«J
UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
RESEARCH TRIANGLE PARK, NC 27711
MAY2020H
OFFICE OF
AIR QUALITY PLANNING
AND STANDARDS
MEMORANDUM
SUBJECT: Guidance for PM2 5 Permit Modeling
FROM: Stephen D.
Director
TO: Regional Air Division Directors, Regions 1-10
This memorandum and attachment, titled "Guidance for PM2 5 Permit Modeling," provides
guidance on demonstrating compliance with the fine particulate matter (PM2 5) National Ambient
Air Quality Standards (NAAQS) and Prevention of Significant Deterioration (PSD) increments,
especially with regard to considerations of the secondarily formed component of PMi 5. This
document reflects the EPA's recommendations for how a major stationary source seeking a PSD
permit may demonstrate that it will not cause or contribute to a violation of the NAAQS and
PSD increments for PM2 5, as required under section 165(a)(3) of the Clean Air Act (CAA) and
40 CFR Sections 51.166(k) and 52.21(k).
A draft version of this guidance document was provided to the public on March 4, 2013, for a
90-day comment period. The document was revised in response to public comments and
additional information provided through on-going interactions with various stakeholders.
Noteworthy changes made to the draft version include:
Clarifications throughout with respect to procedures for adequately addressing primary
and secondarily formed PMi.s.
Inclusion of an example hybrid (qualitative/quantitative) secondary PMi.5 impact
assessment.
Revision of a second tier cumulative PMa.s NAAQS compliance approach.
Revision of Section V and other sections relative to PSD increment for PM2.5.
Please distribute the attached guidance document to state, local, and tribal governments, as
appropriate. If you have any questions regarding this document, please contact Tyler Fox, Air
Quality Modeling Group Leader, Air Quality Assessment Division/OAQPS, at (919) 541-5562,
fox.tyler@epa. gov.
Attachment
Internet Address (URL) http://www.epa.gov
Recycled/Recyclable Printed with Vegetable Oil Based Inks on Recycled Paper (Minimum 25% Postconsumer)
-------
This Page Intentionally Left Blank
-------
Guidance for PM2 5 Permit Modeling
-------
This Page Intentionally Left Blank
-------
EPA-454/B-14-001
May 2014
Guidance for PM2 5 Permit Modeling
U.S. Environmental Protection Agency
Office of Air Quality Planning and Standards
Air Quality Assessment Division
Research Triangle Park, North Carolina
-------
This Page Intentionally Left Blank
-------
TABLE OF CONTENTS
Executive Summary v
Acknowledgements xi
I. Background 1
II. Guidance Overview 9
II. 1 Significant Emissions Rates 15
II.2 Screening and Source Impact Analysis 15
II.3 Cumulative Impact Analysis 21
II.4 Assessment Cases for Source Impacts 23
III. Source Impact Analysis for the PM2 5 NAAQS 27
III.l Assessing Primary PM25 Impacts 29
III.2 Assessing Secondary PM25 Impacts 31
III.2.1 Qualitative Assessments 31
III.2.2 Hybrid Qualitative/Quantitative Assessment 35
III.2.3 Full Quantitative Photochemical Grid Modeling 39
III.3 Comparison to the SIL 45
IV. Cumulative Impact Analysis for the PM25 NAAQS 51
FV.l Modeling Inventory 52
FV.2 Monitored Background 54
IV.3 Comparison to the NAAQS 56
FV.4 Determining Whether Proposed Source Causes or Contributes to Modeled Violations 64
V. PSD Increments for PM2 5 67
V.I Overview of PSD Increments 67
V.2 PM25 Increments Considerations 69
V.3 Screening Analysis for Increments 72
V.4 PM25 Increments Analysis 73
V.4.1 Source Impact Analysis 73
V.4.2 Cumulative Impact Analysis 74
V.4.2.1 Assessing Primary PM25 Impacts from Other Sources 74
V.4.2.2 Assessing Secondary PM25Impacts from Other Sources 75
V.4.2.3 Consideration of PM2 5 Ambient Air Quality Monitoring Data 76
V.5 Determining Significant Contribution to an Increment Violation 77
VI. References 79
-------
Appendix A: Draft Conceptual Description of PM25 Concentrations in the U.S A-l
Appendix B: General Guidance on Use of Dispersion Models for Estimating Primary PM25
Concentrations B-l
Appendix C: Example of a Qualitative Assessment of the Potential for Secondary PM2 5 Formation.... C-l
Appendix D: Example of a Hybrid Qualitative/Quantitative Assessment of the Potential for
Secondary PM2 5 Formation D-l
Appendix E: Example of the background monitoring data calculations for a Second Tier 24-hour
modeling analysis E-l
-------
Disclaimer
This document recommends procedures for permit applicants and permitting authorities
to use to show that they have satisfied the criteria for obtaining or issuing a permit under
applicable regulations. This document is not a rule or regulation, and the guidance it contains
may not apply to a particular situation based upon the individual facts and circumstances. This
guidance does not change or substitute for any law, regulation, or any other legally binding
requirement and is not legally enforceable. The use of non-mandatory language such as
"guidance, " "recommend, " "may, " "should, " and "can, " is intended to describe EPA policies
and recommendations. Mandatory terminology such as "must" and "required" are intended to
describe controlling requirements under the terms of the Clean Air Act and EPA regulations, but
this document does not establish legally binding requirements in and of itself. This document
does not create any rights or obligations enforceable by any party or impose binding,
enforceable requirements on any permit applicant for a PSD permit or PSD permitting authority.
m
-------
This Page Intentionally Left Blank
IV
-------
Executive Summary
The U.S. Environmental Protection Agency (EPA) is providing this "Guidance for PM2.5
Permit Modeling" to fulfill a need for additional guidance on demonstrating compliance with the
fine paniculate matter (PIVb.s) National Ambient Air Quality Standards (NAAQS) and the
Prevention of Significant Deterioration (PSD) increments, especially with regard to
considerations of the secondarily formed components of PIVb.s. This guidance incorporates the
modeling procedures and recommendations from the EPA's March 23, 2010, guidance
memorandum, "Modeling Procedures for Demonstrating Compliance with PM2.5 NAAQS," and
further clarifies procedures for adequately addressing primary and secondarily formed PM2.5 in a
NAAQS and PSD increments compliance demonstration. This guidance is consistent with the
EPA's Guideline on Air Quality Models, also published as Appendix W of Title 40 of the Code
of Federal Regulations (CFR) Part 51. The release of this "Guidance for PM2.5 Permit Modeling"
is also consistent with the commitments contained in the EPA's January 4, 2012, grant of a July
28, 2010, petition filed by the Sierra Club.
Because of the complex chemistry of secondary formation of PM2.5, the EPA's judgment
in the past has been that it was not technically sound to assign with particularity specific models
that must be used to assess the impacts of a single source on PM2.5 concentrations. Instead, the
EPA has determined it was appropriate to satisfy the requirements of Section 165(e)(3)(D) of the
Clean Air Act (CAA) by recommending that the "[c]hoice of methods used to assess the [PM2.s]
impact of an individual source depends on the nature of the source and its emissions," as stated
in Section 5.2.2. I.e. of Appendix W. As such, the appropriate methods for assessing PM2.5
impacts are determined as part of the normal consultation process with the appropriate permitting
authority. A modeling protocol should be developed by the permit applicant and approved by the
-------
appropriate permitting authority to ensure that the analysis conducted will conform to the
recommendations, requirements, and principles of Section 10.2.1 of Appendix W. This guidance
is intended to inform that process through recommendations regarding appropriate methods to
assess secondary PM2.5 impacts from the precursor emissions from the new or modifying source
by providing the permit applicant and the appropriate permitting authority with both focus and
flexibility. As experience is gained with these NAAQS and increments compliance
demonstrations (and as the EPA moves forward to consider single source modeling techniques
pursuant to its grant of the petition from the Sierra Club), this guidance will likely evolve such
that the EPA will be able to provide further specificity on assessing the impacts of a single
source on PIVb.s concentrations.
This guidance document is broken down into five primary sections:
I. Background - The first section provides the relevant regulatory actions and
historical context to this guidance starting with the promulgation of the initial PM2.5
NAAQS in 1997; chronicling the PMio Surrogate Policy that for a period of time was
relied upon for demonstrating compliance with the PM2.5 NAAQS; and arriving at the
present where there is a need for an assessment of both the primary and secondary
PM2.5 impacts, as appropriate, of a new or modifying source for demonstrating
compliance with PM2.5 NAAQS and increments.
II. Guidance Overview - The second section provides a general overview of the steps
that a permit applicant would routinely take under the PSD program for
demonstrating compliance with the PM2.5 NAAQS and increments. The concepts of
significant emissions rates (SERs) and significant impact levels (SILs) are introduced
and then presented in the context of a source impact analysis and a cumulative impact
VI
-------
analysis. The ramifications of the January 22, 2013, decision from U.S. Court of
Appeals for the District of Columbia Circuit on the use of SILs in a source impact
analysis or otherwise are included for reference and consideration throughout the
remaining sections. Four assessment cases (Table ES-1) are then introduced with
respect to assessing the primary and secondary PM2.5 impacts through either the
source impact analysis or the cumulative impact analysis.
III. Source Impact Analysis for the PM2.5 NAAQS - The third section provides a
detailed discussion of a screening assessment of primary and secondary PM2.5
impacts from a new or modifying source using a SIL. The specifics of the four
assessment cases (Table ES-1) are presented along with appropriate approaches for
assessing the primary and secondary impacts of PM2.5. For assessing the primary
PM2.5 impacts from the direct PM2.5 emissions from the new or modifying source, the
typical use of an appropriate preferred dispersion model for near-field PM2.5
modeling listed in Appendix W, currently AERMOD for most applications, or an
approved alternative model is recommended. For assessing the secondary PM2.5
impacts from the precursor emissions from the new or modifying source, three
different approaches are described. These approaches are 1) a qualitative assessment,
2) a hybrid qualitative/quantitative assessment utilizing existing technical work, and
3) a full quantitative photochemical grid modeling exercise.
IV. Cumulative Impact Analysis for the PM2.5 NAAQS - The fourth section provides
a detailed discussion of the assessment of primary and secondary PM2.5 impacts from
a new or modifying source with the inclusion of the primary and secondary PM2.5
impacts of nearby sources and of monitored background. There are specific
Vll
-------
discussions of the modeling inventory and the monitored background. Section IV
concludes with information on determining significant contributions to modeled
violations.
V. PSD Increments for PM2.5 - The fifth section provides a detailed discussion of the
assessment of primary and secondary PM2.5 impacts of a new or modifying source
with respect to the increments.
Table ES-1. EPA Recommended Approaches for Assessing Primary and Secondary PMi.s
Impacts by Assessment Case
As s es s ment Cas e
Case 1:
No Air Quality Analysis
Case 2:
Primary Air Quality
Impacts Only
Case 3:
Primary and Secondary
Air Quality Impacts
Case 4:
Secondary Air Quality
Impacts Only
Description of Assessment Case
Direct PM2.5 emissions < 10 tpy SER
NOxand SO2 emissions < 40 tpy SER
Direct PM2.5 emissions > 10 tpy SER
NOxand SO2 emissions < 40 tpy SER
Direct PM2.5 emissions > 10 tpy SER
NOx an d/or SO2 emissions > 40 tpy SER
Direct PM2.5 emissions < 10 tpy SER
NOx an d/or SO2 emissions > 40 tpy SER
Primary Impacts Approach
N/A
Appendix W preferred or
approved alternative
dispersion model
Appendix W preferred or
approved alternative
dispersion model
N/A
Secondary Impacts
Approach
N/A
N/A
Qualitative
Hybrid qualitative /
quantitative
Full quantitative
photochemical
grid modeling
Qualitative
Hybrid qualitative /
quantitative
Full quantitative
photochemical
grid modeling
In summary, this "Guidance for PM2.5 Permit Modeling" recommends technical
approaches for conducting PM2.5 NAAQS and PSD increments compliance demonstrations
which include adequate accounting for contributions from primary PM2.5 concentrations from a
proposed new or modifying source's direct PM2.5 emissions and from secondarily formed PM2.5
concentrations resulting from the source's PM2.5 precursor emissions. This guidance does not
create any rights or obligations enforceable by any party or impose binding, enforceable
requirements on any permit applicant for a PSD permit or PSD permitting authority. Since each
vm
-------
permitting action will be considered on a case-by-case basis, this document does not limit or
restrict any particular justifiable approach permit applicants and permitting authorities may take
to conduct the required compliance demonstrations. Each individual decision to issue a PSD
permit must be supported by a record sufficient to demonstrate that the proposed construction
and operation of a stationary source will not cause or contribute to a violation of the applicable
PM2.5 NAAQS and PSD increments. While this document illustrates a particular approach that
the EPA considers appropriate and acceptable as a general matter, permit applicants and
permitting authorities should examine all relevant information regarding air quality in the area
that may be affected by a proposed new or modified source and evaluate whether alternative or
additional analysis may be necessary in a given case to demonstrate that the criteria for obtaining
a permit are satisfied. This document does not represent a conclusion or judgment by EPA that
the technical approaches recommended in this document will be sufficient to make a successful
compliance demonstration in every permit application or circumstance.
Permitting authorities retain the discretion to address particular issues discussed in this
document in a different manner than the EPA recommends so long as the approach is adequately
justified, supported by the permitting record and technical literature, and consistent with the
applicable requirements in the CAA and implementing regulations, including the terms of an
approved State Implementation Plan (SIP).
Furthermore, this guidance does not represent final agency action with respect to
applicable legal requirements or the approvability of any particular permit application. To
improve the quality of this guidance, the EPA has solicited public comment and considered the
comments submitted. The EPA has revised the draft guidance in response to many points raised
in public comment, but this document does not reflect a final determination by the EPA as to any
IX
-------
issue raised in public comments. Concerns expressed in public comments about the
permissibility or sufficiency of the approach recommended in this guidance for making the
required demonstration in particular circumstances may be raised in the context of each
individual permit application and should be considered by the permitting authority in light of the
record in each instance before making a final determination to issue or deny a PSD permit.
-------
Acknowledgements
We would like to acknowledge contributions from members of the National Association
of Clean Air Agencies (NACAA) PIVb.s Modeling Implementation Workgroup (NACAA
Workgroup) in providing a detailed set of recommendations (NACAA, 2011) to the EPA with
regards to PM2.5 permit compliance demonstration modeling. This NACAA Workgroup was
comprised of state and local air permitting agency dispersion modelers, permit engineers, and
technical staff from throughout the country. In particular, we recognize Jim Hodina (Linn
County Public Health), Bob Hodanbosi (Ohio EPA, Division of Air Quality), and Clint Bowman
(Washington Department of Ecology) for their roles as Chairpersons for the Emissions
Inventories, Secondary Formation from Project Source, and Representative Background
Concentrations Sub-workgroups, respectively.
We would also like to acknowledge the contributions of the staff of the EPA Office of
Transportation and Air Quality (OTAQ) for their input and assistance in the development of this
document. The EPA's "Transportation Conformity Guidance for Quantitative Hot-spot Analyses
in PM2.5 and PMio Nonattainment and Maintenance Areas" (U.S. EPA, 2013a) guidance
document served as a foundation for many aspects of the modeling guidance contained within
this document.
Finally, there were numerous comments received with respect to issues and concerns of
demonstrating compliance with the NAAQS and increments for PM2.5 during the formal public
comment period for the 10th Conference on Air Quality Modeling and comprehensive comments
received specific to the overall guidance and recommendations presented here through the
comment period for the "Draft Guidance on PM2.5 Permit Modeling" (U.S. EPA 2013b). This
invaluable feedback along with additional information gleaned through ongoing interactions with
XI
-------
various stakeholders have been particularly useful in the consideration of a range of acceptable
options for PM2.5 NAAQS and PSD increments compliance demonstrations and aided the EPA in
the completion of this guidance document
Xll
-------
I. Background
Under Section 165(a)(3) of the CAA, a PSD permit applicant must demonstrate that
emissions from the proposed construction and operation of a facility "will not cause, or
contribute to, air pollution in excess of any (A) maximum allowable increase or maximum
allowable concentration for any pollutant... , [or] (B) national ambient air quality standard..."
This requirement is implemented in the PSD regulations at 40 CFR 52.21(k)(l) (and at 40 CFR
51.166(k)(l) with slightly different wording) as follows:
(k) Source impact analysis(1) Required demonstration. The owner or operator of the
proposed source or modification shall demonstrate that allowable emission increases
from the proposed source or modification, in conjunction with all other applicable
emissions increases or reductions (including secondary emissions), would not cause or
contribute to air pollution in violation of:
(i) Any national ambient air quality standard in any air quality control region; or
(ii) Any applicable maximum allowable increase over the baseline concentration in
any area.
On July 18, 1997, the EPA revised the NAAQS for particulate matter (PM) to add new
annual and 24-hour standards for fine particles using parti culate matter less than 2.5 micrometers
or PM2.5 as the indicator.l The EPA revised the 24-hour NAAQS for PM2.5 on September 21,
2006, by lowering the level of the standard from 65 ug/m3 to 35 ug/m3.2 In the September 21,
2006, action, the EPA also retained the previous 1997 annual standard for PM2.5 and the 24-hour
standard for PMio, and revoked the previous annual standard for PMi0. Subsequently, the
1 See 62 Fed. Reg. 58652.
2 See 71 Fed. Reg. 61144.
-------
Agency revised the PM2.5 standard again on December 14, 2012, by lowering the level of the
annual PM2.5 NAAQS from 15 ug/m3 to 12 ug/m3 and retaining the 24-hour standards for PM2.5
and PMio.3 The annual PM2.5 standard is met when the 3-year average of annual arithmetic mean
concentrations is less than or equal to 12.0 ug/m. The 24-hour PIVb.s standard is met when the 3-
year average of the annual 98* percentile 24-hour concentrations is less than or equal to
35 ug/m3.
On October 20, 2010, EPA established maximum allowable increases for PM2.5.4 These
values are also frequently described as the PSD increments. For Class I areas, the increments for
PM2.5 are 1 ug/m3 for the annual averaging time and 2 ug/m3 for the 24-hour averaging time. In
Class II areas, the increments are 4 ug/m3 for the annual period and 9 ^ig/m3 for the 24-hour
period.
To address the compliance demonstration for the PIVb.s NAAQS, on October 23, 1997,
citing significant technical difficulties with respect to PM2.5 monitoring, emissions estimation,
and modeling, the EPA established a policy known as the PMio Surrogate Policy (U.S. EPA,
1997). This policy allowed permit applicants to use compliance with the applicable PMio
requirements as a surrogate approach for meeting PlV^.s New Source Review (NSR)
requirements until certain technical difficulties were resolved. On May 16, 2008, the EPA
promulgated final rules governing the implementation of the NSR program for PM2.5, which
facilitated phasing out the application of the PMio Surrogate Policy to permits involving PM2.5.5
With regard to nonattainment NSR permits, the rule provided that as of July 15, 2008 (the rule's
effective date), permit applicants and permitting authorities would no longer be able to use the
3 See 78 Fed. Reg. 3086.
4 See 75 Fed. Reg. 64864.
5 See 73 Fed. Reg. 28321.
-------
Surrogate Policy to satisfy the NSR requirements for PM2.5. With regard to PSD permits,
the rule provided that PSD permits issued under the federal PSD program at 40 CFR 52.21
would no longer be allowed to rely on the PMio Surrogate Policy as of the effective date of the
rule. The exception to this outcome was that the rule also provided a "grandfathering provision"
allowing permit applicants for federal PSD permits covered by 40 CFR 52.21, with complete
permit applications submitted as of July 15, 2008, to continue relying on the PMio Surrogate
Policy. The 2008 rule also provided that states with approved PSD programs for PM2.5 could
continue to use the PMio Surrogate Policy until May 201 1 (when SIP revisions containing
provisions to meet the new requirements in the 2008 rule were due), or until the EPA approved
the revised SIP for PM2.5, whichever occurred first.
On June 1, 2009, in response to a petition challenging the continued use of the PMio
Surrogate Policy for issuing PSD permits, the EPA issued a 3-month administrative stay of the
grandfathering provision for PM2.5 affecting federal PSD permits to give the EPA time to
propose repealing the challenged grandfathering provision.6 On September 16, 2009, the original
3 -month stay was extended to June 22, 2010, to allow additional time for the EPA to propose
repealing the grandfathering provision from the PM2.5 NSR implementation rule for federal PSD
permits issued under 40 CFR 52.21.7 On February 11, 2010, the EPA published its proposal to
repeal the grandfathering provision.8 These actions cite the fact that the technical difficulties that
necessitated the PMio Surrogate Policy had been largely, although not entirely, resolved. As part
of the proposed rulemaking to repeal the grandfathering provision contained in the federal PSD
program, the EPA also proposed to require an early end to the use of the PMio Surrogate Policy
6 See 74 Fed. Reg. 26098.
7 See 74 Fed. Reg. 48153.
8 See 75 Fed. Reg. 6827.
-------
for state PSD programs that the EPA had already approved as part of the SIP required by 40 CFR
51.166.
On May 18, 2011, the EPA published a final rule, titled "Implementation of the New
Source Review (NSR) Program for Particulate Matter Less Than 2.5 Micrometers (PM2.s); Final
Rule to Repeal Grandfather Provision" (76 Fed. Reg. 28646), that repealed the grandfathering
provision. In that final action, the EPA ended the use of the PMio Surrogate Policy for PSD
permits under the federal PSD program for sources that were covered by the grandfathering
provision (that is, those sources for which a complete permit application was submitted before
July 15, 2008) and that were not yet issued a permit by the effective date of the final rule. 9 The
final rule also reaffirmed that as of May 2011, states with SIP-approved PSD programs for PM2.5
could no longer use the PMio Surrogate Policy. After the final rule became effective, in order for
any PSD permits to be issued through the federal PSD program or a state SIP, such permit
applications were to be reviewed directly against the PM2.5 requirements. The demonstration
must show, at a minimum, that the source's emissions are controlled to a level that satisfies Best
Available Control Technology (BACT) requirements for PM2.5 and that the emissions (filterable
and condensable10) would not cause or contribute to a violation of any NAAQS for PM2.5.
On March 23, 2010, the EPA issued a guidance memorandum titled "Modeling
Procedures for Demonstrating Compliance with PM2.5 NAAQS" (U.S. EPA, 2010b) to assist
sources and permitting authorities in carrying out the required air quality analysis. The guidance
memorandum recommended certain interim procedures to address the fact that compliance with
the PM2.5 NAAQS is based on a statistical form, and that there are technical complications
9 Sources that applied for a PSD permit under the federal PSD program on or after July 15, 2008, were already
excluded from using the 1997 PMIO Surrogate Policy as a means of satisfying the PSD requirements for PM2.5. See
73 Fed. Reg. 28321.
10 See 40 CFR 51.165(a)(l)(xxxvii)(D), 51.166 (b)(49)(i)(a), and 52.21(b)(50) (i)(a).
-------
associated with the ability of existing models to estimate the impacts of secondarily formed
PM2.5 in the atmosphere resulting from emissions of PM2.5 precursors. For the latter issue, the
EPA recommended that special attention be given to the assessment of monitored background air
quality data since such data account for the contribution of both primary and secondarily formed
PM2.5 in the atmosphere associated with both nearby and regional sources.
On January 7, 2011, the NACAA Workgroup delivered a final report (NACAA, 2011),
including a set of specific recommendations, to the EPA. The NACAA Workgroup was formed
in early 2010 with the objective of providing technical recommendations to the Agency to aid in
further development of PM2.5 permit modeling guidance. The NACAA Workgroup's final report
addressed three specific issues regarding PM2.5 modeling implementation: 1) Emissions
Inventories; 2) Secondary Formation from Project Source; and 3) Representative Background
Concentrations.
The need for additional clarification on addressing both the primary and secondarily
formed PM2.5 in NAAQS compliance demonstrations was heightened following an
administrative action on January 4, 2012, in which the EPA granted a petition submitted on
behalf of the Sierra Club on July 28, 2010 (U.S. EPA, 2012a). The Sierra Club petition requested
that the EPA initiate rulemaking to establish air quality models for ozone and PM2.5 for use by all
major sources applying for a PSD permit. In the petition grant, the EPA committed to engage in
rulemaking to evaluate updates to the Guideline on Air Quality Models as published as
Appendix W of 40 CFRPart 51 and, as appropriate, incorporate new analytical techniques or
models for ozone and secondarily formed PM2.5. As a part of this commitment and in compliance
with Section 320 of the CAA, the EPA conducted the 10th Conference on Air Quality Modeling
-------
(10th Modeling Conference) in March 2012. u At the 10th Modeling Conference, there were
invited presentations of ongoing research of single source plume chemistry and photochemical
grid modeling techniques, an overview presentation on the development of the "Draft Guidance
for PM2.5 Permit Modeling", and several public forums and subsequently written comments
given pertaining to PM2.5 NAAQS modeling.
Based on the EPA's March 23, 2010, guidance memorandum, the NACAA Workgroup
final report recommendations, input from a mixture of stakeholders through numerous forums,
and permit applicant-submitted PM2.5 compliance demonstrations up to that point, the EPA
prepared the "Draft Guidance on PM2.5 Permit Modeling" and released it for public comment on
March 4, 2013. During the course of the public comment period following the release of the draft
guidance, the EPA received numerous comprehensive comments that provided invaluable
feedback on the document and on the newly recommended approaches for PM2.5 NAAQS and
PSD increments compliance demonstrations. This feedback along with additional information
gleaned through ongoing interactions with various stakeholders was particularly useful in the
consideration of a range of acceptable options for PM2.5 NAAQS and PSD increments
compliance demonstrations and aided the EPA in the completion of this guidance document.
This "Guidance for PM2.5 Permit Modeling" recommends appropriate technical
approaches for conducting a PM2.5 NAAQS and PSD increments compliance demonstration
which includes adequate accounting for contributions from primary PM2.5 concentration from a
proposed new or modifying source's direct PM2.5 emissions and from secondarily formed PM2.5
concentrations resulting from the source's PM2.5 precursor emissions. This guidance is consistent
with the EPA's Guideline on Air Quality Models. The release of this "Guidance for PM2.5 Permit
1: Additional information regarding and presentations from the 10th Modeling Conference can be found on the
SCRAM website at: htto://www.epa. gov/ttn/scram/1 Othmodconf.htm.
-------
Modeling" is also consistent with the commitments contained in the EPA's January 4, 2012,
grant of the July 28, 2010, petition filed by the Sierra Club.
Since each permitting action will be considered on a case-by-case basis, this guidance
does not limit or restrict any particular justifiable approach permit applicants and permitting
authorities may take to conduct the required compliance demonstrations. Prospective permit
applicants should recognize the importance of the consultation process with the appropriate
permitting authority. This process will help identify the most appropriate analytical techniques to
be used for conducting a PIVb.s NAAQS and PSD increments compliance demonstration,
including addressing the impacts of individual sources on secondary PM2.5 formation, pursuant to
Section 5.2.2.1.c of Appendix W.
In addition to this guidance, other recently issued EPA guidance of relevance for
consideration in permit modeling for PM2.5 includes:
"Model Clearinghouse Review of Modeling Procedures for Demonstrating
Compliance with PM2.5 NAAQS," February 26, 2010 (U.S. EPA, 2010a);
"Modeling Procedures for Demonstrating Compliance with PM2.5 NAAQS," March
23, 2010 (U.S. EPA, 201 Ob);
"Transportation Conformity Guidance for Quantitative Hot-spot Analyses in PM2.5
and PMio Nonattainment and Maintenance Areas," November 2013 (U.S.EPA, 2013);
and
"Interim Guidance on the Treatment of Condensable Particulate Matter Test Results
in the PSD and NSR Permitting Programs," April 8, 2014 (U.S. EPA, 2014a).
The guidance listed above, in addition to other relevant support documents, can be found on the
SCRAM website at http://www.epa.gov/ttn/scram/.
-------
This Page Intentionally Left Blank
-------
II. Guidance Overview
This modeling guidance provides recommendations on how to conduct a PM2.5 NAAQS
and PSD increments compliance demonstration under the PSD program. It is based on and is
consistent with Appendix W. Appendix W is the primary source of information on the regulatory
application of air quality models for SIP revisions for existing sources and for NSR and PSD
programs for permitting new and modifying sources.
The complexity of secondary PM2.5 formation has historically presented significant
challenges for the identification and establishment of particular models for assessing the PM2.5
impacts of individual stationary sources (NARSTO, 2004; Seinfeld and Pandis, 1998; Cohan and
Napelenok, 2011). Because of these considerations, the EPA's judgment in the past has been that
it was not technically sound to assign with particularity specific models that must be used to
assess the impacts of a single source on PM2.5 concentrations.12 Instead, the EPA has chosen to
satisfy the requirements of the CAA, Section 165(e)(3)(D) through a process of determining
particular models or other analytical techniques that should be used on a case-by-case basis
because the "[c]hoice of methods used to assess the [PM^.s] impact of an individual source
depends on the nature of the source and its emissions," as stated in Section 5.2.2.1c. of Appendix
W. As such, the appropriate methods for assessing PM2.5 impacts are determined as part of the
normal consultation process with the appropriate permitting authority. A modeling protocol
should be developed by the permit applicant and approved by the appropriate permitting
authority to ensure that the analysis conducted will conform to the recommendations,
requirements, and principles of Section 10.2.1 of Appendix W.
12 We note that this technical judgment has no effect on the obligation of sources subject to PSD to conduct a source
impact analysis and demonstrate that a proposed source or modification will not cause or contribute to a violation of
any NAAQS or applicable increment. See 40 CFR 51.166(k); 52.2 l(k). That is, the inclusion of a process rather than
a specific preferred model in Appendix W does not relieve the source of the requirement to make this demonstration,
which necessarily involves an analysis.
