Denial of Petitions for Reconsideration of Certain
Issues: MATS and Utility NSPS
March 2015
-------
This page intentionally left blank.
11
-------
Denial of Petitions for Reconsideration of Certain Issues:
MATS and Utility NSPS
U.S. Environmental Protection Agency
Office of Air Quality Planning and Standards
Sector Policies and Programs Division
Research Triangle Park, North Carolina 27711
March 2015
in
-------
This page intentionally left blank.
iv
-------
Table of Contents
MATS
1.0 Appropriate and Necessary Finding
1.1 Ability to comment on Hg Risk Technical Support Documents (TSDs)
1.2 Reliability of Hg Risk TSD as a basis for "Appropriate and Necessary" finding
1.3 Hg biomarker data
1.4 Fish consumption rates in Hg Risk TSD
1.5 Fish tissue concentrations and samples in Hg Risk TSD
1.6 Mercury MAPs assumption in Hg Risk TSD
1.7 Screening of watersheds with significant non-atmospheric deposition in Hg Risk
1.8 Spatial scale of non-stationary watersheds
1.9 Alternative risk calculations for Hg Risk TSD
1.10 Methylmercury bioavailability and Se/Hg ratio assumptions in Hg Risk TSD
1.11 Health benefits of fish consumption
1.12 Assumptions regarding fish advisories in Hg Risk TSD
1.13 Rounding of EGU contribution thresholds in Hg Risk TSD
1.14 Modeling Hg emissions from final CSAPR in Hg Risk TSD
1.15 Boundary conditions in Hg deposition modeling
1.16 Chemical reactions in the Hg deposition modeling
1.17 Uncertainties in local and regional Hg deposition
1.18 Alternate analysis of Hg deposition
1.19 Emissions in non-Hg risk case studies
1.20 Nickel risk
1.21 EPRF s Multipathway Risk Assessment
2.0 Affirmative Defense
3.0 Averaging
4.0 Beyond-the-floor - Low Rank Virgin Coal Hg Limit
5.0 Bias Against New Cogeneration EGUs
6.0 Common Stack
7.0 Deadlines
7.1 Deadlines for "Newly Applicable Monitoring Requirements"
7.2 January 1, 2012, start date for reporting of performance test results
7.3 Deadline for certification of Hg CEMS and sorbent trap monitoring systems
8.0 Definitions - Applicability
9.0 Definitions - FGD Definition Should Include CFB
-------
10.0 Definitions - "Modification" in Definition of New or Reconstructed
11.0 Definitions - Natural Gas
12.0 Differentiation Between "Coal-fired" and "Solid Oil Derived Fuel-fired" EGUs
13.0 Emission Standards - Based on Non-detectable Test Results
14.0 Emission Standards - Existing IGCC Pb limit
15.0 General Provisions - Applicability
16.0 Heath-based Emission Limit for Acid Gases
17.0 Low Emitting EGU - Bypass Stack
18.0 Monitoring and Testing
18.1 Definition of "deviation"
18.2 Monitoring system requirements
18.3 Citati on correcti on
18.4 Testing
18.5 Bypass stack
19.0 Non-continental Liquid Oil-fired EGUs
19.1 Subcategory size and applicability
19.2 Subcategory data
19.3 Exclusion of data
19.4 Use of outlier data
20.0 PM limits
20.1 Filterable PM as a surrogate
20.2 Requirement for fuel sampling
20.3 Methodology for calculating PM floor
20.4 The Agency improperly assumed that its data captured all best performing EGUs
20.5 Beyond-the-floor analysis for PM
20.6 Final PM "MACT" floor for existing coal-fired EGUs is weaker than existing PM limits
at more than 12 percent of EGUs subject to the rule
20.7 Solid oil-derived fuel-fired EGUs
21.0 Regulatory Impact Analysis and Statutory or Executive Order Analyses
21.1 Regulatory Impact Analysis (RIA) for non-continental liquid oil-fired EGUs
21.2 Regulatory Flexibility Act (RFA) and Small Business Regulatory Enforcement Fairness
Act (SBREFA) issues
21.3 Unfunded Mandate Reform Act (UMRA) and Executive Order 13132 (Federalism) issues
21.4 Methodological issues in estimates of PM2.5 co-benefits
vi
-------
22.0 Reliability and Compliance Time
22.1 Electricity reliability
22.2 Policies and procedure for extending the compliance time beyond 3 years
22.3 EPA made erroneous assumptions with regard to ERCOT
23.0 Reporting
24.0 Statistical Analyses
25.0 Subcategorization - Coal Refuse-fired CFB EGUs
26.0 Subcategorization - IGCC
26.1 IGCC-specific definitions
26.2 Revision of IGCC to specifically exclude the gasifier
26.3 IGCC-specific work practice standards
27.0 Variability
28.0 Work Practice Standards - Organic HAP
Appendix A List of NESHAP Petitioners
vii
-------
NSPS
29.0 Limiting Opacity Standard Exemption for Subpart D Facilities
30.0 Subpart Da Condensable PM Testing for Post-2011 Facilities
31.0 Subpart Da Electronic Reporting Requirements for Facilities with PM CEMS
32.0 Periods of Startup, Shutdown, and Malfunction for EGUs
33.0 Affirmative Defense for EGUs
34.0 Definition of Periods of Out-of-control for EGUs
35.0 Frequency of Subpart D Visible Emissions Testing
36.0 Redundant Subpart Da Recordkeeping
37.0 De Minimis Use of Natural Gas in Subpart Db Steam Generating Units
Appendix B List of NSPS Petitioners
viii
-------
Foreword
On February 16, 2012, pursuant to sections 111 and 112 of the Clean Air Act (CAA), the U.S.
Environmental Protection Agency (EPA) published the final rules titled "National Emission
Standards for Hazardous Air Pollutants from Coal- and Oil-fired Electric Utility Steam
Generating Units and Standards of Performance for Fossil-Fuel-Fired Electric Utility, Industrial -
Commercial-Institutional, and Small Industrial-Commercial-Institutional Steam Generating
Units" (77 FR 9304). The National Emission Standards for Hazardous Air Pollutants (NESHAP)
rule issued pursuant to CAA section 112 is referred to as the Mercury and Air Toxics Standards
(MATS), and the New Source Performance Standards rule issued pursuant to CAA section 111 is
referred to as the Utility NSPS. Following promulgation of the final rules, the Administrator
received petitions for reconsideration of numerous provisions of both MATS and the Utility
NSPS pursuant to CAA section 307(d)(7)(B).1
CAA section 307(d)(7)(B) states that "[o]nly an objection to a rule or procedure which was
raised with reasonable specificity during the period for public comment (including any public
hearing) may be raised during judicial review. If the person raising an objection can demonstrate
to the Administrator that it was impracticable to raise such objection within such time or if the
grounds for such objection arose after the period for public comment (but within the time
specified for judicial review) and if such objection is of central relevance to the outcome of the
rule, the Administrator shall convene a proceeding for reconsideration of the rule and provide the
same procedural rights as would have been afforded had the information been available at the
time the rule was proposed. If the Administrator refuses to convene such a proceeding, such
person may seek review of such refusal in the United States court of appeals for the appropriate
circuit (as provided in subsection (b))." {emphasis added). Thus, the EPA is only required to
grant a CAA section 307(d)(7)(B) petition for reconsideration if the petitioner demonstrates both
(1) that it was impractical to raise the objection during the public comment period, or that the
grounds for such objection arose after the public comment period but within the time specified
for judicial review (i.e., within 60 days after publication of the final rulemaking in the Federal
Register); and (2) that the objection is of central relevance to the outcome of the rule.
On November 30, 2012, the EPA issued a proposed rule reconsidering certain new source
standards, the requirements applicable during periods of startup and shutdown for MATS and the
Utility NSPS (for the PM standard only), certain definitional and monitoring issues in the Utility
NSPS, and additional technical corrections to both MATS and the Utility NSPS. 77 FR 71323.
On April 24, 2013, the EPA issued the final action on reconsideration of the new source MATS
1 The EPA received 20 petitions for reconsideration of the MATS rule and 3 petitions for
reconsideration of the Utility NSPS (see Appendix A. List of NESHAP Petitioners and Appendix
B. List of NSPS Petitioners). In the appendices, the EPA lists the parties that filed petitions for
reconsideration and assigns a unique number to each Petitioner. The EPA uses these numbers to
identify the Petitioners that raised the various reconsideration issues addressed in this document.
Many issues were raised by numerous Petitioners. The EPA has attempted to identify all the
Petitioners that raise each issue; however, given the large number and complexity of the
petitions, it is possible that commenters will be inadvertently misidentified or unidentified for an
issue. Any such inadvertent errors does not affect the substance of the EPA's response to the
issues raised in the petitions and, as such, would be of no consequence.
1
-------
standards, the definitional and monitoring provisions in the Utility NSPS, and the technical
correction in both rules. 78 FR 24073. The EPA issued the final action on reconsideration of the
startup and shutdown provisions in the MATS and Utility NSPS on November 19, 2014. 79 FR
68777. In addition, on February 17, 2015, EPA proposed additional technical corrections to the
final MATS rule and the Utility NSPS. 80 FR 8442.
This document presents the EPA's response to the remaining issues in the petitions for
reconsideration received on the final MATS rule and the Utility NSPS. The EPA carefully
reviewed the petitions and evaluated each issue raised in the petitions for reconsideration to
determine if they meet the CAA section 307(d)(7)(B) criteria for reconsideration. In this action,
the EPA is denying the remaining issues in the petitions for reconsideration because they do not
meet the criteria for reconsideration and/or are moot.
Many of the parties that filed petitions for reconsideration of the final MATS and Utility NSPS
also filed petitions for review of the final rule in the United States Court of Appeals for the
District of Columbia Circuit (Court or D.C. Cir.). Many of the issues raised in the petitions for
reconsideration were also raised in the D.C. Cir. litigation, and other reconsideration issues could
have been raised in that litigation. On April 15, 2014, the Court rejected all petitions for review
of MATS and the Utility NSPS. White Stallion Energy Center v. EPA, 784 F.3d 1222 (D.C. Cir.
2014); cert, granted, State of Michigan v. EPA, No. 14-46 (and consolidated cases).2 As the
Court may only consider issues raised during the period for public comment, issues raised in the
litigation and addressed by the Court clearly do not meet the criteria for reconsideration in CAA
section 307(d)(7)(B). Moreover, parties may not use this final action denying reconsideration as
a basis to litigate issues that could have been raised in the initial litigation.
2 On November 25, 2014, the U.S. Supreme Court granted petitions to hear state and industry
challenges against the Environmental Protection Agency's (EPA) Mercury and Air Toxics
Standards (MATS) on the following question: "Whether the Environmental Protection Agency
unreasonably refused to consider costs in determining whether it is appropriate to regulate
hazardous air pollutants emitted by electric utilities." Oral argument was held on March 25,
2015.
2
-------
MATS
1.0 Appropriate and Necessary Finding
1.1 Ability to comment on Hg Risk Technical Support Documents (TSDs)
Issue 1: Petitioner 20180 asserts that the public was unable to offer meaningful comments on the
National-Scale Mercury (Hg) Risk Assessment accompanying the proposed MATS rule ("draft
Hg Risk TSD")3 due to the allegedly opaque nature of the document.
Response to Issue 1: Petitioner made this same allegation during the public comment period,
and the EPA responded to that in Section IF (pp. 172 - 173) of the December 2011 Response to
Comments document (RTC).4 The EPA is denying the Petition for Reconsideration on this issue
because Petitioner previously submitted comments on this issue, and the EPA responded to those
comments.
Further, the EPA disagrees that the public could not offer meaningful comments on the draft Hg
Risk TSD. The volume of comments and level of technical detail of those comments on the draft
Hg Risk TSD received from Petitioners (EPA-HQ-OAR-2009- 0234-17775, pp. 6 and 59 - 70;
EPA-HQ-OAR-2009-0234-18023, pp. 72 - 99, Attachments B and C) and other commenters
clearly demonstrate that the public had ample opportunity to offer meaningful comments. In
addition, Petitioners did not identify any specific issues on which they claim they were unable to
provide comments.
As noted in Science Advisory Board's (SAB's) peer review letter on the Hg Risk TSD (U.S.
EPA-SAB, 201 la; p. I),5 the EPA provided additional information during the peer review
meeting, and the SAB was able to review the risk assessment and provide their
recommendations. Any commenters attending the public meeting, which included at least one
representative from one of the Petitioners (p. 4),6 had the opportunity to gain the same
3 U.S. Environmental Protection Agency (U.S. EPA). 2011a. Technical Support Document
(TSD): National-scale Mercury Risk Assessment Supporting the Appropriate and Necessary
Finding for Coal- and Oil-fired Electric Generating Units. Office of Air Quality Planning and
Standards. March 2011. EPA-452/D-11-002. EPA-HQ-OAR-2009-0234-3057.
4 U.S. EPA. 201 lb. EPA's Responses to Public Comments on EPA's National Emission
Standards for Hazardous Air Pollutants from Coal- and Oil-Fired Electric Utility Steam
Generating Units. December 2011. Volume 1 of 2. EPA-HQ-OAR-2009-0234-20126.
Available on the Internet at http://www.epa.gov/ttn/atw/utility/mats_rtc_chapters_foreword-l-2-
3-4_121611.pdf.
5 U.S. Environmental Protection Agency-Science Advisory Board (U.S. EPA-SAB). 201 la. Peer
Review of EPA's Draft National-Scale Mercury Risk Assessment. EPA-SAB-11-017. September
2011. EPA-HQ-OAR-2009-0234-19689. Available on the Internet at
http://yosemite.epa.gov/sab/sabproduct.nsf/BCA23C5B7917F5BF852579 lA0072CCAl/$File/E
PA- SAB -11-017-unsigned.pdf.
6 U.S. EPA-SAB. 201 lc. Summary Minutes of the United States Environmental Protection
Agency (U. S. EPA) Science Advisory Board (SAB) Mercury Review Panel, June 15-17, 2011.
3
-------
information and understanding. The EPA provided copies of the additional materials to attendees
and the SAB also posted these materials on their website during the public comment period. In
addition, the EPA indicated in the preamble to the proposal that the Hg Risk TSD would be peer
reviewed (76 FR 25012), and the SAB review process and meetings were announced in the
Federal Register (76 FR 10896 - 10897, 76 FR 17649 - 17650, 76 FR 29746 - 29747, 76 FR
39102 - 39103, 76 FR 50729 - 50730).
The comments that the EPA received on the Hg Risk TSD, including those from Petitioners,
demonstrate that the public had ample opportunity and understanding of the TSD to comment,
and the EPA responded to those comments in Section IF (pp. 123 - 183) of the RTC. Those
comments included comments on the draft SAB report that was made available to the public
before the close of the comment period, and the draft SAB report contained all the substantive
comments included in the final report on September 29, 2011. Because Petitioners have not
demonstrated that it was impracticable to comment on this issue or on the draft Hg Risk TSD
during the public comment period on the proposed MATS rule, the EPA is denying the petition
for reconsideration of this issue.
