, is used to determine compliance with the PM2.5 NAAQS.  Since
the collocated monitor takes fewer readings than the primary monitor,
its average annual values may be unrepresentatively high. See 40 CFR
Part 50, Appendix N, 3(d)(1).

 PAGE   14 

 PAGE   15 

഍഍഍䘍杩牵⁥⸲†敒畳瑬⁳潦⁲慌牷湥散䌠畯瑮⁹潈
灳瑩污猠瑩⁥晡整⁲楦汬湩⁧湩洠獩楳杮搠瑡⁡潦⁲污
⁬畱牡整獲漠⁦〲㠰‬湡⁤污⁬畱牡整獲漠⁦〲㤰഍

$

 

;

K

N

U

V

s

†

‰

‘

©

À

Ã

Ç

È

É

Ê

ø

þ

葠﯈摧ぇK

:

Q

R

S

U

Z

]

̀Ĥ␆ሁā愀Ĥ摧␢ê

฀㟽$␸䠀$葝﷠摧朤i

hY#

hY#

฀ᇽ킄㄂$葝﷠葠ː摧⍙

 hP

萑ː葠ː摧ᡒ

㄀$摧ᇡÆ

摧⓸6

hô

hô

hô

Figure 1 Concentrations at Lawrence County Hospital (site 390870010,
labeled “v_a”) compared to concentrations at Ashland (site
210190017, labeled “v_b”).

The linear relationship between the two monitors is: Concentration at
Lawrence County Hospital = 2.293135 + 0.882184 * concentration at
Ashland.  The correlation coefficient was 0.86384.

Table 2. Correlation and data capture for monitors used to determine a
DV for Lawrence County Hospital site.