-------
Due to the potentially important contribution from secondary formation of PM2.5 and the
more prominent role of ambient monitoring data in the cumulative analysis to represent
background PM2.5 concentrations including secondary formation from precursors from nearby
sources, certain aspects of standard modeling practices used for PMio and other criteria
pollutants may not be appropriate for PM2.5. For example, the contribution from secondary
formation of PIVb.s is not explicitly accounted for by the current preferred dispersion model (i.e.,
AERMOD), which is used to simulate dispersion of direct PM2.5 emissions. Given these issues,
PSD modeling of secondarily formed PM2.5 should currently be viewed as screening-level
analyses under Appendix W, analogous to Section 5.2.4 of Appendix W regarding dispersion
modeling for nitrogen dioxide (NC>2) impacts due to the importance of chemistry in the
conversion of nitric oxide (NO) emissions to ambient NC>2 and lack of a specified "refined"
model.13 The recommendations presented in this guidance for demonstrating compliance with
the PM2.5 NAAQS through dispersion modeling and other techniques have been developed with
the factors listed above in mind.
As with any modeling analysis conducted using approved models identified in
Appendix W, alternative models and methods may be considered on a case-by-case basis, subject
to approval by the EPA Regional Office in accordance with the recommendations in Section 3.2.
Additionally, Section 10.2.2 of Appendix W could potentially be given consideration in select
situations. The provisions of Section 10.2.2 acknowledge that there are circumstances where
there is no applicable model for a particular NAAQS compliance demonstration and that data
13 Section 5.2.4 of Appendix W puts forth a 3-tiered screening approach for NO2 NAAQS compliance
demonstrations to obtain estimates of NO2 for PSD and SIP planning purposes. The level of conservativeness in the
tiered approaches decreases as fewer assumptions are made and a more detailed analysis is applied with the 3rd tier
approach being the use of detailed screening techniques based on dispersion modeling.
10
-------
from an array of ambient monitors surrounding the facility to be permitted could be used in lieu
of modeling if appropriately justified.
Given the complexity of the technical issues that arise in the context of demonstrating
compliance with the PM2.5 NAAQS, we strongly encourage following the recommendations in
Section 10.2.1 of Appendix W that "[e]very effort should be made by the Regional Office to
meet with all parties involved in either a SIP revision or a PSD permit application prior to the
start of any work on such a project. During this meeting, a protocol should be established
between the preparing and reviewing parties to define the procedures to be followed, the data to
be collected, the model to be used, and the analysis of the source and concentration data."
Furthermore, we recommend that the consultative process involve regular communication
between the appropriate permitting authority and the permit applicant at key milestones to ensure
timely resolution of issues that may arise.
As necessary, the EPA Regional Office may seek clarification from the EPA's Office of
Air Quality Planning and Standards (OAQPS) on technical issues and areas of concern in a
modeling protocol or NAAQS compliance demonstration. Through these interactions and
subsequent resolutions of the specific issues, clarifications of preferred modeling procedures can
ultimately become official EPA guidance. This can happen in several ways: 1) the preferred
procedures are published as regulations or guidelines; 2) the preferred procedures are formally
transmitted as guidance to the Air Division Directors in the EPA Regional Offices; 3) the
preferred procedures are formally transmitted as guidance to the EPA Regional Office modeling
contacts as a result of a regional consensus on technical issues; or 4) the preferred procedures are
relied upon in decisions by the EPA's Model Clearinghouse that effectively establish national
precedent that the approach is technically sound. The Model Clearinghouse is the EPA focal
11
-------
point for the review of the technical adequacy of pollutant modeling to satisfy regulatory criteria
and other NAAQS compliance demonstration techniques. Model Clearinghouse memoranda
involving interpretation of modeling guidance for specific applications, as well as clarification
memoranda addressing needs to clarify guidance more generally, are available at the Support
Center for Regulatory Atmospheric Modeling (SCRAM) website at:
http ://www. epa.gov/ttn/scram.
The guidance that follows is appropriate for those new or modifying sources locating or
located in an area classified as attainment or unclassifiable for PM2.5. This document is intended
to provide recommendations on how to conduct PM2.5 NAAQS and PSD increments compliance
demonstrations under the PSD program following the progressive steps shown in Figure II-1
(NAAQS) and Figure II-2 (Increments). The EPA has historically allowed the use of screening
tools to help facilitate the implementation of the PSD program and streamline the permitting
process in circumstances where proposed construction is projected to have an insignificant (or de
minimis) impact on air quality. These screening tools have included SERs, SILs, and significant
monitoring concentrations (SMCs). The use of these screening tools at each progressive step on
the left side (attainment or unclassifiable areas) of Figure II-1 and Figure II-2 are described in
more detail in Sections III, II.2, and II.3.
12
-------
Figure II-1. Overview of PMi.5 NAAQS Compliance Demonstration for New or Modifying
Sources under NSR/PSD Programs
New or Modified
Source
Attainment or Unckssified
Area
Nonattainment Area
tialysis of Ambient Air Quality \
mpacts Not Required for the
Particular Pollutant J
Direct or Interpollutant Offsets
Please reference thePM2.5
NSR Implementation
final rule (73 PR 28321)
Satisfies NAAQS AQ Analysis
Cumulative Impact Analysis
V l J
Satisfies NAAQS AQ Analysis
May Satisfy NAAQS AQ \
Analysis.Please reference
Section IV.4 J
f Compliance Demonstration is Not 1
Y Adequate. J
13
-------
Figure II-2. Overview of PSD Increments Compliance Demonstration for New or
Modifying Sources under NSR/PSD Programs
New or Modified
Source
Attainment or Unclassified
Area
Cumulative Impact Analysis
Satisfies PSD Increment Analys
(May Satisfy PSD Increment \
Analysis.Please reference
Section V.5 J
nonstration is Not |
quate. I
Nonattainment Area
^Analysis of Ambient Air Quality \
1 Impacts Not Required for the I
V Particular Pollutant J \.
Increment Does Not Apply
Satisfies PSD Increment Analysis
/ Compliance Demonstration is Not 1
1 Adequate. I
Compliance Derr
Adeqi
14
-------
II. 1 Significant Emissions Rates
EPA regulations only require an analysis of ambient air quality impacts for pollutants that
a source emits (or that a modification of a source increases) in an amount equal to or greater than
the significant emission rate for that pollutant defined in EPA regulations.14 The EPA
promulgated SERs for PM2.5 and for the PM2.5 precursors, nitrogen oxides (NOX) and sulfur
dioxide (862), in 2008 as part of the first phase of PSD amendments to address PIVb.s.15 (74 Fed.
Reg. 28321 at 28333). The PM2.5 SER for direct emissions of primary PM2.5, defined as 10 tons
per year (tpy) of direct PIVb.s emissions, and the PIVb.s precursor SERs, defined as either 40 tpy
of NOX or 40 tpy of SC>2, are used to determine whether any proposed new major stationary
source or major modification will emit sufficient amounts of direct PM2.5 and/or PM2.5
precursors, i.e., equal to or above the respective SERs, to require review for PM2.5 under the PSD
program.
II.2 Screening and Source Impact Analysis
The EPA has historically supported the use of screening techniques in the PSD program
to determine the extent of the air quality analysis that must be carried out to demonstrate that the
source's emissions will not cause or contribute to a violation of any NAAQS or increment. 16
14 See 40 CFR 51.166(m)(l)(i); 40 CFR 52.21(m)(l)(i).
15 The EPA's final NSR rules for PM2 5 do not require regulation of volatile organic compounds (VOC) or ammonia
(NH3) as precursors to PM2 5 for the PSD program. However, a state may demonstrate to the Administrator's
satisfaction or the EPA may demonstrate that VOC emissions in a specific area are a significant contributor to that
area's ambient PM25 concentrations. See 74 Fed. Reg. 28321. If so, then permit applicants with project sources
having emissions of these pollutants should consult with the appropriate permitting authority and EPA Regional
Office about how to deal with these emissions for the purposes of a NAAQS or PSD increments analysis.
16 This has been consistent with overall support for screening techniques in the modeling guidelines. See, 40 CFR
Part 51, Appendix W, Sections 2.2 and 4.2.1. The Guideline observes that "use of screening techniques followed, as
appropriate, by a more refined analysis is always desirable." 40 CFR Part 51, Appendix W, Section 2.2.c. With
respect to PSD permit review specifically, the Guideline says the following: "The purpose of [screening] techniques
15
-------
Using this screening approach, when a proposed source's modeled impacts are found to be
greater than the level of a SIL identified by the EPA, the EPA has called for a cumulative impact
analysis (considering the combined impact of the proposed source and other sources in the
affected area) to demonstrate that the proposed source will not cause or contribute to a violation
of the NAAQS. On the other hand, the EPA has generally said that if the proposed source's
modeled impacts are found to be below the level of a SIL identified by EPA for the relevant
pollutant, this showing may be sufficient to demonstrate that the source will not cause or
contribute to a modeled violation of the NAAQS.1? However, the EPA has also acknowledged
that there can be circumstances where a showing that the air quality impact of a proposed source
is less than a SIL value identified by the EPA is not sufficient by itself to demonstrate that a
source will not cause or contribute to a violation of the NAAQS or increment.
Prior to 2010, EPA had expressed support in guidance for applying the values in Section
51.165(b)(2) of its regulations as SILs that could be used as part of a demonstration that a source
does not cause or contribute to a violation of the NAAQS. However, when the EPA added SILs
for PM2.5 in 2010 to paragraph (k)(2) of its Section 51.166 and 52.21 regulations, the Agency
observed that "the use of a SIL may not be appropriate when a substantial portion of any
NAAQS or increment is known to be consumed." (75 Fed. Reg. 64894). The EPA also said that
"notwithstanding the existence of a SIL, permitting authorities should determine when it may be
appropriate to conclude that even a de minimis impact will "cause or contribute" to an air quality
problem and to seek remedial action from the proposed new source or modification." (75 Fed.
Reg. 64892).
is to eliminate the need of more detailed modeling for those sources that clearly will not cause or contribute to
ambient concentrations in excess of either the National Ambient Air Quality Standards (NAAQS) or the allowable
prevention of significant deterioration (PSD) concentration increments." Id. Section 2.2.a.
17 See 72 Fed. Reg. 54112 at 54139 and 75 Fed. Reg. 64864 at 64890.
16
-------
In the course of litigation challenging the SILs for PM2.5, the EPA acknowledged that the
regulatory language the EPA adopted in Sections 51.166(k)(2) and 52.21(k)(2) did not provide
sufficient flexibility for permitting authorities to exercise discretion to conduct or require
additional analysis in some circumstances where the EPA had advised doing so. As a result, the
EPA requested that the U.S. Court of Appeals for the District of Columbia Circuit remand and
vacate these provisions so the EPA could take corrective action. On January 22, 2013, the court
granted this request and observed that, under the language in Sections 51.166(k)(2) and
52.21(k)(2), sources in some scenarios would not be required to demonstrate that they would not
cause or contribute to a violation of the NAAQS or increments, even though, based on
Petitioner's arguments, the sources likely would cause or contribute to a violation in such
scenarios. The court concluded this would contravene the statutory command in Section
165(a)(3) of the Act. 705 F.3d at 464-65. The court also said that on remand the EPA may
choose to promulgate regulations that "include SILs that do not allow the construction or
modification of a source to evade the requirements of the Act as do the SILs in the current rule"
and that such regulations would be subject to further review by the court. (Id. at 464).
EPA does not interpret the court's decision to preclude the use of SILs for PM2.5 as part
of a demonstration that a source will not cause or contribute to a violation of the PM2.5 NAAQS.
However, to ensure that PSD permitting decisions meet the requirements of the CAA, permitting
authorities that continue using SILs for PM2.5 must ensure that they select and apply such SILs in
a manner that is consistent with the court's decision and the EPA's statements from the preamble
of the 2010 regulation adopting SILs for PM2.5. The EPA is developing a proposed rule to
address the issues identified by the EPA and the court's decision. If necessary and as appropriate,
this guidance will be amended after this rulemaking is proposed and subsequently finalized. In
17
-------
the interim, permitting authorities may not apply the SIL provisions in the vacated and repealed
Sections 51.166(k)(2) and 52.21(k)(2). Furthermore, permitting authorities should not apply any
state regulations that have not yet been amended to conform to the repeal of these provisions and
still contain regulatory text that is the same as or has a similar effect as the paragraph (k)(2)
language, particularly in the types of scenarios described in the court decision and the EPA's
2010 preamble to the PM2.5 Increments, SILs, and SMC Rule.18 However, with appropriate
safeguards, the EPA believes permitting authorities may continue to select and apply SILs values
for PM2.5 to support PSD permitting decisions and to determine the level of analysis needed to
demonstrate that a source will not cause or contribute to violation of the NAAQS.19 These
safeguards involve two related considerations - the particular values of the SILs to be used and
how those values are used.
The court decision does not preclude the use of SILs for PM2.5, but requires that the EPA
correct the error in the SIL regulations for PM2.5 at 51.166(k)(2) and 52.21(k)(2). As a first step,
on December 9, 2013, the EPA issued a final rule removing these sections of its regulations from
the CFR (78 Fed. Reg. 73698). Until the EPA completes a rulemaking to replace these
provisions, the EPA believes permitting authorities may continue to apply SILs for PM2.5 to
support a PSD permitting decision, but permitting authorities should take care to ensure that SILs
are not used in a manner that is inconsistent with the requirements of Section 165(a)(3) of the
CAA.
Permitting authorities have the discretion to select the particular PM2.5 SIL values that are
18 Prevention of Significant Deterioration (PSD) for Paniculate Matter Less Than 2.5 Micrometers (PM2 5) -
Increments, Significant Impact Levels (SILs) and Significant Monitoring Concentration (SMC). See 75 Fed. Reg.
64864 (October 20, 2010).
19 The topic of the level of analysis needed for PSD increments compliance analysis is discussed in more detail in
Section V.
18
-------
used to support a permitting decision, but the values used should be supported by either a
permitting record or regulation that supports the use of those values in the particular manner they
90
are used. Permitting authorities may not rely on the values contained in the vacated Sections
51.166(k)(2) and 52.21(k)(2) of the EPA's regulations as a screening tool without providing
additional justification in the permitting record. However, with additional justification, it may be
permissible in some cases for a permitting authority to use the same PlVks SIL values as listed in
the vacated Sections 51.166(k)(2) and 52.21(k)(2) to demonstrate that a full cumulative impacts
analysis is not needed to make the NAAQS compliance demonstration
To the extent a permitting authority wishes to use any of the SILs values in the vacated
Sections 51.166(k)(2) or 52.21(k)(2) as a screening tool to determine whether it is necessary to
conduct a cumulative analysis of NAAQS compliance, the permitting authority must first
examine background air quality concentrations to determine whether a substantial portion of the
91
NAAQS has been consumed. For this purpose, the EPA recommends using the preconstruction
monitoring data compiled to meet the requirements of Section 51.166(m) or 52.21(m) of the
EPA's regulations. If the preconstruction monitoring data are sufficiently representative of the
air quality in existence before the increase in emissions from the proposed source and the
difference between the PM2.5 NAAQS and the measured PM2.5 background concentrations in the
20 The EPA has previously observed that the absence of an EPA-promulgated SIL does not preclude PSD permitting
authorities from developing and applying SILs to support permitting decisions. See, Response to Comments,
Implementation of New Source Review (NSR) Program for Paniculate Matter Less Than 2.5 Micrometers in
Diameter (PM25) at 82 (March 2008) [EPA-HQ-OAR-2003-0062-0278]. However, the EPA has also observed that,
"[t]he application of any SIL that is not reflected in a promulgated regulation should be supported by a record in
each instance that shows the value represents a de minimis impact." See, NO2 NAAQS Guidance at 13; and
Mississippi Lime at 41 (granting the petition for review where the permitting authority failed to substantiate in the
record which SIL it applied and its reasons for doing so).
21 The recent court decision vacating the PM2 5 SMC from the PSD regulations will mean that each PSD application
must include ambient monitoring data representative of the area of concern. These data need not be collected by the
PSD permit applicant if existing data are determined by the permitting authority to represent the air quality in the
area of concern over the 12-month period prior to the submittal of a complete PSD application.
19
-------
area is greater than or equal to the SIL value selected from the vacated Sections 51.166(k)(2) and
52.21(k)(2), then the EPA believes it would be sufficient in most cases for permitting authorities
to conclude that a source with an impact equal to or below that SIL value will not cause or
contribute to a violation of the NAAQS and to forego a cumulative modeling analysis for PM2.5
with respect to the NAAQS.
The above comparison of background air quality concentrations and the NAAQS would
not by itself provide adequate justification for foregoing a cumulative modeling analysis for the
PM2.5 increments. Such an approach would be inappropriate because it would not ensure that
there is sufficient "headroom" within the allowable increment to absorb a source contribution
equal to the SIL. However, a permitting authority may still be able to justify reaching a
determination that a new or modified source will not cause or contribute to a violation of the
increments without performing cumulative modeling for increments.
Since the trigger date has only recently been established (i.e., October 20, 2011), for the
next several years, a new or modified source being evaluated for increments compliance will
often be the first source with increment-consuming emissions in the area. As indicated in Figure
II-2, under this situation, a permitting authority may have sufficient reason to conclude that the
impacts of the new or modified source (based on the approach for conducting source impact
analysis described below) may be compared directly to the allowable increments, without the
need for a cumulative modeling analysis. Such a situation would involve the new or modified
source representing the first PSD application in the area after the trigger date, which establishes
the minor source baseline date and baseline area, and confirmation that no relevant major source
construction has already occurred since the major source baseline date.
20
-------
II.3 Cumulative Impact Analysis
As part of a NAAQS compliance demonstration, a cumulative impact analysis for PM2.5
accounts for the combined impacts of direct and precursor emissions from the new or modifying
source, of direct emissions from nearby sources (as appropriate), and of monitored background
levels of PM2.5 that account for secondary PM2.5 impacts from regional transport, secondary
PM2.5 impacts from precursor emissions from nearby sources, and primary PM2.5 impacts from
background sources not included in the modeled inventory. The cumulative impacts are then
compared to the NAAQS to determine whether the source will cause or contribute to a violation
of the NAAQS. Several aspects of the cumulative impact analysis for PM2.5 will be comparable
to analyses conducted for other criteria pollutants, while other aspects will differ due to the
issues identified earlier.
The measured background levels incorporated into a cumulative analysis should be based
on the preconstruction monitoring data gathered in accordance with the requirements of the EPA
regulations. 40 CFR 51.166(m)(l)(iii)-(iv); 40 CFR 52.21(m)(l)(iii)-(iv) (2). The EPA
regulations contain an exemption from the preconstruction monitoring requirements in cases
where ambient concentrations or the predicted impact of the source are less than the SMC. 40
CFR 51.166(i)(5)(i) ; 40 CFR 52.21(i)(5)(i). In the decision mentioned above, aU.S. Court of
Appeals vacated the SMC for PM2.5. Sierra Club v. EPA, 705 F.3d 458. The court concluded that
the PM2.5 SMC provisions (51.166(i)(5)(i)(c) and 52.21(i)(5)(i)(c)) were inconsistent with the
requirements of Section 165(e)(2) of the CAA. The EPA has subsequently removed the PM2.5
99
SMC provisions from the regulation. Thus, permitting authorities may no longer rely on the
SMCs for PM2.5 to exempt permit applicants from compiling preconstruction monitoring data for
: See 78 Fed. Reg. 73698.
21
-------
PM2.5 in accordance with Sections 51.166(m) and 52.21(m) of the EPA's regulation. However,
the EPA believes PSD permit applicants may continue to meet the preconstruction monitoring
requirements in these regulations by gathering for purposes of the permitting analysis data
already available from existing monitors that are determined by the applicable permitting
9^
authority to be representative of background conditions in the affected area.
Where the screening analysis described in Section II.2 above is insufficient to show that a
source will not cause or contribute to a violation of the PSD increments, a cumulative impact
assessment would be necessary to make the demonstration. A cumulative assessment accounts
for the combined impact of the new or modifying source's emissions and those emissions
changes from sources that affect the increment. The cumulative impacts are then compared to the
PSD increments to determine whether the new or modifying source emissions will cause or
contribute to a violation of the PSD increments.
23 "EPA has long implemented the PSD program pursuant to the understanding that representative data may be
substituted where circumstances warrant." (In re: Northern Michigan University Ripley Heating Plant, PSD Appeal
No. 08-02, slip op. at 58 (Feb. 18, 2009));
".. .the prospective PSD source must use existing ... representative air quality data or collect... monitoring data."
(52 Fed. Reg. 24672 (July 1, 1987) at 24686); and
With regard to the PSD requirement for monitoring data, "use of 'monitoring data' refers to either the use of existing
representative air quality data or monitoring the existing air quality." (Ambient Monitoring Guidelines for
Prevention of Significant Deterioration (PSD), EPA-450/4-80-012, November 1980, at page 3).
22
-------
II.4 Assessment Cases for Source Impacts
To support the processes shown in Figure II-1 and Figure II-2, the EPA is recommending
four different assessment cases shown in Table II-1 that define which air quality analyses, if any,
a permit applicant should conduct to demonstrate compliance with the PIVb.s NAAQS and PSD
increments.
Table II-1. EPA Recommended Assessment Cases that Define Needed Air Quality Analyses
of Source Impacts
As s es s ment Cas e
Case 1:
No Air Quality Analysis
Case 2:
Primary Air Quality
Impacts Only
Case 3:
Primary and Secondary
Air Quality Impacts
Case 4:
Secondary Air Quality
Impacts Only
Description of Assessment Case
Direct PM2.5 emissions < 10 tpy SER
NOxand SO2 emissions < 40 tpy SER
Direct PM2.5 emissions > 10 tpy SER
NOxand SO2 emissions < 40 tpy SER
Direct PM2.5 emissions > 10 tpy SER
NOx an d/or SO2 emissions > 40 tpy SER
Direct PM2.5 emissions < 10 tpy SER
NOx an d/or SO2 emissions > 40 tpy SER
Assess Primary Impacts
of Direct PMz.5
Emissions?
NO
YES
YES
NO
Assess Secondary Impacts
of Precursor Emissions of
NOX and/or SO2?
NO
NO
YES
YES
The four assessment cases presented in Table II-l include:
For "Case 1No Air Quality Analysis," if direct PM2.5 emissions are less than
the SER of 10 tpy and both NOX and 862 emissions are individually less than the
respective SERs of 40 tpy, then no modeled compliance demonstration is
required.24
For "Case 2Primary Air Quality Impacts Only," if the direct PM2.5 emissions
are greater than or equal to the SER of 10 tpy and both NOX and SC>2 emissions
are individually less than the respective SERs of 40 tpy, then a modeled PM2.5
1 See 40 CFR 51.166(m)(l)(i); 40 CFR 52.21(m)(l)(i)
23
-------
compliance demonstration is required for only the direct PM2.5 emissions based
on dispersion modeling and no modeling to account for impacts of precursor
emissions from the project source is necessary.
For "Case 3Primary and Secondary Air Quality Impacts," if the direct PM2.5
emissions are greater than or equal to the SER of 10 tpy and NOX and/or 862
precursor emissions are greater than or equal to the respective SERs of 40 tpy,
then a modeled PM2.s compliance demonstration is required for the direct PM2.s
emissions based on dispersion modeling and the permit applicant should also
assess the potential impact of the significant precursor emissions from the project
source. The accounting of the precursor emissions impact on secondary PM2.5
formation may be: a) qualitative in nature; b) based on a hybrid of qualitative and
quantitative assessments utilizing existing technical work; or c) a full quantitative
photochemical grid modeling exercise. The EPA anticipates only a few situations
would require explicit photochemical grid modeling.
For "Case 4Secondary Air Quality Impacts Only," if the direct PM2.5 emissions
are less than the SER of 10 tpy, but the NOX and/or SO2 precursor emissions are
greater than or equal to the respective SERs of 40 tpy, then a modeled PM2.5
compliance demonstration for the direct PM2.5 emissions is not required, but the
permit applicant should assess the potential impact of the significant precursor
emissions from the project source. Similar to "Case 3," the accounting of the
precursor emissions impact on secondary PIVb.s formation may be: a) qualitative
in nature; b) based on a hybrid of qualitative and quantitative assessments
utilizing existing technical work; or c) a full quantitative photochemical grid
24
-------
modeling exercise. Again, the EPA anticipates that only a limited number of
situations would require explicit photochemical grid modeling.
Details regarding the source impact analysis and cumulative impact analysis associated
with Cases 2, 3, and 4, where project emissions are equal to or greater than the respective SERs
for direct PM2.5 emissions only (Case 2), for both direct PM2.5 and precursor emissions of NOX
and/or SC>2 (Case 3), or for precursor emissions of NOX and/or SC>2 only (Case 4), are provided in
Sections III and IV (NAAQS) and Section V (Increments).
25
-------
This Page Intentionally Left Blank
26
-------
III. Source Impact Analysis for the PM2.5 NAAQS
This section provides details regarding the recommended approaches for conducting the
source impact analysis associated with each of the four assessment cases presented in Table III-l
so long as the SIL has been appropriately justified for use in each NAAQS compliance
demonstration as described in Section 11.2. In each of the assessment cases, the analysis should
begin by evaluating the impacts of direct PM2.5 emissions and/or PM2.5 precursor emissions
based upon the total amount of these emissions as compared to the respective SERs.
Table III-l. EPA Recommended Approaches for Assessing Primary and Secondary
Impacts by Assessment Case
As s es s ment Cas e
Case 1:
No Air Quality Analysis
Case 2:
Primary Air Quality
Impacts Only
Case 3:
Primary and Secondary
Air Quality Impacts
Case 4:
Secondary Air Quality
Impacts Only
Description of Assessment Case
Direct PM2.5 emissions < 10 tpy SER
NOxand SO2 emissions < 40 tpy SER
Direct PM2.5 emissions > 10 tpy SER
NOxand SO2 emissions < 40 tpy SER
Direct PM2.5 emissions > 10 tpy SER
NOx and/or SO2 emissions > 40 tpy SER
Direct PM2.5 emissions < 10 tpy SER
NOx and/or SO2 emissions > 40 tpy SER
Primary Impacts Approach
N/A
Appendix W preferred or
approved alternative
dispersion model
Appendix W preferred or
approved alternative
dispersion model
N/A
Secondary Impacts
Approach
N/A
N/A
Qualitative
Hybrid qualitative /
quantitative
Full quantitative
photochemical
grid modeling
Qualitative
Hybrid qualitative /
quantitative
Full quantitative
photochemical
grid modeling
A modeled NAAQS compliance demonstration is not required for Case 1 since neither
direct PM2.5 emissions nor PM2.5 precursor (NOX and/or 802) emissions are equal to or greater
than the respective SERs. Case 1 is the only assessment case that does not require a modeled
NAAQS compliance demonstration. Each of the remaining three assessment cases would
necessitate a source impact analysis.
The simplest or most traditional assessment case is Case 2 where only direct PM2.5
27
-------
emissions are greater than or equal to the SER. For Case 2, the permit applicant would only need
to demonstrate that ambient PIVb.s impacts associated with its increase in direct PM2.5 emissions
are below a SIL based on dispersion modeling using AERMOD or other appropriate preferred
model listed in Appendix A of Appendix W, or an alternative model subject to the provisions of
Section 3.2 of Appendix W.
Since both direct PIVb.s emissions and NOX and/or 862 precursor emissions are equal to
or greater than the respective SERs for Case 3, this will likely be the most challenging of the four
assessment cases. As with Case 2, the ambient PIVb.s impacts associated with direct PIVb.s
emissions can be estimated based on application of an appropriate preferred dispersion model for
near-field PM2.5 modeling listed in Appendix W, currently AERMOD for most applications, or
an approved alternative model. However, AERMOD does not account for secondary formation
of PM2.5 associated with the source's precursor emissions. Since the source also emits quantities
of PM2.5 precursors above the respective SERs for Case 3, some assessment of their potential
contribution to secondary PM2.5 is necessary. The assessment of NOX and/or SO2 precursor
emission impacts on secondary PM2.5 formation may be: a) qualitative in nature; b) based on a
hybrid of qualitative and quantitative assessments utilizing existing technical work; or c) a full
quantitative photochemical grid modeling exercise. The EPA anticipates that only a limited
number of situations would require explicit photochemical grid modeling.
Since direct PM2.5 emissions are below the applicable SER for Case 4, the source impact
analysis in this case would only address the potential contribution to secondary PM2.5 from NOX
and/or SO2 precursor emissions, and would not require any modeling of direct PM2.5 emissions.
As discussed above for Case 3, the assessment of the precursor emission impacts on secondary
PM2.5 formation for Case 4 may be: a) qualitative in nature; b) based on a hybrid of qualitative
28
-------
and quantitative assessments utilizing existing technical work; or c) a full quantitative
photochemical grid modeling exercise. As with Case 3, the EPA anticipates that only a few
situations would require explicit photochemical grid modeling.
III.l Assessing Primary PMi.5 Impacts
The assessment of primary PM2.5 impacts from the proposed new or modifying source is
generally the same for the NAAQS and increments and should be consistent with Appendix W.
As noted above, Appendix W recommends specific models as "preferred" for specific types of
applications, based on model performance evaluations and other criteria. The purpose of
recommending the use of a particular preferred model is to ensure that the best-performing
model is used in assessing PM impacts from a particular project and is employed in a consistent
fashion.25 In 2005, the EPA promulgated AERMOD as the Agency's preferred near-field
dispersion model for a wide range of regulatory applications in all types of terrain based on
extensive developmental and performance evaluation.26 For NSR/PSD modeling for the PM2.5
NAAQS, the AERMOD modeling system should be used to model direct PM2.5 emissions unless
another preferred model is more appropriate, such as the Buoyant Line and Point source
dispersion model (BLP), or the use of an alternative model can be justified consistent with
Section 3.2 of Appendix W.
25 The best performing model is one that best predicts regulatory design values for a particular pollutant. The EPA's
Protocol for Determining the Best Performing Model (U.S. EPA, 1992) defines appropriate methodologies and
statistical criteria for this evaluation. According to the document, "For a pollutant... for which short-term ambient
standards exist, the statistic of interest involves the network-wide highest concentration... the precise time, location,
and meteorological condition is of minor concern compared to the magnitude of the highest concentration actually
occurring."
26 The final rule can be found at: http://www.epa.gov/scram001/guidance/guide/appw 05.pdf. Extensive
documentation is available describing the various components of AERMOD, including user guides, model
formulation, and evaluation papers. See EPA's SCRAM website for AERMOD documentation:
www.epa.gov/scram001/dispersion prefrec.htm#aermod
29
-------
As noted in the EPA's March 23, 2010, PM2.5 guidance memorandum, although dry
and/or wet deposition may be important processes when estimating ambient concentrations of
PM in general, these factors are expected to be minor for PM2.5 due to the small particle size. In
addition, there may be additional uncertainty associated with deposition modeling for PM2.5 due
to the fact that deposition properties may vary depending on the constituent elements of PM2.5.