Furthermore, in the While Stallion decision, the Court found that "EPA's 'appropriate and
necessary' determination in 2000, and its reaffirmation of that determination in 2012, are amply
supported by EPA's findings regarding the health effects of mercury exposure." White Stallion,
748 F.3d at 1245-46. The Court cited to the SAB decision in affirming the finding and noted the
Petitioners' claims that there were procedural irregularities with the Hg peer review process,
which issues the Court did not reach. Id. The Court's decision to affirm without addressing the
alleged procedural irregularities further demonstrates that this reconsideration issue is not of
central relevance.
Issue 2: Petitioner 20180 asserts that they were unable to comment on the national-scale Hg risk
assessment accompanying the final MATS rule 7 ("revised Hg Risk TSD") during the public
comment period because SAB completed their peer review after the public comment period
closed; thus, the EPA was unable to publish a response to the SAB prior to final rulemaking, as
promised in the proposed MATS rulemaking (76 FR 25012). Petitioner further claims that EPA
substantially revised its Hg Risk TSD with new factual information, including extensive new fish
tissue data. Petitioners also state that no independent peer review was conducted on the revised
Hg Risk TSD.
Response to Issue 2: In the preamble to the proposed MATS rule (76 FR 25012), the EPA
indicated that the draft Hg Risk TSD would be peer-reviewed prior to the final MATS rule and
EPA-HQ-OAR-2009-0234-19689. Available on the Internet at
http://yosemite.epa.gov/sab/sabproduct.nsf/MeetingCal/4A60092A413F56608525783F0050F14
8/$File/Minutes.06.015-17.11-final.pdf.
7 U.S. EPA. 201 Id. Revised Technical Support Document: National-Scale Assessment of
Mercury Risk to Populations with High Consumption of Self-caught Freshwater Fish In Support
of the Appropriate and Necessary Finding for Coal- and Oil-Fired Electric Generating Units.
Office of Air Quality Planning and Standards. November 2011. EPA-452/R-11-009. EPA-HQ-
OAR-2009-0234-19913.
4
-------
that the EPA would respond to any recommendations from that peer review. Because the
revisions to the Hg Risk TSD, including the additional fish tissue data, were in response to
comments that the EPA received from the SAB on the draft Hg Risk TSD and did not change the
manner in which the risk assessment was conducted or the conclusions of the risk assessment,
the EPA was not required to provide additional opportunity to comment on the risk assessment.
First, it is important to distinguish the peer review itself from the final documentation of that
review. Although the EPA agrees that the SAB did not sign their peer review letter until
September 29, 2011, the peer review itself was conducted in a public meeting during the public
comment period on June 15 - 16, 2011 (EPA-HQ-OAR-2009-0234-19689). On June 17, 2011,
the Mercury Review Panel began drafting their comments, and they discussed the final wording
of that letter during two subsequent public teleconferences held on July 20, 2011, and September
7, 2011. The SAB sent their final letter to EPA on September 29, 2011. The content and the
recommendations in the final letter closely matched the June and July drafts of the SAB letter.
The public had ample opportunity to participate in the peer review process (EPA-HQ-OAR-
2009-0234-19689), including the nomination of panel members, comment opportunities during
the public meeting and teleconferences, and access to the SAB's draft comments prior to the
issuance of the final letter and close of the MATS comment period. For example, as noted in the
RTC (pp. 172 - 173), Petitioner 20180 cited the SAB's draft letter in their comments on the
proposed rule, which provides evidence that the public was able to access and review the SAB's
comments. These comments also demonstrate that the public had ample opportunity to comment
on any issues raised in the peer review, and the EPA responded to those comments. Because
Petitioner has not demonstrated that it was impracticable to comment on issues raised in the peer
review during the public comment period on the proposed MATS rule, the EPA is denying the
petition for reconsideration of this issue.
Second, as the SAB stated in its letter, "the SAB supports the overall design of and approach to
the risk assessment" (U.S. EPA-SAB, 2011a; EPA-HQ-OAR-2009-0234-19689, p. 2). The SAB
specifically encouraged the EPA to contact states with high-deposition watersheds to determine
if additional fish tissue methylmercury (MeHg) data were available. The EPA did so,
incorporating the additional fish tissue data in the same manner as the original data, and,
therefore, the expansion of the fish tissue data was a logical outgrowth of the risk assessment
based on SAB's comments. Other revisions to the Hg Risk TSD in response to the SAB's
comments were clarifications of the methods used in the Hg Risk TSD, and the EPA did not
revise the overall analytical approach in the revised Hg Risk TSD. See NRDC v. Thomas, 838
F.2d 1224, 1242 (D.C. Cir. 1988) and Small Refiner Lead Phase-Down Task Force v. EPA, 705
F.2d at 547 (agency may make changes to proposed rule without triggering new round of
comments, where changes are logical outgrowth of proposal and comments). Further, because
Petitioner cited the SAB's draft letter in their comments on the proposed MATS rule (EPA-HQ-
OAR-2009- 0234-17775, p. 59), Petitioner does not demonstrate that it was impracticable to
comment on the suggested revisions or the SAB's recommendations, as evidenced by the
comments received on these recommendations.
5
-------
Third, the EPA did not find a second round of peer review to be appropriate for the Hg Risk
TSD. The EPA followed the requirements in EPA's Peer Review Handbook8 for highly
influential scientific assessments (76 FR 9312), which does not require multiple rounds of peer
review. In addition, the revised Hg Risk TSD responded to all of the SAB's recommendations
(76 FR 9313 - 9316). Further, the revisions, which did not change the overall analytical
approach, only strengthened the conclusions of the Hg Risk TSD. Therefore, Petitioner's
arguments that an additional round of peer review and public comment period should have been
conducted is not of central relevance to the "Appropriate and Necessary" finding.
Furthermore, in the While Stallion decision, the Court found that "EPA's 'appropriate and
necessary' determination in 2000, and its reaffirmation of that determination in 2012, are amply
supported by EPA's findings regarding the health effects of mercury exposure." White Stallion,
748 F.3d at 1245-46. The Court cited to the SAB decision in affirming the finding and noted the
Petitioners' claims that there were procedural irregularities with the Hg peer review process,
which issues the Court did not reach. Id. The Court's decision to affirm without addressing the
alleged procedural irregularities further demonstrates that this reconsideration issue is not of
central relevance.
For all these reasons, the EPA is denying the petition for reconsideration of this issue.
1.2 Reliability of Hg Risk TSD as a basis for "Appropriate and Necessary" finding
Issue 3: Petitioner 20180 alleges that the Hg Risk TSD does not provide a reliable scientific
basis for the "Appropriate and Necessary" finding because (1) the Hg Risk TSD contains
questionable conservative assumptions that result in unrealistic risk estimates, and (2) the EPA
did not respond to the peer review comments from the SAB.
Response to Issue 3: Issues regarding the reliability of the Hg Risk TSD for providing a basis
for the "Appropriate and Necessary" finding were raised in public comments submitted in
response to the proposed MATS rule. The EPA's responses to these comments on the draft Hg
Risk TSD are in Section IF (pp. 123 - 183) of the RTC. Responses to Petitioners' specific
"conservative assumptions" allegations regarding the Hg Risk TSD are provided in the responses
to Issues 4 to 44 below.
Furthermore, in the White Stallion decision, the Court found that "EPA's 'appropriate and
necessary' determination in 2000, and its reaffirmation of that determination in 2012, are amply
supported by EPA's findings regarding the health effects of mercury exposure." White Stallion,
748 F.3d at 1245-46. The Court's decision directly rebuts the first issue raised in this petition.
Concerning the second issue, the Court cited to the SAB decision in affirming the finding and
noted the Petitioners' claims that there were procedural irregularities with the Hg peer review
process, which issues the Court did not reach. Id. As the Court may only consider issues raised
during the period for public comment, issues raised in the litigation and addressed by the Court
8 U.S. EPA. 2012. Peer Review Handbook, 3rd edition. EPA/100/B-06/002, Science Policy
Council, Washington, DC. Available on the Internet at
http://www.epa.gov/peerreview/pdfs/peer_review_handbook_2012.pdf.
6
-------
clearly do not meet the criteria for reconsideration in CAA section 307(d)(7)(B). Because
Petitioner has not demonstrated that it was impracticable to comment on this issue during the
public comment period on the proposed MATS rule, we are denying reconsideration of this
issue.
As stated in the RTC (p. 171), the SAB "supports the overall design of and approach to the risk
assessment and finds that it should provide an objective, reasonable, and credible determination
of the potential for a public health hazard from mercury emitted from U.S. EGUs." (U.S. EPA-
SAB, 2011a; EPA-HQ-OAR-2009-0234-19689, p. 2). The SAB made this conclusion based on
the draft Hg Risk TSD (EPA-HQ-OAR-2009-0234-3057), which indicates that the SAB did not
consider the Hg Risk TSD to contain questionable conservative assumptions as Petitioner
alleges. Although the SAB recommended that the EPA add descriptions of the key analytical
methods and findings to improve the clarity of the TSD (p. 2), the SAB's comments did not
require changes to the overall analytical approach of the Hg Risk TSD. Contrary to Petitioner's
assertion, the EPA fully responded to all of the SAB's recommendations (76 FR 9313 - 9316).
Further, Petitioner does not raise any methodological concerns regarding the design of the risk
assessment or interpretation of conclusions that could not have been raised to the SAB or to the
EPA during the public comment period.
The EPA relies on the consensus conclusions from the Mercury Panel of the SAB as the body of
scientific experts nominated to review the Hg Risk TSD, regarding the reliability of the
assessment for its intended purpose. For this reason, the issue is not of central relevance.
For all of these reasons, we are denying the petition for reconsideration of this issue.
Issue 4: Petitioners 20183 and 20180 claim that the revised Hg Risk TSD continues to combine
multiple highly conservative assumptions that lead to unrealistic exposure estimates that
overestimate risks. Petitioner 20183 (att. 2) and Petitioner 20180 (att. 3) also assert that the
EPA's assumptions are more conservative than the EPA's risk guidance (U.S. EPA, 1989)9 and
that use of 99th percentile exposure assumptions, including the fish consumption rate, results in
exposure estimates beyond the range of possible exposures. Further, Petitioner 20183 (att. 2) and
Petitioner 20180 (att. 3) allege that the SAB may not have fully understood the combination of
conservative assumptions, including the combination of the 99th percentile fish consumption rate
with the 75th percentile fish size after removing fish less than seven inches.
Response to Issue 4: Issues regarding the potential conservatism of the exposure estimates were
raised in public comments submitted in response to the proposed MATS rule. The EPA
responded to these comments in Section IF (pp. 142, 171, 177, and 180) of the RTC.
As stated in the RTC (p. 135), "the design of the risk assessment is particular to this statutory
context." Further, the draft Hg Risk TSD reviewed by the SAB clearly noted that the EPA
9 U.S. EPA. 1989. Risk Assessment Guidance for Superfund. Vol. 1, Human Health Evaluation
Manual (Part A). EPA/540/1-89/002. U.S. Environmental Protection Agency, Office of
Emergency and Remedial Response. Washington, D.C. December.
7
-------
excluded fish less than 7 inches in length from the fish tissue dataset (p.71), applied the 75th
percentile fish size to that dataset (p. 72), and applied the 99th percentile consumption rates (pp.
73). Thus, Petitioners' assertions that the SAB did not comprehend the Hg Risk TSD are without
foundation. Furthermore, in the White Stallion decision, the Court found that "EPA's
'appropriate and necessary' determination in 2000, and its reaffirmation of that determination in
2012, are amply supported by EPA's findings regarding the health effects of mercury exposure."
White Stallion, 748 F.3d at 1245-46. As the Court may only consider issues raised during the
period for public comment, issues raised in the litigation and addressed by the Court clearly do
not meet the criteria for reconsideration in CAA section 307(d)(7)(B). Because Petitioners have
not demonstrated that it was impracticable to comment on these issues during the public
comment period on the proposed MATS rule, the EPA is denying the petition for reconsideration
of this issue.
The SAB supported the overall design and approach of the Hg Risk TSD, including the fish
consumption rates used (U.S. EPA-SAB, 2011a; EPA-HQ-OAR-2009-0234-19689, pp. 2 and 15
- 16). The SAB panel inquired about the fish size during the public SAB peer review meeting,
and in response during the peer review process the EPA presented a detailed example of an
exposure calculation that specifically highlighted the use of both the high-end consumption rate
and the 75th percentile fish tissue Hg value (EPA-HQ-OAR-2009-0234-19689). Because the
SAB ultimately expressed support for the overall design of the analysis in their comments
submitted after this presentation, which clearly indicated the combination of assumptions within
the calculation, Petitioners have not demonstrated that this issue is of central relevance or that the
SAB was unaware of the Agency's approach. For these reasons also, the EPA is denying the
petition for reconsideration on this issue.
Responses to Petitioners' specific issues regarding fish consumption rates and fish tissue samples
in the Hg Risk TSD are provided in the responses to Issues 10 to 17 below.
Issue 5: Petitioner 20183 (att. 2) and Petitioner 20180 (att. 3) claim the EPA did not follow its
own policies regarding "weight of the evidence", selected only reports that support the EPA's
preset outcome that Hg exposures due to EGU emissions are a big problem and ignored all
contradictory evidence.
Response to Issue 5: The EPA is denying reconsideration on the issue because this issue is
substantively similar to an issue raised in public comments, and the EPA responded to those
comments. In Section IF (p. 168) of the RTC, commenters suggested that the EPA did not use
the latest data in the Hg Risk TSD (specifically toxicological data) and that the EPA should
consider the latest information. In response to that comment, the EPA stated, "It is the policy of
the EPA to use the most current peer reviewed, publicly available data and methodologies in its
risk assessments." The EPA did so here. Specifically, the EPA considered the substantive peer-
reviewed studies submitted during the public comment period that were relevant to the design
and application of the Hg Risk TSD and did not rely upon studies that were not peer-reviewed.
Further, in the peer review of the draft Hg Risk TSD, the SAB panel noted, "the SAB supports
the overall design of and approach to the risk assessment and finds that it should provide an
objective, reasonable, and credible determination of the potential for a public health hazard from
8
-------
Hg emitted from U.S. EGUs" (U.S. EPA-SAB, 201 la, p. 2) In this statement of support for the
overall design of the Hg Risk TSD, the SAB agreed that the EPA included objective evidence.
The comments that the EPA received on these issues demonstrate that the public had ample
opportunity to comment on this issue, and the EPA responded to those comments. Because
Petitioners have not demonstrated that it was impracticable to comment on this issue during the
public comment period on the proposed MATS rule, the EPA is denying reconsideration of this
issue. The EPA is also denying this issue because it is not of central relevance because Petitioner
provides no information that would change the Hg Risk TSD in light of the SAB support for the
Agency's approach and the fact that the EPA complied with its general approach and explained
its reasoning regarding the information used in the Hg Risk TSD.
Furthermore, in the While Stallion decision, the Court found that "EPA's 'appropriate and
necessary' determination in 2000, and its reaffirmation of that determination in 2012, are amply
supported by EPA's findings regarding the health effects of mercury exposure." White Stallion,
748 F.3d at 1245-46. As the Court may only consider issues raised during the period for public
comment, issues raised in the litigation and addressed by the Court clearly do not meet the
criteria for reconsideration in CAA section 307(d)(7)(B). Moreover, parties may not use this
final action denying reconsideration as a basis to litigate issues that could have been raised in the
initial litigation.
1.3 Hg biomarker data
Issue 6: Petitioners 20183 and 20180 allege that the EPA did not respond to or acknowledge
public comments from Tetra Tech on behalf of Petitioner 20183 regarding Dellinger (2004),10
specifically concerning the discrepancy in that study between Hg biomarker data, such as blood
Hg levels, and Hg exposure based on surveys of fish consumption.
Response to Issue 6: Petitioners raise a procedural challenge to the final rule by alleging that the
EPA failed to respond to an allegedly significant comment. As a preliminary matter, the EPA
responded to comments concerning the Dellinger study in Section IF (Vol. 1, pp. 134 - 136) of
the RTC, and those responses and responses to other comments concerning studies of Hg blood
levels and fish consumption surveys in Section IF (Vol. 1, pp. 149 - 160) of the RTC address the
substance of the Tetra Tech comments even if the EPA did not identify those comments in the
relevant responses.
In any case, the EPA maintains that the comments regarding the discrepancy between biomarker
data and exposure estimates were not raised with "reasonable specificity" and petitioner has not
demonstrated that it was impracticable to raise those issues with reasonable specificity during the
public comment period. For these reasons, EPA is denying the petition for reconsideration of
this issue.
10 Dellinger, J. A. 2004. "Exposure assessment and initial intervention regarding fish
consumption of tribal members of the Upper Great Lakes Region in the United States,"
Environmental Research 95, 325-340.
9
-------
The EPA does not find Tetra Tech's comments regarding the discrepancy between
biomarker data and exposure estimates in the Dellinger study to be relevant to the Hg
Risk TSD. Specifically, at the time of the final MATS rule, the EPA did not consider a
recommendation to collect Hg blood data from the subsistence fisher populations
modeled in the Hg Risk TSD to be a significant comment because it was well outside the
scope of the risk assessment. The EPA did respond to comments regarding the blood Hg
levels in the National Health and Nutrition Examination Survey (NHANES) (see Section
IF, pp. 149 - 150, of the RTC, Vol. 1).
Upon further review and with the aid of the more specific information provided in the
reconsideration petition regarding this issue, the EPA has determined that a fuller response to the
discrepancy between biomarker data and exposure estimates based on fish surveys including the
Dellinger study may be appropriate. The EPA provides responses to these specific issues in the
responses to Issues 7 to 9 below.
As fully explained in the responses to Issues 7 to 9 below, issues regarding biomarker data are
not of central relevance to the "Appropriate and Necessary" finding because the core risk
estimates in the Hg Risk TSD are based on comparisons of modeled Hg intake from fish
consumption to the reference dose for methylmercury (MeHg RfD) in order to generate a hazard
quotient (HQ), which does not involve modeling biomarker levels (such as blood Hg levels).11
Specifically, the exposure estimates used in this comparison represent daily-average intake rates
for MeHg from the consumption of fish, which is consistent with the units for the MeHg RfD.
Because the RfD is not stated in terms of an equivalent biomarker level (e.g., hair Hg, or blood
Hg), there is no need to model biomarker levels in order to calculate the HQ. In addition, the
available biomarker studies were not designed to capture the high-end subsistence fishers
consuming fish with higher levels of MeHg, as modeled in the Hg Risk TSD.
Because the EPA responded to issues regarding Hg blood levels and consumptions surveys, and
because Petitioners have not established that it was impracticable to raise the issue with
reasonable specificity during the public comment period, we are denying reconsideration on this
issue. We are also denying reconsideration because the issue is not of central relevance to the
"Appropriate and Necessary" finding as explained above and in response to other comments on
Hg blood levels submitted during the comment period on the MATS proposal.
Furthermore, in the White Stallion decision, the Court found that "EPA's 'appropriate and
necessary' determination in 2000, and its reaffirmation of that determination in 2012, are amply
supported by EPA's findings regarding the health effects of mercury exposure." White Stallion,
748 F.3d at 1245-46. As the Court may only consider issues raised during the period for public
comment, issues raised in the litigation and addressed by the Court clearly do not meet the
criteria for reconsideration in CAA section 307(d)(7)(B). Moreover, parties may not use this
11 In Appendix B of the revised Hg Risk TSD, the EPA simulated biomarker levels in order to
model IQ loss. In response to SAB comments, the EPA moved this risk metric to an appendix
because of concerns that IQ reduction did not capture all of the neurodevelopmental endpoints
potentially associated with Hg exposure (U.S. EPA-SAB, 201 la, p. 2).
10
-------
final action denying reconsideration as a basis to litigate issues that could have been raised in the
initial litigation.
Issue 7: Petitioners 20183 and 20180 allege that the EPA overestimated Hg exposure and
ignored evidence that suggests one or more of the EPA's exposure assumptions yield
unrealistically high risk estimates compared to biomarker data from four studies (Dellinger,
2004; Hibbeln, et al., 2007;12 Daniels, et al., 2004;13 Xue, et al., 201214). Specifically, Petitioners
claim that biomarker data provide strong evidence that Hg exposures are not as high as the EPA
predicted in the Hg Risk TSD. For example, Petitioners assert that the EPA's highest exposure
scenario for a female subsistence fisher would correspond to an Hg blood concentration of
almost 150 parts per billion (ppb), whereas the data from the NHANES report the highest blood
concentration to be less than 40 ppb. Petitioner speculates that this disparity results from the EPA
overestimating fish consumption, fish Hg concentrations, bioavailability of fish MeHg, or a
combination of all three assumptions.
Response to Issue 7: The EPA is denying reconsideration of this issue because (1) blood Hg
levels from the NHANES report were raised in public comments and the EPA responded to those
comments, (2) none of these issues are of central relevance to the "Appropriate and Necessary"
finding, and (3) specific allegations that biomarker data from several studies undermine the
EPA's Hg risk estimates are new comments that Petitioners could have previously raised in
public comment.
First, issues regarding the blood Hg levels in the NHANES data were raised in public comments
submitted in response to the proposed MATS rule. The EPA responded to these comments in
Section IF (pp. 149 - 155) of the RTC.
As stated in the RTC (p. 149), because the NHANES survey is designed to reflect the general
U.S. population, it is unlikely to include the type of subsistence fishers modeled in the Hg Risk
TSD; studies focusing on high fish consuming individuals have found blood Hg levels that are
substantially higher than the range reported in the NHANES survey. For example, as cited in the
RTC, the Hightower and Moore (2003) study, which evaluated 89 adults identified as likely
having elevated blood Hg levels from consuming fish with high Hg levels, found mean blood Hg
levels for women that were ten times higher than an earlier general population survey (CDC,
2001).15
12 Hibbeln, JR, Davis, JM, Steer, C, Emmett, C, Rogers, I, Williams, C, Golding, J. 2007.
"Maternal seafood consumption in pregnancy and neurodevelopmental outcomes in childhood
(ALSPAC study): an observational cohort study," Lancet 369:578-85.
13 Daniels, JL, Longnecker, MP, Rowland, AS, Golding, J, and The ALSPAC Study Team-
University of Bristol Institute of Child Health. 2004. "Fish Intake During Pregnancy and Early
Cognitive Development of Offspring." Epidemiology 15:394-402.
14 Xue, J, Zartarian, VG, Liu, SV, and Geller, AM. 2012. "Methylmercury exposure from fish
consumption in vulnerable racial/ethnic populations: Probabilistic SHEDS-Dietary model
analyses using 1999-2006 NHANES and 1990-2002 TDS data." Sci Tot Environ 414:373-379.
15 CDC. 2001. First National Report on Human Exposure to Environmental Chemicals Results:
Mercury. U.S. Centers for Disease Control and Prevention.
11
-------
The EPA disagrees that the NHANES dataset provides evidence that the EPA's exposure
estimates are too high. The NHANES survey is designed to be representative of the general U.S.
population and samples approximately 4,200 females for blood Hg levels per year (CDC,
2012).16 Although this sampling approach provides reliable estimates for the general population,
it is unlikely to capture small subgroups such as high-end subsistence fish consumers that are the
focus of the Hg Risk TSD. Because many high-end fish consumers have low socio-economic
status, it may be challenging to reach them through typical survey methods (Weiss and Bailar,
2002).17 As discussed in the MATS proposal (76 FR 25020) and in the revised Hg Risk TSD (pp.
36 - 40), even though it can be difficult to capture this high-consuming subpopulation in national
surveys, multiple studies identify subsistence-level fish consumption at various locations across
the country (Burger, et al., 1999;18 Burger, et al., 2002;19 Corburn, et al., 2002;20 Dellinger,
2004; Moya, et al., 2008;21 Shilling, et al., 201022).
The comments that the EPA received on blood Hg levels from NHANES demonstrate that the
public had ample opportunity to comment on this issue, and the EPA responded to those
comments. Because Petitioners have not demonstrated that it was impracticable to comment on
this issue during the public comment period on the proposed MATS rule, the EPA is denying the
petition for reconsideration of this issue.
Second, issues regarding biomarker data are not of central relevance to the "Appropriate and
Necessary" finding because the core risk estimates in the Hg Risk TSD are based on
comparisons of modeled Hg intake from fish consumption to the MeHg RfD, which does not
involve modeling biomarker levels (such as blood Hg levels). In addition, the EPA disagrees that
16 CDC. 2012. Fourth National Report on Human Exposure to Environmental Chemicals
Updated Tables, February 2012: Total Blood Mercury. U.S. Centers for Disease Control and
Prevention. Available on the internet at
http://www.cdc.gov/exposurereport/pdf/FourthReport_UpdatedTables_Feb2012.pdf.
17 Weiss, C., and Bailar, B. A. 2002. "High response rates for low-income population in-person
surveys." In M. Ver Ploeg, R. A. Moffitt, & C. F. Citro (Eds.), Studies of welfare populations:
Data collection and research issues (pp. 86-104). Washington, DC: The National Academies
Press.
18 Burger, J., Stephens, W. L., Boring, C. S., Kuklinski, M., Gibbons, J. W., Gochfeld M. 1999.
"Factors in Exposure Assessment: Ethnic and Socioeconomic Differences in Fishing and
Consumption of Fish Caught along the Savannah River." Risk Analysis 19, No. 3, p. 427.
19 Burger, J. 2002. "Daily consumption of wild fish and game: Exposures of high end
recreationalists " International Journal of Environmental Health Research 12:4, p. 343-354.
20 Corburn, J. 2002. "Combining community-based research and local knowledge to confront
asthma and subsistence-fishing hazards in Greenpoint/Williamsburg, Brooklyn, New York."
Environmental Health Perspectives, 110(2).
21 Moya. J., Itkiin, C., Selevan, S.G., Rogers, J.W., Clickner, R. P. 2008. "Estimates of fish
consumption rates for consumers of bought and self-caught fish in Connecticut, Florida,
Minnesota, and North Dakota." Science of the Total Environment 403, issue 1-3, p. 89-98.
22 Shilling, Fraser, Aubrey White, Lucas Lippert, MarkLubell. 2010. "Contaminated fish
consumption in California's Central Valley Delta." Environmental Research 110, p. 334-344.
12
-------
the biomarker studies cited by Petitioner (Dellinger, 2004; Hibbeln, et al., 2007; Daniels, et al.,
2004; Xue, et al., 2012) are directly relevant to the Hg Risk TSD because they do not represent
the high-end subsistence fish consumption scenarios that the EPA modeled. Specifically, the
individuals in those studies do not necessarily consume subsistence levels of fish at watersheds
with relatively high levels of fish tissue Hg. High consumption rates do not necessarily translate
into high blood Hg levels unless the fish consumed at high rates have elevated Hg
concentrations. Both of these factors must occur in tandem for an individual to have elevated
blood Hg levels, and either factor alone is not sufficient. Because the biomarker studies cited by
Petitioner do not reflect the scenarios that were modeled in the Hg Risk TSD, the studies
themselves are not relevant and the measured blood Hg levels in those studies, which are lower
than those associated with the scenarios modeled in the Hg Risk TSD, are also not relevant.
Third, Petitioners have raised new issues regarding biomarker data, which Petitioners allege
provides evidence that biomarker data undermine the exposure estimates in the Hg Risk TSD,
they cite several studies previously cited by Petitioners during the public comment period
(Dellinger, 2004; Hibbeln, et al., 2007; Daniels, et al., 2004) as well as a new study by Xue, et
al., (2012). The Xue, et al., study was published after the close of the public comment period;
however, the study does not provide information that is substantively different from the
information in the earlier biomarker studies. Specifically, Petitioners claim that the Xue, et al.,
study shows substantially lower blood Hg levels than associated with the scenarios in the Hg
Risk TSD, and Petitioners allege that this discrepancy is likely due to an error in the EPA's
exposure factors (e.g., Hg bioavailability). The EPA disagrees that the Xue, et al., study is
relevant because the exposure estimates in the study do not reflect a combination of high-end
fish consumption rates and high-end fish tissue Hg levels as evaluated in the Hg Risk TSD.
In particular, Petitioners contrast the 99th percentile blood Hg level from the Xue, et al., study
(6.63 ug/L or 6.63 ppb) for adult females with the highest blood Hg levels associated with the
scenarios modeled in the Hg Risk TSD (-150 ug/L or 150 ppb). The level in the Xue, et al.,
study is the 99th percentile blood Hg concentration for the group of individuals included in the
study, which likely includes individuals who obtain fish from a combination of self-caught and
commercial sources. However, this study was not designed to capture the subsistence fishers
consuming fish with higher levels of MeHg that were assessed in the Hg Risk TSD.
Because issues regarding the blood Hg levels were raised in public comments and the EPA
responded to these comments, Petitioners have not demonstrated that it was impracticable to
comment on this issue during the public comment period on the proposed MATS rule. In
addition, because the Xue, et al., study does not provide estimates relevant to the populations
modeled in the Hg Risk TSD, the EPA would not have been able to use the results of this study
even if it were available during the public comment period. For this reason, the Xue, et al., study
is not of central relevance to the "Appropriate and Necessary" finding. In addition, issues
regarding biomarker data are not of central relevance to the "Appropriate and Necessary" finding
because the core risk estimates in the Hg Risk TSD do not involve modeling biomarker levels
and because the available biomarker studies are not likely to capture the high-end subsistence
fishers consuming fish with higher levels of Hg, as modeled in the Hg Risk TSD. For all of these
reasons and the reasons cited in responses to Issues 6, 8 and 9, the EPA is denying
reconsideration of this issue.
13
-------
Furthermore, in the White Stallion decision, the Court found that "EPA's 'appropriate and
necessary' determination in 2000, and its reaffirmation of that determination in 2012, are amply
supported by EPA's findings regarding the health effects of mercury exposure." White Stallion,
748 F.3d at 1245-46. As the Court may only consider issues raised during the period for public
comment, issues raised in the litigation and addressed by the Court clearly do not meet the
criteria for reconsideration in CAA section 307(d)(7)(B). Moreover, parties may not use this
final action denying reconsideration as a basis to litigate issues that could have been raised in the
initial litigation.
Issue 8: Petitioners 20183 and 20180 assert that using biomarker data would remove some of the
uncertainties associated with using estimates from "dietary recall surveys" of fish consumption,
in which participants estimate their fish consumption based on their memory of past
consumption. Petitioners assert that the Dellinger (2004) study shows that recall surveys tend to
overestimate consumption compared to "actual ingestion rates" based on participants weighing
the fish they consumed during meals. Petitioners further assert that the blood Hg levels in
Dellinger study were about one-tenth of what would be expected had the survey participants
actually consumed fish at the higher rates reported in the recall surveys and that the EPA did not
discuss this discrepancy.