Therefore, use of deposition algorithms to account for depletion in estimating ambient PM2.5
concentrations should be done with caution and only when clear documentation and justification
of the deposition parameters is provided.
The AERMOD modeling system includes the following components:
AERMOD: the dispersion model (U.S. EPA, 2004a; U.S. EPA, 2014b);
AERMAP: the terrain processor for AERMOD (U.S. EPA, 2004b, U.S. EPA, 201 la);
and
AERMET: the meteorological data processor for AERMOD (U.S. EPA, 2004c; U.S.
EPA, 2014c).
Other components that may be used, depending on the application, are:
BPIPPRIME: the building input processor (U. S. EPA, 2004d);
AERSURFACE: the surface characteristics processor for AERMET (U.S. EPA, 2008);
AERSCREEN: a screening version of AERMOD (U.S. EPA, 201 Ib; U.S. EPA, 201 Ic);
and
AERMINUTE: a pre-processor to calculate hourly average winds from ASOS 2-minute
observations (U.S. EPA, 201 Id).
Before running AERMOD, the user should become familiar with the user's guides
30
-------
associated with the modeling components listed above and the most recent version of the
AERMOD Implementation Guide (U.S. EPA, 2009). In addition to these documents, detailed
guidance on the use of the AERMOD modeling system for estimating primary PM2.5 impacts is
provided in Appendix B. Because AERMOD is limited to modeling only direct PM2.5 emissions,
additional or alternative approaches must be used to provide an assessment of the secondary
PM2.5 impact from the proposed new or modifying source, as discussed in more detail in the
following sections.
III.2 Assessing Secondary PMi.5 Impacts
This section provides more detail on the recommended approaches for assessing the
impacts of precursor emissions on secondary PM2.5 formation for Cases 3 and 4 presented in
Table III-l including:
a qualitative assessment;
a hybrid of qualitative and quantitative assessments utilizing existing technical work; and
a full quantitative photochemical grid modeling exercise.
III.2.1 Qualitative Assessments
In a number of NAAQS compliance demonstrations requiring an assessment of the
impact from secondary PM2.5 formation, it is anticipated that a holistic qualitative analysis of the
new or modifying emissions source and the atmospheric environment in which the emissions
source is to be located will suffice for determining that secondary PM2.5 impacts associated with
the source's precursor emissions will not cause or contribute to a violation of the 24-hour or
annual PM2.5 NAAQS. Each NAAQS compliance demonstration will be unique and may require
31
-------
multiple factors to be considered and assumptions to be thoroughly justified as a part of the
qualitative assessment. A well-developed modeling protocol that includes a detailed conceptual
description of the current air pollution concentrations in the area (see Appendix A for examples
of elements of a conceptual description) and of the nature of the emissions sources surrounding
the new or modifying emissions source is paramount for determining the necessary components
of an acceptable qualitative assessment of the impact from secondary PM2.5 formation.27 With
appropriate consultation, submittal, and subsequent approval of the modeling protocol by the
appropriate permitting authority, many potential problems and unintended oversights in the
qualitative assessment can be resolved early in the process or avoided all together.
In the development of an appropriate conceptual description of PM2.5 to support a
qualitative assessment of the impact from secondary PM2.5 formation, it is important to fully
characterize the current PM2.5 concentrations in the region where the new or modifying
emissions source is to be located. This characterization should take into consideration not only
the most current 24-hour and annual PIVb.s design values, which would typically be used as
background concentrations in a cumulative modeling demonstration, but should also include an
understanding of the seasonality and speciated composition of the current PM2.5 concentrations
and any long term trends that may be occurring. Understanding whether or not PM2.5
27 For more detailed information on the development of such conceptual descriptions for an area, please refer to the
following:
Chapter 10 of "Particulate Matter Assessment for Policy Makers: A NARSTO Assessment." P. McMurry, M.
Shepherd, and J. Vickery, eds. Cambridge University Press, Cambridge, England (NARSTO, 2004).
Section 11, "How Do I Get Started? 'A Conceptual Description'" of "Guidance on the Use of Models and Other
Analyses for Demonstrating Attainment of Air Quality Goals for Ozone, PM2 5, and Regional Haze." U.S.
Environmental Protection Agency, Research Triangle Park, North Carolina (U.S. EPA, 2007a).
In addition, relevant regional examples include: "Conceptual Model of PM25 Episodes in the Midwest", January
2009, Lake Michigan Air Directors Consortium; and "Conceptual Model of Particulate Matter Pollution in the
California San Joaquin Valley," Document Number CP045-1-98, September 8, 1998.
32
-------
concentrations are higher or lower in certain seasons or fairly uniform throughout a year and
determining whether there are particular component species (e.g., sulfates, nitrates, and
elemental or organic carbon) that dominate the makeup of high, low, and average PM2.5
concentrations will help guide the degree of analysis and ultimately the justification that will be
required in the qualitative assessment based on the magnitude and characteristics of any
significant precursor emissions from the source. It may also be important to describe the typical
background concentrations of certain chemical species that participate in the photochemical
reactions that form secondary PlV^.s, such as NHs, VOC, and ozone. It is possible that there are
mitigating factors for secondary PM2.5 formation given limitations of other chemical species
important in the photochemical reactions, e.g., minimal NHa in the ambient environment that
could limit any precursor pollutant from readily reacting to form secondary PM2.5. The
qualitative assessment should include a narrative explaining how any identified significant
precursor emissions and subsequent secondary PM2.5 formation could contribute to the existing
PM2.5 concentration environment in the region.
A good conceptual description will also characterize the meteorological conditions that
are representative of the region and are associated with periods and/or seasons of higher and
lower ambient 24-hour PM2.5 concentrations. Identification of meteorological phenomena that
typically occur during periods of high 24-hour PM2.5 concentrations, such as low-level
temperature inversions, stagnant high pressure systems, etc., can be extremely important in
understanding the importance, or lack thereof, of photochemistry and secondary PIVb.s formation
for the higher ambient PM2.5 concentrations. The analysis and understanding of meteorological
conditions will also inform the assessment of the seasonality of the 24-hour PM2.5 concentrations
in the region. The qualitative assessment should expand upon the characterization of
33
-------
meteorology described in the conceptual description to explain any meteorological factors that
could limit or enhance the formation of secondary PM2.5 from any significant precursor
emissions.
Analysis of existing photochemical grid modeling developed for regional haze, ozone,
and PM2.5 SIPs or other photochemical grid modeling used in related sensitivity projects or
analysis to support prior air quality rules may also be considered to help understand the general
response of secondary PM2.5 formation to certain magnitudes of a precursor pollutant in that
region. While the new or modifying emissions source may emit a significant level of a precursor
pollutant under PSD regulations, that level of emission may be extremely small when compared
against the total emissions of that precursor pollutant throughout the region. The qualitative
assessment of the impact from secondary PM2.5 formation can be strengthened if substantial
regional decreases or increases of that precursor pollutant have been demonstrated through
photochemical grid modeling exercises do not cause significant decreases or increases of
secondary PM2.5.
An example of a thoroughly developed qualitative assessment of the potential for
secondary PM2.5 formation to cause or contribute to a violations of the NAAQS was provided by
the EPA Region 10 Office through a response to public comments document regarding a CAA
permit issued for Shell's Discoverer drill ship and support fleet to explore for oil and gas in the
Chukchi Sea off Alaska. While the environment in and around the Chukchi Sea and North Slope
of Alaska is unique when compared to the rest of the United States, the various components
contained within this qualitative assessment provide a template that could be followed, with
appropriate modifications, in the development of other case-specific qualitative assessments. An
excerpt from this response to public comments document is provided in Appendix C.
34
-------
As shown in the EPA Region 10 example, the qualitative assessment of the potential for
secondary PM2.5 formation by Shell's Discoverer drill ship and support fleet was developed in a
narrative manner integrating numerous factors specific to the North Slope region of Alaska that
provided sufficient evidence that the PM2.5 NAAQS would not be violated in this particular case.
The qualitative assessment examined the regional background PM2.5 monitoring data and aspects
of secondary PM2.5 formation from existing sources; the relative ratio of the combined modeled
primary PM2.5 impacts and background PM2.5 concentrations to the level of the NAAQS; the
spatial and temporal correlation of the primary and secondary PIVb.s impacts; meteorological
characteristics of the region during periods of precursor pollutant emissions; the level of
conservatism associated with the modeling of the primary PM2.5 component and other elements
of conservatism built into the overall NAAQS compliance demonstration; aspects of the
precursor pollutant emissions in the context of limitations of other chemical species necessary
for the photochemical reactions to form secondary PM2.s; and an additional level of NAAQS
protection through a post-construction monitoring requirement. While each of the components of
the EPA Region 10 example may or may not be necessary, this example should provide a useful
template for other qualitative assessments under this guidance, recognizing that additional
components may be essential in other qualitative assessments of the potential for secondary
PM2.5 formation.
III.2.2 Hybrid Qualitative/Quantitative Assessment
The qualitative assessment discussed above is largely focused on a determination that the
proposed new or modifying source precursor emissions, in combination with the estimated
primary PM2.5 impacts (if applicable for Case 3), will not cause or contribute to a violation of the
35
-------
24-hour and/or annual PM2.5 NAAQS. However, it may not always be possible to provide such a
justification without some quantification of the potential secondary PM2.5 impacts from the
proposed new or modifying source's precursor emissions. In such cases, the EPA expects that
existing air quality model-based information regarding the potential for 862 and NOX precursor
emissions to form secondary PM2.5 concentrations may be used to establish an appropriate
estimate of secondary PM2.5 impacts from the proposed new or modifying source. As described
above, there may be situations where the proposed new or modifying source's total ambient
impact (i.e., primary and secondary impacts) is less than a SIL, and the record demonstrates that
no further air quality assessment would be needed to demonstrate that the source would not
cause or contribute to a violation of any NAAQS. Otherwise, a cumulative impact assessment
would be necessary, which is discussed in Section IV.
To inform a hybrid qualitative/quantitative assessment, the existing air quality model-
based information would need to be appropriate in terms of representing the type of source, its
precursor emissions, and its geographic location, in addition to those elements of the conceptual
description discussed above for the qualitative assessment. The quantitative modeling
information may be available from past or current SIP attainment demonstration modeling,
published modeling studies, or peer-review literature with estimates of model responsiveness to
precursor emissions in contexts that are relevant to the new or modifying source. The estimates
of model responsiveness, such as impact on PM2.5 concentrations per ton of SC>2 emissions, could
then be used in conjunction with the precursor emissions estimates for the proposed new or
modifying source to provide a quantitative estimate of the impact of such precursor emissions on
the formation of secondary PM2.5 concentrations. The estimates should be technically credible in
representing such impacts and it may be advisable for the estimate to reflect an upper bound of
36
-------
potential impacts.
The NACAA Workgroup final report (NACAA, 2011) provides details on potential
approaches to quantify the secondary PM2.5 impacts from a proposed new or modifying source
that may be appropriate to inform a hybrid qualitative/quantitative assessments of PM2.5 impacts
(See Appendix C and D of NACAA, 2011). One suggested method in the final report is to
convert emissions of precursors into equivalent amounts of direct PM2.5 emissions using
"pollutant offset ratios" and then use a dispersion model to assess the impacts of the combination
of direct PIVb.s emissions and the equivalent direct PM2.5 emissions. The "pollutant offset ratios"
referenced in the final report were those put forth by the EPA in the 2008 "Implementation of the
New Source Review (NSR) Program for Particulate Matter Less Than 2.5 Micrometers (PIVk.s)"
final rule (73 Fed. Reg. 28321) concerning the development and adoption of interpollutant
trading (offset) provisions for PM2.5 under state nonattainment area NSR programs for PM2.5.28
The EPA's July 23, 2007, technical analysis titled "Details on Technical Assessment to Develop
Interpollutant Trading Ratios for PM2.5 Offsets," describes the method used to establish the
original "preferred" precursor offset ratios (U.S. EPA, 2007b).
We do not support using the specific results from the EPA's 2007 technical assessment in
this context without additional technical demonstration specific to the source(s) and area(s) for
which the ratios would be applied. However, we expect that the EPA Regional Offices, with
assistance from the OAQPS, may assist state/local air permitting agencies, as necessary, to
In the preamble to the 2008 final rule (73 Fed. Reg. 28321), the EPA included preferred or presumptive offset
ratios, applicable to specific PM2 5 precursors that state/local air agencies may adopt in conjunction with the new
interpollutant offset provisions for PM2 5, and for which the state could rely on the EPA's technical work to
demonstrate the adequacy of the ratios for use in any PM25 nonattainment area. In a July 21, 2011 memorandum,
EPA changed its policy and stated that it no longer supported the ratios provided in the preamble to the 2008 final
rule as presumptively approvable ratios for adoption in SIPs containing nonattainment NSR programs for PM2 5.
Memorandum from Gina McCarthy, Assistant Administrator to Regional Air Division Directors, "Revised Policy to
Address Reconsideration of Interpollutant Trading Provisions for Fine Particles (PM25)" (U.S. EPA, 2011e).
37
-------
structure appropriate technical demonstrations leading to the development of appropriate source
and area-specific offset ratios for PM2.5 that may be appropriate for the purposes of estimating
potential secondary PM2.5 impacts. As described in the EPA's July 21, 2011, memorandum
addressing reconsideration of the interpollutant trading provisions for the 2008 final rule, the
EPA acknowledged that existing models and techniques are adequate to "conduct local
demonstrations leading to the development of area-specific ratios for PIVb.s nonattainment areas"
and provided a general framework for efforts that may be relevant in developing appropriate
"pollutant offset ratios" for use in hybrid qualitative/quantitative assessment of secondary PM2.5
impacts (U.S. EPA, 201 le).
An example of a hybrid qualitative/quantitative assessment of secondary PM2.5 impacts
was developed by a permit applicant, Sasol, for a major facility expansion in Southern Louisiana
through close coordination with the EPA Region 6 Office and the Louisiana Department of
Environmental Quality (LDEQ). Sasol and LDEQ worked closely with Region 6 to ensure that
the ambient impacts analysis was robust and defendable. In this particular hybrid assessment,
Sasol took an approach of using the formerly presumptive interpollutant trading ratios for NOX
and 862 to PM2.5 offsets and conservatively applied them in an illustrative example to
demonstrate how relatively inconsequential the impacts of secondary PM2.5 formation would be
in the area of significant impact surrounding their facility. Sasol did not seek to directly apply the
formerly presumptive interpollutant trading ratios in an absolute sense. Rather, the intention was
to present the analysis in a manner to determine if further technical justification would be
required or if the application of the formerly presumptive interpollutant trading ratios was
adequate in a hybrid qualitative/quantitative sense. A more detailed discussion of Sasol's hybrid
assessment is provided in Appendix D.
38
-------
The EPA also notes that the NACAA Workgroup "considered, but rejected, other
methods for assessing secondary PM2.5 impacts, including use of a simple emissions divided by
distance (Q/D) metric and use of AERMOD with 100 percent conversion of SC>2 and NOX
concentrations to (NFL^SC^ and (NFL^NCb." The EPA has reviewed the detailed discussion
provided in Appendix E of the NACAA Workgroup final report and agrees with these
conclusions.
III.2.3 Full Quantitative Photochemical Grid Modeling
In those rare cases where it is deemed necessary to estimate secondary PM2.5 impacts
with full quantitative photochemical grid modeling, the candidate model for use in estimating
single source impacts on secondarily formed PM2.5 should meet the general criteria for an
"alternative model" outlined in Section 3.2.2 of 40 CFR 51.112 and 40 CFR Part 51,
Appendix W, for condition (3) where "the preferred model is less appropriate for the specific
application, or there is no preferred model," i.e.,
i. The model has received a scientific peer review;
ii. The model can be demonstrated to be applicable to the problem on a theoretical
basis;
iii. The databases that are necessary to perform the analysis are available and
adequate;
iv. Appropriate performance evaluations of the model have shown that the model is
not biased toward underestimates; and
iv. A protocol on methods and procedures to be followed has been established.
39
-------
Section 3.2.1 of Appendix W also discusses appropriate methodologies for evaluating
performance of models for regulatory applications, including the EPA's "Protocol for
Determining the Best Performing Model" (U.S. EPA, 1992). The determination of acceptability
of a particular model and approach for such an alternative model application is an EPA Regional
Office responsibility that may also include consultation with the EPA Headquarters, if
appropriate.
As noted in the NACAA Workgroup final report, photochemical grid models provide a
complete characterization of emissions, chemical transformation, transport, and deposition using
time and space variant meteorology. The EPA's modeling guidance for PM2.5 attainment
demonstrations (U.S. EPA, 2007a) identifies both the Comprehensive Air Quality Model with
Extensions (CAMx) (ENVIRON, 2011; Nobel et al., 2001; Russell, 2008) and the Community
Multiscale Air Quality (CMAQ) model (Byun and Schere, 2006; Foley et al., 2010). These state-
of-the-science photochemical grid models have been used by the EPA for air quality modeling to
support federal rulemaking and by state/local air permitting agencies for their air quality
planning efforts. Some photochemical grid models have been instrumented with extensions that
allow for the identification of impacts from specific sources to important receptor locations.
These extensions generally fall in the categories of source apportionment and source sensitivity,
and of sub-grid plume treatment and sampling, as described below.
Based on the current capabilities of photochemical grid models and consistent with the
NACAA Workgroup report, the EPA recommends the following approaches be considered to
estimate secondary PM2.5 impacts from a proposed new or modifying source using this type of
model:
"Brute force zero-out" or difference method where two model simulations are conducted,
40
-------
one with all existing sources and a second, counterfactual simulation with all existing
sources and the new source emissions, with the difference being taken as the contribution
from the new or modifying source.
Instrumented techniques such as
o Source apportionment tools where the precursor emissions from the new or
modifying source are tracked to provide a contribution estimate for that individual
source, or
o Higher-order decoupled direct method (HDDM) which tracks the sensitivity of
results to the emissions from a new or modifying source to provide coefficients
relating source emissions to air quality response.
The NACAA Workgroup final report notes that these approaches represent
fundamentally different methods and may result in different estimates for secondary PM2.5
impacts depending on the non-linear chemical processes. The EPA, state/local permitting
agencies, and others within the atmospheric modeling community continue to apply these
techniques to test and evaluate their suitability for estimating single source impacts on
secondarily formed PM2.5. These efforts are critically important to inform current application of
these models and techniques for purposes of assessing the secondary PIVb.s impacts from a
proposed new or modifying source, as well as to inform efforts to evaluate updates to
Appendix W with new analytical techniques or models for ozone and secondary PM2.5 per the
commitments contained in the EPA's January 4, 2012, grant of the July 28, 2010, petition filed
by the Sierra Club.29
29 Several photochemical grid modeling approaches that allow for estimation of the secondary PM25 impacts from a
proposed new or modifying source were presented during the Emerging Models / Techniques Session of the 10th
41
-------
Photochemical grid models that have been instrumented with source apportionment
techniques track emissions from specific sources through the chemical transformation, transport,
and deposition processes to estimate the source's contribution to predicted air quality at
downwind receptors (Baker and Foley, 2011). Source sensitivity approaches provide information
about how model predicted concentrations change based on an increase or decrease in emissions
from a specific source. The difference in air quality between the original baseline simulation and
the simulation where emissions are perturbed provides a quantitative estimate of that source's
contribution to the cumulative impact estimate.
Another approach to differentiate the contribution of single sources on changes in model
predicted air quality is the higher-order decoupled direct method (HDDM), which tracks the
sensitivity of model results to emissions for a specific source through all chemical and physical
processes in the modeling system (Bergin et al., 2008). Sensitivity coefficients relating source
emissions to air quality are estimated during the model simulation and output at the resolution of
the photochemical grid model. An important difference between source apportionment and
source sensitivity is that source apportionment answers the "contribution" question, "How much
did a source contribute overall to modeled air quality?" and source sensitivity answers the
"responsiveness" question, "How will modeled air quality change if the source's emissions
change?"
In some instances where the source and critical receptors are in very close proximity, the
source and receptors may be located in the same photochemical grid model cell. Since physical
and chemical processes simulated in the model represent a volume average, this may not
adequately (or appropriately) represent the gradients of pollution that may exist between the
Modeling Conference. Additional information regarding and presentations from the 10th Modeling Conference can
be found on the SCRAM website at: http://www.epa.gov/ttn/scram/1 Othmodconf.htm.
42
-------
source and receptors. One approach to more explicitly represent the spatial gradient in source-
receptor relationships when they are in close proximity would be to use smaller sized grid cells.
Grid resolution would be defined such that the source and receptors are no longer in the same
grid cell. Ideally, there would also be several grid cells between the source and receptors to best
resolve near-source pollution gradients.
In these situations of close proximity between the source and receptors, a photochemical
grid model instrumented with sub-grid plume treatment and sampling may be an alternative
approach for characterizing these relationships. Sub-grid plume treatment extensions in
photochemical grid models typically solve for in-plume chemistry and use a set of physical and
chemical criteria for determination of when puff mass is merged back into the host model grid.
However, accounting for source specific impacts both at the sub-grid and grid levels is
challenging and enhancements to traditional implementations of this approach may be necessary
to fully capture source impacts for permit applications.
For this guidance, the EPA is not prescribing in detail how photochemical grid models
(or their instrumented extensions) should be applied for the purposes of conducting a NAAQS
compliance demonstration since these details may involve case-specific factors that would need
to be part of the consultative process with the appropriate permitting authority and reflected in
the agreed-upon modeling protocol. With this in mind, we recommend that the modeling
protocols for this purpose should include the follow elements:
1. Overview of Modeling/Analysis Project
Participating organizations
Schedule for completion of the project
Description of the conceptual model for the project source/receptor area
43
-------
Identify how modeling and other analyses will be archived and documented
Identify specific deliverables to the appropriate permitting authority
2. Model and Modeling Inputs
Rationale for the selection of air quality, meteorological, and emissions models
Modeling domain
Horizontal and vertical resolution
Specification of initial and boundary conditions
Episode selection and rationale for episode selection
Rationale for and description of meteorological model setup
Basis for and development of emissions inputs
Methods used to quality assure emissions, meteorological, and other model inputs
3. Details on the approach for comparison to the SIL and/or NAAQS
4. Model Performance Evaluation
Describe ambient database(s)
Describe evaluation procedures and performance metrics
As stated previously, we expect that the EPA Regional Offices, with assistance from the
OAQPS, may assist states, as necessary, to structure appropriate technical demonstrations
leading to the development of appropriate photochemical grid modeling applications for the
purposes of estimating potential secondary PM2.5 impacts.
44
-------
III.3 Comparison to the SIL
Where a permit applicant wishes to compare the proposed source's total ambient PM2.5
impacts to a SIL in order to make the required demonstration that a source does not cause or
contribute to a violation of the NAAQS, the compliance demonstration will vary depending on
whether Case 2, 3, or 4, where direct PM2.5 and/or precursor emissions are equal to or greater
than the respective SERs, is applicable.
For Case 2, where only direct PM2.5 emissions are equal to or greater than the applicable
(10 tpy) SER, the SIL may be compared to the modeled estimates of ambient primary PM2.5
concentrations due to direct emissions using the preferred AERMOD dispersion model (or
acceptable preferred or alternative model). The modeling methods used in this initial source
impact assessment phase of the PM2.5 analysis for Case 2 are similar to the methods used for
other pollutants, including the use of maximum allowable emissions, following Table 8-2 of
Appendix W. However, due to the form of the PM2.5 NAAQS, we recommend that a SIL be
compared to either of the following, depending on the meteorological data used in the analysis:
The highest of the 5-year averages of the maximum modeled 24-hour or annual PM2.5
concentrations predicted each year at each receptor, based on 5 years of
representative National Weather Service (NWS) data; or
The highest modeled 24-hour or annual PM2.5 concentrations predicted across all
receptors based on 1 year of site-specific meteorological data, or the highest of the
multi-year averages of the maximum modeled 24-hour or annual PM2.5 concentrations
predicted each year at each receptor, based on 2 or more years, up to 5 complete years
of available site-specific meteorological data.
These metrics represent the maximum contribution that project emissions could make to the air
45
-------
quality impact at any receptor, given the form of the NAAQS, and therefore provide an
appropriate part of the basis for determining whether a cumulative modeling analysis would be
needed.
For Case 3, where the source's direct PIVb.s emissions and emissions of at least one
precursor are equal to or greater than the respective SERs, the comparison of the SIL would need
to address both primary and secondary PIVb.s ambient impacts associated with the proposed
source. As with Case 2, the ambient impacts due to direct PM2.5 emissions would be estimated
using the preferred AERMOD dispersion model (or acceptable alternative model). However, the
comparison to the SIL will depend on the type of assessment conducted for the secondary PM2.5
impacts from the source. As noted above, the assessment of the precursor emission impacts on
secondary PM2.5 formation may be: a) qualitative in nature; b) based on a hybrid of qualitative
and quantitative assessments utilizing existing technical work; or c) a full quantitative
photochemical grid modeling exercise.
Since any SIL that is used should represent a specific insignificant (or de minimis)
ambient concentration of PM2.5 that may be used to demonstrate that a source will not cause or
contribute to a NAAQS violation without conducting a cumulative impact assessment, basing the
initial source impact analysis for Case 3 on a qualitative assessment (or a hybrid of qualitative
and quantitative assessments) of secondary PM2.5 ambient impacts may be difficult to justify.
This is because there would be no specific quantitative estimate of total PM2.5 impacts for
comparison to the SIL, unless a valid argument can be made that secondary PM2.5 impacts
associated with the source's precursor emissions will be very small (e.g., precursor emissions
barely exceed the respective SERs and/or the chemical environment is not conducive to
secondary formation). As such, when using either of these approaches, it may be appropriate to
46
-------
forego the SIL assessment and focus on the NAAQS compliance demonstration using a
cumulative impact analysis.
For cases where a full quantitative photochemical grid modeling assessment of secondary
PM2.5 is conducted, the SIL comparison for Case 3 should be based on the combined ambient
impacts of primary and secondary PM2.5. However, the primary and secondary PM2.5 impacts
may be combined in various ways which may entail greater or lesser degrees of conservatism.
For example, combining the peak estimated primary PM2.5 impact with the peak estimated
secondary PM2.5 impact, unpaired in time and space would likely result in a conservative
estimate of combined impacts since, as noted above, peak impacts associated with a source's
direct PIVb.s and precursor emissions are not likely well-correlated in time or space. On the other
hand, the conservatism associated with combining peak estimated primary and secondary
impacts for comparison to a SIL would likely make such an approach easier to justify than other
approaches for combining estimated primary and secondary PM2.5 impacts.
The other extreme for combining primary and secondary PM2.5 impacts for comparison to
a SIL for Case 3, relative to combining peak primary and peak secondary impacts unpaired in
time and space, would be full temporal and spatial pairing of estimated primary and secondary
PM2.5 impacts. Such an approach may not be feasible in many cases, given that the dispersion
modeling and photochemical grid modeling may be based on different data periods. Furthermore,
full temporal and spatial pairing of primary and secondary PM2.5 impacts may not be appropriate
in many cases due to the fact that photochemical grid modeling represents gridded concentration
estimates whereas dispersion modeling produces estimates at discrete receptor locations and
given the limitations in the skill of both the dispersion model and the photochemical grid model
to accurately predict impacts on a paired in time and space basis. On the other hand, some degree
47
-------
of temporal pairing of primary and secondary PM2.5 impacts on a seasonal or monthly basis
should be appropriate in most cases, recognizing the general lack of correlation between primary
and secondary impacts.
The permitting authority and the permit applicant should thoroughly discuss the details
regarding combining modeled primary and secondary PM2.5 impacts for Case 3 and should reach
agreement on a protocol during the initial review of the modeling protocol. It may be appropriate
for the protocol to specifically identify multiple tiers for combining the modeled primary and
secondary PM2.5 impacts with the more conservative approaches being easier to justify. The
permitting authority should ensure that any approach for combining estimated primary and
secondary PM2.5 impacts for comparison to a SIL for Case 3 conforms to the recommendations
described above for Case 2 regarding the form of the modeled estimate. Accordingly, the
approach should be based on the highest of the multi-year averages of the maximum modeled
24-hour or annual PM2.5 concentrations predicted each year at each receptor, which represents
the maximum contribution that the source's emissions could make in a cumulative impact
assessment.
For Case 4, where the source's precursor emissions are equal to or greater than the
respective SERs but direct PM2.5 emissions are not, the SIL comparison would only address
secondary PM2.5 ambient impacts associated with the proposed source. The assessment of the
precursor emission impacts on secondary PM2.5 formation may be: a) qualitative in nature; b)
based on a hybrid of qualitative and quantitative assessments utilizing existing technical work; or
c) a full quantitative photochemical grid modeling exercise. As discussed above for Case 3, since
a SIL should represent a specific insignificant (or de minimis) ambient concentration of PM2.5
that may be used to demonstrate that a source will not cause or contribute to a violation without a
48
-------
cumulative impact assessment, basing the source impact analysis on a purely qualitative
assessment of secondary PM2.5 ambient impacts or a hybrid of qualitative and quantitative
assessments, utilizing existing technical work for Case 4, may be difficult to justify unless a
demonstrably conservative estimate of the secondary PM2.5 contribution can be made that is
below a SIL. As such, when using either of these approaches, it may be appropriate for the
permitting authority to recommend the permit applicant to forego the SIL assessment and focus
on the NAAQS compliance demonstration using a cumulative impact analysis. However, it may
be more feasible for the permitting authority to allow the permit applicant to apply a SIL to full
photochemical grid model estimates of secondary PM2.5 for Case 4 than for Case 3 since the
issues associated with combining modeled estimates of primary and secondary PM2.5 would not
apply for Case 4. In these cases, the highest of the multi-year averages of the maximum modeled
24-hour or annual PM2.5 concentrations predicted each year at each receptor should be compared
to a SIL, since these metrics represent the maximum contribution that the source could make.