Response to Issue 8: Issues regarding the Dellinger study were raised in public comments
submitted in response to the proposed MATS rule. The EPA responded to these comments in
Section IF (p. 136) of the RTC.
Contrary to Petitioners' assertion, the EPA did discuss the differences between the fish
consumption rates based on recall surveys and actual ingestion rates in the revised Hg Risk TSD.
Specifically, the EPA noted (see Table 1-6, p. 40) that because of the small sample size
associated with the actual ingestion rates in the Dellinger study (i.e., 147 individuals had "actual"
rates based on weighing fish consumed compared to 822 individuals in the recall survey), it is
unlikely that the "actual" Hg intake rates in this survey would capture high-end consumers. In
addition, the Dellinger study made no direct comparisons between recall amounts and actual
amounts for the individuals whose blood measurements were also collected. Further, the families
whose "actual" fish consumption was measured are not representative of the typical study
participant because they consumed significantly fewer fish meals (12 to 34 meals per year
compared to 47 to 118). We also note the Dellinger study did not measure Hg concentrations in
the actual fish consumed by the participants, instead reporting average Hg levels in fish in
locations where the tribal participants lived. Therefore, the participants in the Dellinger study
may not have met both criteria necessary for individuals to experience elevated Hg exposure
through fish consumption: high consumption of fish with high Hg concentrations. Thus, the
Dellinger study does not provide information to determine the exact relationship between Hg
concentrations in actual fish consumed and measured blood Hg levels in study participants,
particularly those participants who would meet both criteria for elevated blood Hg from fish
consumption. Although we question the relevancy of the information on "actual" consumption
rates provided in the Dellinger study, we note that the recall-based consumption rates in the
Dellinger study are consistent with high-end subsistence rates reported in other studies (Burger,
et al., 2002; Shilling, et al., 2010) and with those used in the Hg Risk TSD. In addition, the SAB
14
-------
concluded that alternative approaches for estimating fish consumption rates for the Hg Risk TSD
were not recommended (U.S. EPA-SAB, 201 la, p. 16).
Because Petitioners have not demonstrated that it was impracticable to comment on this issue
during the public comment period on the proposed MATS rule, and for all of the reasons cited in
the response to Issue 7, the EPA is denying reconsideration of this issue.
Furthermore, in the While Stallion decision, the Court found that "EPA's 'appropriate and
necessary' determination in 2000, and its reaffirmation of that determination in 2012, are amply
supported by EPA's findings regarding the health effects of mercury exposure." White Stallion,
748 F.3d at 1245-46. As the Court may only consider issues raised during the period for public
comment, issues raised in the litigation and addressed by the Court clearly do not meet the
criteria for reconsideration in CAA section 307(d)(7)(B). Moreover, parties may not use this
final action denying reconsideration as a basis to litigate issues that could have been raised in the
initial litigation.
Issue 9: Petitioner 20183 (att. 2) and Petitioner 20180 (att. 3) cite the Hibbeln, et al., (2007) and
Daniels, et al., (2004) studies as evidence that reported consumption rates from fish consumption
surveys are unreliable and that biomarkers are more reliable measures of exposure.
Response to Issue 9: Issues regarding the Hibbeln, et al., study were raised in public comments
submitted in response to the proposed MATS rule. The EPA responded to those comments in
Section IF (pp. 146 - 174, 159 - 160, and 167) of the RTC.
Commenters did not raise issues regarding the Daniels, et al., study, which reported biomarkers
results from the same population as the Hibbeln, et al., study, but the EPA cited the Daniels, et
al., study in the same response in the RTC. Because Petitioners have not demonstrated that it was
impracticable to comment on this issue during the public comment period on the proposed
MATS rule, as evidenced by the comments received on the issue, and for all of the reasons cited
in the response to other petitions on similar issues, the EPA is denying reconsideration of this
issue.
The relationship between fish consumption and Hg biomarker levels is dependent, in part, on two
critical factors: the fish ingestion rate and the Hg concentration in the fish consumed. Although
the studies cited by Petitioner provide data on reported fish consumption rates, they do not
provide data on Hg levels in those fish. Petitioner acknowledges this data gap. Contrary to
Petitioner's assertion, because the studies cannot account for fish tissue Hg levels, it is not
possible to evaluate the degree to which these studies capture the high-end scenarios reflected in
the Hg Risk TSD. Although these studies may reflect high fish consumption rates, the EPA
cannot determine whether these rates reflect high-end fish consumption targeting fish with
substantially elevated Hg levels due to a lack of information on fish tissue Hg data in these
studies. Therefore, these studies cannot be meaningfully compared to the Hg Risk TSD.
Further, the Hg Risk TSD did not generate population-representative estimates of exposure.
Although all surveys have some inherent uncertainty regarding the representativeness of the
surveyed sample, the fish consumption surveys support the existence of high-end consumers.
15
-------
Although Petitioners argue that the EPA's 99th percentile consumption rate (i.e., 373 gram per
day (g/day)) is too high to be reliable, this consumption rate is equivalent to one 13-ounce
fishmeal a day. Although this is a high consumption rate, it is not an unreasonable rate for a
woman who consumes a relatively large amount of self-caught fish (Burger, et al., 1999; Burger,
et al., 2002; Corburn, et al., 2002; Dellinger, 2004; Moya, et al., 2008; Shilling, et al., 2010). In
fact, as cited in the Utility Study (p. 7-36, U.S. EPA, 1998), the 99th percentile fish consumption
rate for adults in the Columbia River Tribes was 389 g/day (Columbia River inter-Tribal
Commission, 1994).23 Further, the NHANES survey reports that women (age 13 to 49) consume
5.1 grams per kilogram (g/kg) per day of meat at the 99th percentile, which is equivalent to 12
ounces of meat per day for the typical 64 kg woman (U.S. EPA, 201 le).24
As discussed in the revised Hg Risk TSD (pp. 5-10) and the proposed MATS rule (76 FR
25007), the EPA's scenario-based risk assessment was designed specifically to identify
watersheds with potentially at-risk populations due to elevated Hg exposures attributable to U.S.
EGUs, and thus appropriately focused on subsistence populations at the higher percentiles of fish
consumption that may fish in watersheds impacted by U.S. EGU-attributable Hg deposition.
For all these reasons, the EPA is denying reconsideration of this issue.
Furthermore, in the While Stallion decision, the Court found that "EPA's 'appropriate and
necessary' determination in 2000, and its reaffirmation of that determination in 2012, are amply
supported by EPA's findings regarding the health effects of mercury exposure." White Stallion,
748 F.3d at 1245-46. As the Court may only consider issues raised during the period for public
comment, issues raised in the litigation and addressed by the Court clearly do not meet the
criteria for reconsideration in CAA section 307(d)(7)(B). Moreover, parties may not use this
final action denying reconsideration as a basis to litigate issues that could have been raised in the
initial litigation.
1.4 Fish consumption rates in Hg Risk TSD
Issue 10: Petitioner 20183 (att. 4) and Petitioner 20180 (att. 5) claim that the EPA should not use
the 99th percentile rate for fish consumption because it is inconsistent with consumption rates
used in the EPA's guidance and analyses. Specifically, Petitioners cite lower consumption rates
used in (a) the 1998 Utility Study (U.S. EPA, 1998),25 (b) the Pulp and Paper RTR (U.S. EPA,
23 Columbia River Inter-Tribal Fish Commission. 1994. A Fish Consumption Survey of the
Umatilla, Nez Perce, Yakima and Warm Springs Tribes of the Columbia River Basin. Technical
Report 94-3. October.
24 U.S. EPA. 201 le. Exposure Factors Handbook: 2011 Edition, EPA/600/R-09/052F.
September 2011. Table 11-3. Available on the Internet at
http: //cfpub. epa. gov/ncea/ri sk/recordi spl ay. cfm? dei d=23 6252.
25 U.S. EPA. 1998. Study of Hazardous Air Pollutant Emissions from Electric Utility Steam
Generating Units - Final Report to Congress. Volume 1. Office of Air Quality Planning and
Standards. February. EPA-HQ-OAR-2009-0234-3052.
16
-------
201 If),26 and (c) several Total Maximum Daily Load (TMDLs) assessments (U.S. EPA, 2001).27
In addition, Petitioner 20183 (att. 2) and Petitioner 20180 (att. 3) claim that the fish consumption
rates in the Hg Risk TSD are more conservative than the EPA's risk guidance (U.S. EPA, 1989)
and instead the EPA should have used the Reasonable Maximum Exposure (RME) approach,
which EPA has typically interpreted as the 90th to 99th percentile exposure (with the 95th
percentile as the typical value used).
Response to Issue 10: Issues regarding the fish consumption rates in other EPA analyses were
raised in public comments submitted in response to the proposed MATS rule. The EPA
responded to these comments in Section IF (p. 139) of the RTC. In addition, in the White
Stallion decision, the Court found that "EPA's 'appropriate and necessary' determination in
2000, and its reaffirmation of that determination in 2012, are amply supported by EPA's findings
regarding the health effects of mercury exposure." White Stallion, 748 F.3d at 1245-46. As the
Court may only consider issues raised during the period for public comment, issues raised in the
litigation and addressed by the Court clearly do not meet the criteria for reconsideration in CAA
section 307(d)(7)(B). Moreover, parties may not use this final action denying reconsideration as
a basis to litigate issues that could have been raised in the initial litigation.
As stated in the RTC, "the Hg Risk TSD (including modeling of the 99th percentile fish
consumption rate by subsistence fishers) reflects consideration for the provisions of the CAA
addressing the 'Appropriate and Necessary' finding for U.S. EGUs, and is consistent with
treatment of other HAP under CAA section 112, which focuses on maximally exposed
individuals. In that context, the design of the risk assessment is particular to this statutory
context." (U.S. EPA-SAB, 201 la). Therefore, use of the alternative consumption rates in the
1998 Utility Study, Pulp and Paper RTR, and the TMDL assessments are not required or
appropriate in this regulatory context, and the SAB supported the rates the EPA used in the Hg
Risk TSD (U.S. EPA-SAB, 201 la, pp. 3 - 4). In addition, although the Utility Study (p., 7-35,
U.S. EPA, 1998) applied the mean consumption rate for subsistence adults, the report noted that
the mean rate was below the EPA's recommended consumption rate and that the 99th percentile
fish consumption rate for adults in the Columbia River Tribes was 389 g/day (Columbia River
inter-Tribal Commission, 1994). Further, the Hg Risk TSD includes a range of high-end
percentiles for fish consumption rates, including the 90th, 95th, and 99th percentiles (see Tables 2-
7 to 2-10).
Because Petitioners have not demonstrated that it was impracticable to comment on this issue
during the public comment period on the proposed MATS rule, as evidenced by the comments
received on this issue, the EPA is denying the petition for reconsideration of this issue.
Issue 11: Petitioner 20183 (att. 2) and Petitioner 20180 (att. 3) claim that the EPA should not use
a 99th percentile fish consumption rate in the Hg Risk TSD due to alleged specific limitations in
26 U.S. EPA. 201 If. Draft Residual Risk Assessment for the Pulp & Paper Source Category.
Office of Air Planning and Standards. December 2011.
27 U. S. EPA. 2001. Total Maximum Daily Load (TMDL) Development for Total Mercury Fish
Tissue Residue in Brier Creek (Located in the Savannah River Basin). U.S. EPA Region 4.
August.
17
-------
the derivation of this rate and because such a high percentile is allegedly inherently unreliable.
Petitioners further assert that using the Burger, et al., (2002) study as the basis of the EPA's 99th
percentile rates is allegedly problematic because of that study's limited geography (i.e., South
Carolina only), failure to describe the derivation of the 99th percentile rate, absence of
information on specific fish species, absence of supporting biomarker data collection, and
inadequate sample size to generate stable 99th percentile estimates. In addition, Petitioner 20183
(att. 4) and Petitioner 20180 (att. 5) state that the small sample size in the Burger, et al., (2002)
study provides uncertain fish consumption rates at the high-end percentiles. Petitioner 20183 (att.
2) and Petitioner 20180 (att. 3) also state that it is not clear how the EPA used consumption rates
from the Dellinger (2004) and Shilling (2010) studies.
Response to Issue 11: Issues regarding using 99th percentile fish consumption rates were raised
in public comments submitted in response to the proposed MATS rule. The EPA responded to
these comments in Section IF (pp. 135 - 145) of the RTC.
Specifically, the EPA addressed the limitations in the Burger, et al., study (pp. 136 - 137).
Because Petitioners have not demonstrated that it was impracticable to comment on this issue
during the public comment period on the proposed MATS rule, and Petitioners have provided no
evidence to change the EPA's previous response, the EPA is denying reconsideration of this
issue. In addition, in the White Stallion decision, the Court found that "EPA's 'appropriate and
necessary' determination in 2000, and its reaffirmation of that determination in 2012, are amply
supported by EPA's findings regarding the health effects of mercury exposure." White Stallion,
748 F.3d at 1245-46. As the Court may only consider issues raised during the period for public
comment, issues raised in the litigation and addressed by the Court clearly do not meet the
criteria for reconsideration in CAA section 307(d)(7)(B). Moreover, parties may not use this
final action denying reconsideration as a basis to litigate issues that could have been raised in the
initial litigation.
Further, the EPA disagrees with Petitioners' assertions. As noted in the Hg Risk TSD (p. 8), the
EPA modeled risk for high-end subsistence fishers expected to experience the greatest risk from
U.S. EGU-attributable Hg. As described in section 1.4.2.1 and Table 1-6 of the Hg Risk TSD,
several studies support the conclusion that high-end fish subsistence consumption could range
between 100 and 400 g/day for high-end consumers, which is approximately an 8-ounce (oz.)
fishmeal every 2 to 3 days to a larger fishmeal (14 oz.) every day. These rates are not
unreasonable consumption rates for a high-end subsistence fisher because (a) they are supported
by multiple studies cited in the Hg Risk TSD and (b) they represent levels of fish consumption
that would contribute a substantial portion of dietary protein intake obtained through meat
consumption (U.S. EPA, 201 Id). Indeed, the SAB concluded "the consumption rates and
locations for fishing activity are supported by data" (U.S. EPA-SAB, 201 la, p. 15). The EPA
maintains this position, and Petitioners have provided no compelling evidence to establish
otherwise.
Although the EPA acknowledges potential uncertainty in food recall surveys, as well as the
specific limitations in the Burger, et al., (2002) study (e.g., small sample size), these limitations
are primarily applicable to risk assessments focused on the entire population of subsistence
fishers. However, because the Hg Risk TSD is focused on a specific high-end subsistence fisher
18
-------
scenario, rather than all subsistence fishers, these issues are not of central relevance to the
"Appropriate and Necessary" finding. Although the EPA agrees that the small sample size in the
Burger, et al., study adds uncertainty to specific high-end fish consumption rates, this study
together with the several other studies and in the Hg Risk TSD and Utility Study (U.S. EPA,
1998) provide support for these high-end consumption rates (in the range of 100 - 400 g/day).
Regarding the use of rates from the Dellinger (2004) and Shilling, et al., (2010) studies, the EPA
explained how the EPA used the consumption rates from these studies. As stated in the Hg Risk
TSD in section 1.4.3, both studies supported the consumption rates used in modeling the typical
female fisher as well as the scenarios for additional socioeconomic groups. Specifically, the
Dellinger study provided 95th and 99th percentile fish consumption rates of 213.1 and 492.8 g/day
respectively, while the Shilling, et al., study provided 95th percentile fish consumption rates up to
265.8 g/day. Consequently, both studies confirmed that the high-end percentile rates used in the
typical female fisher scenario (i.e., 173 g/day for the 95th percentile and 373 g/day for the 99th
percentile) in the Hg Risk TSD were reasonable. In addition, the Dellinger, et al., and Shilling, et
al., studies provided consumption rates for the other socioeconomic subsistence fisher population
scenarios in the Hg Risk TSD.
Because Petitioners have not demonstrated that it was impracticable to comment on this issue
during the public comment period, as evidenced by the comments and the EPA's responses to
those comments, and because the issue is not of central relevance to the "Appropriate and
Necessary" finding, the EPA is denying reconsideration of this issue.
Issue 12: Petitioner 20183 (att. 4) and Petitioner 20180 (att. 5) comment that the EPA states that
the Hg Risk TSD can be viewed as a public health hazard screening analysis. Petitioner further
states that the EPA's Residual Risk Report to Congress (U.S. EPA, 1999) concludes that
screening analyses can be used to eliminate low-risk categories but should not serve as the basis
for establishing additional risk reduction requirements under CAA section 112 (f). Petitioners
state that because screening analyses are typically designed to overestimate risks and the Hg Risk
TSD is a screening analysis, the Hg Risk TSD (and the 99th percentile rate) is subject to
significant uncertainty. Petitioners also note that the Burger, et al., (2002) and Moya, et al.,
(2008) studies identify the limited geographic scope for high-end subsistence level fish
consumption rates. For these reasons, Petitioners recommend that (a) the EPA used the 95th
percentile rate from the Burger, et al., study and (b) that the EPA "should conduct a more refined
analysis of the fish consumption rate in order to decrease the uncertainty associated with the fish
consumption rate. Petitioners also point to the lower recreational fish consumption rates used in
the MATS Regulatory Impact Analysis (RIA) as support for using lower fish consumption rates
in the Hg Risk TSD.
Response to Issue 12: Issues regarding the fish consumption rates were raised in public
comments submitted in response to the proposed MATS rule. The EPA responded to these
comments in Section IF (pp. 135 - 145) of the RTC.
Because Petitioners have not demonstrated that it was impracticable to comment on this issue
during the public comment period on the proposed MATS rule, the EPA is denying
reconsideration of this issue. In addition, in the White Stallion decision, the Court found that
19
-------
"EPA's 'appropriate and necessary' determination in 2000, and its reaffirmation of that
determination in 2012, are amply supported by EPA's findings regarding the health effects of
mercury exposure." White Stallion, 748 F.3d at 1245-46. As the Court may only consider issues
raised during the period for public comment, issues raised in the litigation and addressed by the
Court clearly do not meet the criteria for reconsideration in CAA section 307(d)(7)(B).
Moreover, parties may not use this final action denying reconsideration as a basis to litigate
issues that could have been raised in the initial litigation.
Under CAA section 112(f), eight years after MATS promulgation, the EPA is required to
complete a residual risk assessment to determine whether additional emissions reductions
beyond the MACT level of control from coal- and oil-fired EGUs are warranted to protect public
health. Thus, the statute only contemplates a residual risk evaluation after application of MACT
standards. At that time, the EPA will complete a two-stage risk assessment described by
Petitioner (i.e., a screening-level analysis followed, if warranted, by a more refined risk
assessment) as part of the residual risk analysis. Because the statute requires a residual risk
evaluation after application of MACT standards, EPA will have by that time collected as much
as 5 years' worth of HAP emissions data from EGUs complying with the MACT level of control
and that considerable data can be used to better evaluate the facility specific risks from all EGUs
that section 112(f) requires. Had Congress intended for EPA to use the residual risk analysis it is
reasonable to conclude it would done so expressly. See CAA section 112(n)(l)(A) (using the
words "section", "subsection", and "subparagraph"; and specifically directing the Agency to
regulate EGUs "under this section [112]" if the Agency determines regulation is appropriate and
necessary).
But statutory language in CAA section 112(n) does not support Petitioners' assertion that the
EPA should have used a two-stage risk assessment such as used in the residual risk
determination to support the "Appropriate and Necessary" finding, and the EPA disagrees with
this assertion. In contrast, the "Appropriate and Necessary" finding is based in part on an
assessment of the public health hazard associated with Hg emitted from U.S. EGUs and non-Hg
HAP pollutants emitted from U.S. EGUs, supported both by the studies required by CAA section
112(n)(l), and by the updated technical analyses described in the preamble to the proposed and
final MATS rules. The EPA also disagrees with Petitioners' recommendation to use the 95th
percentile rates because it does not target the subpopulation at greatest risk of Hg exposure from
U.S. EGU-attributable Hg emissions. See CAA section 112(n)(l)(C) (expressing concern for
risks posed by exposure to mercury as a result of fish "consumption by sensitive populations")
Although the EPA acknowledges that the MATS RIA uses lower fish consumption rates to
model exposure for recreational fishers in the general population, the EPA disagrees that it
should have used the rates from the RIA in the Hg Risk TSD. Because the RIA is focused on the
general population, whereas the Hg Risk TSD is focused on high-end subsistence fishers (i.e.
"sensitive populations"), it is appropriate for the RIA to use lower fish consumption rates.
Further, the RIA was not used to support the "Appropriate and Necessary" finding because it is
an analysis of the specific MATS standards and is not an analysis of the need for regulation
under CAA section 112, nor is it an assessment of the public health hazard associated with U.S.
EGU HAP emissions absent regulation under section 112. As such, the RIA does not reflect the
specific requirements of CAA section 112(n), and is therefore neither relevant nor useful in
20
-------
making the appropriate and necessary finding. In addition, issues regarding consumption rates in
the RIA were raised in public comments submitted in response to the proposed MATS rule, and
the EPA responded to those comments in Section IF (p. 140) of the RTC.
Because the EPA disagrees that it should have used the rates from the RIA, because Petitioners
have not demonstrated that it was impracticable to comment on this issue during the public
comment period, and because the RIA is not of central relevance to the "Appropriate and
Necessary" finding, the EPA is denying reconsideration of this issue.
Issue 13: Petitioner 20183 (att. 4) and Petitioner 20180 (att. 5) assert that the SAB in a prior
review of an EPA action stated that the EPA cannot make reliable predictions at the 99.5th
percentile (Henderson, 2007).28 Petitioners cite this statement to support their position that the
EPA should use a lower percentile for fish consumption in the Hg Risk TSD.
Response to Issue 13: Issues regarding the fish consumption rates applied in the Hg Risk TSD
were raised in public comments submitted in response to the proposed MATS rule. The EPA
responded to these comments in Section IF (pp. 135 - 145) of the RTC.
Because Petitioners have not demonstrated that it was impracticable to comment on fish
consumption rate issues during the public comment period on the proposed MATS rule, the EPA
is denying reconsideration of this issue. In addition, in the White Stallion decision, the Court
found that "EPA's 'appropriate and necessary' determination in 2000, and its reaffirmation of
that determination in 2012, are amply supported by EPA's findings regarding the health effects
of mercury exposure." White Stallion, 748 F.3d at 1245-46. As the Court may only consider
issues raised during the period for public comment, issues raised in the litigation and addressed
by the Court clearly do not meet the criteria for reconsideration in CAA section 307(d)(7)(B).
Moreover, parties may not use this final action denying reconsideration as a basis to litigate
issues that could have been raised in the initial litigation.
Moreover, Petitioners did not raise the specific issue regarding the SAB's prior reviews of other
EPA actions in comments during the public comment period on the proposed MATS rule despite
providing extensive comments on other aspects of the Hg Risk TSD, and they have not provided
a reason why they were unable to do so. Because Petitioners have not demonstrated that it was
impracticable to comment on the issue of SAB's prior comments on an unrelated study during
the public comment period on the proposed MATS rule, the EPA is denying the petition for
reconsideration of this issue.
Further, the EPA disagrees with Petitioners' assertion that the SAB's previous comment supports
a lower percentile for fish consumption. The SAB's comments cited by Petitioner were specific
to the context of the lead NAAQS risk assessment, which is a fundamentally different risk
analysis involving a different pollutant, different exposure pathways, and different health
28 Henderson, R. 2007. Letter from Dr. Rogene Henderson, Chair, Clean Air Scientific Advisory
Committee, to Administrator Stephen L. Johnson. Clean Air Scientific Advisory Committee
(CASA C) Lead Review of the 2nd Draft Lead Human Exposure and Health Risk Assessments
Document. EPA-CASAC-07-007. September 27, 2007.
21
-------
endpoints than the Hg Risk TSD. Importantly, the EPA assessed lead risk for the general
population, whereas the Hg Risk TSD uses a scenario-based approach for high-end subsistence
fishers. For these reasons, the SAB's comments, although appropriate for risk assessments for
general populations such as the lead NAAQS risk assessment, are not relevant to risk
assessments for particular sensitive populations (e.g. high-end subsistence fishers considered in
the Hg Risk TSD). The EPA further notes that in the peer review of the Hg Risk TSD, the SAB
supported the overall design and approach of the Hg Risk TSD (p. 2), which includes the fish
consumption percentiles (U.S. EPA-SAB, 2011a).
Because Petitioners have not demonstrated that it was impracticable to comment on these issues
during the public comment period and because the EPA disagrees with Petitioners' interpretation
of the SAB's prior comments, and because the SAB letter identified is not of central relevance,
the EPA is denying the petition for reconsideration of this issue.
1.5 Fish tissue concentrations and samples in Hg Risk TSD
Issue 14: Petitioner 20183 (att. 4) and Petitioner 20180 (att. 5) claim that using the 75th
percentile fish tissue concentration is likely to overestimate Hg exposure compared to the 95th
percentile "upper confidence limit" (UCL) concentration recommended in the EPA's risk
guidance (U.S. EPA, 1992,29 200230). Petitioner 20183 (att. 4) and Petitioner 20180 (att. 5)
further assert that the EPA's guidance recommends that a minimum of 10 to 15 samples are need
to obtain meaningful results for the UCL (U.S. EPA, 2010),31 but 41 percent of the watersheds in
the Hg Risk TSD have only 1 to 2 samples per watershed.
Response to Issue 14: Issues regarding the 75th percentile fish tissue level were raised in public
comments and the EPA responded to those comments in Section IF of the RTC (pp. 140 - 145
and 176 - 178).
In the RTC (p. 140), the EPA acknowledged that some subsistence fishers could consume fish
without consideration for size (given dietary necessity) and that some subsistence fishers could
target larger fish in order to maximize the volume of edible fish. The Hg Risk TSD assumes that
a subset of subsistence fishers likely target larger fish, which is represented by the 75th percentile
fish tissue value. The EPA has focused the Hg Risk TSD on this subset of subsistence fishers
targeting larger fish, to represent those subsistence fishers likely to experience the highest
exposure to Hg emitted by U.S. EGUs. Further, as stated in the RTC (p. 141), the SAB supported
the use of the 75th percentile fish tissue concentration for this purpose, and concluded that
"[u]sing the 75th percentile of fish tissue values as a reflection of consumption of larger, but not
the largest, fish among sport and subsistence fishers is a reasonable approach and is consistent
29 U.S. EPA. 1992. Supplemental Guidance to RAGS: Calculating the Concentration Term.
Office of Solid Waste and Emergency Response. Publication 9285.7-081. Washington, D.C. May.
30 U.S. EPA. 2002. Calculating Upper Confidence Limits for Exposure Point Concentrations at
Hazardous Waste Sites. Office of Emergency and Remedial Response. Washington, D.C.
December.
31 U.S. EPA. 2010. ProUCL Version 4.00.05 User Guide. Office of Research and Development.
EPA/600/R-07/038, May.
22
-------
with published and unpublished data on predominant types of fish consumed" (U.S. EPA-SAB,
201 la, p. 12). In addition, the EPA specifically addressed the issue of low sample size associated
with the 75th percentile fish tissue concentration for each watershed. In the RTC (p. 141), the
EPA stated, "[t]he EPA disagrees with the commenter that it is not reasonable to use watersheds
where only a single fish sample is available. Although it is generally preferred to have multiple
samples, the SAB noted that using a single sample is likely to underestimate the 75th percentile
fish MeHg concentration and is therefore likely to underestimate the risk estimates for those
watersheds." Because Petitioners have not demonstrated that it was impracticable to comment on
this issue during the public comment period on the proposed MATS rule, the EPA is denying the
petition for reconsideration of this issue. In addition, in the White Stallion decision, the Court
found that "EPA's 'appropriate and necessary' determination in 2000, and its reaffirmation of
that determination in 2012, are amply supported by EPA's findings regarding the health effects
of mercury exposure." White Stallion, 748 F.3d at 1245-46. As the Court may only consider
issues raised during the period for public comment, issues raised in the litigation and addressed
by the Court clearly do not meet the criteria for reconsideration in CAA section 307(d)(7)(B).
Moreover, parties may not use this final action denying reconsideration as a basis to litigate
issues that could have been raised in the initial litigation.
In addition, Petitioners did not raise the specific issue regarding the use of the UCL in comments
during the public comment period on the proposed MATS rule despite providing extensive
comments on other aspects of the Hg Risk TSD, and they have not provided a reason why they
were unable to do so. Further, the EPA disagrees with Petitioners' assertion that the EPA should
use the 95th percentile UCL instead of the 75th percentile level. The 95th percentile UCL provides
a high-confidence estimate of the average fish tissue concentration within a given watershed, not
an estimate of fish tissue concentrations in larger fish. Use of the 95th percentile UCL on the
average fish tissue Hg concentration would not address the relevant sensitive populations
considered in the Hg Risk TSD. Petitioners have not demonstrated that it was impracticable to
comment on this specific issue during the public comment period on the proposed MATS rule
and the SAB supported the approach. In addition, this issue is not of central relevance to the
"Appropriate and Necessary" finding as explained above.
For all these reasons, the EPA is denying the petition for reconsideration of this issue.
Issue 15: Petitioner 20183 (att. 4) and Petitioner 20180 (att. 5) state that the EPA included fish
tissue samples in the Hg Risk TSD from large lakes including Lake Michigan, Lake Ontario,
Lake Erie, Lake Champlain, and Lake Pontchartrain. Petitioners assert that fish tissue samples
from these waterbodies do not meet the basic criteria for application of the Mercury MAPs
assumption.
Response to Issue 15: Petitioners did not raise the issue regarding fish tissue samples from large
lakes during the public comment period on the proposed MATS rule. In any case, the SAB
recommended augmentations to the fish tissue data set (U.S. EPA-SAB, 201 la, p. 3), and the
EPA revised the dataset for the revised Hg Risk TSD accordingly. The EPA also stated in the
revised Hg Risk TSD (pp. 7 - 9), that large lakes were excluded from the risk assessment because
of uncertainty in modeling the linkage between EGU-attributable Hg deposition and Hg
concentrations in fish, and this exclusion likely results in an underestimate of the number of
23
-------
watersheds with populations potentially at risk in the U.S. In response to this newly raised issue,
the EPA reviewed the data in the revised spreadsheet and associated data files (EPA-HQ-OAR-
2009-0234-19898) to determine if any of the watersheds in the revised Hg Risk TSD included
fish tissue samples from the Great Lakes, Lake Champlain, or Lake Pontchartrain. This review
revealed that, contrary to the assertion in the revised Hg Risk TSD (pp. 7 - 9), a small subset of
the HUC-12 watersheds immediately adjacent to these large lakes included a small fraction of
fish tissue samples taken within those larger lakes. Specifically, this review identified 48
watersheds out of the 754 watersheds with potentially at-risk populations due to total risk and 5
percent Hg contribution from U.S. EGUs were immediately adjacent to one of the Great Lakes
and Lake Pontchartrain. In addition, this review identified 15 watersheds out of the 327
watersheds with potentially at-risk populations from U.S. EGU emissions alone were
immediately adjacent to one of the Great Lakes and Lake Pontchartrain.