49
-------
This Page Intentionally Left Blank
50
-------
IV. Cumulative Impact Analysis for the PMi.5 NAAQS
Where the screening analysis described in Section II is insufficient to show that a source
will not cause or contribute to a violation of the NAAQS, a cumulative impact assessment would
be necessary to make the NAAQS compliance demonstration. A cumulative assessment accounts
for the combined impact of the new or modifying source's emissions, emissions from other
nearby sources, and representative background levels of PM2.5 within the modeling domain. The
cumulative impacts are then compared to the NAAQS to determine whether the new or
modifying source emissions will cause or contribute to a violation of the NAAQS. This section
provides details on conducting an appropriate cumulative impact assessment for the PM2.5
NAAQS.
The cumulative impact assessment should include the following components of PM2.5
impacts, as appropriate, for comparison to the NAAQS:
Proposed new or modifying source
o Primary impacts on PM2.5, i.e., from direct PM2.5 emissions
o Secondary impacts on PM2.5, i.e., from precursor (NOX and/or 862)
emissions
Nearby sources
o Primary impacts on PM2.5, as appropriate
Monitored background of PM2.5 that accounts for secondary PIVb.s impacts from
regional transport, secondary PM2.5 impacts from nearby sources, and primary
PM2.5 impacts from background sources not included in the modeled inventory.
As with the source impact analysis discussed previously, the primary impacts related to
direct PM2.5 emissions from the proposed new or modifying source and nearby sources should be
51
-------
estimated based on the AERMOD dispersion model (or other acceptable preferred model or an
approved alternative model) while the estimate of secondary PM2.5 impacts from the proposed
new or modifying source will vary depending on whether the assessment of the proposed
source's precursor emission impacts on secondary PIVb.s formation are: a) qualitative in nature;
b) based on a hybrid of qualitative and quantitative assessments utilizing existing technical work;
or c) based on a full quantitative photochemical grid modeling exercise. As noted above,
secondary impacts on PM2.5 from regional transport and precursor emissions from nearby
sources should be accounted for through representative monitored background concentrations.
I V.I Modeling Inventory
The current guidelines on emission inventories for purposes of NAAQS compliance
modeling contained in Section 8.1 of Appendix W will generally be applicable for the PM2.5
modeling inventory. The guidelines in Appendix W address the appropriate emission level to be
modeled, which in most cases is the maximum allowable emission rate under the proposed
permit. The remainder of this section will focus on the modeling inventory of direct PM2.5
emissions that should be used in dispersion modeling of primary PM2.5 impacts. Although the
EPA's "Guidance on the Use of Models and Other Analyses for Demonstrating Attainment of
Air Quality Goals for Ozone, PIVb.s, and Regional Haze" (U.S. EPA, 2007a) provides some
guidance relevant to applications involving full quantitative photochemical grid modeling,
additional considerations and guidance regarding modeling inventories for such analyses in
support of PM2.5 NAAQS compliance demonstrations in PSD permitting under this guidance will
be provided by EPA on a case-by-case basis.
As discussed in more detail in the EPA's March 1, 2011, clarification memorandum
52
-------
regarding Appendix W modeling guidelines for the 1-hour NC>2 NAAQS (U.S. EPA, 201 If),
Section 8.2.3 of Appendix W emphasizes the importance of professional judgment in the
identification of nearby and other sources to be included in the modeled emission inventory and
establishes "a significant concentration gradient in the vicinity of the [proposed] source" as the
main criterion for this selection. Appendix W also suggests that "the number of such [nearby]
sources is expected to be small except in unusual situations." (Section 8.2.3.b). The EPA's
March 1, 2011, guidance also includes a detailed discussion of the significant concentration
gradient criterion included in Section 8.2.3.b of Appendix W, indicating that the significant
concentration gradient criterion suggests that the emphasis on determining which nearby sources
to include in the cumulative modeling analysis should focus on the area within about 10
kilometers of the project location in most cases. However, several application-specific factors
should be considered when determining the appropriate inventory of nearby sources to include in
the cumulative modeling analysis, including the potential influence of terrain characteristics on
concentration gradients and the availability and adequacy of ambient monitoring data to account
for background sources.
Consistent with the March 1, 2011, guidance, the EPA cautions against the application of
very prescriptive procedures for identifying which nearby sources should be included in the
modeled emission inventory for NAAQS compliance demonstrations, such as the procedures
described in Chapter C, Section IV.C.l of the draft "New Source Review Workshop Manual"
(U.S. EPA, 1990). This caution should not be taken to imply that the procedures outlined in the
draft "New Source Review Workshop Manual" are flawed or inappropriate. Cumulative impact
assessments based on following such procedures will generally be acceptable as the basis for
permitting decisions, contingent on an appropriate accounting for the monitored contribution.
53
-------
Our main concern is that following such procedures in a literal and uncritical manner may
increase the likelihood of double-counting modeled and monitored concentrations in many cases,
resulting in cumulative impact assessments that are overly conservative and would unnecessarily
complicate the permitting process in some cases. The identification of which sources to include
in the modeled emissions inventory should be addressed in the modeling protocol and, as
necessary, discussed in advance with the permitting authority.
Since modeling of direct PM2.5 emissions has not been frequently conducted to date, the
availability of an adequate direct PM2.5 emission inventory for nearby sources may not exist in
all cases. Recommendations for developing PM2.5 emission inventories for use in PSD
applications will be addressed separately, but existing SIP inventories for PIVb.s or statewide
PSD inventories of sources for refined modeling may provide a useful starting point for this
effort.
IV.2 Monitored Background
Sections 8.2.2 and 8.2.3 of Appendix W provide recommendations for determination of
background concentrations for inclusion in cumulative impact assessments for NAAQS
compliance, which should account for impacts from existing sources that are not explicitly
included in the modeled inventory and natural sources. From newly-acquired pre-construction
monitoring data and/or existing representative air quality data gathered for purposes of a
permitting analysis, permit applicants should assess and document what the background
monitoring data represent to the extent possible, including any information that may be available
54
-------
from the state or other agency responsible for siting and maintaining the monitor.30 It is also
worth noting that the relative makeup of PM2.5 components and temporal patterns associated
with the highest 24-hour PM2.5 levels may differ considerably from the relative amounts of PM2.5
components associated with annual average PM2.5 levels, especially in western states.
The determination of monitored background concentrations of PM2.5 to include in the
PM2.5 cumulative impact assessment may entail different considerations from those for other
criteria pollutants and may also depend on whether the application involves full quantitative
photochemical grid modeling. An important aspect of the monitored background concentration
for PM2.5 is that the ambient monitoring data should, in most cases account for the contribution
of secondary PM2.5 formation associated with existing sources impacting the modeling domain in
addition to the background levels of primary PM2.5 associated with background sources that are
not included in the modeled inventory. As with other criteria pollutants, consideration should
also be given to the potential for some double-counting of the impacts from modeled emissions
that may be contributing to the background monitored concentrations, but this should generally
be of less importance for PM2.5 than the representativeness of the monitor for secondary
contributions, unless the monitor is located relatively close to nearby sources of primary PIVb.s
that could be impacting the monitor. Also, the nature of secondary PM2.5, monitored background
concentrations of PM2.5 are likely to be more homogeneous across the modeling domain in most
cases compared to most other pollutants, although this will also depend on the potential for local
sources of primary PIVb.s to be contributing to the monitored concentrations.
Depending on the nature of local PM2.5 levels within the modeling domain, it may be
30 Please note in the case of an existing source seeking a permit for a modification, there is potential overlap across
secondary contributions from monitored background and from precursor emission from the existing source. In such
cases, recommendations for excluding monitored values when the source in question is impacting the monitor in
Section 8.2.2.b of Appendix W may need to be modified to avoid overcompensating in cases where the monitored
concentrations are also intended to account for the existing project source's contributions to secondary PM2 5.
55
-------
appropriate to account for seasonal variations in monitored background PM2.5 levels which may
not be correlated with seasonal patterns of the modeled primary PM2.5 levels. For example,
maximum modeled primary PM2.5 impacts associated with fugitive or other low-level emission
sources are likely to occur during winter months due to longer periods of stable atmospheric
conditions, whereas maximum ambient levels of secondary PM2.5 in the eastern United States
typically occur during spring and summer months due to high levels of sulfates. The use of
temporally-varying monitored background concentrations in a cumulative impact analysis is
discussed in more detail in Section IV.3.
IV.3 Comparison to the NAAQS
Combining the modeled and monitored concentrations of PM2.5 for comparison to the
PM2.5 NAAQS entails considerations that differ from those for other criteria pollutants due to the
issues identified above. The discussion below addresses comparisons to the NAAQS in the
context of dispersion modeling of direct PM2.5 emissions only (for Case 2), and also provides
guidance regarding NAAQS comparisons for applications involving qualitative, hybrid
qualitative/quantitative, or full quantitative photochemical grid modeling assessments of
secondary PM2.5 impacts (for Cases 3 and 4).
Given the importance of secondary contributions for PM2.5 and the potentially high
background levels relative to the PM2.5 NAAQS, greater emphasis is generally placed on the
monitored background contribution relative to the modeled inventory for PM2.5 than for other
pollutants. This is true for both NAAQS and increments assessments. Also, given the
probabilistic form of the PM2.5 NAAQS, careful consideration should be given to how the
monitored and modeled concentrations are combined to estimate the cumulative impact levels.
56
-------
The representative monitored PM2.5 design value, rather than the overall maximum
monitored background concentration, should generally be used as the monitored component of
the cumulative analysis. The PM2.5 design value for the annual averaging period is based on the
3-year average of the annual average PIVb.s concentrations. The PM2.5 design value for the 24-
hour averaging period is based on the 3-year average of the annual 98* percentile 24-hour
average PM2.5 concentrations. Details regarding the determination of the annual 98th percentile
monitored 24-hour value based on the number of days sampled during the year are provided in
the data interpretation procedures for the PIVb.s NAAQS, Appendix N to 40 CFR Part 50.
It should be noted here that although the monitored design values for the PM2.5 standards
are defined in terms of 3-year averages, this definition does not preempt or alter the Appendix W
requirement for use of 5 years of representative NWS meteorological data or at least 1 year of
site-specific data for air quality modeling purposes.31 The 5-year average based on use of
representative NWS meteorological data, or an average across one or more (up to 5) complete
years of available site-specific data, serves as an unbiased estimate of the 3-year average for
purposes of modeling demonstrations of compliance with the NAAQS. Modeling of "rolling 3-
year averages," using years 1 through 3, years 2 through 4, and years 3 through 5 as
recommended in the EPA's "Guidance on the Use of Models and Other Analyses for
Demonstrating Attainment of Air Quality Goals for Ozone, PM2.5, and Regional Haze", is not
required.32
The EPA's March 23, 2010, clarification memo recommended as a First Tier that the
modeled annual (or 24-hour) concentrations of primary PM2.5 to be added to the monitored
31 See 40 CFR Part 51, Appendix W, Section 8.3.1.2.b.
32 The "Guidance on the Use of Models and Other Analyses for Demonstrating Attainment of Air Quality Goals for
Ozone, PM2.5, and Regional Haze" can be found on the SCRAM website at:
http://www.epa.gov/ttn/scram/guidance/guide/fmal-03-pm-rh-guidance.pdf
57
-------
annual (or 24-hour) design value for comparison to the NAAQS should be based on the highest
average of the modeled annual (or 24-hour) averages across 5 years for representative NWS
meteorological data or the highest modeled annual (or 24-hour) average for one year (or multi-
year average of 2 up to 5 complete years) of site-specific meteorological data using the same
procedures recommended for the initial source impact analysis. The memo cited several issues,
especially the importance of the contribution from secondary formation of PIVb.s from precursor
emissions and the fact that such contributions are not explicitly accounted for by the dispersion
model, as the basis for viewing modeling of PM2.5 as screening-level analyses, analogous to the
screening nature of the guidance in Section 5.2.4 of Appendix W regarding dispersion modeling
for NC>2 impacts, given the importance of chemistry in the conversion of NO emissions to
ambient NC>2.
Recognizing that the primary focus and motivation for this guidance is to provide
recommendations on appropriate tools and methodologies to account for the potential
contribution from a new or modifying source's precursor emissions on ambient PM2.5 levels, it is
appropriate to reassess the EPA's March 23, 2010, guidance under this broadened paradigm.
Since each of the four cases outlined above, based on comparisons of the project's direct PM2.5
and precursor emissions with the respective SERs, involves some assessment of the source's
potential secondary PIVb.s impacts, we recommend as a new First Tier that the modeled design
value be added to the monitored design value from a representative monitor. This represents no
fundamental change with respect to the modeled annual concentration. However, the modeled
24-hour concentration to be added to the monitored design value would now be based on the
multi-year average of the 98th percentile of modeled annual 24-hour concentrations rather than
the multi-year average of the highest (100* percentile) of modeled annual 24-hour
58
-------
concentrations.
For Case 2, where only the project's direct PM2.5 emissions are equal to or greater than
the SER, the modeled design value should be based on AERMOD (or other acceptable preferred
or alternative model) estimates of primary PM2.5 impacts combined with the monitored design
value. The monitor should be representative in that it accounts for the contribution of secondary
PM2.5 formation associated with existing sources within the modeling domain, in addition to the
background levels of primary PM2.5 associated with background sources that are not included in
the modeled inventory. For Case 3, where both the project's direct PM2.5 emissions and precursor
emissions are equal to or greater than the respective SERs, the cumulative impact for comparison
to the NAAQS should be based on the sum of the modeled design value for primary PM2.5
impacts (from dispersion model estimates based on the project's and other nearby source's direct
PM2.5 emissions), the modeled design value for secondary PIVb.s impacts (from a qualitative,
hybrid, or quantitative assessment accounting for the project's precursor PM2.5 emissions), and
the monitored design value (same representativeness caveats as with Case 2). For Case 4, where
only the project's precursor emissions are equal to or greater than the respective SERs, the
cumulative impact for comparison to the NAAQS should be based on the sum of the modeled
design value for secondary PM2.5 impacts (from a qualitative, hybrid, or quantitative assessment
as with Case 3) and the monitored design value (same representativeness caveats as with Cases 2
and 3). The resulting cumulative PM2.5 concentrations would then be compared to the annual
PM2.5 NAAQS of 12 ug/m3 and 24-hour PM2.5 NAAQS of 35 ug/m3.
The recommendations provided above constitute a First Tier modeling analysis for PM2.5
NAAQS compliance demonstrations that should be acceptable without further justification. For
applications where impacts from primary PM2.5 emissions are not temporally correlated with
59
-------
background PM2.5 levels, combining the modeled and monitored contributions as described
above may be overly conservative in some situations. For example, there are areas of the country
where background PM2.5 levels are substantially higher on average during the summer months as
compared to the winter months; however, the projected modeled impacts from the new or
modified source may be substantially greater in the winter rather than in the summer. In such
cases, a Second Tier modeling analysis that would involve combining the monitored and
modeled PM2.5 concentrations on a seasonal (or quarterly) basis may be considered. The use of a
seasonally-varying monitored background component is likely to be a more important factor for
the 24-hour NAAQS analysis than for the annual NAAQS. Careful evaluation of when model
projections of PM2.5 impacts and background PM2.5 levels peak throughout the year is strongly
advised before embarking on a Second Tier modeling analysis. This is because the First Tier
approach may adequately capture the temporal correlation and would otherwise make a Second
Tier modeling analysis unnecessary. As a part of this evaluation process, consultation with the
appropriate permitting authority is advised.
The AERMOD model provides several options for specifying the monitored background
concentration for inclusion in the cumulative impact assessment. The options that are most
relevant to PM2.5 analyses include an option to specify a single annual background concentration
that is applied to each hour of the year (appropriate for the First Tier annual and 24-hour
analyses described above), and an option to specify four seasonal background values that are
combined with modeled concentrations on a seasonal basis (appropriate for a Second Tier 24-
hour analysis). The AERMOD model also allows the user to track the contribution from
background concentrations to the cumulative modeled design value.
For the Second Tier 24-hour modeling analyses, it is recommended that the distribution
60
-------
of monitored data equal to and less than the annual 98* percentile be appropriately divided into
seasons (or quarters) for each of the three years that are used to develop the monitored design
value. This results in data for each year (for three years) which contains one season (quarter)
with the 98th percentile value and three seasons (quarters) with the maximum values which are
less than or equal to the 98* percentile value. The maximum concentration from each of the
seasonal (or quarterly) subsets should then be averaged across these three years of monitoring
data. The resulting average of seasonal (or quarterly) maximums should then be included as the
four seasonal background values within the AERMOD model. Therefore, the monitored
concentrations greater than the 98* percentile in each of the three years would not be included in
the seasonal (or quarterly) subsets. These excluded monitored concentrations are the same
values that are excluded when determining the monitored design value. An example of the
calculations for a Second Tier 24-hour modeling analysis is provided in Appendix E.
For a monitor with a daily (1-in-l day monitor) sampling frequency and 100% data
completeness, this would mean that the top seen monitored concentrations for each year would
be excluded from the seasonal (or quarterly) subdivided datasets. Similarly, for a monitor with
every third day (l-in-3 day monitor) sampling frequency and 100% data completeness, the top
two monitored concentrations for each year would be excluded from the seasonal (or quarterly)
subdivided datasets. The monitored concentrations excluded from the subdivided datasets could
primarily come from one or two seasons (or quarters) each year or could be evenly distributed
across all four seasons (or quarters) each year. Additionally, the monitored concentrations not
included in the subdivided datasets could shift seasonally (or quarterly) from one year to the
next. Given the reasoning for considering a Second Tier 24-hour analysis (lack of temporal
correlation between modeled and monitored concentrations), it is likely that the monitored data
61
-------
greater than the 98* percentile would be concentrated in one or two season as opposed to evenly
distributed throughout the year. As mentioned earlier, one should reference Appendix N of 40
CFR Part 50 to determine the appropriate 98 percentile rank of the monitored data based on the
monitor sampling frequency and valid number of days sampled during each year.
Since several recent permit applications have come to our attention proposing to combine
monitored background and modeled concentrations on an hour-by-hour basis, using hourly
monitored background data collected concurrently with the meteorological data period being
processed by the model, we feel compelled to include a discussion of the potential merits and
concerns regarding such an approach in the context of PM2.5 NAAQS compliance
demonstrations. On the surface, the hourly pairing or "paired sums" approach could be perceived
as being a more "refined" method than what is recommended in the First or Second Tier methods
and, therefore, more appropriate for assessing the impacts from primary PIVb.s emissions.
However, the implicit assumption underlying this approach is that the background monitored
levels for each hour are spatially uniform and that the monitored values are fully representative
of background levels at each receptor for each hour. Such an assumption clearly ignores the
many factors that contribute to the temporal and spatial variability of ambient PM2.5
concentrations across a typical modeling domain on an hourly basis.
The complexities of the PM2.5 ambient monitoring network also present special
challenges with a "paired sum" approach that are not present with the other NAAQS pollutants.
The Federal Reference Method (FRM) PM2.5 monitoring network is based on 24-hour samples
that are taken on average every third day at the l-in-3 day monitors. The frequency of daily or 1-
in-1 day PIVb.s monitors is steadily increasing but is relatively limited to the largest cities and
metropolitan regions of the U.S. Various methods to "data fill" the l-in-3 day monitoring
62
-------
database to create a pseudo-daily dataset have been explored in a few situations , but none of
these data filling methods have been demonstrated to create a representative daily PM2.5 dataset
that the EPA would consider acceptable for inclusion in a PM2.5 NAAQS compliance
demonstration. The use of continuous PM2.5 monitors, which are more limited in number
compared to the FRM monitors and may require careful quality assurance of individual hourly
measurements, may be an option but should be discussed in advance with the appropriate
permitting authority.
Considering the spatial and temporal variability throughout a typical modeling domain on
an hourly basis and the complexities and limitations of hourly observations from the current
PM2.5 ambient monitoring network, we do not recommend a "paired sums" approach on an hour-
by-hour basis. Furthermore, the pairing of daily monitored background and 24-hour average
modeled concentrations is not recommended except in rare cases of relatively isolated sources
where the available 1-in-l day FRM/FEM monitor can be shown to be representative of the
ambient concentration levels in the areas of maximum impact from the proposed new source. In
most cases, the seasonal (or quarterly) pairing of monitored and modeled concentrations
previously described in the Second Tier method should sufficiently address situations to which
the impacts from primary PM2.5 emissions are not temporally correlated with background PM2.5
levels. Any monitor-model pairing approach aside from the First or Second Tier methods should
be justified on a case-by-case basis in consultation with the appropriate permitting authority and
the appropriate EPA Regional Office.
63
-------
IV.4 Determining Whether Proposed Source Causes or Contributes to Modeled
Violations
If the cumulative impact assessment following these recommendations results in modeled
violations of the PM2.5 NAAQS, then the permit applicant will need to determine whether the
project's emissions cause or contribute to the modeled violations. The EPA has previously
supported showing the proposed source does not cause or contribute by showing that the source
does not make a "significant contribution" to the modeled violation based on a comparison of the
modeled impacts from the project emissions associated with the modeled violation, paired in
time and space, to the SIL for the relevant pollutant and averaging period contained in 40 CFR
51.165(b) of the EPA's regulations. The EPA has interpreted this regulation to support the
conclusion that a source with an impact below the relevant value in section 51.165(b)(2) does not
significantly contribute to either an existing violation of the NAAQS in a nonattainment area or
violations predicted in an attainment area based on a cumulative analysis.33
The January 22, 2013, court decision did not vacate the PM2.5 SIL value in section
51.165(b) of the EPA's regulations. However, the court recognized that the language in section
51.165(b)(2) operates in a manner different from sections 51.166(k)(2) and 52.21(k)(2), which
were vacated by the court. The court observed that section 51.165(b)(2) "simply states that a
source may be deemed to violate the NAAQS if its exceeds the SILs in certain situations." (705
F.3d at 465-66). For this reason, the court did not see the need to resolve the Petitioner's
challenge to the EPA's methodology for determining the PM2.5 values in section 51.165(b)(2) of
the regulations, which are the same as the Class II area values in the vacated sections
51.166(k)(2) and 52.21(k)(2). The court decision did not directly address the use of the values in
33 See 75 Fed. Reg. at 64,890; 61 Fed. Reg. 38,250, 38,293 (July 23, 1996); In re Prairie State Generating Co., 13
E.A.D. 1, 103-09 (EAB 2006). EPA has sometimes described this step as a "culpability analysis."
64
-------
section 51.165(b)(2) to determine whether a source causes or significantly contributes to a
modeled violation. However, in light of other elements of the court decision, the EPA advises
permitting authorities to consult with the EPA before using the SIL value for PM2.5 in section
51.165(b)(2) as the basis for concluding that a source with an impact below this value does not
cause or contribute to a modeled violation.
A demonstration that a proposed source does not make a significant contribution should
be based on a comparison of the modeled concentrations at the receptor location showing the
violation to a SIL, across 5 years for representative NWS meteorological data and the modeled
concentration for 1 year, or multiyear average of 2 up to 5 complete years, of site-specific
meteorological data. For a violation of the annual PJVb.s NAAQS, the average of the predicted
annual concentrations at the affected receptor(s) should be compared to a SIL, while the average
of the predicted annual 98th percentile 24-hour average concentrations at the affected receptor(s)
should be used for the 24-hour PM2.5 NAAQS.
65
-------
This Page Intentionally Left Blank
66
-------
V. PSD Increments for PM2.5
As cited in Section II of this guidance, section 165(a)(3) of the CAA requires that
proposed new and modified major stationary sources seeking a PSD permit must demonstrate
that their proposed emissions increases will not cause or contribute to a violation of any NAAQS
or PSD increment. Based on the flow diagram presented in Figure II-2 in Section II, this section
describes the EPA's recommendations for completing the required analysis of the PSD
increments for PM2.5.
V.I Overview of PSD Increments
The term "increment" generally refers to what the CAA calls the "maximum allowable
increase" of an air pollutant that is allowed to occur above the applicable baseline air quality
concentration for that pollutant. Thus, by establishing the maximum allowable increase for a
particular pollutant and averaging period, any cumulative increase in the ambient concentration
of that pollutant that is greater than the amount allowed is considered "significant deterioration."
In order to apply the increment concept as part of a PSD permit review, it is necessary to
identify the affected geographic area in which the increment will be tracked and the emissions
changes that affect increment. The relevant geographic area for determining the amount of
increment consumed is known as the "baseline area." 34 The baseline area may be comprised of
one or more attainment or unclassifiable areas for a particular pollutant that are in a particular
state. In accordance with the definition of "baseline area," the area is an "intrastate area" and
does not include any area in another state. At a minimum, the baseline area is the attainment or
unclassifiable area in which a PSD source will locate. Within any baseline area, three key dates
34 "Baseline area" is defined in the PSD regulations at 40 CFR 51.166(b)(15) and 52.21(b)(15).
67
-------
will apply in order to track increment: (1) trigger date; (2) minor source baseline date; and (3)
major source baseline date. The trigger date is a fixed date, which is the earliest date after which
proposed sources must track increment in the baseline area. In turn, the minor source baseline
date is the date on which the first PSD application in a baseline area is submitted to the PSD
permitting authority after the trigger date. Depending upon the number of separate attainment
areas that exist for a particular pollutant in the state, there may be a number of minor source
baseline dates that apply to different baseline areas established in that state. Beginning with the
PSD source whose complete application has established the minor source baseline date in a
particular area, any increase or decrease in actual emissions from any major or minor source
henceforth will consume or expand the available PSD increments for that baseline area. Finally,
the major source baseline date is a fixed date, which precedes the trigger date, after which
construction related emissions solely from major stationary sources affect increment, as further
explained below.
PM2.5 emissions changes occurring before the minor source baseline date generally do not
impact increment in an area, but are considered to contribute to the baseline air quality level also
known as the baseline concentration, as described in more detail below. However, it is important
to note that the CAA provides an exception for certain emissions changes that occur specifically
at major stationary sources regardless of when those emissions changes actually occur. This date,
as explained above, is the "major source baseline date." Specifically, for projects at major
stationary sources on which construction commenced at a date prior to the major source baseline
date, the emissions increases from such projects should be considered to contribute to the
baseline air quality level even though the emissions change may not actually occur until after the
minor source baseline date. Alternately, for projects at major stationary sources on which
68
-------
construction commenced after the major source baseline date, the project emissions will be
considered to affect increment, even if the project actually begins operation before the minor
source baseline date.
V.2 PMi.s Increments Considerations
In its 2010 PM2.5 Increments, SILs, and SMC Rule, the EPA established PM2.5
increments at the levels shown in Table V-l, as follows:
Table V-l.
Increments
Increments, (Jg/m3
Annual arithmetic mean
24-hour maximum
Class I
l
2
Class II
4
9
Class III
8
18
Source: Prevention of Significant Deterioration (PSD) for Participate Matter Less Than 2.5 Micrometers (PM2.5) - Increments,
Significant Impact Levels (SILs) and Significant Monitoring Concentration (SMC) final rule (75 FR 64864)
The PM2.5 increments analysis includes many of the same technical considerations in
assessing source impacts as discussed earlier in this guidance for PM2.5 NAAQS compliance
demonstrations, specifically the assessment cases described in Section II-4 and detailed in
Table III-l. However, there are some important differences. The main difference is that the
increments compliance demonstration is based on calculating the change in ambient PM2.5
concentrations over the applicable baseline concentration, which includes proposed emissions
increases from the new or modified source, increment-consuming emissions from other sources
that affect increment consumption in the baseline area, and increment-expanding decreases in
emissions from the same sources. Another key difference is that the cumulative impact analysis
for increments is based on the actual emission changes occurring after a prescribed minor source
baseline date (with the stated exception related to major sources commencing construction after
69
-------
the major source baseline date), whereas NAAQS analyses are generally based on the cumulative
impact associated with the maximum allowable emissions from the new or modifying source and
other nearby sources. Finally, it is important to note that the PM2.5 NAAQS and increments for
the 24-hour averaging period are defined in different forms and therefore must be analyzed
differently.35 The 24-hour PM2.5 NAAQS is defined based on the 3-year average of the annual
98th percentile of the 24-hour average concentrations, while the 24-hour PM2.5 increments are
based on the second highest maximum 24-hour concentration.
The 2010 "PM2.5 Increments, SILs, and SMC Rule" established October 20, 2011, as the
"trigger date" and October 20, 2010, as the "major source baseline date" for PM2.5 increments.
The EPA developed the increment system for PM2.5 generally following the same concepts that
were previously applied for the increments for PMio, SO2, and NO2. In each case, the framework
reflects the statutory concepts set forth in the definition of "baseline concentration" contained in
the CAA at section 169(4), which reads as follows:
The term "baseline concentration" means, with respect to a pollutant, the ambient
concentration levels which exist at the time of the first application for a permit in an area
subject to this part, based on air quality data available in the Environmental Protection
Agency or a State air pollution control agency and on such monitoring data as the permit
applicant is required to submit. Such ambient concentration levels shall take into account
all projected emissions in, or which may affect, such area from any major emitting
facility on which construction commenced prior to January 6, 1975, but which has not
begun operation by the date of the baseline air quality concentration determination (i.e.,
the minor source baseline date). Emissions of sulfur oxides and particulate matter from
35 The annual NAAQS and increments for PM2 5 are both measured as annual arithmetic mean values.
70
-------
any major emitting facility on which construction commenced after January 6, 1975, shall
not be included in the baseline and shall be counted against the maximum allowable
increases in pollutant concentrations established under this part.
Thus, from this definition, it can be seen that it is conceptually possible to measure "significant
deterioration" in at least two separate ways. That is, either as (1) a direct modeled projection of
the change in air quality after the applicable baseline date caused by all increment-consuming or
expanding emissions compared to the maximum allowable increase of the air pollutant
concentration (increment) in the baseline area, or (2) a determination of whether the ambient air
quality concentration in a baseline area will exceed an allowable ambient air quality ceiling,
determined by adding the maximum allowable pollutant concentration increase (increment) to
the baseline air quality concentration (baseline concentration) for the baseline area.
Historically, because of various limitations associated with the use of ambient air quality
monitoring data for measuring increment consumption,36 the EPA elected to determine
significant deterioration exclusively on the basis of the first approach, which models only the
increment-related emissions increases or decreases to determine the resulting ambient air quality
change and compares this value with maximum allowable pollutant concentration increases
(increments) for a particular pollutant. However, the present technical challenges associated with
the ability to estimate the impacts of secondarily formed PM2.5 in the atmosphere resulting from
emissions of PM2.5 precursors make it necessary to consider alternative methods of assessing
increments where the increments are affected by both direct PM2.5 emissions and PM2.5 precursor
36 The EPA described certain limitations associated with the use of ambient air quality monitoring data for
measuring increment consumption in the preamble to its proposed PSD regulations in 1979. For example, the CAA
provided that certain emissions changes should not be considered to be increment consuming. These limitations
generally continue to apply to the extent that certain emissions changes detected by an ambient monitor are not
considered to consume increment. See 44 Fed. Reg. 51924 at 51944 (September 5, 1979).
71
-------
emissions that form secondary PM2.5. Accordingly, the remainder of this section provides
recommendations for accomplishing the PM2.5 increments analysis.
V.3 Screening Analysis for Increments
The comparison of background air quality concentrations and the NAAQS, as
recommended in Section II of this document as an initial step for the NAAQS compliance
demonstration, would not by itself provide adequate justification for foregoing a cumulative
modeling analysis for the PM2.5 increments. Such an approach would be inappropriate because it
would not ensure that there is sufficient "headroom" within the allowable increment to absorb a
source contribution equal to the SIL. However, a permitting authority may still be able to justify
reaching a determination that a new or modified source will not cause or contribute to a violation
of the increments without performing cumulative modeling for increments.
The EPA recommends that a justification for not performing cumulative modeling for
PM2.5 increments compliance should be based on (1) a comparison of the predicted impacts of
the new or modified source and the allowable increment values, (2) information on the extent to
which, if any, increment has already been consumed since either the major source baseline date
(for major source construction prior to the minor source baseline date) or minor source baseline
date by nearby sources that have been permitted prior to the source under analysis, and (3)
information on increment consumption or expansion by more distant sources.
Since the trigger date has only recently been established (i.e., October 20, 2011), for the
next several years a new or modified source being evaluated for increments compliance will
often be the first source with increment-consuming emissions in the area. As indicated in Figure
II-2, under this situation, a permitting authority may have sufficient reason to conclude that the
72
-------
impacts of the new or modified source (based on the approach for conducting source impact
analysis described below) may be compared directly to the allowable increments, without the
need for a cumulative modeling analysis. Such an approach would be appropriate when the new
or modified source represents the first PSD application in the area after the trigger date, which
establishes the minor source baseline date and baseline area, and no relevant major source
construction has already occurred since the major source baseline date.
V.4 PMi.s Increments Analysis
The guidance provided under Sections III and IV regarding NAAQS compliance
demonstrations should generally be applicable for PIVb.s increments analyses, with the primary
distinction that actual emission increases (or decreases) from only increment-affecting sources
may be used instead of maximum allowable emissions in the cumulative impact analysis.
V.4.1 Source Impact Analysis
The EPA's recommendations on conducting the source impact analysis for PM2.5
increments rely upon the same four assessment cases for NAAQS, as described in Section II.4.
As shown in Table V-2, a modeled compliance demonstration is not required for Case 1 since
neither direct PIVb.s emissions nor PIVb.s precursor (NOX and/or 862) emissions are equal to or
greater than the respective SERs. Case 1 is the only assessment case that does not require a
modeled compliance demonstration for PIVb.s, whereas each of the remaining three assessment
cases would necessitate a source impact analysis that should be conducted following the detailed
recommendations provided in Section III for NAAQS analysis.
73
-------
Table V-2. EPA Recommended Approaches for Assessing Primary and Secondary PMi.5
Impacts by Assessment Case
As s es s ment Cas e
Case 1:
No Air Quality Analysis
Case 2:
Primary Air Quality
Impacts Only
Case 3:
Primary and Secondary
Air Quality Impacts
Case 4:
Secondary Air Quality
Impacts Only
Description of Assessment Case
Direct PM2.5 emissions < 10 tpy SER
NOxand SO2 emissions < 40 tpy SER
Direct PM2.5 emissions > 10 tpy SER
NOxand SO2 emissions < 40 tpy SER
Direct PM2.5 emissions > 10 tpy SER
NOx an d/or SO2 emissions > 40 tpy SER
Direct PM2.5 emissions < 10 tpy SER
NOx an d/or SO2 emissions > 40 tpy SER
Primary Impacts Approach
N/A
Appendix W preferred or
approved alternative
dispersion model
Appendix W preferred or
approved alternative
dispersion model
N/A
Secondary Impacts
Approach
N/A
N/A
Qualitative
Hybrid qualitative /
quantitative
Full quantitative
photochemical
grid modeling
Qualitative
Hybrid qualitative /
quantitative
Full quantitative
photochemical
grid modeling
V.4.2 Cumulative Impact Analysis
Where the screening analysis described above is insufficient to show that a source will
not cause or contribute to a violation of the PSD increments, a cumulative impact assessment
would be necessary to make the demonstration. A cumulative assessment accounts for the
combined impact of the new or modifying source's emissions and those emissions changes from
sources that affect the increments. The cumulative impacts are then compared to the PSD
increments to determine whether the new or modifying source emissions will cause or contribute
to a violation of the PSD increments. This section provides details on conducting an appropriate
cumulative impact assessment for PM2.5.
V.4.2.1 Assessing Primary PMi.5 Impacts from Other Sources
To assess direct PM2.5 emissions from increment-consuming or increment-expanding
sources, the PM2.5 increments analysis would follow the traditional approach involving modeling
of only PM2.5 emissions changes that affect the increment, and should be based on application of
74
-------
AERMOD (or other appropriate preferred or approved alternative model), using actual emission
changes associated with any increment-consuming or increment-expanding sources. The
AERMOD model allows for inclusion of these emissions (represented as negative emissions for
the sources expanding increment) in the same model run that includes the allowable increase in
emissions from the project source, and will therefore output the net cumulative concentrations
(although the "maximum" cumulative impacts will be output as zero if the cumulative impacts
computed in the model are less than zero).
V.4.2.2 Assessing Secondary PMi.5 Impacts from Other Sources
To assess changes in PM2.5 precursor emissions from increment-consuming or increment-
expanding sources, the assessment of potential impacts of secondary PM2.5 due to those
emissions changes may be: a) qualitative in nature; b) based on a hybrid of qualitative and
quantitative assessments utilizing existing technical work; or c) a full source-specific quantitative
photochemical modeling exercise.
Several promulgated rules have resulted in reductions in precursor emissions affecting
ambient PM2.5 concentrations across most areas in recent years.37 This is particularly true in the
Eastern U.S. As a result, in many cases, the potential for increment consumption due to
secondary PM2.5 impacts from existing sources may easily be addressed through a qualitative
assessment, supported by data that generally confirms a downward trend in precursor emissions
occurring after the applicable PIVb.s minor source baseline date (or the major source baseline
date). In such cases, the PM2.5 increments modeling analysis may be simplified to focus solely on
potential increment consumption associated with direct PM2.5 emissions. For areas where PM2.5
37 Such rules would include the Clean Air Interstate Rule (CAIR), Mercury and Air Toxics Standards Rule (MATS),
NOx SIP Call and multiple federal mobile source rules.
75
-------
precursor emissions increases from other sources are known to contribute to PM2.5 concentration
increases within the baseline area and thus consume PM2.5 increment, the photochemical grid
modeling methods discussed in Section III may be appropriate for estimating the portion of
PM2.5 increment consumed due to secondary PIVb.s impacts associated with those increases in
precursor emissions.
V.4.2.3 Consideration of PMi.s Ambient Air Quality Monitoring Data
In light of the current technical complications associated with the ability to model
precursor emissions to estimate secondarily formed PM2.5 in the atmosphere, the EPA believes it
may be possible under certain circumstances to use ambient air quality monitoring data for PIVb.s
as part of the cumulative impact analysis. This involves using ambient monitoring data as the
primary means of assessing increment consumption or expansion for PM2.5 by measuring
ambient air quality on the minor source baseline date (baseline concentration) and thereafter to
determine changes in air quality resulting from direct PM2.5 emissions and PM2.5 precursors. This
document does not provide detailed recommendations for conducting the PM2.5 increments
analysis in this manner, but simply acknowledges that it may be possible in certain
circumstances to use this approach for PSD permitting. There would continue to be a need to
model projected impacts as part of the PM2.5 increments analysis to include consideration of
increment consumed by emissions that have not yet occurred. One should also consider the
extent to which the available monitoring data adequately reflect the air quality changes caused by
direct PM2.5 and precursor emissions from sources impacting the baseline area.
Where the PSD permit applicant believes that this approach is potentially useful for
conducting the PM2.5 increments analysis for a particular PSD permit review, early coordination
76
-------
with the permitting authority is strongly encouraged to establish the appropriate baseline
concentration(s) within the applicable baseline area and how subsequent ambient monitoring
data in the area, when compared to the baseline air quality data, can be used to assess cumulative
increment consumption. The EPA will work with air agencies to support this approach on a case
specific basis. Based on these experiences, it is our intention to provide additional guidance
setting forth more specific recommendations on this particular approach at a future date.
V.5 Determining Significant Contribution to an Increment Violation
As previously explained, the EPA does not anticipate the need to complete a cumulative
increments analysis in most situations due to the recent setting of the trigger date for PM2.5.
Therefore, most PM2.5 increments analyses will need to consider the emissions increases
resulting only from the proposed new source or modification that establishes the minor source
baseline date for an area. Consequently, we believe that permitting authorities will encounter
few, if any, situations over the next several years in which there is a predicted increment
violation.
Nevertheless, there may be situations where a cumulative increments analysis is
necessary and that analysis projects a modeled increment violation. This guidance recommends
that such violations be addressed in a manner similar to the NAAQS analysis described in
Section IV of this document; that is, when a PSD applicant elects to use a SIL to show to the
permitting authority that the source's emissions do not make a significant contribution to a
modeled violation, the EPA advises permitting authorities to consult with the EPA before
allowing the use of a SIL value, including those PM2.5 values contained in section 51.165(b)(2),
as the basis for concluding that a source with an impact below this value does not cause or
77
-------
contribute to a modeled violation of the PM2.5 increment.
78
-------
VI. References
U.S. EPA, 1990: New Source Review Workshop Manual: Prevention of Significant
Deterioration and Nonattainment Area Permitting - DRAFT. U.S. Environmental
Protection Agency, Research Triangle Park, North Carolina 27711.
http://www.epa.gov/ttn/nsr/gen/wkshpman.pdf
U.S. EPA, 1992: Protocol for Determining the Best Performing Model. September 1992. EPA-
454/R-92-025. U.S. Environmental Protection Agency, Research Triangle Park, North
Carolina 27711.
U.S. EPA, 1997: Interim Implementation of New Source Review Requirements for PM2.5. John
Seitz Memorandum dated October 23, 1997. U.S. Environmental Protection Agency,
Research Triangle Park, North Carolina, 27711.
http://www.epa.gov/region7/air/nsr/nsrmemos/pm25.pdf
Seinfeld and Pandis, 1998. Atmospheric Chemistry and Physics: From Air Pollution to Climate
Change. J. Seinfeld and S. Pandis. Wiley Interscience. New York, New York. ISBN 0 47
1178152.
Nobel et al., 2001: Accounting for spatial variation of ozone productivity inNOx emission
trading. C. Nobel, E. McDonald-Buller, Y. Kimura, and D. Allen. Environmental Science
& Technology. 2001; 35, 4397-4407.
NARSTO, 2004. Particulate Matter Assessment for Policy Makers: A NARSTO Assessment. P.
McMurry, M. Shepherd, and J. Vickery, eds. Cambridge University Press, Cambridge,
England. ISBN 0 52 184287 5.
U.S. EPA, 2005. Guideline on Air Quality Models. 40 CFRPart 51 Appendix W.
http://www.epa.gov/ttn/scram/guidance/guide/appw_05.pdf
Byun and Schere, 2006: Review of the governing equations, computational algorithms, and other
components of the models-3 Community Multiscale Air Quality (CMAQ) modeling
system. D. Byun and K. Schere. Applied Mechanics Reviews. 2006; 59, 51-77'.
U.S. EPA, 2007a: Guidance on the Use of Models and Other Analyses for Demonstrating
Attainment of Air Quality Goals for Ozone, PM2.s, and Regional Haze. April 2007. EPA-
454/B-07-002. U.S. Environmental Protection Agency, Research Triangle Park, North
Carolina 27711. http://www.epa.gov/ttn/scram/guidance/guide/fmal-03-pm-rh-
guidance.pdf
U.S. EPA, 2007b: Details on Technical Assessment to Develop Interpollutant Trading Ratios for
PM2.5 Offsets. Technical Analysis dated July 23, 2007. U.S. Environmental Protection
Agency, Research Triangle Park, North Carolina 27711.
Bergin et al., 2008. Single Source Impact Analysis Using Three-Dimensional Air Quality
Models. M. Bergin, A. Russell, T. Odman, D. Cohan, and W. Chameldes. Journal of the
Air & Waste Management Association. 2008; 58, 1351-1359.
Russell, 2008: EPA Supersites Program-related emissions-based particulate matter modeling:
Initial applications and advances. A. Russell. Journal of the Air & Waste Management
Association. 2008; 58, 289-302.
Foley et al., 2010: Incremental testing of the Community Multiscale Air Quality (CMAQ)
modeling system version 4.7. K. Foley, S. Roselle, K. Appel, P. Bhave, J. Pleim, T. Otte,
R. Mathur, G. Sarwar, J. Young, R. Gilliam, C. Nolte, J. Kelly, A. Gilliland, and J, Bash.
Geoscientific Model Development. 2010; 3, 205-226.
79
-------
U.S. EPA, 2010a: Model Clearinghouse Review of Modeling Procedures for Demonstrating
Compliance with PM2.5 NAAQS. Tyler Fox Memorandum dated February 26, 2010. U.S.
Environmental Protection Agency, Research Triangle Park, North Carolina 27711.
http://www.epa.gov/ttn/scram/guidance/mch/new mch/MCmemo Region6 PM25 NAA
QS_Compliance.pdf
U.S. EPA, 2010b: Modeling Procedures for Demonstrating Compliance with PM2.5 NAAQS.
Stephen Page Memorandum dated March 23, 2010. U.S. Environmental Protection
Agency, Research Triangle Park, North Carolina, 27711.
http://www.epa.gov/ttn/scram/Official%20Signed%20Modeling%20Proc%20for%20De
mo%20Compli%20w%20PMM.pdf
ENVIRON, 2011: User's Guide Comprehensive Air Quality Model with Extensions. ENVIRON
International Corporation, Novato, California, http://www.camx.com.
U.S. EPA, 201 la: Addendum - User's Guide for the AERMOD Terrain Preprocessor
(AERMAP). EPA-454/B-03-003. U.S. Environmental Protection Agency, Research
Triangle Park, North Carolina 27711.
http://www.epa.gov/ttn/scram/models/aermod/aermap/aermap userguide.zip
U.S. EPA, 201 Ib: AERSCREEN User's Guide. EPA-454-/B-11-001. U.S. Environmental
Protection Agency, Research Triangle Park, North Carolina 27711.
http://www.epa.gov/ttn/scram/models/screen/aerscreen_userguide.pdf
U.S. EPA, 201 Ic: AERSCREEN Released as the EPA Recommended Screening Model. Tyler
Fox Memorandum dated April 11, 2011. U.S. Environmental Protection Agency,
Research Triangle Park, North Carolina 27711.
http://www.epa.gov/ttn/scram/20110411 AERSCREEN Release Memo.pdf
U.S EPA, 201 Id AERMINUTE User's Guide. U.S. Environmental Protection Agency, Research
Triangle Park, North Carolina 27711.
http://www.epa.gov/ttn/scram/7thconf/aermod/aerminute vll325.zip
U.S. EPA, 201 le: Revised Policy to Address Reconsideration of Interpollutant Trading
Provisions for Fine Particles (PM2.s). Gina McCarthy Memorandum dated July 21, 2011.
U.S. Environmental Protection Agency, Washington, District of Columbia 20460.
http://www.epa.gov/region7/air/nsr/nsrmemos/pm25trade.pdf.
U.S. EPA, 201 If: Additional Clarification Regarding Application of Appendix W Modeling
Guidance for the 1-hour NO2 National Ambient Air Quality Standard. Tyler Fox
Memorandum dated March 1, 2011. U.S. Environmental Protection Agency, Research
Triangle Park, North Carolina 27711.
http://www.epa.gov/ttn/scram/guidance/clarifi cation/Additional_Clarifications_Appendix
W Houiiy-NO2-NAAQS FINAL 03-01-2011.pdf
Baker and Foley, 2011. A nonlinear regression model estimating single source concentrations of
primary and secondarily formed PM2.5. K. Baker and K. Foley. Atmospheric
Environment. 2011; 45:3758-67.
Cohan and Napelenok, 2011. Atmospheric Response Modeling for Decision Support. D. Cohan
and S. Napelenok. Atmosphere. 2011; 2(3): 407-425.
80
-------
NACAA, 2011: PIVb.s Modeling Implementation for Projects Subject to National Ambient Air
Quality Demonstration Requirements Pursuant to New Source Review. Report from
NACAA PM2.5 Modeling Implementation Workgroup dated January 7, 2011.
Washington, District of Columbia 20001.
http ://www. epa.gov/ttn/scram/1 Othmodconf/review_material/01072011-
NACAAPM2.5ModelingWorkgroupReport-FINAL.pdf
U.S. EPA, 2012a: "Sierra Club Petition Grant". Gina McCarthy Administrative Action dated
January 4, 2012. U.S. Environmental Protection Agency, Washington, District of
Columbia 20460.
http://www.epa.gov/ttn/scram/10thmodconf/review material/Sierra Club Petition OAR-
ll-002-1093.pdf
U.S. EPA, 2012b: Summary of Public Comments, 10th Conference on Air Quality Modeling.
U.S. Environmental Protection Agency, Research Triangle Park, North Carolina, 27711.
http://www.epa.gov/ttn/scram/10thmodconf/10thMC Summary of Comments-
Revised 10-05-2012.pdf
U.S. EPA, 2013a: Transportation Conformity Guidance for Quantitative Hot-spot Analyses in
PM2.5 and PMio Nonattainment and Maintenance Areas. November 2013. EPA-420-B-
10-040. U.S. Environmental Protection Agency, Ann Arbor, Michigan 48105.
http://www.epa.gov/oms/stateresources/transconf/policy/420bl3053-sec.pdfand
http://www.epa.gov/oms/stateresources/transconf/policy/420bl3053-appx.pdf.
U.S. EPA, 2013b: Draft Guidance for PM2.5 Permit Modeling. March 4, 2013. EPA-454-/B-11-
001. U.S. Environmental Protection Agency, Research Triangle Park, North Carolina
27711.
http://www.epa.gov/ttn/scram/guidance/guide/Draft Guidance for PM25 Permit Model
ing.pdf
U.S. EPA, 2014a: Interim Guidance on the Treatment of Condensable Parti culate Matter Test
Results in the PSD and NSR Permitting Programs. Stephen Page Memorandum dated
April 8, 2014. U.S. Environmental Protection Agency, Research Triangle Park, North
Carolina, 27711. http://www.epa.gov/ttn/emc/methods/psdnsrinterimcmpmemo4814.pdf
U.S. EPA, 2014b: Addendum - User's Guide for the AMS/EPA Regulatory Model - AERMOD.
EPA-454/B-03-001. U.S. Environmental Protection Agency, Research Triangle Park,
North Carolina 27711.
http://www.epa.gov/ttn/scram/models/aermod/aermod_userguide.zip
U.S. EPA, 2014c: Addendum - User's Guide for the AERMOD Meteorological Preprocessor
(AERMET). EPA-454/B-03-002. U.S. Environmental Protection Agency, Research
Triangle Park, NC 27711.
http://www.epa.gov/ttn/scram/7thconf/aermod/aermet userguide.zip
81
-------
This Page Intentionally Left Blank
82
-------
Appendix A: Draft Conceptual Description of PM2.5 Concentrations in the U.S.
This appendix provides a brief summary of the current PM2.5 monitoring networks and
characterizes PM air quality in terms of its chemical composition, concentration levels, and
spatial and temporal patterns across the nation based largely on ambient data and analyses
contained in the EPA's The Particle Pollution Report,38 Particulate Matter Staff Paper,39 and new
ambient data summaries based on 2008-2010 PM2.5 mass and speciation data. It also discusses
regional and local source contributions to urban PM2.5 concentrations. Such information may be
useful for permit applicants in preparing conceptual descriptions, as discussed in Section III.2.1
of this guidance.
1. PMi.s Monitoring Networks
1.1. PM2.s, PMio and PMio-2.5 Mass Networks
The 1997 promulgation of a fine parti culate NAAQS (EPA, 1997) led to deployment of
over 1500 PM2.5 sites (about 1000 currently) used to determine whether an area complies with
the standard. These sites use a Federal Reference Method (FRM) or Federal Equivalent Method
(FEM), daily sampling over 24-hours, or every third or sixth day. Nearly 300 additional
measurements not meeting FRM or FEM specifications are provided by the chemical speciation
sites (Figure A-l). Approximately 600 stations provide indirect measurements of continuous
(hourly resolution) PM2.5 mass using a variety of techniques.
1.2. Interagency Monitoring of Protected Visual Environments (IMPROVE) Program
The IMPROVE network, with over 100 sites, has provided nearly a two-decade record of
major components of PM2.5 (sulfate, nitrate, organic and elemental carbon fractions, and trace
metals) in pristine areas of the United States (Figure A-l). IMPROVE is led by the National Park
Service; various federal and state agencies support its operations. The primary focus of the
network is to track visibility and trends in visibility.
1.3. PMi.5 Chemical Speciation Monitoring
In addition to the IMPROVE network, over 300 EPA speciation sites were added from
2000 - 2002 in urban areas of the United States to assist PM2.5 assessment efforts. No FRM exists
for particulate speciation, which is not directly required to determine attainment, and there are
slight differences between monitors and methods used in the Speciation Trends Network (STN).
However, the network's coverage (Figure A-l) across urban and rural areas has proved essential
for a wide range of research and analysis. The speciation networks typically collect a 24-hour
sample every three, and sometimes six, days.
38 The Particle Pollution Report: Current Understanding of Air Quality and Emissions through 2003.
http://www.epa.gov/airtrends/aqtrnd04/pmreport03/pmcover 24Q5.pdf#page=l.
39 Particulate Matter Staff Paper: Review completed in 2012.
http: //www. epa. gov/ttn/naaqs/standards/pm/s_pm_cr_sp. html.
A-l
-------
Daily 24-hour speciation collection is limited to occasional efforts in the SEARCH (see
below) network. Similarly, only a handful of sites provide near continuous speciation data,
usually limited to some combination of sulfate, carbon (organic and elemental splits) and nitrate.
This enables insight to diurnal patterns for diagnosing various cause-effect phenomena related to
emissions characterization, source attribution analysis and model evaluation.
Figure A-l. Locations of chemical speciation sites delineated by program type
» Speciation SLAMS
. IMPROVE
D Speciation Trends Sites
A Continuous Speciation Sites with Multiple Measurements
1.4. South Eastern Aerosol Research and Characterization (SEARCH) Study
This study experiment is an industry-funded network of 8 sites that originally emerged
from the Southern Oxidants Study (SOS) in the 1990s and has operated for over a decade in
response to the 1997 revisions to the national ambient air quality standards for ground-level
ozone and particulate matter. SEARCH is part of a public-private collaboration that provides an
array of standard criteria pollutant measurements but also includes daily 24-hour PM speciation
at selected times and locations, gaseous ammonia, reactive nitrogen (NOy), and true nitrogen
dioxide (i.e., a measurement of NC>2 concentration unaffected by other nitrogen oxides, which
contaminate FRM NC>2 measurements). These measurements had not been available in major
government-funded routine networks and in order to identify sources of ozone precursors and
fine particulate matter and to attribute health effects to specific components, the SEARCH
project sponsors believe that it is necessary to measure pollutant composition as well as mass.
A-2
-------
1.5. PM Supersites Program
This program provided highly resolved aerosol measurements at eight U.S. cities for
several time periods from 1999 through 2004, with some sites collecting data after 2004.40 A
number of instrument configurations were deployed, ranging from additional locations for
standard speciation monitors, to systems capturing near-continuous size-dependent speciation
profiles.
2. Composition of PMi.5
Particulate matter (PM) is a highly complex mixture of solid particles and liquid droplets
distributed among numerous atmospheric gases which interact with solid and liquid phases.
Particles range in size from those smaller than 1 nanometer (10~9 meter) to over 100 microns (1
micron is 10"6 meter) in diameter (for reference, a typical strand of human hair is 70 microns and
particles less than about 20 microns generally are not detectable by the human eye). Particles are
classified as PM2.5 and PMio-2.5, corresponding to their size (diameter) range in microns and
referring to total particle mass under 2.5 and between 2.5 and 10 microns, respectively.
Particles span many sizes and shapes and consist of hundreds of different chemicals.
Particles are emitted directly from sources and also are formed through atmospheric chemical
reactions and often are referred to as primary and secondary particles, respectively. Particle
pollution also varies by time of year and location and is affected by several aspects of weather
such as temperature, clouds, humidity, and wind. Further complicating particles is the shifting
between solid/liquid and gaseous phases influenced by concentration and meteorology,
especially temperature.
Particles are made up of different chemical components. The major components, or
species, are carbon, sulfate and nitrate compounds, and crustal materials such as soil and ash
(Figure A-2). The different components that make up particle pollution come from specific
sources and are often formed in the atmosphere. Particulate matter includes both "primary" PM,
which is directly emitted into the air, and "secondary" PM, which forms indirectly from fuel
combustion and other sources. Primary PM consists of carbon (soot)emitted from cars, trucks,
heavy equipment, forest fires, and burning wasteand crustal material from unpaved roads,
stone crushing, construction sites, and metallurgical operations. Secondary PM forms in the
atmosphere from gases. Some of these reactions require sunlight and/or water vapor. Secondary
PM includes:
Sulfates formed from sulfur dioxide emissions from power plants and industrial
facilities;
Nitrates formed from nitrogen oxide emissions from cars, trucks, industrial facilities,
and power plants; and
40 Solomon, P.A., P.K. Hopke, J. Froines, and R. Scheffe, 2008: Key Scientific and Policy and Health-Relevant
Findings from the U.S. EPA's Particulate Matter Supersites Program and Related Studies: An Integration and
Synthesis of Results, J. Air & Waste Manage. Assoc., 58, S-l - S-92.
A-3
-------
Carbon formed from reactive organic gas emissions from cars, trucks, industrial
facilities, forest fires, and biogenic sources such as trees.
Figure A-2. National Average of Source Contribution to Fine Particle Levels
Cars, trucks, heavy equipment,
wildfires, wood/waste burning,
and biogenics
Cars, trucks,
industrial combustion, and
power generation
Suspended soil
and industrial metallurgical
operations
Industrial combustion and power
generation
Source: The Paniculate Matter Report, EPA-454-R-04-002, Fall 2004. Carbon reflects both organic carbon and
elemental carbon. Organic carbon accounts for automobiles, biogenics, gas-powered off-road, and wildfires.
Elemental carbon is mainly from diesel powered sources.
In addition, ammonia from sources such as fertilizer and animal feed operations
contributes to the formation of sulfates and nitrates that exist in the atmosphere as ammonium
sulfate and ammonium nitrate. Note that fine particles can be transported long distances by wind
and weather and can be found in the air thousands of miles from where they were formed.
The chemical makeup of particles varies across the United States (as shown in Figure A-
3).41 For example, fine particles in the eastern half of the United States contain more sulfates
than those in the West, while fine particles in southern California contain more nitrates than
other areas of the country. Organic carbon is a substantial component of fine particle mass
everywhere. This figure represents the composition of PM2.5 as measured by the PM2.5 FRM.42
41 The 15 cities are the same ones included in the Integrated Science Assessment for Paniculate Matter (2009) which
includes a similar map based on 2005-2007 PM2 5 data.
42 Frank, N. H., Retained Nitrate, Hydrated Sulfates, and Carbonaceous Mass in Federal Reference Method Fine
Paniculate Matter for Six Eastern U.S. Cities, T Air& Waste Manage. Assoc.' 2006, '56', 500-511.
A-4
-------
Figure A-3. Annual Average PMi.s Composition in 15 Urban Areas: 2008-2010
3. Seasonal and Daily Patterns of PMi.5
Fine particles often have a seasonal pattern. Both daily values and quarterly average of
PM2.5 also reveal patterns based on the time of year. Unlike daily ozone levels, which are usually
elevated in the summer, daily PM2.5 values at some locations can be high at any time of the year.
As shown in Figure A-4, PM2.5 values in the eastern half of the United States are typically higher
in the third calendar quarter (July-September) when sulfates are more readily formed from sulfur
dioxide (802) emissions from power plants in that region and when secondary organic aerosol is
more readily formed in the atmosphere. Fine particle concentrations tend to be higher in the first
calendar quarter (January through March) in the Midwest in part because fine particle nitrates are
more readily formed in cooler weather. PM2.5 values are high during the first (January through
March) and fourth calendar quarter (October through December) in many areas of the West, in
part because of fine particle nitrates and also due to carbonaceous particles which are directly
emitted from wood stove and fireplace use. Average concentration from all locations reporting
PM2.5 with valid design values is shown.
A-5
-------
Figure A-4. Quarterly Averages of PMi.5 Concentration: 2008-2010
PM2.5 - Qtr1
*»* .'{..
k^ " « fc cW
V . !
*,-. ,y > e
PM2.5 - Qtr3
t"-%* :% . ' /...
.
PM2.5mass, . <=g
PM2.5 - Qtr2
PM2.5 - Qtr4
15.1-18
IS
The composition of PM2.s also varies by season and helps explain why mass varies by
season. Figure A-5 shows the average composition by season (spring, summer, fall and winter)
for PM2.5 data collected during 2008-10. In the eastern United States, sulfate are high in the
spring (March-May) and summer (July-September). Nitrates are most evident in the midwest and
western cities where its percentage is moderately high in the spring and fall (October-and highest
during the winter.) Organic mass (OM) is high throughout the year.