Next, considering the limitations in applying the proportionality assumption for fish from large
lakes acknowledged in the revised Hg Risk TSD (pp. 7 - 9), the EPA conducted a new sensitivity
analysis to determine the impact of excluding these watersheds on the risk estimates. The results
of this sensitivity analysis shows that removing the 47 watersheds from the risk analysis reduces
the percentage of watersheds with potentially at-risk populations due to total risk and 5 percent
Hg contribution from U.S. EGUs from 24 percent to 23 percent. Removing the 15 watersheds
from the risk analysis does not change the percentage of watersheds with potentially at-risk
populations from U.S. EGUs alone (i.e., remains at 10 percent). In summary, this new sensitivity
analysis shows that, even when those watersheds with some fish samples taken within large lakes
are excluded, the impact on risk estimates is negligible. Because the issue raised by Petitioners
would not, even if accepted by the EPA, change the conclusions of the revised Hg Risk TSD, this
issue is not of central relevance to the "Appropriate and Necessary" finding.
Because the issue is not of central relevance to the "Appropriate and Necessary" finding, the
EPA is denying reconsideration of this issue.
Issue 16: Petitioner 20183 (att. 4) and Petitioner 20180 (att. 5) assert that the State of Florida
sampled Hg concentrations in fish in lakes in 2006 and sampled in lakes and streams in 2010 to
2011, covering 275 additional watersheds. Petitioner asserts that these additional fish tissue data
should have been incorporated into the Hg Risk TSD.
Response to Issue 16: Issues regarding the fish tissue samples from Florida were raised in
public comments submitted in response to the proposed MATS rule. The EPA responded to these
comments in Section IF (pp. 175 - 176) of the RTC. Furthermore, in the White Stallion decision,
the Court found that "EPA's 'appropriate and necessary' determination in 2000, and its
reaffirmation of that determination in 2012, are amply supported by EPA's findings regarding
the health effects of mercury exposure." White Stallion, 748 F.3d at 1245-46. As the Court may
only consider issues raised during the period for public comment, issues raised in the litigation
and addressed by the Court clearly do not meet the criteria for reconsideration in CAA section
307(d)(7)(B). Moreover, parties may not use this final action denying reconsideration as a basis
to litigate issues that could have been raised in the initial litigation.
24
-------
In the RTC (pp. 175 - 176), the EPA stated that, "[t]he SAB concluded, 'Although the SAB
considers the number of watersheds included in the assessment adequate, some watersheds in
areas with relatively high Hg deposition from U.S. EGUs were under-sampled due to lack of fish
tissue methy[l]mercury data. The SAB encourages the Agency to contact states with these
watersheds to determine if additional fish tissue MeHg data are available to improve coverage of
the assessment.' (U.S. EPA-SAB, 201 la). As a result of the SAB advice, the EPA obtained
additional fish tissue MeHg sample data from several states, particularly Pennsylvania,
Wisconsin, Minnesota, New Jersey, and Michigan." Consistent with this advice, the EPA
incorporated additional fish tissue data only for areas with relatively elevated Hg deposition from
U.S. EGUs.
Contrary to Petitioners' implication, the EPA did not willfully ignore available fish tissue data
from Florida. In response to SAB's concern regarding the apparent lack of fish tissue data in
"states that receive... 'relatively elevated' mercury deposition from U.S. EGU emissions and
have limited fish methylmercury measurements" (U.S. EPA-SAB, 201 la; p. 11), the EPA added
data for states likely to have higher Hg impacts or were specifically identified by the SAB. The
EPA did not conduct an exhaustive search beyond those states, and Petitioners did not submit
fish tissue data from Florida or any other state to the EPA, so we are unable to assess the
legitimacy of Petitioners' assertions concerning the data. In addition, because modeled Hg
deposition from U.S. EGUs in Florida was relatively low compared with the deposition in other
states identified by the SAB (revised Hg Risk TSD, p. 59), obtaining additional fish tissue data
from Florida was considered a lower priority, especially considering the limited impact the data
could have on the results.
Given that the number of watersheds that Petitioners assert could have been added from Florida
(i.e., 275 additional watersheds based on Petitioner's information) would have comprised less
than 9 percent of the total number of watersheds modeled in the Hg Risk TSD, the overall
percentage of at-risk watersheds would not be significantly affected. At most, hypothetically
assuming that none of the 275 additional watersheds in Florida would have populations
potentially at risk, the percentage of watersheds could decrease from 29 percent to 27 percent,
which would remain unacceptable. Because the EPA has no direct information to suggest the 275
additional watersheds in Florida would have populations potentially at risk, this two percentage
point difference reflects an upper bound on the potential impact of this issue. Therefore, this
issue is not of central relevance to the "Appropriate and Necessary" finding.
Because Petitioners have not demonstrated that it was impracticable to comment on this issue
during the public comment period on the proposed MATS rule, and because the issue is not of
central relevance to the "Appropriate and Necessary" finding since it would not alter our
conclusion even if the most conservative assumptions were made, the EPA is denying the
petition for reconsideration of this issue.
Issue 17: Petitioner 20183 (att. 4) and Petitioner 20180 (att. 5) assert that fish sampling
programs conducted by states often target larger top predator species that can have higher Hg
levels, which Petitioner asserts leads to overestimates of Hg concentrations in fish.
25
-------
Response to Issue 17: Petitioners did not raise the issue regarding state fish sampling programs
during the public comment period on the proposed MATS rule despite providing extensive
comments on other aspects of the Hg Risk TSD, and they have not provided a reason why they
were unable to do so.
The SAB (U.S. EPA, 201 la, p. 14) raised the issue regarding state fish sampling programs and
recommended that the EPA add a discussion of how state programs contribute to uncertainty and
variability in the dataset. Accordingly, in the uncertainty discussion in the revised Hg Risk TSD
(p. 96), the EPA noted the differences in state-level protocols regarding fish sampling could
result in high-bias in the characterization of fish tissue Hg levels due both to potential targeting
of watersheds suspected of having higher fish tissue Hg levels as well as the types of species
targeted for sampling. Further, the EPA stated that, "[bjecause the goal of the risk assessment is
to capture risk for subsistence fishers that are likely to experience elevated U.S. EGU-
attributable risk, having fish tissue Hg concentrations that likely reflect targeting of more highly
impacted waterbodies (i.e., biased towards higher risk locations) is not problematic and in fact, is
preferable." This same observation holds for the targeting of larger fish species with potentially
higher Hg levels. The fact that this issue was identified by the SAB demonstrates that Petitioners
could have commented on the issue during the public comment period, but the Petitioner did not
do so. Further, Petitioner provided no new information that would support a change to the Hg
Risk TSD because this issue was described in the uncertainty section of the Hg Risk TSD, and
the EPA added the discussion in response to the positive SAB review. There is no indication that
this uncertainty undermined the SAB's confidence in the Hg Risk TSD.
Because Petitioners have not demonstrated that it was impracticable to comment on this issue
during the public comment period on the proposed MATS rule and because this issue is not of
central relevance to the "Appropriate and Necessary" finding, the EPA is denying the petition for
reconsideration of this issue.
1.6 Proportionality assumption in Hg Risk TSD
Issue 18: Petitioner 20183 (att. 4) and Petitioner 20180 (att. 5) assert that several studies do not
support the Mercury MAPs assumption (proportionality assumption) that a change in Hg
deposition over a given watershed will result in a proportional change in fish tissue Hg
concentrations (Watras and Morrison 2008,32 Dittman and Driscoll 2009,33 Johansson, et al.,
32 Watras, C.J., and K.A. Morrison. 2008. "The response of two remote, temperature lakes to
changes in atmospheric mercury deposition, sulfate and the water cycle" Can. J. Fish. Aquat. Sci.
65: 100-116.
33 Dittman, J. A., C.T. Driscoll. 2009. "Factors influencing changes in mercury concentrations in
lake water and yellow perch (Perca flavescens) in Adirondack Lakes." Biogeochemistry 93: 179-
196.
26
-------
2001,34 MPCA 2007,35 Chalmers, et al., 2011,36 Butler, et al., 200837). Based on these studies,
Petitioners assert that (1) changes in fish Hg concentrations in response to changes in
atmospheric deposition vary across water bodies; (2) data from Minnesota, Wisconsin and the
Adirondacks show inconsistent responses in fish Hg concentrations to similar decreases in
atmospheric deposition, (e.g., some lakes show increases in fish Hg concentrations and some
decreases); (3) some lakes showed less decrease in fish Hg or an increase, which may suggest the
possibility of relatively slow response from large watersheds or that other factors are controlling;
(4) the fish Hg concentrations in drainage lakes are influenced by both changes in direct
deposition and fluvial inputs of Hg and microbial substrates; and (5) different responses in fish
Hg concentrations were observed in lakes from the same region that experience similar changes
in atmospheric deposition. Petitioners also state that the METAALICUS study (Harris, et al.,
2007)38 shows that in the early years of the study, there was a proportional response of fish tissue
Hg to direct Hg deposition, but no response in the fish tissue Hg from the watershed deposition.
Petitioners assert that these factors undermine the EPA's use of the proportionality assumption in
the Hg Risk TSD.
Response to Issue 18: The EPA denies the Petitions for Reconsideration on the issue because (a)
many of the issues are similar to proportionality issues raised in public comments and the EPA
responded to those comments; (b) Petitioners however did not raise these issues during the public
comment period and have not asserted or explained that it was impractical to do so; and (c)
Petitioners provide no new information that would support a change to the EPA's responses to
these issues. Furthermore, in the White Stallion decision, the Court found that "EPA's
'appropriate and necessary' determination in 2000, and its reaffirmation of that determination in
2012, are amply supported by EPA's findings regarding the health effects of mercury exposure."
White Stallion, 748 F.3d at 1245-46. As the Court may only consider issues raised during the
period for public comment, issues raised in the litigation and addressed by the Court clearly do
not meet the criteria for reconsideration in CAA section 307(d)(7)(B). Moreover, parties may not
use this final action denying reconsideration as a basis to litigate issues that could have been
raised in the initial litigation.
34 Johansson, K. Bergback, B. and Tyler, G. 2001. "Impact of atmospheric long range transport
of lead, mercury and cadmium on the Swedish forest environment." Water Air Soil Pollution.
Focus 1, 279-297.
35 Minnesota Pollution Control Agency (MPCA). 2007. Minnesota Statewide Mercury Total
Maximum Daily Load. March 27.
36 Chalmers, A.T., D.M. Argue, D.A.Gay, M.E. Brigham, C.J. Schmitt, D.L. Lorenz. 2011.
"Mercury trends in fish from rivers and lakes in the United States, 1969-2005," Environmental
Monitoring and Assessment 175: 175-191.
37 Butler, T.J., M. D. Cohen, F. M. Vermeylen, G. E. Likens, D. Schmeltz, R. S. Artz. 2008.
"Regional precipitation mercury trends in the eastern USA, 1998-2005: Declines in the
Northeast and Midwest, no trend in the Southeast," Atmospheric Environment 42: 1582-1592.
38 Harris, R.C., et al., 2007. "Whole-ecosystem study shows rapid fish-mercury response to
changes in mercury deposition." PNAS 104 (42): 16586-16591.
27
-------
Issues regarding the proportionality assumption were raised in public comments submitted in
response to the proposed MATS rule. The EPA responded to these comments in Section IF (pp.
123 - 129) oftheRTC.
As noted in the RTC (p. 123), the EPA stated that many factors can affect the lag time between a
change in Hg deposition over watersheds and the associated change in fish tissue Hg
concentrations (e.g., pH, sulfate deposition, topography including nature of the surrounding
watershed, which influences importance of erosion/runoff as a source of loading). The
differences in the response of watersheds to changes in Hg deposition (as described by
Petitioner) reflect these differences in lag-times rather than a potential lack of applicability of the
proportionality assumption. Different rates of change in fish tissue Hg levels will be reflected as
different Hg levels in fish across watersheds if those measurements are taken before all
watersheds have had sufficient time to reach steady-state.
Differences in lag time can also reflect the multi-phase impacts of deposited Hg on fish tissue Hg
concentrations, with Hg deposited directly to the waterbody having a quicker response and Hg
deposited to the surrounding watershed having a lagged response given the time needed for the
erosion/runoff of that Hg to reach the waterbody (Harris, et al., 2007). Petitioner's comment
regarding the METAALICUS study points to this multi-phase phenomena for fish Hg response
and highlights the potential for different response times across watersheds. However, the multi-
phase nature of the response of fish Hg concentrations to Hg deposition over watersheds does not
contradict the applicability of the proportionality assumption provided that sufficient time is
allowed for the full impact of Hg deposition to be realized in the fish.
Further, the SAB supported using the proportionality assumption in the Hg Risk TSD. As noted
in the RTC (p. 125), the SAB stated that, "[s]ince the Hg Maps approach was developed, several
recent publications have supported the finding of a linear relationship between Hg loading and
accumulation in aquatic biota (Orihel, 2007; Orihel, 2008; Harris, 2007). These studies suggested
that Hg deposited directly to aquatic ecosystems can become quickly available to biota and
accumulated in fish, and that reductions in atmospheric Hg deposition should lead to decreases in
MeHg concentrations in biota. These results substantiate EPA's assumption that proportionality
between air deposition changes and fish tissue MeHg level changes is sufficiently robust for its
application in this risk assessment" (U.S. EPA-SAB, 201 la). Given the SAB support for the
proportionality assumption and the fact that differences in observed fish tissue Hg levels across
watersheds in a region can reflect differences in the lag time associated with those Hg deposition
changes, the issues raised by Petitioner have provided no new information that would support a
change to the EPA's response on this issue.
In addition, the EPA noted in the Hg Risk TSD that specific regions of the country with similar
general trends in Hg deposition could have notable differences in Hg deposition rates across
watersheds within that region. For example, Figure 2-2 of the Hg Risk TSD shows that
Minnesota, Wisconsin and the Adirondacks (i.e., the areas referenced by Petitioner) have spatial
variability in total Hg deposition. Because of this spatial variability, any comparison of fish
tissue Hg concentrations would need to consider watershed-level differences in Hg deposition
rather than general patterns of deposition. However, apparent differences in the response of fish
28
-------
tissue Hg levels to changes in Hg deposition likely reflect different lag periods for this linkage,
rather than a lack of applicability of the proportionality assumption.
Because Petitioners have not demonstrated that it was impracticable to comment on this issue
during the public comment period on the proposed MATS rule, as evidenced by comments on
these issues from commenters and the SAB, and because the issue is not of central relevance, the
EPA is denying reconsideration of these issues.