A-6
-------
Figure A-5. Seasonal Speciation Profiles of PM^sfor Select Urban Areas: 2008-2010
- = .'1 Ft:'1 = =(-"= nt -r -
Q Q
FRM PM? h ;pFr a-inri - Spri
Q Q
Q Q
The composition of the highest daily PM2.s values may be different than that for the
annual average. Table A-l provides 2008-10 data on daily PIVb.s values and their composition on
high mass days for various sites within large metropolitan areas (in the east: Birmingham, AL;
Atlanta, GA; New York City, NY; Cleveland, OH; Chicago, IL; and St. Louis, MO; in the west:
Salt Lake City, UT; and Fresno, CA). Mass is proportioned into five components: sulfates,
nitrates, OM, elemental carbon (EC) and crustal material. For each site, the table shows the
2008-2010 annual average speciation profile, the breakdown for the top 10 percent of days per
year and corresponding FRM mass. The table shows some notable differences in the percentage
contribution of each of the species to total mass when looking at the high end of the distribution
versus the annual average. Except for the southeast (where there is little nitrate in PIVb.s), nitrates
are slightly higher in the top 10 percent of the PM2.5 days. For the 2008-2010 measurements, the
percent of sulfates is currently similar or slightly less on the top 10 percent of the days as
compared to the annual averages. The portion of OM appears to be similar on the high days
compared to the annual averages.
A-7
-------
Table A-l. PM2.5 Composition on High PM2.5 Mass Days in Select Urban Areas: 2008-2010
Urban Area
Atlanta
Birmingham
New York
City
Cleveland
Chicago
St Louis
Salt Lake City
Fresno
San Diego
Tacoma
Metric
Ann Mean
Top 10%
Ann Mean
Top 10%
Ann Mean
Top 10%
Ann Mean
Top 10%
Ann Mean
Top 10%
Ann Mean
Top 10%
Ann Mean
Top 10%
Ann Mean
Top 10%
Ann Mean
Top 10%
Ann Mean
Top 10%
Composition Percents (%)
Sulfate
35
35
37
34
44
41
40
42
34
36
40
37
13
8
14
8
26
16
17
8
Nitrate
1
1
1
1
9
13
13
15
18
29
13
20
27
55
21
43
4
14
2
3
OM
46
49
42
47
26
30
27
26
33
25
29
30
36
25
50
43
54
58
62
74
EC
9
8
7
8
12
10
7
7
6
5
7
6
7
5
5
4
7
7
9
10
Crustal
5
5
9
9
5
4
9
7
4
3
7
4
12
6
6
1
5
2
4
3
PM2.5
Mass
(ug/m3)
12.2
22.3
14.1
27.4
11.3
24.2
13.4
28.5
11.7
25.2
12.3
24.2
10.2
35.3
15.4
45.6
12.4
23.2
9.4
25.1
Ann. Top
Avg 10 %
*^v-
*S,S.
*-t-
*?W
^^*%^
^*T^
i
^^
^^
*^*~
£|£
tote: The percentages do not add to 100% due to a small amount of passively i i suifiti.nDJi ^H miuii-mis
ollected fine particle mass included in the measurement of PM2 5 by the FRM l^^
4. Regional and Local Sources
Both local and regional sources contribute to particle pollution. Figure A-6 shows how
much of the PM2.5 mass can be attributed to local versus regional sources for 13 selected urban
areas. In each of these urban areas, monitoring sites were paired with nearby rural sites. When
the average rural concentration is subtracted from the measured urban concentration, the
estimated local and regional contributions become apparent. Urban and nearby rural PM2.5
concentrations suggest substantial regional contributions to fine particles in the East. The
measured PM2.5 concentration is not necessarily the maximum for each urban area. Regional
A-8
-------
concentrations are derived from the rural IMPROVE monitoring network.
43
Figure A-6. "Urban excess" of locally generated PMi.5 mass for four chemical components:
sulfate, nitrate, organic mass (OM) and elemental carbon (EC)
Note: derived as the interpolated difference between urban CSN concentrations (squares) compared with nearby
IMPROVE site concentrations within 150 km (circles). Annual mean concentrations from 2005-2008 are used. CSN
sites not used in the analyses are shown as triangles.44
As shown in Figure A-6, we observe a large urban excess across the United State for
most PM2.5 species but especially for elemental carbon (EC) and organic mass (OM). Large
excess for OM is observed in California, throughout the Northwest, and in the Southeast. The
prevalence of urban excess in EC is seen more widely. Large urban excess of nitrates is seen in
California. These results indicate that local sources of these pollutants are indeed contributing to
the PM2.s air quality problem in these areas. As expected for a predominately regional pollutant,
only a modest urban excess is observed for sulfates.
In the East, regional pollution contributes more than half of total PM2.5 concentrations.
Rural background PM2.s concentrations are high in the East and are somewhat uniform over large
geographic areas. These regional concentrations come from emission sources such as power
plants, natural sources, and urban pollution and can be transported hundreds of miles. The local
and regional contributions for the major chemical components that make up urban PM2.s:
sulfates, carbon, and nitrates.
43 Information regarding the IMPROVE monitoring network can be found at the following website:
http ://vista. cira. colostate.edu/improve
44 Hand et. al, Spatial and Seasonal Patterns and Temporal Variability of Haze and its Constituents in the United
States: Report V, 2011 (http://vista.cira.colostate.edu/improve/Publications/Reports/2011/2011 .htm)
A-9
-------
This Page Intentionally Left Blank
A-10
-------
Appendix B: General Guidance on Use of Dispersion Models for Estimating Primary
Concentrations
This appendix provides general guidance on the application of dispersion models for
estimating ambient concentrations of PM25 associated with direct emissions of primary PM25.
This guidance is based on and is consistent with the EPA's Guideline on Air Quality Models,
published as Appendix W of 40 CFR Part 51, and focuses primarily on the application of
AERMOD, the EPA's preferred dispersion model for most situations. Appendix W is the
primary source of information on the regulatory application of air quality models for State
Implementation Plan (SIP) revisions for existing sources and for New Source Review (NSR) and
Prevention of Significant Deterioration (PSD) programs. There will be applications of dispersion
models unique to specific areas, (i.e., there may be areas of the country where it is necessary to
model unique specific sources or types of sources). In such cases, there should be consultation
with the state or appropriate permitting authority with the appropriate EPA Regional Office
modeling contact to discuss how best to model a particular source.
Recently issued EPA guidance of relevance for consideration in modeling for PM2.5
includes:
"Model Clearinghouse Review of Modeling Procedures for Demonstrating Compliance
with PM2.5NAAQS" February 26, 2010 (U.S. EPA, 2010a);
"Modeling Procedures for Demonstrating Compliance with PM2 5 NAAQS" March 23,
2010 (U.S. EPA, 201 Ob); and
"Transportation Conformity Guidance for Quantitative Hot-spot Analyses in PM2 5 and
PMio Nonattainment and Maintenance Areas" November 2013 (U.S. EPA, 20 13 a).
The guidance listed above, in addition to other relevant support documents can be found on the
SCRAM website at: http ://www. epa. gov/ttn/scram/.
The following sections will refer to the relevant sections of Appendix W and other
existing guidance with summaries as necessary. Please refer to those original guidance
documents for full discussion and consult with the appropriate EPA Regional Office modeling
contact if questions arise about interpretation on modeling techniques and procedures.45
1. Model selection
Preferred air quality models for use in regulatory applications are addressed in Appendix
A of the EPA's Guideline on Air Quality Models. If a model is to be used for a particular
application, the user should follow the guidance on the preferred model for that application.
These models may be used without an area specific formal demonstration of applicability as long
as they are used as indicated in each model summary of Appendix A. Further recommendations
for the application of these models to specific source problems are found in Appendix W. In
45 A list of EPA Regional Office modeling contacts is available on the SCRAM website at:
http://www.epa.gov/ttn/scram/guidance_cont_re gions.htm.
B-l
-------
2005, the EPA promulgated the American Meteorological Society/Environmental Protection
Agency Regulatory Model (AERMOD) as the Agency's preferred near-field dispersion model
for a wide range of regulatory applications in all types of terrain based on extensive
developmental and performance evaluation. For PSD/NSR modeling under the PM2.5 NAAQS,
AERMOD should be used to model primary PM2.5 emissions unless use of an alternative model
can be justified (Section 3.2, Appendix W), such as the Buoyant Line and Point Source
Dispersion Model (BLP).
The AERMOD modeling system includes the following components:
AERMOD: the dispersion model (U.S. EPA, 2004a; U.S. EPA, 2014a);
AERMAP: the terrain processor for AERMOD (U.S. EPA, 2004b, U.S. EPA, 201 la);
and
AERMET: the meteorological data processor for AERMOD (U.S. EPA, 2004c; U.S.
EPA, 2014b).
Other components that may be used, depending on the application, are:
BPIPPREVIE: the building input processor (U. S. EPA, 2004d);
AERSURFACE: the surface characteristics processor for AERMET (U.S. EPA, 2008);
AERSCREEN: a screening version of AERMOD (U.S. EPA, 201 Ib; U.S. EPA, 201 Ic);
and
AERMINUTE: a pre-processor to calculate hourly average winds from ASOS 2-minute
observations (U.S. EPA, 201 Id).
Before running AERMOD, the user should become familiar with the user's guides associated
with the modeling components listed above and the AERMOD Implementation Guide (AIG)
(U.S. EPA, 2009). The AIG lists several recommendations for applications of AERMOD that
would be applicable for SIP and PSD permit modeling.
1.2. Receptor grid
The model receptor grid is unique to the particular situation and depends on the size of
the modeling domain, the number of modeled sources, and complexity of the terrain. Receptors
should be placed in areas that are considered ambient air (i.e., where the public generally has
access) and placed out to a distance such that areas of violation can be detected from the model
output to help determine the size of nonattainment areas. Receptor placement should be of
sufficient density to provide resolution needed to detect significant gradients in the
concentrations with receptors placed closer together near the source to detect local gradients and
placed farther apart away from the source. In addition, the user may want to place receptors at
key locations such as around facility fence lines (which define the ambient air boundary for a
particular source) or monitor locations (for comparison to monitored concentrations for model
evaluation purposes). The receptor network should cover the modeling domain. States may
already have existing receptor placement strategies in place for regulatory dispersion modeling
under NSR/PSD permit programs.
B-2
-------
If modeling indicates elevated levels of PM2.5 (near the standard) near the edge of the
receptor grid, consideration should be given to expanding the grid or conducting an additional
modeling run centered on the area of concern. As noted above, terrain complexity should also be
considered when setting up the receptor grid. If complex terrain is included in the model
calculations, AERMOD requires that receptor elevations be included in the model inputs. In
those cases, the AERMAP terrain processor (U.S. EPA, 2004b; U.S EPA, 201 la) should be used
to generate the receptor elevations and hill heights. The latest version of AERMAP (version
09040 or later) can process either Digitized Elevation Model (DEM) or National Elevation Data
(NED) data files. The AIG recommends the use of NED data since it is more up to date than
DEM data, which is no longer updated (Section 4.3 of the AIG).
2. Source inputs
This section provides guidance on source characterization to develop appropriate inputs
for dispersion modeling with the AERMOD modeling system. Section 2.1 provides guidance on
use of emission, Section 2.2 covers guidance on Good Engineering Practice (GEP) stack heights,
Section 2.3 provides details on source configuration and source types, Section 2.4 provides
details on urban/rural determination of the sources, and Section 2.5 provides general guidance on
source grouping, which may be important for design value calculations.
2.1. Emissions
Consistent with Appendix W, dispersion modeling for the purposes of PSD permitting
should be based on the use of continuous operation at maximum allowable emissions or federally
enforceable permit limits (see Table 8-2 of Appendix W) for the project source for all applicable
averaging periods. Also consistent with past and current guidance, in the absence of maximum
allowable emissions or federally enforceable permit limits, potential to emit emissions (i.e.,
design capacity) should be used. Maximum allowable emissions and continuous operation should
also be assumed for nearby sources included in the modeled inventory for the 24-hr PM2.5
NAAQS, while maximum allowable emissions and the actual operating factor averaged over the
most recent 2 years should be used for modeled nearby sources for the annual PM2.5 NAAQS.
2.2. Good Engineering Practice (GEP) stack height
Consistent with previous modeling guidance and Section 6.2.2 of Appendix W, for stacks
with heights that are within the limits of Good Engineering Practice (GEP), actual heights should
be used in modeling. Under the EPA's regulations at 40 CFR 51.100, GEP height, Hg, is
determined to be the greater of:
65 m, measured from the ground-level elevation at the base of the stack;
for stacks in existence on January 12, 1979, and for which the owner or operator had
obtained all applicable permits or approvals required under 40 CFR Parts 51 and 52
Hg=2.5H
provided the owner or operator produces evidence that this equation was actually relied
B-3
-------
on in designing the stack or establishing an emission limitation to ensure protection
against downwash;
for all other stacks,
Hg=H+1.5L,
where H is the height of the nearby structure(s) measured from the ground-level elevation
at the base of the stack and L is the lesser dimension of height or projected width of
nearby structure(s); or
the height demonstrated by a fluid model or a field study approved by the EPA or the
state/local permitting agency which ensures that the emissions from a stack do not result
in excessive concentrations of any air pollutant as a result of atmospheric downwash,
wakes, eddy effects created by the source itself, nearby structures or nearby terrain
features.
For more details about GEP, see the Guideline for Determination of Good Engineering Practice
Stack Height Technical Support Document (U.S. EPA, 1985).
If stack heights exceed GEP, then GEP heights should be used with the individual stack's
other parameters (temperature, diameter, exit velocity). For stacks modeled with actual heights
below GEP that may be subject to building downwash influences, building downwash should be
considered as this can impact concentrations near the source (Section 6.2.2b, Appendix W). If
building downwash is being considered, the BPIPPREVIE program (U.S. EPA, 2004d) should be
used to input building parameters for AERMOD. More information about buildings and stacks is
provided in Section 6.5.
2.3. Source configurations and source types
An accurate characterization of the modeled facilities is critical for refined dispersion
modeling, including accurate stack parameters and physical plant layout. Accurate stack
parameters should be determined for the emissions being modeled. Since modeling would be
done with maximum allowable or potential emissions levels at each stack, the stack's parameters
such as exit temperature, diameter, and exit velocity should reflect those emissions levels.
Accurate locations (i.e.. latitude and longitude or Universal Transverse Mercator (UTM)
coordinates and datum)46 of the modeled emission sources are also important, as this can affect
the impact of an emission source on receptors, determination of stack base elevation, and relative
location to any nearby building structures. Not only are accurate stack locations needed, but
accurate information for any nearby buildings is important. This information would include
location and orientation relative to stacks and building size parameters (height, and corner
coordinates of tiers) as these parameters are input into BPIPPREVIE to calculate building
parameters for AERMOD. If stack locations and or building information are not accurate,
46 Latitudes and longitudes to four decimal places position a stack within 30 feet of its actual location and five
decimal places position a stack within three feet of its actual location. Users should use the greatest precision
available.
B-4
-------
downwash will not be accurately accounted for in AERMOD.
Emission source type characterization within the modeling environment is also important.
As stated in the AERMOD User's Guide (U.S. EPA, 2004a; U.S. EPA, 2012a), emissions
sources can be characterized as several different source types: POINT sources, capped stacks
(POINTCAP), horizontal stacks (POINTHOR), VOLUME sources, OPENPIT sources, LINE
sources, rectangular AREA sources, circular area sources (AREACIRC), and irregularly shaped
area sources (AREAPOLY). Note that POINTCAP and POINTHOR are not part of the
regulatory default option in AERMOD because the user must invoke the BETA option in the
model options keyword MODELOPT while not including the "DFAULT" modeling option for
these options to work properly. While most sources can be characterized as POINT sources,
some sources, such as fugitive releases or nonpoint sources (emissions from ports/ships, airports,
or smaller point sources with no accurate locations), may be best characterized as VOLUME or
AREA type sources. Sources such as flares can be modeled in AERMOD using the parameter
input methodology described in Section 2.1.2 of the AERSCREEN User's Guide (U. S. EPA,
201 Ib). If questions arise about proper source characterization or typing, users should consult the
appropriate EPA Regional Office modeling contact.
2.4. Urban/rural determination
For any dispersion modeling exercise, the urban or rural determination of a source is
important in determining the boundary layer characteristics that affect the model's prediction of
downwind concentrations. Figure B-l gives example maximum 24-hour concentration profiles
for a 10 meter stack (Figure B-la) and a 100 m stack (Figure B-lb) based on urban vs. rural
designation. The urban population used for the examples is 100,000. In Figure B-la, the urban
concentration is much higher than the rural concentration for distances less than 750 m from the
stack but then drops below the rural concentration beyond 750 m. For the taller stack in Figure
B-lb, the urban concentration is much higher than the rural concentration even as distances
increase from the source. These profiles show that the urban or rural designation of a source can
be quite important.
Determining whether a source is urban or rural can be done using the methodology
outlined in Section 7.2.3 of Appendix W and recommendations outlined in Sections 5.1 through
5.3 in the AIG (U.S. EPA, 2009). In summary, there are two methods of urban/rural
classification described in Section 7.2.3 of Appendix W.
The first method of urban determination is a land use method (Appendix W, Section
7.2.3c). In the land use method, the user analyzes the land use within a 3 km radius of the source
using the meteorological land use scheme described by Auer (1978). Using this methodology, a
source is considered urban if the land use types II (heavy industrial), 12 (light-moderate
industrial), Cl (commercial), R2 (common residential), and R3 (compact residential) are 50
percent or more of the area within the 3 km radius circle. Otherwise, the source is considered a
rural source. The second method uses population density and is described in Section 7.2.3d of
Appendix W. As with the land use method, a circle of 3 km radius is used. If the population
density within the circle is greater than 750 people/km2, then the source is considered urban.
Otherwise, the source is modeled as a rural source. Of the two methods, the land use method is
B-5
-------
considered more definitive (Section 7.2.3e, Appendix W).
Caution should be exercised with either classification method. As stated in Section 5.1 of
the AIG (U.S. EPA, 2009), when using the land use method, a source may be in an urban area
but located close enough to a body of water or other non-urban land use category to result in an
erroneous rural classification for the source. The AIG in Section 5.1 cautions users against using
the land use scheme on a source by source basis, but advises considering the potential for urban
heat island influences across the full modeling domain. When using the population density
method, Section 7.2.3e of Appendix W states, "Population density should be used with caution
and should not be applied to highly industrialized areas where the population density may be low
and thus a rural classification would be indicated, but the area is sufficiently built-up so that the
urban land use criteria would be satisfied..." With either method, Section 7.2.3(f) of Appendix W
recommends modeling all sources within an urban complex as urban, even if some sources
within the complex would be considered rural using either the land use or population density
method.
B-6
-------
Figure B-l. Urban (red) and rural (blue) concentration profiles for (a) 10 m buoyant stack
release, and (b) 100 m buoyant stack release
600
.
s
a
S
1 2
O
!
1.5
05
10m
500 1000 1500 2000 2500 3000 3500 4000 4500 5000
Distance (m)
100m
500 1000 1500 2000 2500 3000 3500 4000 4500 5000
Di stance (m)
B-7
-------
Another consideration that may need attention by the user, and is discussed in Section 5.1
of the AIG, relates to tall stacks located within or adjacent to small to moderate size urban areas.
In such cases, the stack height or effective plume height for very buoyant sources may extend
above the urban boundary layer height. The application of the urban option in AERMOD for
these types of sources may artificially limit the plume height. The use of the urban option may
not be appropriate for these sources, since the actual plume is likely to be transported over the
urban boundary layer. Section 5.1 of the AIG gives details on determining if a tall stack should
be modeled as urban or rural based on comparing the stack or effective plume height to the urban
boundary layer height. The 100 m stack illustrated in Figure B-lb, may be such an example as
the urban boundary layer height for this stack would be 189 m (based on a population of
100,000) and equation 104 of the AERMOD formulation document (Cimorelli, et al., 2004). This
equation is:
(B-l)
where z;uo is a reference height of 400 m corresponding to a reference population P0 of 2,000,000
people.
Given that the stack is a buoyant release, the plume may extend above the urban
boundary layer and may be best characterized as a rural source, even if it were near an urban
complex. Exclusion of these elevated sources from application of the urban option would need to
be justified on a case-by-case basis in consultation with the appropriate permitting authority.
AERMOD requires the input of urban population when utilizing the urban option.
Population can be entered to one or two significant digits (i.e., an urban population of 1,674,365
can be entered as 1,700,000). Users can enter multiple urban areas and populations using the
URBANOPT keyword in the runstream file (U.S. EPA, 2004a; U.S. EPA, 2012a). If multiple
urban areas are entered, AERMOD requires that each urban source be associated with a
particular urban area or AERMOD model calculations will abort. Urban populations can be
determined by using a method described in Section 5.2 of the AIG (U.S. EPA, 2009).
2.5. Source groups
In AERMOD, individual emission sources' concentration results can be combined into
groups using the SRCGROUP keyword (Section 3.3.11 of the AERMOD User's Guide (U.S,
EPA, 2004a). The user can automatically calculate a total concentration (from all sources) using
the SRCGROUP ALL keyword. For the purposes of design value calculations, source group
ALL should be used, especially if all sources in the modeling domain are modeled in one
AERMOD run. Design values should be calculated from the total concentrations (all sources and
background). Individual source contributions outputs to the total concentration may be necessary
to determine the culpability to any NAAQS violations.
B-8
-------
3. Meteorological data
This section gives guidance on the selection of meteorological data for input into
AERMOD. Much of the guidance from Section 8.3 of Appendix W is applicable to SIP and PSD
permit modeling and is summarized here. In Section 7.2.1, the use of a new tool, AERMINUTE
(U.S. EPA, 201 Id), is introduced. AERMINUTE is an AERMET pre-processor that calculates
hourly averaged winds from ASOS (Automated Surface Observing System) 1-minute winds.
3.1. Surface characteristics and representativeness
The selection of meteorological data that are input into a dispersion model should be
considered carefully. The selection of data should be based on spatial and climatological
(temporal) representativeness (Appendix W, Section 8.3). The representativeness of the data is
based on: 1) the proximity of the meteorological monitoring site to the area under consideration,
2) the complexity of terrain, 3) the exposure of the meteorological site, and 4) the period of time
during which data are collected. Sources of meteorological data are: National Weather Service
(NWS) stations, site-specific or onsite data, and other sources such as universities, Federal
Aviation Administration (FAA), military stations, and others. Appendix W addresses spatial
representativeness issues in Sections 8.3.a and 8.3.c.
Spatial representativeness of the meteorological data can be adversely affected by large
distances between the source and receptors of interest and the complex topographic
characteristics of the area (Appendix W, Section 8.3.a and 8.3.c). If the modeling domain is large
enough such that conditions vary drastically across the domain, then the selection of a single
station to represent the domain should be carefully considered. Also, care should be taken when
selecting a station if the area has complex terrain. While a source and meteorological station may
be in close proximity, there may be complex terrain between them such that conditions at the
meteorological station may not be representative of the source. An example would be a source
located on the windward side of a mountain chain with a meteorological station a few kilometers
away on the leeward side of the mountain. Spatial representativeness for off-site data should also
be assessed by comparing the surface characteristics (albedo, Bowen ratio, and surface
roughness) of the meteorological monitoring site and the analysis area. When processing
meteorological data in AERMET (U.S. EPA, 2004c; U.S. EPA, 2014b), the surface
characteristics of the meteorological site should be used (Section 8.3.c of Appendix W and the
AERSURFACE User's Guide (U.S. EPA 2008)). Spatial representativeness should also be
addressed for each meteorological variable separately. For example, temperature data from a
meteorological station several kilometers from the analysis area may be considered adequately
representative, while it may be necessary to collect wind data near the plume height (Section
8.3.c of Appendix W).
Surface characteristics can be calculated in several ways. For details see Section 3.1.2 of
the AIG (U.S. EPA, 2009). The EPA has developed a tool, AERSURFACE (U.S. EPA, 2008) to
aid in the determination of surface characteristics. The current version of AERSURFACE uses
the 1992 National Land Cover Data. Note that the use of AERSURFACE is not a regulatory
requirement but the methodology outlined in Section 3.1.2 of the AIG should be followed unless
an alternative method can be justified.
B-9
-------
3.2. Meteorological inputs
Appendix W states in Section 8.3.1.1 that the user should acquire enough meteorological
data to ensure that worst-case conditions are adequately represented in the model results.
Appendix W states that 5 years of NWS meteorological data or at least 1 year of site-specific
data should be used(Section 8.3.1.2, Appendix W) and should be adequately representative of the
study area. If 1 or more years of site-specific data are available, those data are preferred. While
the form of the PM2.5 NAAQS contemplates obtaining 3 years of monitoring data, this does not
preempt the use of 5 years of NWS data or at least 1 year of site-specific data in the modeling.
The 5-year average based on the use of NWS data, or an average across 1 or more years of
available site specific data, serves as an unbiased estimate of the 3-year average for purposes of
modeling demonstrations of compliance with the NAAQS.
3.2.1. NWS data
NWS data are available from the National Climatic Data Center (NCDC) in many
formats, with the most common one in recent years being the Integrated Surface Hourly data
(ISH). Most available formats can be processed by AERMET. As stated in Section 3.1, when
using data from an NWS station alone or in conjunction with site-specific data, the data should
be spatially and temporally representative of conditions at the modeled sources. Key points
regarding the use of NWS data can be found in the EPA's March 8, 2013 clarification memo
"Use of ASOS meteorological data in AERMOD dispersion modeling" (U.S. EPA, 2013b). The
key points are:
The EPA has previously analyzed the effects of ASOS implementation on dispersion
modeling and found that generally AERMOD was less sensitive than ISCST3 to the
implementation of ASOS.
The implementation of the ASOS system over the conventional observation system
should not preclude the consideration of NWS stations in dispersion modeling.
The EPA has implemented an adjustment factor (0.5 knots) in AERMET to adjust for
wind speed truncation in ASOS winds
The EPA has developed the AERMINUTE processor (U.S. EPA, 201 Id) to process 2-
minute ASOS winds and calculate an hourly average for input into AERMET. The use of
hourly averaged winds better reflect actual conditions over the hour as opposed to a
single 2-minute observation.
While the EPA's March 8, 2013, memo states that ASOS should not preclude the use of
NWS data in dispersion modeling, and Section 8.3.1.2 of Appendix W recommends the most
recent five years of NWS data, Section 8.3.1.2 also recognizes cases where professional
judgment indicates that ASOS data are inadequate and pre-ASOS, or observer based data may be
considered for use. The appropriate permitting authority and EPA Regional Office modeling
contact should be consulted when questions arise about the representativeness or applicability of
NWS data.
3.2.2. Site-specific data
B-10
-------
The use of site-specific meteorological data is the best way to achieve spatial
representativeness. AERMET can process a variety of formats and variables for site-specific
data. The use of site-specific data for regulatory applications is discussed in detail in Section
8.3.3 of Appendix W. Due to the range of data that can be collected onsite and the range of
formats of data input to AERMET, the user should consult Appendix W, the AERMET User's
Guide (U.S. EPA, 2004c; U. S. EPA, 2014b), and Meteorological Monitoring Guidance for
Regulatory Modeling Applications (U.S. EPA, 2000). Also, when processing site-specific data
for an urban application, Section 3.3 of the AERMOD Implementation Guide offers
recommendations for data processing. In summary, the guide recommends that site-specific
turbulence measurements should not be used when applying AERMOD's urban option in order
to avoid double counting the effects of enhanced turbulence due to the urban heat island.
3.2.3. Upper air data
AERMET requires full upper air soundings to calculate the convective mixing height. For
AERMOD applications in the U.S., the early morning sounding, usually the 1200 UTC
(Universal Time Coordinate) sounding, is typically used for this purpose. Upper air soundings
can be obtained from the Radiosonde Data of North America CD for the period 1946-1997.
Upper air soundings for 1994 through the present are also available for free download from the
Radiosonde Database Access website. Users should choose all levels or mandatory and
significant pressure levels47 when selecting upper air data. Selecting mandatory levels only
would not be adequate for input into AERMET as the use of just mandatory levels would not
provide an adequate characterization of the potential temperature profile.
4. Running AERMOD and implications for design value calculations
Recent enhancements to AERMOD include options to aid in the calculation of design
values for comparison with the PM2.5 NAAQS and to aid in determining whether emissions from
the project source contributed significantly to any modeled violations. These enhancements
include:
The MAXDCONT option, which shows the contribution of each user-specified source
group to the high ranked values for a specified target source group paired in time and
space. The user can specify a range of ranks to analyze or specify an upper bound rank,
i.e. 8th highest, corresponding to the 98th percentile for the 24-hour PM2.s NAAQS, and a
lower threshold concentration value, such as the NAAQS for the target source group. The
model will process each rank within the range specified, but will stop after the first rank
(in descending order of concentration) that is below the threshold value if specified by the
user. A warning message will be generated if the threshold is not reached within the
range of ranks analyzed (based on the range of ranks specified on the RECTABLE
keyword). This option may be needed to aid in determining which sources should be
considered for controls.
47 By international convention, mandatory levels are in millibars: 1,000, 850, 700, 500, 400, 300, 200, 150, 100, 50,
30, 20, 10, 7 5, 3, 2, and 1. Significant levels may vary depending on the meteorological conditions at the upper-air
station.
B-ll
-------
For more details about the enhancements, see the AERMOD User's guide Addendum (U. S.
EPA, 2014a).
Ideally, all explicitly modeled sources, receptors, and background should be modeled in
one AERMOD run for all modeled years. In this case, one of the above output options can be
used in AERMOD to calculate design values for comparison to the NAAQS and determine the
area's attainment status and/or inform attainment/nonattainment boundaries. The use of these
options in AERMOD allows AERMOD to internally calculate concentration metrics that can be
used to calculate design values and, therefore, lessen the need for large output files, i.e. hourly
POSTFILES.
However, there may be situations where a single AERMOD run with all explicitly
modeled sources is not possible. These situations often arise due to runtime or storage space
considerations during the AERMOD modeling. Sometimes separate AERMOD runs are done for
each facility or group of facilities, or by year, or the receptor network is divided into separate
sub-networks. In some types of these situations, the MAXDCONT output option may not be an
option for design value calculations, especially if all sources are not included in a single run. If
the user wishes to utilize one of the three output options, then care should be taken in developing
the model inputs to ensure accurate design value calculations.
Situations that would effectively preclude the use of the MAXDCONT option to calculate
meaningful AERMOD design value calculations include the following examples:
Separate AERMOD runs for each source or groups of sources.
o SIP modeling includes 10 facilities for 5 years of NWS data and each facility is
modeled for 5 years in a separate AERMOD run, resulting in ten separate AERMOD
runs.
Separate AERMOD runs for each source and each modeled year.
o 10 facilities are modeled for 5 years of NWS data. Each facility is modeled separately
for each year, resulting in fifty individual AERMOD runs.
In the two situations listed above, the MAXDCONT option would not be useful as the
different AERMOD runs do not include a total concentration with contributions from all
facilities. In these situations, the use of 24-hour POSTFILES, which can be quite large, and
external post-processing would be needed to calculate design values.
Situations in which the MAXDCONT options may be used but may necessitate some
external post-processing afterwards to calculate a design value include:
The receptor network is divided into sections and an AERMOD run, with all sources and
years, is made for each sub-network.
o A receptor network of 1,000 receptors is divided into four 250 receptor sub-
networks. 10 facilities are modeled with 5 years of NWS data in one AERMOD
run for each receptor network, resulting in four AERMOD runs. After the
B-12
-------
AERMOD runs are complete, the MAXDCONT results for each network can be
re-combined into the larger network.
All sources and receptors are modeled in an AERMOD run for each year.
Ten facilities are modeled with 5 years of NWS data. All facilities are modeled with all
receptors for each year individually, resulting in five AERMOD runs. MAXDCONT
output can be used and post-processed to generate the necessary design value
concentrations. The receptor network is divided and each year is modeled separately for
each sub-network with all sources.
Ten facilities are modeled with 5 years of NWS data for 1,000 receptors. The receptor
network is divided into four 250 receptor networks. For each sub-network, all ten
facilities are modeled for each year separately, resulting in twenty AERMOD runs.
MAXDCONT output can be used and post-processed to generate the necessary design
value concentrations.
B-13
-------
5. References
Auer, Jr., A.H., 1978: Correlation of Land Use and Cover with Meteorological Anomalies.
Journal of Applied Meteorology, 17(5), 636-643.
Erode, R., K. Wesson, J. Thurman, and C. Tillerson, 2008: AERMOD Sensitivity to the Choice
of Surface Characteristics, Paper 811, Air And Waste Management Association Annual
Conference.
Cimorelli, A. J., S. G. Perry, A. Venkatram, J. C. Weil, R. J. Paine, R. B. Wilson, R. F. Lee, W.
D. Peters, R. W. Erode, and J. O. Paumier, 2004. AERMOD: Description of Model
Formulation, EPA-454/R-03-004. U.S. Environmental Protection Agency, Research
Triangle Park, NC. http://www.epa.gov/ttn/scram/7thconf/aermod/aermod_mfd.pdfU.S
U.S. EPA, 1985: Guideline for Determination of Good Engineering Practice Stack Height
(Technical Support Document for the Stack Height Regulations), Revised. EPA-450/4-
80-023R. U.S. Environmental Protection Agency, Research Triangle Park, NC 27711.
http ://www. epa.gov/ttn/scram/guidance/guide/gep .pdf
U.S. EPA, 1992: Screening Procedures for Estimating the Air Quality Impact of Stationary
Sources. EPA-454/R-92-019. U.S. Environmental Protection Agency, Research Triangle
Park, NC 27711. http://www.epa.gov/ttn/scram/guidance/guide/scrng.wpd
U.S. EPA, 1994: SO2 Guideline Document. EPA-452/R-95-008. U.S. Environmental Protection
Agency, Research Triangle Park, NC 27711.
U.S. EPA, 2000: Meteorological Monitoring Guidance for Regulatory Modeling Applications.
EPA-454/R-99-005. U.S. Environmental Protection Agency, Research Triangle Park, NC
27711. http://www.epa.gov/ttn/scram/guidance/met/mmgrma.pdf
U.S. EPA, 2004a: User's Guide for the AMS/EPA Regulatory Model - AERMOD. EPA-454/B-
03-001. U.S. Environmental Protection Agency, Research Triangle Park, NC 27711.
http://www.epa.gov/ttn/scram/models/aermod/aermod userguide.zip
U.S. EPA, 2004b: User's Guide for the AERMOD Terrain Preprocessor(AERMAP). EPA-
454/B-03-003. U.S. Environmental Protection Agency, Research Triangle Park, North
Carolina 27711.
http://www.epa.gov/ttn/scram/models/aermod/aermap/aermap userguide.zip
U.S. EPA, 2004c: User's Guide for the AERMOD Meteorological Preprocessor (AERMET).
EPA-454/B-03-002. U.S. Environmental Protection Agency, Research Triangle Park, NC
27711. http://www.epa.gov/ttn/scram/7thconf/aermod/aermet_userguide.zip
U.S. EPA, 2004d: User's Guide to the Building Profile Input Program. EPA-454/R-93-038. U.S.
Environmental Protection Agency, Research Triangle Park, North Carolina 27711.
U.S. EPA, 2005. Guideline on Air Quality Models. 40 CFRPart 51 Appendix W.
http://www.epa.gov/ttn/scram/guidance/guide/appw_05.pdf
U.S. EPA, 2008: AERSURFACE User's Guide. EPA-454/B-08-001. U.S. Environmental
Protection Agency, Research Triangle Park, North Carolina 27711.
http://www.epa.gov/ttn/scram/7thconf/aermod/aersurface userguide.pdf
U.S. EPA, 2009: AERMOD Implementation Guide. U.S. Environmental Protection Agency,
Research Triangle Park, North Carolina 27711.
http://www.epa.gov/ttn/scram/7thconf/aermod/aermod_implmtn_guide_19March
2009.pdf
B-14
-------
U.S. EPA, 2010a: Model Clearinghouse Review of Modeling Procedures for Demonstrating
Compliance with PM2.5 NAAQS. Tyler Fox Memorandum dated February 26, 2010. U.S.
Environmental Protection Agency, Research Triangle Park, North Carolina 27711.
http://www.epa.gov/ttn/scram/guidance/mch/new mch/MCmemo Region6 PM25 NAA
QS_Compliance.pdf
U.S. EPA, 2010b: Modeling Procedures for Demonstrating Compliance with PM2.5 NAAQS.
Stephen Page Memorandum dated March 23, 2010. U.S. Environmental Protection
Agency, Research Triangle Park, North Carolina, 27711.
http://www.epa.gov/ttn/scram/Official%20Signed%20Modeling%20Proc%20for%20De
mo%20Compli%20w%20PMM.pdf
U.S. EPA, 201 la: Addendum - User's Guide for the AERMOD Terrain Preprocessor
(AERMAP). EPA-454/B-03-003. U.S. Environmental Protection Agency, Research
Triangle Park, North Carolina 27711.
http://www.epa.gov/ttn/scram/models/aermod/aermap/aermap userguide.zip
U.S. EPA, 201 Ib: AERSCREEN User's Guide. EPA-454-/B-11-001. U.S. Environmental
Protection Agency, Research Triangle Park, North Carolina 27711.
http://www.epa.gov/ttn/scram/models/screen/aerscreen_userguide.pdf
U.S. EPA, 201 Ic: AERSCREEN Released as the EPA Recommended Screening Model. Tyler
Fox Memorandum dated April 11, 2011. U.S. Environmental Protection Agency,
Research Triangle Park, North Carolina 27711.
http://www.epa.gov/ttn/scram/20110411 AERSCREEN Release Memo.pdf
U.S EPA, 201 Id AERMINUTE User's Guide. U.S. Environmental Protection Agency, Research
Triangle Park, North Carolina 27711.
http://www.epa.gov/ttn/scram/7thconf/aermod/aerminute vll325.zip
U.S. EPA, 2013a: Transportation Conformity Guidance for Quantitative Hot-spot Analyses in
PM2.5 and PMio Nonattainment and Maintenance Areas. November 2013. EPA-420-B-
10-040. U.S. Environmental Protection Agency, Ann Arbor, Michigan 48105.
http://www.epa.gov/oms/stateresources/transconf/policy/420bl3053-sec.pdfand
http://www.epa.gov/oms/stateresources/transconf/policy/420bl3053-appx.pdf.
U.S. EPA, 2013b: Use of ASOS meteorological data in AERMOD dispersion modeling. Tyler
Fox Memorandum dated March 8, 2013. U.S. Environmental Protection Agency,
Research Triangle Park, North Carolina 27711.
http://www.epa.gov/ttn/scram/guidance/clarificati on/20130308_Met_Data_Clarificati on.
rjdf
U.S. EPA, 2014a: Addendum - User's Guide for the AMS/EPA Regulatory Model - AERMOD.
EPA-454/B-03-001. U.S. Environmental Protection Agency, Research Triangle Park,
North Carolina 27711.
http://www.epa.gov/ttn/scram/models/aermod/aermod userguide.zip
U.S. EPA, 2014b: Addendum - User's Guide for the AERMOD Meteorological Preprocessor
(AERMET). EPA-454/B-03-002. U.S. Environmental Protection Agency, Research
Triangle Park, NC 27711.
http://www.epa.gov/ttn/scram/7thconf/aermod/aermet userguide.zip
B-15
-------
This Page Intentionally Left Blank
B-16
-------
Appendix C: Example of a Qualitative Assessment of the Potential for Secondary PMi.5
Formation
In late 2011, the EPA Region 10 Office developed a qualitative assessment of the
potential for secondary fine particulate matter (PM2.5) formation to cause or contribute to a
violation of the PM2.5 National Ambient Air Quality Standard (NAAQS) through a response to
public comments document regarding a Clean Air Act permit issued for Shell's Discoverer drill
ship and support fleet to explore for oil and gas in the Chukchi Sea off Alaska. While the
environment in and around the Chukchi Sea and North Slope of Alaska is unique when
compared to the rest of the United States, the various components contained within this
qualitative assessment provide a template that could be followed, with appropriate modifications,
in the development of other case-specific qualitative assessments. An excerpt from this response
to public comments document is provided below for reference.
As shown in the EPA Region 10 example, the qualitative assessment of the potential for
secondary PM2.5 formation by the Shell's Discoverer drill ship and support fleet was developed
in a narrative manner integrating numerous factors specific to the North Slope region of Alaska
that provided sufficient evidence that the PIVb.s NAAQS would not be violated in this particular
case. The qualitative assessment examined the regional background PM2.5 monitoring data and
aspects of secondary PM2.5 formation from existing sources; the relative ratio of the combined
modeled primary PM2.5 impacts and background PM2.5 concentrations to the level of the
NAAQS; the spatial and temporal correlation of the primary and secondary PM2.5 impacts;
meteorological characteristics of the region during periods of precursor pollutant emissions; the
level of conservatism associated with the modeling of the primary PM2.5 component and other
elements of conservatism built into the overall NAAQS compliance demonstration; aspects of
the precursor pollutant emissions in the context of limitations of other chemical species
necessary for the photochemical reactions to form secondary PM^.s; and an additional level of
NAAQS protection through a post-construction monitoring requirement. While each of the
components of the EPA Region 10 example may or may not be necessary, this example should
provide a useful template for other qualitative assessments under this guidance, recognizing that
additional components may be essential in other qualitative assessments of the potential for
secondary PM2.5 formation.
Additional information regarding this EPA Region 10 Office permit action can be found
through the following web link: http://vosemite.epa.gov/R10/airpage.nsf/Permits/chukchiap/.
Region 10 example:
In support of the 2011 Revised Draft Permits, Region 10 provided a detailed
explanation for why it believes that modeling secondary PM2.5 emissions is not
needed in order to determine that emissions of PM2.5 precursors from the Discoverer
and Associated Fleet would not, together with emissions of primary PM2.5, cause or
contribute to a violation of the 24-hour PM2.5 NAAQS. The factors Region 10 relied
on to reach this conclusion include:
C-l
-------
1) The background PM2.5 monitoring data considered in the air quality analysis is
quality assured, quality controlled data from monitors operating for more than one
year that Region 10 believes will have accounted for much of the secondary
formation from existing regional emission sources that will occur in the Chukchi
Sea and Beaufort Sea regions. Monitoring data show low levels of daily PM2.5,
generally in the range of 2 ug/m3, with the higher PM2.5 values generally
occurring on days where windblown dust or fires are believed to be contributing
factors. Thus, there is no indication that secondary formation of PM2.5 from
existing sources in the North Slope is currently causing or contributing to
exceedances or a violation of the PM2.5 NAAQS in the onshore communities.
2) Modeled primary PM2.5 impacts from the Discoverer and Associated Fleet that,
when using a conservative "First Tier" approach to combining modeled primary
PM2.s impacts with monitored background PM2.s concentrations, are less than 67
percent of the PM2.5 NAAQS. Thus, although not expected, considerable
formation of secondary PM2.5 emissions could occur before the NAAQS would be
threatened.
3) Secondary PM2.5 impacts associated with Discoverer and Associated Fleet
precursor emissions are expected to be low near the emission release points where
modeled concentrations associated with primary PM2.5 emissions are highest,
because there has not been enough time for the secondary chemical reactions to
occur. Conversely, secondary PM2.5 impacts are more likely to be higher farther
from the Discoverer and the Associated Fleet where impacts from primary PM2.5
emissions from the Discoverer and the Associated Fleet are expected to be lower.
This makes it unlikely that maximum primary PM2.5 impacts and maximum
secondary PM2.5 impacts from the Discoverer and the Associated Fleet will occur
at the same time (paired in time) or location (paired in space). See March 23,
2010 PM2.5 Guidance Memo at 9.
4) The relatively small amount of NOx emissions (a PM2.5 precursor) that will be
authorized under these permits in comparison to existing NOx emissions in the
North Slope area in general, together with the generally low levels of PM2.5
recorded at monitoring stations in the area, make it unlikely that NOx emissions
from the Discoverer and the Associated Fleet would cause or contribute to a
violation of the PM2.5 NAAQS.
5) The background concentrations of certain chemical species that participate in
photochemical reactions to form secondary PM2.5, including ammonia and volatile
organic compounds, are expected to be negligible in the offshore air masses
where the Discoverer will be permitted to operate. The emissions authorized
under the permits of approximately 43 tons per year of VOC and 0.52 tons per
year of ammonia [citation omitted] would also not be expected to result in the
conversion of significant quantities of NOx emissions to secondary particles in
the areas impacted by primary PM2.5 emissions.
C-2
-------
6) There are several other conservative assumptions incorporated in the modeling
of primary PM2.5 emissions. These include the conservatism inherent in using a
"First Tier" approach to combining modeled primary PM2.5 impacts with
monitored background PM2.5 concentrations; assuming that the Discoverer will be
operating in a single drilling location for 3 years, when it is more likely that the
Discoverer will operate in a different location each year (if not more frequently);
orienting the Associated Fleet with hourly modeled wind direction and using
emission release characteristics based on actual meteorological conditions; and
the fact that the background monitored data used to represent offshore conditions
was collected onshore, where it is influenced by local sources, and is, therefore
likely to be a conservative estimate of background PIVb.s levels in the area of
maximum impact near the Discoverer.
7) With respect to the Chukchi Sea impacts, the predominant easterly wind
directions in the Chukchi Sea along with the distance between the project location
and the existing sources in the North Slope oil and gas fields are such that
emissions from the Discoverer and Associated Fleet are not likely to significantly
contribute to the maximum ambient concentrations resulting from the existing
source emissions.
8) Region 10 required post-construction monitoring in the previous permits
because the conservative screening modeling resulted in predicted levels that were
just below the 24-hour PM2.5 NAAQS. With the additional emission reductions in
direct PM2.5 emissions and the use of a refined model, predicted PM2.5
concentrations are now well below the NAAQS. However, Region 10 has decided
to retain the post-construction monitoring requirement in order to obtain better
information on the quantity of secondary particles in the North Slope
communities.
Based on these factors, and consistent with current guidance, Region 10 believes that
an adequate assessment has been made to demonstrate that the PM2.5 NAAQS will be
protected, accounting for primary PM2.5 impacts and potential contributions due to
PM2.5 precursors from the Discoverer and the Associated Fleet, and that it is not
necessary to use a photochemical model to further evaluate secondary PM2.5 formation
in these permitting actions.
C-3
-------
This Page Intentionally Left Blank
C-4
-------
Appendix D: Example of a Hybrid Qualitative/Quantitative Assessment of the Potential
for Secondary PMi.s Formation
In late 2013/early 2014, a permit applicant, Sasol, engaged and closely coordinated with
the EPA Region 6 Office and the Louisiana Department of Environmental Quality (LDEQ) in the
development of a hybrid qualitative/quantitative assessment of the potential for secondary fine
particulate matter (PM^.s) formation to cause or contribute to a violation of the PM2.5 National
Ambient Air Quality Standard (NAAQS) for their proposed major facility expansions in
Southern Louisiana. Sasol's expansion and new plant are a very large investment (up to $18
Billion), and Sasol and LDEQ worked closely with Region 6 to ensure that the ambient impacts
analysis was robust and defendable. In this particular hybrid assessment, Sasol took an approach
of using the formerly presumptive interpollutant trading ratios for NOX and SC>2 to PM2.5 offsets
and conservatively applied them in an illustrative example to demonstrate how relatively
inconsequential the impacts of secondary PM2.5 formation would be in the area of significant
impact surrounding their facility. In Sasol's case, the projected emissions increases of the direct
PM2.5 emissions and both PM2.5 precursors of NOX and SC>2 were above their respective
Significant Emissions Rates (SERs). Sasol also performed an analysis of PM2.5 speciated
monitoring data to further support the amount of impacts of nitrates on high PM2.5 values in the
area is relatively small and corroborate the ratio based analysis. Thus, this situation is an
example of a Case 3 assessment as presented in Table III-l of this guidance.
It is important to note that the EPA revised the provisions of the interpollutant trading
policy for PM2.5 on July 21, 2011, as described in Section III.2.2 of this guidance, to remove the
general presumptiveness of the interpollutant trading ratios without further technical
justification. Sasol is located in Southwestern Louisiana near the coast but chose to use the
western state value of 100 for NOx as a more conservative assessment. Sasol did not seek to
directly apply the formerly presumptive interpollutant trading ratios in an absolute sense. Rather,
the intention was to present the analysis in a manner to determine if further technical justification
would be required or if the application of the formerly presumptive interpollutant trading ratios
where adequate in a hybrid qualitative/quantitative sense.
Using the formerly presumptive interpollutant trading ratios resulted in total projected
secondary PM2.5 formation of 0.18 ng/m3 for the 24-hour PM2.5NAAQS and of 0.04 ng/m3 for
the annual PM2.5 NAAQS. When considered along with the primary PM2.5 impacts and
representative background data, the secondary PM2.5 impacts with respect to the 24-hour
NAAQS would have to be on the order of 32 times higher and to the annual NAAQS would have
to be at least 15 times greater before a potential projected violation might occur. This also
assumes that the maximum secondary PM2.5 impacts from the NOX and SC>2 precursor emissions
would occur at the same place and time as the maximum primary PM2.5 impacts. Based upon
Sasol's PM2.5 primary modeling projecting maximum concentrations very close to the facility
and decreasing 60% within three kilometers of the facility, it is very unlikely that the primary
and secondary maximums would ever occur at the same receptors. So, it would take a
considerable and unreasonably conservative change to the interpollutant trading ratios used in
this example before the NAAQS could be threatened based on the total proposed emissions
increases from this facility.
D-l
-------
At the same time Sasol also conducted an analysis of speciated data at a nearby monitor
to further corroborate the ratio analysis. There is a PM2.s monitor within 1A mile of the Sasol's
property line, but it does not have speciated data collection. Sasol utilized a representative PIVb.s
monitor approximately 25 miles away that did have long-term speciated data. Sasol evaluated the
PM2.5 speciated data from the nearby monitor to support that nitrate is not a large contributer to
high PM2.5 values on an annual basis or even on the higher daily values in the colder months
when nitrates would be expected to yield more secondary PM2.5.
Given the close coordination with the respective permitting authorities, it was determined
that a more thorough technical demonstration with respect to interpollutant ratios specific to this
source and area was not warranted and that the illustrative use of the formerly presumptive
interpollutant trading ratios was sufficient to demonstrate that secondary PM2.5 formation would
not cause or contribute to a violation of the NAAQS. The permit applicant's corroborative
analysis of the PM2.5 speciated data further supported that the main increase of emissions (NOX)
would not be expected to yield significant levels of secondary PIVb.s and the applicable ambient
standards will not be exceeded by this project.
Region 6 example:
Justification on Secondary PMi.5 Approach
At a December 13, 2013, meeting and on a January 17, 2014, conference call EPA
Region 6 requested an analysis to examine the fraction of sulfate and nitrate in the
PM2.5 measurements in the study area and additional justification on the modeling
approach for secondary PM2.5. This document presents the results of the requested
analysis.
Secondary PM2.5 is formed primarily from reaction of sulfur dioxide (802) emissions
to form particulate sulfate and from nitrogen oxides (NOX) reacting to form
particulate nitrate. The approach used to estimate the secondary particulate is
described in Section [Secondary Particulate Estimate (listed below)}.
With regard to NOX. and SC>2, the Sasol project emissions are dominated by NOX with
annual emissions of 1,595 tons per year compared to 862 emissions of only 121 tons
per year. EPA requested additional information on the secondary PM2.5 formation
from nitrate in the colder months.
The PM2s background monitor is the Westlake monitor located very near the project
site. However, this monitor did not record speciated PM2.5 data, so it is not possible to
compute the sulfate and nitrate fractions at this location. Monitors with speciated
PM2.5 data include the Port Arthur Memorial School (AIRS: 48-245-0021) in Port
Arthur, Texas, the Capitol Baton Rouge (AIRS: 22-033-0009), and the Shreveport
(AIRS: 22-015-0008) monitors. The Port Arthur monitor was chosen as being the
most representative because it is the closest monitor to the Sasol site and is in a less
urban area than the Capitol Baton Rouge monitor and is not as far inland as the
Shreveport monitor. The Port Arthur area is also located along the gulf coast and
most closely represents the combination of a metropolitan size and industrial presence
when compared to the Calcasieu Parish area where Sasol is located.
D-2
-------
The Port Arthur monitor, located in Port Arthur, Texas, is approximately 25 miles
west of the project. Given the regional nature of PIVb.s concentrations, this monitor
should be representative of the study area. The most recently available five years of
data for this site was for 2006-2010 and was obtained from the USEPA. The data
shows that nitrate makes up 2.6 percent of the average of the 24-hour concentrations
of PM2.5 and 2.3 percent of the 5-year average concentration. On the day with the
highest 24-hour average PM2.5 measurement, nitrate was 2.2 percent of the PM2.5
concentration.
In general, the generation of PM2.5 occurs more from nitrate during colder winter
months than during the summer. Examination of the worst 10% of PM2.5 days during
the colder months (November through February) at the Port Arthur monitor for 2006-
2010 reveals that the average nitrate contribution is 2.9 percent, only slightly higher
than the 5-year average concentration. Thus, even on days with high PM2.5
concentrations in the colder months, particulate nitrate is still a relatively small
portion of the total PM2.5 concentrations.
Based on this relatively low fraction of particulate nitrate in the observed PM2.5, and
the magnitude of existing NOX emissions in the area, it is clear that secondary
formation of particulate nitrate is not significant in the project area.
Particulate sulfate makes up 29.6 percent of the 5-year average of the 24-hour
concentrations and 29.0 percent of the 5-year average concentrations. On the day with
the highest 24-hour average PM2.5 observation, sulfate was 10.6 percent of the PM2.5
concentration.
Table 1 presents the total PM2.5 ambient air impact estimated using the formerly
approved interpollutant trading ratios. The nitrate equivalent ratio (1.026) is [6.5]
times greater than the sulfate equivalent ratio ([1.004]). While sulfate does make up a
significant portion of the total PIVb.s mass, the projected increase in SO2 emissions
(121 tpy) from the Sasol GTL and LCCP projects are a very small fraction of the total
SO2 emissions in the large industrial area impacting Port Arthur (i.e. Beaumont/Port
Arthur, Lake Charles, Houston/Galveston).
An implicit conservatism to the ratio approach that was used by Sasol is that the
primary and secondary impacts occur at the same location at the same time. The 24-
hour average modeled PM2.5 concentration is presented in Figure 1. Examination of
this figure reveals that the highest impact occurs very near the Sasol project border.
Within a few kilometers of the project site, the concentrations fall significantly from
the peak of modeled concentration of 9 |J,g/m3 to less than 3 |J,g/m3. Formation of
secondary sulfate and nitrate particulate is a fairly slow process with conversion rates
taking many hours to days. Thus, the peak secondary impacts are expected to occur
well downwind of the peak primary impacts.
Given this information, the study team is comfortable that the ambient ratio analysis
D-3
-------
presented in the ozone and secondary PM2.5 modeling report is an appropriate
approach to estimating the secondary PM2.s impacts for the project.
Table 1: Total PM25 (Primary 4 Secondary) Ambient Air Impacts with Comparison with
NAAQS and PSD Increment Limits
NAAQS 24-hour
NAAQS Annual
PSD increment 24-hour
PSD Increment Annual
Modeled
Primary
PM2J
(|ig/m3)
6
1.6
7.6
1.5
Sulfate
Equivalent
Ratio
1.004
1.004
1.004
1.004
Nitrate
Equivalent
Ratio
1.026
1.026
1.026
1.026
Primary and
Secondary PMjj
(^g/m3)
6.2423=29.2
1.6+9.8 = 11.4
7.8
1.5
PM2.5
Standard
(M/m3)
35
12
9
4
Monitored background concentrations for PM,3 2i-hour is 23 Ug/rn5 and for PM2.5 annual is9.S Ug/'m5.
D-4
-------
Figure l; Sasol Primary 24-hour Maximum PM2.5 Concentration Isopleth (j^g/m3). Peak value i
8,6
33600OQ-
£ 3350000-
Dt
1
3340000-
3330000
3320000
Primary PM2.5
24-hour Concenlralions
440000 45000C
460000
480000 490000 500000
UTM - Easting (meters)
D-5
-------
[Secondary Particulate Estimate]
Recent EPA guidance (EPA March 2013) has suggested the need to examine
secondary particulate formation. Directly emitted sulfur or nitrogen compounds are
likely to react with available water and other pollutants in secondary reactions to form
particulate ammonium sulfate -(NFL^SC^ or ammonium nitrate -NELjNOs. These
latter compounds are formed primarily downwind of the specific sources of concern,
given reaction times, ambient temperature and other environmental factors. The sulfur
compounds emitted by the two projects are in the form of SC>2. The nitrogen
compounds emitted by the two projects are in the form of NOX. The Sasol projects
(GTL and LCCP combined) would have 1,595 tpy of NOX, 121 tpy SO2, and 612 tpy
of direct PM2.5 emissions.
The NACAA/EPA recommendation to account for secondary PM2.5 formation is to
divide the projected emissions by a region average offset ratio. The national ratio for
SC>2 is 40 and for NOX is 100 for western states and 200 for eastern states. To be
conservative, the western value was used in the analysis since it estimates a higher
secondary ratio. The total PM2.5 emissions are calculated by multiplying the primary
PM2.5 modeled concentration by the ratio obtained from the secondary equivalent
PM2.5 calculation.
For the Sasol combined project emissions the formulas are:
Total Equivalent PM2.5 = Primary PM2.5 + (SO2/40) + (NOX/100) = 612 + (121/400) +
(1,595/100) = 631.0 ton/year
Total PM2.5 Impact (ng/m3) = Primary PM2.5 Impact (ng/m3) * (Total Equivalent
Primary PIVb.s (tpy) / Primary PIVb.s (tpy))
Total Equivalent PM2 5 / Primary PM2 5 = 631.0 tpy / 612 tpy = 1.03
Hence the modeled impacts for PM2.5 could be increased by a factor of 1.03 [(1.004 for SO2 and
1.026 for NOX)] to account for the secondary formation for those sources emitting significant
amounts of secondary PM2.5 precursor emissions.
D-6
-------
Appendix E: Example of the background monitoring data calculations for a Second Tier
24-hour modeling analysis
This appendix provides an illustrative example of the calculations and data sorting
recommendations for the background monitoring data to be used in a Second Tier 24-hour PM2.5
modeling analysis. In this example, it was determined through discussion and coordination with
the appropriate permitting authority that the impacts from the project source's primary PM2.5
emissions were most prominent during the cool season and were not temporally correlated with
background PM2.5 levels that were typical highest during the warm season. So, combining the
modeled and monitored contributions through a First Tier 24-hour PM2.5 modeling analysis was
determined to be potentially overly conservative. Extending the compliance demonstration to a
Second Tier analysis allows for a more refined and appropriate assessment of the cumulative
impacts on the primary PM2.5 emissions in this particular situation.
The example provided is from an idealized Federal Reference Method (FRM) PM2.5
monitoring site that operates on a daily (1-in-l day) frequency with 100% data completeness. In
this case, the annual 98th percentile concentration is the 8th highest concentration of the year. In
most cases, the FRM monitoring site will likely operate on a l-and-3 day frequency and will also
likely have missing data due to monitor maintenance or collected data not meeting all of the
quality assurance criteria. Please reference Appendix N to 40 CFR Part 50 to determine the
appropriate 98th percentile rank of the monitored data based on the monitor sampling frequency
and valid number of days sampled during each year.
The appropriate seasonal (or quarterly) background concentrations to be included as
inputs to the AERMOD model per a Second Tier 24-hour PM2.s modeling analysis are as
follows:
Step 1 - Start with the most recent 3-years of representative background PM2.5 ambient
monitoring data that are being used to develop the monitored background PM2.5 design
value. In this example, the 3-years of 2008 to 2010 are being used to determine the
monitored design value.
Step 2 - For each year, determine the appropriate rank for the daily 98th percentile PM2.5
concentration. Again, this idealized example is from a 1-in-l day monitor with 100% data
completeness. So, the 8th highest concentration of each year is the 98th percentile PM2.s
concentration. The 98th percentile PM2.5 concentration for 2008 is highlighted in Table E-
1. The full concentration data from 2009 and 2010 are not shown across the steps in this
Appendix for simplicity but would be similar to that of 2008.