Issue 19: Petitioner 20183 (att. 4) and Petitioner 20180 (att. 5) state that "[t]he SAB
recommended a quantitative evaluation of the uncertainty associated with the proportionality
assumption" and they noted that such an analysis was not completed by the EPA.
Response to Issue 19: The EPA disagrees that the revised Hg Risk TSD did not respond to a
significant comment from the SAB. In fact, the revised Hg Risk TSD (pp. 4, 100 - 103) directly
responded to this SAB recommendation by expanding the uncertainty discussion for the
proportionality assumption, including references to specific studies on the proportionality
assumption.
Further, Petitioners have misrepresented the SAB's comment as recommending a quantitative
rather than qualitative uncertainty assessment. The SAB stated, "[a]lthough the Panel was
generally satisfied with the presentation of uncertainties and limitations associated with the
application of the Mercury Maps [proportionality] approach in qualitative terms, it recommended
that the document include quantitative estimates of uncertainty that are available in the existing
literature" (U.S. EPA-SAB, 201 la, p. 4). The SAB also reiterated their support for qualitative
uncertainty assessments in the Hg Risk TSD, concluding that "[t]he qualitative nature of the
discussion is appropriate since this is a conditional analysis" (p. 22). Considering these
statements together, the EPA concluded that the SAB was not asking the EPA to conduct a
quantitative uncertainty assessment regarding the proportionality assumption. Rather, the SAB
requested that the Hg Risk TSD include a qualitative uncertainty discussion that included
quantitative information to the extent it is available in the literature. Because the revised Hg Risk
TSD fully responded to the SAB's recommendation and the assessment recommended by
Petitioners would not have changed the core results of the Hg Risk TSD, this issue is not of
central relevance to the "Appropriate and Necessary" finding.
In addition, the Hg Risk TSD included several quantitative sensitivity analyses to assess the
potential impact of the application of the proportionality assumption. Specifically, the EPA
conducted an assessment that excluded fish tissue data from four states with higher potential for
non-air Hg and an assessment that excluded flowing waterbodies. Because the results of these
sensitivity analyses did not change the overall conclusions in the Hg Risk TSD, this issue is not
of central relevance to the "Appropriate and Necessary" finding. Because this issue is not of
central relevance to the "Appropriate and Necessary" finding, the EPA is denying the petition for
reconsideration of this issue.
1.7 Screening of watersheds with significant non-atmospheric deposition in Hg Risk TSD
29
-------
Issue 20: Petitioner 20183 (att. 4) and Petitioner 20180 (att. 5) found the EPA's method for
screening out watersheds with significant non-atmospheric deposition to be inadequate.
Petitioner 20183 (att. 4) and Petitioner 20180 (att. 5) identified several watersheds with very high
Hg fish tissue measurements to be associated with historic mines, Superfund sites, and industrial
facilities. Petitioner 20183 (att. 4) and Petitioner 20180 (att. 5) recommend the EPA apply a
more thorough screening to remove watersheds with known or likely sources of terrestrial
contamination from the Hg Risk TSD.
Response to Issue 20: Issues regarding the method for excluding watersheds were raised in
public comments submitted in response to the proposed MATS rule. The EPA responded to these
comments in Section IF (pp. 128 - 129) of the RTC. In the RTC (p. 129), the EPA states that (in
relation to the methods EPA used to exclude watersheds that were likely significantly influenced
by non-atmospheric sources), the SAB concluded, "[t]he technique used to exclude watersheds
that may have substantial non-air inputs is sound. Although additional criteria could be applied,
they are unlikely to substantially change the results" (U.S. EPA-SAB, 201 la).
In addition, in the revised Hg Risk TSD, the EPA conducted two sensitivity analyses addressing
the potential issue of not excluding all watersheds with significant non-air Hg impacts (pp. 87 -
91). Specifically, the EPA assessed two scenarios: (a) excluding watersheds located in four states
with potential for significant non-air Hg releases and/or increased methylation potential (i.e., LA,
SC, MN, and ME) and (b) including only watersheds where atmospheric Hg deposition likely
plays a dominant role (i.e., watersheds in the upper 25th percentile of total Hg deposition). The
results of the sensitivity analyses suggest that excluding watersheds in states with potential for
significant non-air Hg loading does not substantially affect risk and that including watersheds
with the highest total Hg deposition moderately increased risk. Together, these results suggest
that the potential inclusion of a (likely small) subset of watersheds with significant non-air Hg
loading in modeling risk is unlikely to bias the results of the Hg Risk TSD. Therefore, this issue
is not of central relevance to the "Appropriate and Necessary" finding.
The EPA denies the Petition for Reconsideration on this issue because EPA provided an
opportunity to comment on these issues and, in fact, substantively similar issues were raised in
public comments and the EPA responded to those comments. Because Petitioners have not
demonstrated that it was impracticable to comment on this issue during the public comment
period on the proposed MATS rule and because the issue is not of central relevance, the EPA is
denying the petition for reconsideration of this issue.
Furthermore, in the While Stallion decision, the Court found that "EPA's 'appropriate and
necessary' determination in 2000, and its reaffirmation of that determination in 2012, are amply
supported by EPA's findings regarding the health effects of mercury exposure." White Stallion,
748 F.3d at 1245-46. As the Court may only consider issues raised during the period for public
comment, issues raised in the litigation and addressed by the Court clearly do not meet the
criteria for reconsideration in CAA section 307(d)(7)(B). Moreover, parties may not use this
final action denying reconsideration as a basis to litigate issues that could have been raised in the
initial litigation.
1.8 Spatial scale of non-stationary watersheds
30
-------
Issue 21: Petitioner 20183 (att. 4) and Petitioner 20180 (att. 5) asserted that using HUC-12
watersheds might be too small for applying the proportionality assumption to streams, which can
reflect Hg loading from the larger upstream watersheds.
Response to Issue 21: Petitioners did not raise the issue regarding the spatial scale of flowing
water bodies such as streams and rivers during the public comment period on the proposed
MATS rule despite providing extensive comments on other aspects of the Hg Risk TSD, and
they have not demonstrated that they were unable to do so.
In addition, in the draft Hg Risk TSD, the core risk results reflected both stationary waterbodies
and flowing waterbodies. The EPA also conducted a sensitivity analysis excluding flowing
waterbodies to assess potential impact of the application of the proportionality assumption in
these watersheds (pp. 58 - 62). This sensitivity analysis demonstrated (a) the risk estimates for
the average watershed did not change from the core results including both stationary and flowing
waterbodies, (b) risk estimates for the 90th to 95th percentile watersheds remained largely
unchanged, and (c) risk estimates for the 99th percentile watershed were actually higher for
stationary waterbodies. Further, as stated in the revised Hg Risk TSD (pp. 87 and 102), because
the SAB stated that the proportionality assumption can be readily applied for stationary and
flowing waterbodies (U.S. EPA-SAB, 201 la, p. 18), and the prior sensitivity analysis
demonstrated that this assumption would not substantially change the risk estimates, the EPA did
not repeat this sensitivity analysis in the revised Hg Risk TSD.
The EPA believes that it is appropriate to include flowing waterbodies in the analysis and the
SAB agrees. In addition, based on our sensitivity analysis, we believe the exclusion of flowing
water bodies would actually demonstrate a higher percentage of at risk watersheds, which
appears to be the opposite of the assumption made by Petitioners, and for these reasons, we do
not believe the issue is of central relevance.
Because Petitioners have not demonstrated that it was impracticable to comment on this issue
during the public comment period on the proposed MATS rule and because we do not believe the
issue is of central relevance based on the record, the EPA is denying reconsideration of this
issue.
1.9 Alternative risk calculations for Hg Risk TSD
Issue 22: Petitioner 20183 (att. 4) and Petitioner 20180 (att. 5) generated risk estimates using
alternative parameters for fish ingestion rates, fish tissue concentrations, and the cooking
adjustment factor. Using these alternative parameters in different scenarios, the Petitioner 20183
(att. 4) and Petitioner 20180 (att. 5) found either no watersheds or a substantially smaller fraction
of watersheds exceeding the HQ risk criteria. Based on these findings, Petitioners assert that the
EPA has substantially overestimated actual risk through the use of multiple upper-bound values
for exposure parameters used in their risk calculations.
Response to Issue 22: Issues regarding alternate parameters and scenarios in the analysis
conducted on behalf of Petitioners were raised in public comments submitted in response to the
31
-------
proposed MATS rule. Specifically, the analysis submitted by Petitioners acknowledges that
similar comments were submitted on the proposed rule (Petitioner 20183 (att. 4) and Petitioner
20180 (att. 5), p. 3).The EPA responded to these comments in Section IF (pp. 171 - 172) of the
RTC. In the White Stallion decision, the Court found that "EPA's 'appropriate and necessary'
determination in 2000, and its reaffirmation of that determination in 2012, are amply supported
by EPA's findings regarding the health effects of mercury exposure." White Stallion, 748 F.3d at
1245-46. As the Court may only consider issues raised during the period for public comment,
issues raised in the litigation and addressed by the Court clearly do not meet the criteria for
reconsideration in CAA section 307(d)(7)(B). Moreover, parties may not use this final action
denying reconsideration as a basis to litigate issues that could have been raised in the initial
litigation.
As stated in the RTC (pp. 171 - 172), the EPA disagreed with the commenters' description of the
purpose of the Hg Risk TSD and the EPA asserts that the revised analysis submitted by
Petitioner 20183 (att. 4) and Petitioner 20180 (att. 5) still does not provide coverage for
subsistence fishers likely to experience elevated U.S. EGU-related Hg exposure. Based on
review of Petitioner 20183's (att. 4) and Petitioner's 20180 (att. 5) revised alternative risk
calculations submitted with the reconsideration petition, the EPA maintains that the scenario
calculations submitted for the reconsideration petition are substantively similar to the
calculations submitted during the comment period still do not meet the stated goals of the Hg
Risk TSD, which were to estimate risk for the subset of subsistence fishers expected to
experience higher-end risk (pp. 5-10). Petitioner's scenarios represent (a) typical risk for all
subsistence fishers and (b) recreational anglers. Neither of these two populations is the focus of
the Hg risk assessment. Petitioners' approach would only be relevant if the EPA were trying to
generate a representative range of risk experienced by all subsistence fishers and recreational
anglers. For these reasons, and because the EPA addressed similar comments and explained that
we were not changing our approach in light of the goals of the SAB-supported Hg risk TSD, this
issue is not of central relevance.
The EPA denies the Petition for Reconsideration on this issue because this issue was raised in
public comments and the EPA responded to those comments. Thus, Petitioners have not
demonstrated that it was impracticable to provide their analysis during the public comment
period on the proposed MATS rule. In addition, for the reasons stated above, we are also denying
reconsideration because the new information provided by the Petitioners is not of central
relevance to the "Appropriate and Necessary" finding.
1.10 MeHg bioavailability and Se/Hg ratio assumptions in Hg Risk TSD
Issue 23: Petitioner 20183 (att. 2) and Petitioner 20180 (att. 3) assert that the EPA overestimates
exposure and risk because of flawed assumptions regarding MeHg absorption. Specifically, they
assert that the EPA's assumption that 95 percent of MeHg in fish is absorbed is based on older,
poorly documented studies, and recent studies demonstrate that MeHg bio-accessibility from fish
32
-------
is typically much lower (Cabanero, et al., 2004,39 2007;40 Canuel, et al., (2006);41 He, et al.,
2010;42 Kwasniak, et al., 2012;43 Laird, et al., 2009;44 Maulvault, et al., 2011;45 Metian, et al.,
2009;46 Ouedraogo and Amyot, 2011;47 Shim, et al., 2009;48 Torres-Escribano, et al., 2011a,49
201 lb50). They also assert that the EPA did not account for factors that affect the bioavailability
of MeHg, including fish species, size, cooking method and MeHg concentration.
Response to Issue 23: Issues regarding Hg bioaccessibility and absorption were not raised
during the public comment period on the proposed MATS rule despite extensive comments on
other aspects of the Hg Risk TSD, and Petitioners have not provided a reason why they were
unable to raise these issues.
39 Cabanero Al, Madrid Y, Camara, C. 2004. "Selenium and mercury bioaccessibility in fish
samples: an in vitro digestion method." Anal Chim Acta 526:51-61.
40 Cabanero Al, Madrid Y, Camara C. 2007. "Mercury-selenium species ratio in representative
fish samples and their bioaccessibility by an in vitro digestion method." Biol Trace Elem Res
119:195-211.
41 Canuel R, Boucher de Grosbois S, Atikesse L, Lucotte M, Arp P, Ritchie C, et al., 2006. "New
evidence on variations of human body burden of methylmercury from fish consumption."
Environ Health Perspect 114:3 02-3 06.
42 He, M, Ke, CH, Wang, W.X., 2010. "Effects of cooking and subcellular distribution on the
bioaccessibility of trace elements in two marine fish species." J Agric Food Chem 58:3517-3523.
43 Kwasniak J, Falkowska L, Kwasniak M. 2012. "The assessment of organic mercury in Baltic
fish by use of an in vitro digestion model." Food Chem 132:752-758 (available online 15 Nov
2011).
44 Laird BD, Shade C, Gantner N, Man Chan H, Siciliano SD. 2009. "Bioaccessibility of mercury
from traditional northern country foods measured using an in vitro gastrointestinal model is
independent of mercury concentration." Sci Tot Environ. 407:6003-6008.
45 Maulvault AL, Machado R, Afonso C, Louren<;o HM, Nunes ML, Coelho I, Langerholc T,
Marques A. 2011. "Bioaccessibility of Hg, Cd and As in cooked black scabbard fish and edible
crab." Food Chem Tox 49: 2808-2815.
46 Metian M, Charbonnier L, Oberhaensli F, Bustamante P, Jeffree R, Amiard JC, Warnau M.
2008. "Assessment of metal, metalloid, and radionuclide bioaccessibility from mussels to human
consumers, using centrifugation and simulated digestion methods coupled with radiotracer
techniques." EcotoxEnviron Safe 72:1499-1502.
47 Ouedraogo O, Amyot M. 2011. "Effects of various cooking methods and food components on
bioaccessibility of mercury from fish." Environ Res 111: 1064-1069.
48 Shim SM, Ferruzzi MG, Kim YC, Janle EM, Santerre CR. 2009. "Impact of phytochemical-
rich foods on bioaccessibility of mercury from fish." Food Chem 112:46-50.
49 Torres-Escribano S, Denis S, Blanquet-Diot S, CalatayudM, Barrios L, Velez D, Alric M,
Montoro R. 201 la. "Comparison of a static and a dynamic in vitro model to estimate the
bioaccessibility of As, Cd, Pb and Hg from food reference materials Fucus sp. (IAEA-140/TM)
and Lobster hepatopancreas (TORT-2)." Sci Total Environ 409: January, 604-611.
50 Torres-Escribano S, Ruiz A, Barrios L, Velez D, Montoro R. 201 lb. "Influence of mercury
bioaccessibility on exposure assessment associated with consumption of cooked predatory fish in
Spain." J Sci Food Agric 91: April, 981-986.