Step 3 - Remove from further consideration in this analysis the PM2.5 concentrations
from each year that are greater than the 98* percentile PM2.5 concentration. In the case
presented for a 1-in-l day monitor, the top 7 concentrations are removed. If the monitor
were a l-in-3 day monitor, only the top 2 concentrations would be removed. The resultant
dataset after the top 7 concentrations have been removed from further consideration in
this analysis for 2008 is presented in Table E-2.
E-l
-------
Step 4 - For each year, divide the resultant annual dataset of the monitored data equal to
or less than the 98th percentile PM2.5 concentration into each season (or quarter). For
2008, the seasonal subsets are presented in Table E-3.
Step 5 - Determine the maximum PM2.5 concentration from each of the seasonal (or
quarterly) subsets created in Step 4 for each year. The maximum PM2.5 concentration
from each season for 2008 is highlighted in Table E-3.
Step 6 - Average the seasonal (or quarterly) maximums from Step 5 across the three
years of monitoring data to create the four seasonal background PM2.5 concentrations to
be included as inputs to the AERMOD model. These averages for the 2008 to 2010
dataset used in this example are presented in Table E-4. As noted above, the full
concentration data from 2009 and 2010 are not shown across the steps in this Appendix
for simplicity, but the seasonal maximums from 2009 and 2010 presented in Table E-4
were determined by following the previous five steps similar to that of 2008.
E-2
-------
Table E-l. 2008 Daily PM2.5 Concentrations
Date
1-Jan
2-Jan
3-Jan
4-Jan
5-Jan
6-Jan
7-Jan
8-Jan
9-Jan
10- Jan
11-Jan
12- Jan
13- Jan
14-Jan
15- Jan
16-Jan
17- Jan
18- Jan
19- Jan
20-Jan
21-Jan
22-Jan
23-Jan
24-Jan
25-Jan
26-Jan
27-Jan
28-Jan
29-Jan
30-Jan
31 -Jan
1-Feb
2-Feb
3-Feb
4-Feb
5-Feb
6-Feb
7-Feb
8-Feb
9-Feb
10-Feb
1 1-Feb
1 2-Feb
1 3-Feb
14-Feb
1 5-Feb
Cone.
10.4
5.4
10.0
16.4
11.2
11.1
10.2
11.4
8.1
9.4
5.7
8.9
18.1
11.0
11.8
10.7
10.0
15.6
18.0
6.6
7.4
13.5
16.0
9.4
12.6
13.6
16.1
10.0
10.4
6.9
4.9
5.4
7.1
10.9
12.1
17.1
10.3
4.0
9.7
11.5
3.0
5.5
18.9
17.6
11.2
14.4
Date
1 6-Feb
1 7-Feb
1 8-Feb
1 9-Feb
20-Feb
21-Feb
22-Feb
23-Feb
24-Feb
25-Feb
26-Feb
27-Feb
28-Feb
29-Feb
1-Mar
2-Mar
3-Mar
4-Mar
5-Mar
6-Mar
7-Mar
8-Mar
9-Mar
10-Mar
11 -Mar
1 2-Mar
1 3-Mar
14-Mar
1 5-Mar
1 6-Mar
1 7-Mar
1 8-Mar
1 9-Mar
20-Mar
21 -Mar
22-Mar
23-Mar
24-Mar
25-Mar
26-Mar
27-Mar
28-Mar
29-Mar
30-Mar
3 1-Mar
1-Apr
Cone.
15.1
11.8
3.4
4.5
4.8
11.9
20.1
11.4
19.3
18.2
12.8
5.5
9.7
12.1
9.6
5.6
12.5
7.1
4.9
9.9
11.2
5 5
8.8
11.0
12.1
9.7
15.1
21.6
16.6
7.9
9.6
10.3
8.4
4.9
8.7
13.3
12.2
10.3
11.9
20.1
22.5
18.2
10.8
6.4
3.3
7.8
Date
2-Apr
3-Apr
4-Apr
5-Apr
6-Apr
7-Apr
8-Apr
9-Apr
10-Apr
11-Apr
12-Apr
13-Apr
14-Apr
15-Apr
16-Apr
17-Apr
18-Apr
19-Apr
20-Apr
21-Apr
22-Apr
23-Apr
24-Apr
25-Apr
26-Apr
27-Apr
28-Apr
29-Apr
30-Apr
1-May
2-May
3-May
4-May
5-May
6-May
7-May
8-May
9-May
10-May
11 -May
1 2-May
1 3-May
14-May
1 5-May
1 6-May
1 7-May
Cone.
10.5
8.2
9.7
6.9
6.3
7.9
9.8
16.5
13.3
11.0
8.8
6.3
5.1
7.9
8.2
14.7
22.5
12.8
6.9
7.5
6.0
9.1
10.3
12.0
12.5
11.3
7.6
7.4
11.4
12.6
10.0
11.2
10.4
15.7
16.1
16.8
14.5
11.7
9.0
6.7
7.9
8.3
12.2
13.1
8.8
8.2
Date
1 8-May
1 9-May
20-May
21 -May
22-May
23-May
24-May
25-May
26-May
27-May
28-May
29-May
30-May
3 1-May
1-Jun
2-Jun
3-Jun
4-Jun
5-Jun
6-Jun
7-Jun
8-Jun
9-Jun
10-Jun
11-Jun
12-Jun
13-Jun
14-Jun
15-Jun
16-Jun
17-Jun
18-Jun
19-Jun
20-Jun
21-Jun
22-Jun
23-Jun
24-Jun
25-Jun
26-Jun
27-Jun
28-Jun
29-Jun
30-Jun
1-Jul
2-Jul
Cone.
11.1
7.7
13.6
12.1
10.0
13.3
11.2
17.7
14.2
15.4
13.9
9.3
14.5
20.5
15.3
11.5
17.9
21.1
17.9
17.6
15.0
22.3
27.9
21.6
19.4
21.2
29.1
15.6
14.8
17.8
12.6
10.5
15.0
22.7
18.7
15.2
16.8
15.1
20.7
23.0
17.8
12.4
12.7
8.9
7.1
13.8
Date
3-Jul
4-Jul
5-Jul
6-Jul
7-Jul
8-Jul
9-Jul
10-Jul
11-Jul
12-Jul
13-Jul
14-Jul
15-Jul
16-Jul
17-Jul
18-Jul
19-Jul
20-Jul
21-Jul
22-Jul
23-Jul
24-Jul
25-Jul
26-Jul
27-Jul
28-Jul
29-Jul
30-Jul
31-Jul
1-Aug
2-Aug
3-Aug
4-Aug
5-Aug
6-Aug
7-Aug
8-Aug
9-Aug
10-Aug
11-Aug
12-Aug
13-Aug
14-Aug
15-Aug
16-Aug
17-Aug
Cone.
17.1
19.8
14.3
11.5
14.3
12.2
11.1
9.7
16.4
21.5
25.1
11.7
18.9
28.9
27.6
12.8
6.2
20.1
26.5
16.9
12.8
7.9
15.7
24.9
22 2
17.5
19.1
21.1
18.0
16.3
19.3
17.9
25.1
29.3
19.1
14.0
10.8
15.0
21.7
14.3
14.7
13.0
13.5
17.5
23.9
18.4
Date
18-Aug
19-Aug
20-Aug
21-Aug
22-Aug
23-Aug
24-Aug
25-Aug
26-Aug
27-Aug
28-Aug
29-Aug
30-Aug
31-Aug
1-Sep
2-Sep
3-Sep
4-Sep
5-Sep
6-Sep
7-Sep
8-Sep
9-Sep
10-Sep
11-Sep
12-Sep
13-Sep
14-Sep
15-Sep
16-Sep
17-Sep
18-Sep
19-Sep
20-Sep
21-Sep
22-Sep
23-Sep
24-Sep
25-Sep
26-Sep
27-Sep
28-Sep
29-Sep
30-Sep
1-Oct
2-Oct
Cone.
18.7
21.5
20.1
18.4
16.7
13.8
19.0
17.6
15.4
12.6
12.1
10.1
17.2
19.9
19.4
18.2
24.0
15.4
12.4
12.5
15.8
23.4
11.5
6.0
11.8
10.7
7.6
7.5
7.1
7.7
11.3
16.8
14.8
8.0
10.8
14.5
21.2
8.6
1.2
16.0
12.1
18.0
17.8
16.4
12.3
8.2
Date
3-Oct
4-Oct
5-Oct
6-Oct
7-Oct
8-Oct
9-Oct
10-Oct
11-Oct
12-Oct
13-Oct
14-Oct
15-Oct
16-Oct
17-Oct
18-Oct
19-Oct
20-Oct
21-Oct
22-Oct
23-Oct
24-Oct
25-Oct
26-Oct
27-Oct
28-Oct
29-Oct
30-Oct
31-Oct
1-Nov
2-Nov
3-Nov
4-Nov
5-Nov
6-Nov
7-Nov
8-Nov
9-Nov
10-Nov
11-Nov
12-Nov
13-Nov
14-Nov
15-Nov
16-Nov
17-Nov
Cone.
12.3
19.5
23.7
19.8
21.7
12.2
5.1
10.2
10.7
5.6
5.9
9.7
12.8
16.4
12.0
7.9
6.6
8.1
12.2
4.6
6.1
4.6
4.5
10.5
6.4
4.6
5.6
7.6
11.2
16.2
17.3
18.3
8.9
5.8
8.6
15.0
8.3
10.0
12.8
11.8
14.8
14.5
7.7
3.6
4.6
7.8
Date
18-Nov
19-Nov
20-Nov
21-Nov
22-Nov
23-Nov
24-Nov
25-Nov
26-Nov
27-Nov
28-Nov
29-Nov
30-Nov
1-Dec
2-Dec
3-Dec
4-Dec
5-Dec
6-Dec
7-Dec
8-Dec
9-Dec
10-Dec
11 -Dec
1 2-Dec
1 3-Dec
14-Dec
1 5-Dec
1 6-Dec
1 7-Dec
1 8-Dec
1 9-Dec
20-Dec
21 -Dec
22-Dec
23-Dec
24-Dec
25-Dec
26-Dec
27-Dec
28-Dec
29-Dec
30-Dec
31-Dec
Cone.
4.4
8.2
11.1
5.3
8.9
14.0
12.7
9.7
12.8
16.6
17.2
16.6
4.5
7.5
10.6
16.7
12.5
7.3
10.4
13.4
10.5
9.3
6.5
3.0
3.5
10.2
17.6
12.4
9.7
7.0
7.9
6.9
8.1
4.9
7.7
7.7
10.5
6.5
7.6
13.3
6.4
3.7
4.7
4.4
Annual 98th Percentile Concentration = 21.5 ng/
E-3
-------
th
Table E-2. 2008 Daily PM2.5 Concentrations Less Than or Equal to the 98m Percentile
Date
1-Jan
2-Jan
3-Jan
4-Jan
5-Jan
6-Jan
7-Jan
8-Jan
9-Jan
10- Jan
11-Jan
12- Jan
13- Jan
14-Jan
15- Jan
16- Jan
17- Jan
18-Jan
19- Jan
20-Jan
21-Jan
22-Jan
23-Jan
24-Jan
25-Jan
26-Jan
27-Jan
28-Jan
29-Jan
30-Jan
31 -Jan
1-Feb
2-Feb
3-Feb
4-Feb
5-Feb
6-Feb
7-Feb
8-Feb
9-Feb
10-Feb
1 1-Feb
1 2-Feb
1 3-Feb
14-Feb
1 5-Feb
Cone.
10.4
5.4
10.0
16.4
11.2
11.1
10.2
11.4
8.1
9.4
5.7
8.9
18.1
11.0
11.8
10.7
10.0
15.6
18.0
6.6
7.4
13.5
16.0
9.4
12.6
13.6
16.1
10.0
10.4
6.9
4.9
5.4
7.1
10.9
12.1
17.1
10.3
4.0
9.7
11.5
3.0
5.5
18.9
17.6
11.2
14.4
Date
1 6-Feb
1 7-Feb
1 8-Feb
1 9-Feb
20-Feb
21-Feb
22-Feb
23-Feb
24-Feb
25-Feb
26-Feb
27-Feb
28-Feb
29-Feb
1-Mar
2-Mar
3-Mar
4-Mar
5-Mar
6-Mar
7-Mar
8-Mar
9-Mar
10-Mar
11 -Mar
1 2-Mar
13-Mar
14-Mar
15-Mar
16-Mar
1 7-Mar
1 8-Mar
1 9-Mar
20-Mar
21-Mar
22-Mar
23-Mar
24-Mar
25-Mar
26-Mar
27-Mar
28-Mar
29-Mar
30-Mar
3 1-Mar
1-Apr
Cone.
15.1
11.8
3.4
4.5
4.8
11.9
20.1
11.4
19.3
18.2
12.8
5 5
9.7
12.1
9.6
5.6
12.5
7.1
4.9
9.9
11.2
5.5
8.8
11.0
12.1
9.7
15.1
21.6
16.6
7.9
9.6
10.3
8.4
4.9
8.7
13.3
12.2
10.3
11.9
20.1
22 5
18.2
10.8
6.4
3.3
7.8
Date
2-Apr
3-Apr
4-Apr
5-Apr
6-Apr
7-Apr
8-Apr
9-Apr
10-Apr
11 -Apr
12-Apr
13-Apr
14-Apr
15-Apr
16-Apr
17-Apr
18-Apr
19-Apr
20-Apr
21 -Apr
22-Apr
23-Apr
24-Apr
25-Apr
26-Apr
27-Apr
28-Apr
29-Apr
30-Apr
1-May
2-May
3-May
4-May
5-May
6-May
7-May
8-May
9-May
10-May
11 -May
1 2-May
13-May
14-May
1 5-May
1 6-May
1 7-May
Cone.
10.5
8.2
9.7
6.9
6.3
7.9
9.8
16.5
13.3
11.0
8.8
6.3
5.1
7.9
8.2
14.7
22.5
12.8
6.9
7.5
6.0
9.1
10.3
12.0
12.5
11.3
7.6
7.4
11.4
12.6
10.0
11.2
10.4
15.7
16.1
16.8
14.5
11.7
9.0
6.7
7.9
8.3
12.2
13.1
8.8
8.2
Date
1 8-May
1 9-May
20-May
21 -May
22-May
23-May
24-May
25-May
26-May
27-May
28-May
29-May
30-May
3 1-May
1-Jun
2-Jun
3-Jun
4-Jun
5-Jun
6-Jun
7-Jun
8-Jun
9-Jun
10-Jun
11-Jun
12-Jun
13-Jun
14-Jun
15-Jun
16-Jun
17-Jun
18-Jun
19-Jun
20-Jun
21-Jun
22-Jun
23-Jun
24-Jun
25-Jun
26-Jun
27-Jun
28-Jun
29-Jun
30-Jun
1-Jul
2-Jul
Cone.
11.1
7.7
13.6
12.1
10.0
13.3
11.2
17.7
14.2
15.4
13.9
9.3
14.5
20.5
15.3
11.5
17.9
21.1
17.9
17.6
15.0
22.3
RC
21.6
19.4
21.2
RC
15.6
14.8
17.8
12.6
10.5
15.0
22.7
18.7
15.2
16.8
15.1
20.7
23.0
17.8
12.4
12.7
8.9
7.1
13.8
Date
3-Jul
4-Jul
5-Jul
6-Jul
7-Jul
8-Jul
9-Jul
10-Jul
11-Jul
12-Jul
13-Jul
14-Jul
15-Jul
16-Jul
17-Jul
18-Jul
19-Jul
20-Jul
21-Jul
22-Jul
23-Jul
24-Jul
25-Jul
26-Jul
27-Jul
28-Jul
29-Jul
30-Jul
31-Jul
1-Aug
2-Aug
3-Aug
4-Aug
5-Aug
6-Aug
7-Aug
8-Aug
9-Aug
10-Aug
11-Aug
12-Aug
13-Aug
14-Aug
15-Aug
1 6-Aug
17-Aug
Cone.
17.1
19.8
14.3
11.5
14.3
12.2
11.1
9.7
16.4
21.5
RC
11.7
18.9
RC
RC
12.8
6.2
20.1
RC
16.9
12.8
7.9
15.7
24.9
22 2
17.5
19.1
21.1
18.0
16.3
19.3
17.9
25.1
RC
19.1
14.0
10.8
15.0
21.7
14.3
14.7
13.0
13.5
17.5
23.9
18.4
Date
18-Aug
19-Aug
20-Aug
21-Aug
22-Aug
23-Aug
24-Aug
25-Aug
26-Aug
27-Aug
28-Aug
29-Aug
30-Aug
31-Aug
1-Sep
2-Sep
3-Sep
4-Sep
5-Sep
6-Sep
7-Sep
8-Sep
9-Sep
10-Sep
11-Sep
12-Sep
13-Sep
14-Sep
15-Sep
16-Sep
17-Sep
18-Sep
19-Sep
20-Sep
21-Sep
22-Sep
23-Sep
24-Sep
25-Sep
26-Sep
27-Sep
28-Sep
29-Sep
30-Sep
1-Oct
2-Oct
Cone.
18.7
21.5
20.1
18.4
16.7
13.8
19.0
17.6
15.4
12.6
12.1
10.1
17.2
19.9
19.4
18.2
24.0
15.4
12.4
12.5
15.8
23.4
11.5
6.0
11.8
10.7
7.6
7.5
7.1
7.7
11.3
16.8
14.8
8.0
10.8
14.5
21.2
8.6
1.2
16.0
12.1
18.0
17.8
16.4
12.3
8.2
Date
3-Oct
4-Oct
5-Oct
6-Oct
7-Oct
8-Oct
9-Oct
10-Oct
11-Oct
12-Oct
13-Oct
14-Oct
15-Oct
16-Oct
17-Oct
18-Oct
19-Oct
20-Oct
21-Oct
22-Oct
23-Oct
24-Oct
25-Oct
26-Oct
27-Oct
28-Oct
29-Oct
30-Oct
31-Oct
1-Nov
2-Nov
3-Nov
4-Nov
5-Nov
6-Nov
7-Nov
8-Nov
9-Nov
10-Nov
11-Nov
12-Nov
13-Nov
14-Nov
15-Nov
16-Nov
17-Nov
Cone.
12.3
19.5
23.7
19.8
21.7
12.2
5.1
10.2
10.7
5.6
5.9
9.7
12.8
16.4
12.0
7.9
6.6
8.1
12.2
4.6
6.1
4.6
4.5
10.5
6.4
4.6
5.6
7.6
11.2
16.2
17.3
18.3
8.9
5.8
8.6
15.0
8.3
10.0
12.8
11.8
14.8
14.5
7.7
3.6
4.6
7.8
Date
18-Nov
19-Nov
20-Nov
21-Nov
22-Nov
23-Nov
24-Nov
25-Nov
26-Nov
27-Nov
28-Nov
29-Nov
30-Nov
1-Dec
2-Dec
3-Dec
4-Dec
5-Dec
6-Dec
7-Dec
8-Dec
9-Dec
10-Dec
11 -Dec
1 2-Dec
1 3-Dec
14-Dec
1 5-Dec
1 6-Dec
1 7-Dec
1 8-Dec
19-Dec
20-Dec
21 -Dec
22-Dec
23-Dec
24-Dec
25-Dec
26-Dec
27-Dec
28-Dec
29-Dec
30-Dec
31-Dec
Cone.
4.4
8.2
11.1
5.3
8.9
14.0
12.7
9.7
12.8
16.6
17.2
16.6
4.5
7.5
10.6
16.7
12.5
7.3
10.4
13.4
10.5
9.3
6.5
3.0
3.5
10.2
17.6
12.4
9.7
7.0
7.9
6.9
8.1
4.9
7.7
7.7
10.5
6.5
7.6
13.3
6.4
3.7
4.7
4.4
Annual 98th Percentile Concentration = 21.5 |ag/m
RC = Above 98th Percentile and Removed from Consideration
E-4
-------
Table E-3. 2008 Daily PM2.5 Concentrations Less Than or Equal to the 98
th
Season/ Quarter 1
Date
1-Jan
2-Jan
3-Jan
4-Jan
5-Jan
6-Jan
7-Jan
8-Jan
9-Jan
10-Jan
11-Jan
12-Jan
13-Jan
14-Jan
15-Jan
16-Jan
17-Jan
18- Jan
19-Jan
20-Jan
21 -Jan
22-Jan
23-Jan
24-Jan
25-Jan
26- Jan
27-Jan
28-Jan
29-Jan
30- Jan
31-Jan
1-Feb
2-Feb
3-Feb
4-Feb
5-Feb
6-Feb
7-Feb
8-Feb
9-Feb
10-Feb
1 1-Feb
12-Feb
13-Feb
14-Feb
15-Feb
Cone.
10.4
5.4
10.0
16.4
11.2
11.1
10.2
11.4
8.1
9.4
5.7
8.9
18.1
11.0
11.8
10.7
10.0
15.6
18.0
6.6
7.4
13.5
16.0
9.4
12.6
13.6
16.1
10.0
10.4
6.9
4.9
5.4
7.1
10.9
12.1
17.1
10.3
4.0
9.7
11.5
3.0
5.5
18.9
17.6
11.2
14.4
Date
16-Feb
17-Feb
1 8-Feb
19-Feb
20-Feb
21-Feb
22-Feb
23-Feb
24-Feb
25-Feb
26-Feb
27-Feb
28-Feb
29-Feb
1-Mar
2-Mar
3-Mar
4-Mar
5-Mar
6-Mar
7-Mar
8-Mar
9-Mar
10-Mar
11 -Mar
12-Mar
13-Mar
14-Mar
15-Mar
16-Mar
17-Mar
1 8-Mar
19-Mar
20-Mar
21 -Mar
22-Mar
23-Mar
24-Mar
25-Mar
26-Mar
27-Mar
28-Mar
29-Mar
30-Mar
31-Mar
Seasonal / Quarterly Maximum
Cone.
15.1
11.8
3.4
4.5
4.8
11.9
20.1
11.4
19.3
18.2
12.8
5.5
9.7
12.1
9.6
5.6
12.5
7.1
4.9
9.9
11.2
5.5
8.8
11.0
12.1
9.7
15.1
21.6
16.6
7.9
9.6
10.3
8.4
4.9
8.7
13.3
12.2
10.3
11.9
20.1
22.5
18.2
10.8
6.4
3.3
22.5
Seas on /Quarter 2
Date Cone.
1-Apr
2-Apr
3-Apr
4-Apr
5-Apr
6-Apr
7-Apr
8-Apr
9-Apr
10-Apr
11 -Apr
12-Apr
13-Apr
14-Apr
15-Apr
16-Apr
17-Apr
18-Apr
19-Apr
20-Apr
21-Apr
22-Apr
23-Apr
24-Apr
25-Apr
26-Apr
27-Apr
28-Apr
29-Apr
30-Apr
1-May
2-May
3-May
4-May
5-May
6-May
7-May
8-May
9-May
10-May
11 -May
1 2-May
1 3-May
1 4-May
1 5-May
1 6-May
7.8
10.5
8.2
9.7
6.9
6.3
7.9
9.8
16.5
13.3
11.0
8.8
6.3
5.1
7.9
8.2
14.7
22.5
12.8
6.9
7.5
6.0
9.1
10.3
12.0
12.5
11.3
7.6
7.4
11.4
12.6
10.0
11.2
10.4
15.7
16.1
16.8
14.5
11.7
9.0
6.7
7.9
8.3
12.2
13.1
8.8
Date Cone.
1 7-May
1 8-May
1 9-May
20-May
21-May
22-May
23-May
24-May
25-May
26-May
27-May
28-May
29-May
30-May
3 1-May
1-Jun
2-Jun
3-Jun
4-Jun
5-Jun
6-Jun
7-Jun
8-Jun
9-Jun
10-Jun
11-Jun
12-Jun
13-Jun
14-Jun
15-Jun
16-Jun
17-Jun
18-Jun
19-Jun
20-Jun
21-Jun
22-Jun
23-Jun
24-Jun
25-Jun
26-Jun
27-Jun
28-Jun
29-Jun
30-Jun
Seasonal / Quarterly Maximum
8.2
11.1
7.7
13.6
12.1
10.0
13.3
11.2
17.7
14.2
15.4
13.9
9.3
14.5
20.5
15.3
11.5
17.9
21.1
17.9
17.6
15.0
22.3
RC
21.6
19.4
21.2
RC
15.6
14.8
17.8
12.6
10.5
15.0
22.7
18.7
15.2
16.8
15.1
20.7
23.0
17.8
12.4
12.7
8.9
23.0
Seas on /Quarter 3
Date Cone.
1-Jul
2-Jul
3-Jul
4-Jul
5-Jul
6-Jul
7-Jul
8-Jul
9-Jul
10-Jul
11-Jul
12-Jul
13-Jul
14-Jul
15-Jul
16-Jul
17-Jul
18-Jul
19-Jul
20-Jul
21-Jul
22-Jul
23-Jul
24-Jul
25-Jul
26-Jul
27-Jul
28-Jul
29-Jul
30-Jul
31-Jul
1-Aug
2-Aug
3-Aug
4-Aug
5-Aug
6-Aug
7-Aug
8-Aug
9-Aug
10-Aug
11-Aug
12-Aug
13-Aug
14-Aug
15-Aug
7.1
13.8
17.1
19.8
14.3
11.5
14.3
12.2
11.1
9.7
16.4
21.5
RC
11.7
18.9
RC
RC
12.8
6.2
20.1
RC
16.9
12.8
7.9
15.7
24.9
22.2
17.5
19.1
21.1
18.0
16.3
19.3
17.9
25.1
RC
19.1
14.0
10.8
15.0
21.7
14.3
14.7
13.0
13.5
17.5
Date Cone.
16-Aug
17-Aug
18-Aug
19-Aug
20-Aug
21-Aug
22-Aug
23-Aug
24-Aug
25-Aug
26-Aug
27-Aug
28-Aug
29-Aug
30-Aug
31-Aug
1-Sep
2-Sep
3-Sep
4-Sep
5-Sep
6-Sep
7-Sep
8-Sep
9-Sep
10-Sep
11-Sep
12-Sep
13-Sep
14-Sep
15-Sep
16-Sep
17-Sep
18-Sep
19-Sep
20-Sep
21-Sep
22-Sep
23-Sep
24-Sep
25-Sep
26-Sep
27-Sep
28-Sep
29-Sep
30-Sep
Seasonal / Quarterly Maximum
23.9
18.4
18.7
21.5
20.1
18.4
16.7
13.8
19.0
17.6
15.4
12.6
12.1
10.1
17.2
19.9
19.4
18.2
24.0
15.4
12.4
12.5
15.8
23.4
11.5
6.0
11.8
10.7
7.6
7.5
7.1
7.7
11.3
16.8
14.8
8.0
10.8
14.5
21.2
8.6
1.2
16.0
12.1
18.0
17.8
16.4
25.1
Percentile by Quarter
Season /Quarter 4
Date Cone.
1-Oct
2-Oct
3-Oct
4-Oct
5-Oct
6-Oct
7-Oct
8-Oct
9-Oct
10-Oct
11-Oct
12-Oct
13-Oct
14-Oct
15-Oct
16-Oct
17-Oct
18-Oct
19-Oct
20-Oct
21-Oct
22-Oct
23-Oct
24-Oct
25-Oct
26-Oct
27-Oct
28-Oct
29-Oct
30-Oct
31-Oct
1-Nov
2-Nov
3-Nov
4-Nov
5-Nov
6-Nov
7-Nov
8-Nov
9-Nov
10-Nov
11-Nov
12-Nov
13-Nov
14-Nov
15-Nov
12.3
8.2
12.3
19.5
23.7
19.8
21.7
12.2
5.1
10.2
10.7
5.6
5.9
9.7
12.8
16.4
12.0
7.9
6.6
8.1
12.2
4.6
6.1
4.6
4.5
10.5
6.4
4.6
5.6
7.6
11.2
16.2
17.3
18.3
8.9
5.8
8.6
15.0
8.3
10.0
12.8
11.8
14.8
14.5
7.7
3.6
Date Cone.
16-Nov
17-Nov
18-Nov
19-Nov
20-Nov
21-Nov
22-Nov
23-Nov
24-Nov
25-Nov
26-Nov
27-Nov
28-Nov
29-Nov
30-Nov
1-Dec
2-Dec
3-Dec
4-Dec
5-Dec
6-Dec
7-Dec
8-Dec
9-Dec
10-Dec
11-Dec
12-Dec
13-Dec
14-Dec
15-Dec
16-Dec
17-Dec
18-Dec
19-Dec
20-Dec
21-Dec
22-Dec
23-Dec
24-Dec
25-Dec
26-Dec
27-Dec
28-Dec
29-Dec
30-Dec
31-Dec
Seasonal / Quarterly Maximum
4.6
7.8
4.4
8.2
11.1
5.3
8.9
14.0
12.7
9.7
12.8
16.6
17.2
16.6
4.5
7.5
10.6
16.7
12.5
7.3
10.4
13.4
10.5
9.3
6.5
3.0
3.5
10.2
17.6
12.4
9.7
7.0
7.9
6.9
8.1
4.9
7.7
7.7
10.5
6.5
7.6
13.3
6.4
3.7
4.7
4.4
23.7
Seasonal / Quarterly Maximum Concentration
RC = Abow 98th Percentile andRemowdfrom Consideration
E-5
-------
Table E-4. Resulting Average of Seasonal (or Quarterly) Maximums for Inclusion into AERMOD
Seasonal / Quarterly Average Highest Monitored Concentration
(From Annual Datasets Equal To and Less Than the 98th Percentile)
2008
2009
2010
Average
Ql
22.5
21.1
20.7
21.433
Q2
23.0
20.7
22.6
22.100
Q3
25.1
21.2
23.5
23.267
Q4
23.7
19.8
20.7
21.400
(Note, the complete datasetsfor 2009 and 2010 are not shown in Appendix E but would follow the same steps as for 2008)
E-6
-------
-------
United States Office of Air Quality Planning and Standards Publication No. EPA-454/B-14-001
Environmental Protection Air Quality Assessment Division [May 2014]
Agency Research Triangle Park, NC
------- |