33
-------
All of the studies cited by Petitioner were published before the public comment period with the
exception of two studies (Ouedraogo and Amyot, 2011, and Maulvault, et al., 2011), and these
two studies provide substantively similar information regarding bioaccessibility and absorption
as the previously published studies, which use in vitro digestion methods for estimation of
bioavailability, as discussed in more detail below. The other comments that the EPA received on
the Hg Risk TSD, including comments on the cooking loss factor, demonstrate that the public
had ample opportunity to comment on this issue, and the EPA responded to the comments
received in Section IF of the RTC. In the White Stallion decision, the Court found that "EPA's
'appropriate and necessary' determination in 2000, and its reaffirmation of that determination in
2012, are amply supported by EPA's findings regarding the health effects of mercury exposure."
White Stallion, 748 F.3d at 1245-46. As the Court may only consider issues raised during the
period for public comment, issues raised in the litigation and addressed by the Court clearly do
not meet the criteria for reconsideration in CAA section 307(d)(7)(B). Moreover, parties may not
use this final action denying reconsideration as a basis to litigate issues that could have been
raised in the initial litigation.
In responding to comments regarding the cooking adjustment factor (pp. 131 - 134), the EPA
addressed the issue of Hg bioaccessibility, stating that the EPA's Hg risk model for generating
IQ loss estimates, which was de-emphasized in the revised Hg Risk TSD at the suggestion of the
SAB, is currently parameterized for total Hg intake not measurements of bioaccessible Hg. In
contrast, the approach used to generate HQ estimates does not require consideration of
bioaccessibility because the risk metric is based on an estimate of the daily-ingested intake of
MeHg, which is then compared against the MeHg RfD.
The core analysis in the Hg Risk TSD was based on comparisons of Hg ingestion from fish
consumption to the U.S. EPA MeHg RfD, which does not involve estimating mercury
bioavailability and absorption of MeHg from food. Because Hg bioavailability/absorption
considerations are reflected within the RfD (U.S. EPA, 2001)51, it was not necessary to include
bioavailability as a separate factor in the HQ calculation. In contrast, issues regarding
bioavailability and absorption of MeHg from food are relevant to the IQ loss analysis, which was
de-emphasized in the revised Hg Risk TSD in response to SAB comments (U.S. EPA-SAB,
201 la, p. 2), because the IQ loss analysis does not rely on the RfD. As stated in the RTC (pp.
157 - 159), the U.S. EPA is neither reviewing nor revising its 2001 RfD for MeHg at this time.
During peer review of the RfD for MeHg, both the NRC (2001)52 and an independent peer
review panel affirmed the validity of an absorption factor for MeHg over 90 percent (Response
to RfD peer reviewers available at http://www.epa.gov/iris/supdocs/methpr.pdf). For example,
one study that the EPA used to determine the absorption rate was Miettinen, et al., (1971),53
51 U.S. Environmental Protection Agency (U.S. EPA). 2001. Integrated Risk Information System
(IRIS). Methylmercury (MeHg) (CASRN 22967-92-6). Available at http://www.epa.gov/iris/
subst/0073.htm.
52 NRC. 2000. Toxicological effects of methylmercury. Committee on the Toxicological Effects of
Methylmercury, Board on Environmental Studies and Toxicology, Commission on Life Sciences,
National Research Council. Washington, DC: National Academy Press.
53 Miettinen JK, Rahola T, Hattula T, Rissanen K, Tillander M. 1971. "Elimination of 203Hg-
methylmercury in man." Ann Clin Res. Apr 3(2): 116-22.
34
-------
which estimated that the absorption of the MeHg in cooked fish ranged from 91.2 percent to 97.0
percent with a mean of 94 percent. Contrary to Petitioner's assertions, the EPA did not cite the
Charbonneau, et al., (1976)54 study in cats in the derivation of the MeHg RfD. Although
Petitioners cite the Canuel, et al., (2006) study as evidence of the inconsistency between
measured hair Hg levels and a food frequency questionnaire, the authors state, "the relationship
between MeHg oral dose and body burden - expressed as human MeHg exposure through fish
consumption versus Hg levels in hair - may vary among certain ethnic groups." Canuel, et al.,
propose several hypotheses for this variation, including absorption, but this study did not
measure MeHg absorption as implied by Petitioner.
Further, the EPA considers that studies in humans of measured MeHg absorption from ingested
cooked fish are more relevant to risk assessment for MeHg than are in vitro studies; these in vitro
studies cited by Petitioners are laboratory tests that simulate dissolution of chemicals in the
gastrointestinal tract. Petitioner cites several in vitro studies to support their assertion of lower
bioaccessibility and absorption rates (Cabanero, et al., 2004, 2007; He, et al., 2010; Kwasniak, et
al., 2012; Laird, et al., 2009; Maulvault, et al., 2011; Metian, et al., 2009; Ouedraogo and Amyot,
2011; Shim, et al., 2009; Torres-Escribano, et al., 201 la, 201 lb). Although the EPA agrees that
these studies can provide measures of in vitro bioaccessibility that would be useful in generating
hypotheses regarding the potential for cooking methods to affect MeHg absorption from food,
these studies do not provide evidence that undermines the absorption factor that the EPA used in
the derivation of the RfD for MeHg. Therefore, this issue is not of central relevance to the
"Appropriate and Necessary" finding. In addition, as stated above, all but two of these studies
were published prior to the public comment period.
Petitioners have not demonstrated that it was impracticable to comment on these issues during
the public comment period on the proposed MATS rule. The fact that other commenters
commented on similar issues during the comment period demonstrates that it was not
impracticable to comment during the comment period. Further, even if the two new studies had
been available prior to the end of the public comment period, the cited studies are not of central
relevance to the "Appropriate and Necessary" finding. For all of these reasons, the EPA is
denying the petitions for reconsideration on this issue.
Issue 24: Petitioner 20183 (att. 2) and Petitioner 20180 (att. 3) assert that the EPA overestimates
exposure and risk because of flawed assumptions regarding varying selenium (Se)/Hg ratios.
Specifically, Petitioners assert that Se/Hg ratios may be higher in the smaller fish with lower Hg
levels that are modeled in the Hg Risk TSD. Petitioner cites the Burger and Gochfeld (2012)55
study as showing that the ratio between Se and Hg in fish varies with fish size, with lower ratios
as fish size increases. Based on this argument, Petitioner concludes that Hg exposures may be
reduced in subsistence populations that consume smaller fish.
54 Charbonneau S, Munro I, Nera E, et al., 1976. "Chronic toxicity of methylmercury in the adult
cat." Interim Report Tox 5:337-349.
55 Burger J and Gochfeld M. 2012. "Selenium and mercury molar ratios in saltwater fish from
New Jersey: Individual and species variability complicate use in human health fish consumption
advisories." Environ Res. 114, 12-23.
35
-------
Response to Issue 24: Issues regarding fish size-dependent Se and Hg ratios were not raised
during the public comment period on the proposed MATS rule despite extensive comments on
other aspects of the Hg Risk TSD, including the role of Se in MeHg toxicity (p. 166), and
Petitioners have not provided a reason why they were unable to raise these issues.
However, the SAB did discuss the more general issue of the potential interaction between Se and
Hg in fish during the peer review public meeting. Specifically, the SAB panel discussed the
potential for amelioration of MeHg effects by co-exposure to Se. After consideration of studies
on Se and Hg as well as other nutrients in U.S. fish, the SAB concluded that nutrients such as Se
could mitigate the concentration-effect relationship, but there is not enough information to
conduct a quantitative analysis of this impact (U.S. EPA-SAB, 201 la, EPA-HQ-OAR-2009-
0234-19689, p. 20). The SAB further noted, "[s]ince the subsistence fish consumers that form the
focus of this study are at notable risk of having poor nutrition, Hg exposures may be non-linearly
related to toxicity risks. Other SAB members note that effects of Se on MeHg toxicity are based
primarily on observations in animals, and there is disagreement in the scientific community
regarding the significance of these observations to humans." (p. 21). In addition, in the RTC (pp.
162 - 163), the EPA addressed the potential for nutrients such as Se in fish (particularly marine
fish) to ameliorate some of the observed adverse effects of MeHg when co-exposure occurs.
Studies indicate that it is more appropriate to determine potential amelioration of MeHg effects
by assessing co-exposure to MeHg and nutrients in people, not by the relative concentrations of
MeHg and nutrients in the fish they consume. Petitioners have provided no data to indicate that
humans consuming MeHg in fish have reduced exposure or no measured exposure to MeHg as a
consequence of the Se also found in the fish.
Regarding the conclusions of the new Burger and Gochfeld (2012) study, Petitioners fail to note
that this study found substantial variation in these ratios among saltwater fish species and that
several fish species (i.e., tautog, sea bass, cunner, summer and winter flounder, weakfish, scup)
did not show a lower ratio in smaller fish. Further, another new study (Burger, et al., 2012)56 by
the same authors on freshwater fish in Tennessee found no relationship between the Se/Hg ratio
and fish size. The authors concluded, "it would be difficult to use the molar ratio in predicting
either the risk from Hg toxicity or in devising consumption advisories." The authors noted the
inconsistency in the results from their two studies without providing a rationale for the
inconsistency. The Hg Risk TSD is focused on freshwater fish consumption. For these reasons,
this issue is not of central relevance to the "Appropriate and Necessary" finding.
Because Petitioners have not demonstrated that it was impracticable to comment on this issue
during the public comment period on the proposed MATS rule, and because the issue is not of
central relevance, the EPA is denying the petitions for reconsideration of this issue.
Issue 25: Petitioner 20183 (att. 2) and Petitioner 20180 (att. 3) assert that the EPA assumed a
linear dose-response relationship for MeHg in the derivation of the RfD.
56 Burger, Joanna, Michael Gochfeld, C. Jeitner, M. Donio, and T. Pittfield. 2012 "Selenium:
Mercury Molar Ratios in Freshwater Fish from Tennessee: Individual, Species, and
Geographical Variations have Implications for Management." EcoHealth DOI: 10.1007/sl0393-
012-0761-y
36
-------
Response to Issue 25: Issues regarding the derivation and use of the RfD were raised in public
comments submitted in response to the proposed MATS rule. The EPA responded to these
comments in Section IF (pp. 157 - 159) of the RTC.
As stated in the RTC (pp. 157 - 159), the U.S. EPA is neither reviewing nor revising its 2001
RfD for MeHg at this time. In the White Stallion decision, the Court found that "EPA's
'appropriate and necessary' determination in 2000, and its reaffirmation of that determination in
2012, are amply supported by EPA's findings regarding the health effects of mercury exposure."
White Stallion, 748 F.3d at 1245-46. As the Court may only consider issues raised during the
period for public comment, issues raised in the litigation and addressed by the Court clearly do
not meet the criteria for reconsideration in CAA section 307(d)(7)(B). Moreover, parties may not
use this final action denying reconsideration as a basis to litigate issues that could have been
raised in the initial litigation.
Further, Petitioner's assertion regarding the basis of the RfD is incorrect. As noted in the RTC
(pp. 150 - 151 and 158), a K-power model with K = 1 (a linear model) provided the best fit to the
data; thus, the EPA made no assumptions about the shape of the dose-response relationship for
MeHg health effects. Because the RfD was not revised in the proposed or final MATS rule and
the Petitioners provide no new information that, if accepted, could have changed the EPA's
previous response or Hg risk TSD, this issue is not of central relevance to the "Appropriate and
Necessary" finding.
Because Petitioners have not demonstrated that it was impracticable to comment on this issue
during the public comment period on the proposed MATS rule, and because this issue is not of
central relevance to the "Appropriate and Necessary" finding, the EPA is denying the petition for
reconsideration of this issue.
1.11 Health benefits of fish consumption
Issue 26: Petitioners 20183 (att. 2) and Petitioner 20180 (att. 3) assert that the EPA should use
caution assessing MeHg risks in order to not discourage the substantial health benefits from fish
consumption without a commensurate reduction in Hg-related health risks.
Response to Issue 26: Issues regarding the benefits of fish consumption were raised in public
comments submitted in response to the proposed MATS rule. The EPA responded to these
comments in Section IF (p. 146) of the RTC.
These comments demonstrate that the public had ample opportunity to comment on this issue,
and the EPA responded to those comments. In the White Stallion decision, the Court found that
"EPA's 'appropriate and necessary' determination in 2000, and its reaffirmation of that
determination in 2012, are amply supported by EPA's findings regarding the health effects of
mercury exposure." White Stallion, 748 F.3d at 1245-46. As the Court may only consider issues
raised during the period for public comment, issues raised in the litigation and addressed by the
Court clearly do not meet the criteria for reconsideration in CAA section 307(d)(7)(B).
37
-------
Moreover, parties may not use this final action denying reconsideration as a basis to litigate
issues that could have been raised in the initial litigation.
Specifically, the EPA noted that "many studies show that beneficial effects of fish on both
cardiovascular and neurodevelopmental health are decreased by concomitant exposure to
MeHg." This means that reducing Hg in fish both reduces Hg-related risk and likely increases
the beneficial aspects of fish consumption. Because the purpose of the Hg Risk TSD was to
evaluate risk from EGU attributable Hg, the identified concerns about the Hg Risk TSD
potentially causing the public to reduce fish consumption is not of central relevance.
Because Petitioners have not demonstrated that it was impracticable to comment on this issue
during the public comment period on the proposed MATS rule, as evidenced by the comments
received on this issue, and because the issue is not of central relevance, the EPA is denying the
petition for reconsideration of this issue.
1.12 Assumptions regarding fish advisories in Hg Risk TSD
Issue 27: Petitioner 20183 (att. 2) and Petitioner 20180 (att. 3) observe that the EPA stated in the
RTC (p. 142) that because some states have set minimum size limits for catches, consumption of
fish shorter than seven inches would violate state or local fish and game laws. Petitioners assert
that if the EPA assumes that subsistence fishers observe such laws, it would be appropriate for
the EPA to assume that they also observe fish advisories.
Response to Issue 27: Issues regarding fish advisories and fish size were raised in public
comments submitted in response to the proposed MATS rule. The EPA responded to these
comments in Section IF (pp. 166 - 168) of the RTC.
These comments demonstrate that the public had ample opportunity to comment on the issue,
and the EPA responded to those comments. Although Petitioner mentions this issue specifically
in the context of the EPA's response in the RTC, Petitioner has provided no evidence that it was
impracticable to comment on this issue during the public comment period on the proposed
MATS rule. Furthermore, in the White Stallion decision, the Court found that "EPA's
'appropriate and necessary' determination in 2000, and its reaffirmation of that determination in
2012, are amply supported by EPA's findings regarding the health effects of mercury exposure."
White Stallion, 748 F.3d at 1245-46. As the Court may only consider issues raised during the
period for public comment, issues raised in the litigation and addressed by the Court clearly do
not meet the criteria for reconsideration in CAA section 307(d)(7)(B). Moreover, parties may not
use this final action denying reconsideration as a basis to litigate issues that could have been
raised in the initial litigation.
In addition, as noted in the RTC (p. 142), the focus of the Hg Risk TSD is on modeling risk for
the subset of subsistence fishers whose behavior places them at greater risk. That behavior
includes higher consumption rates as well as targeting fish that are larger. Although some
subsistence fishers could consume fish without consideration for size, some subsistence fishers
likely target larger fish to maximize the volume of edible fish. The EPA excluded fish smaller
than 7 inches to reflect the minimum size limit for a number of key edible freshwater fish and
38
-------
took the 75th percentile to reflect this focus on subsistence fishers who target larger fish to
maximize their diet. The EPA cited the fact that states such as Pennsylvania use 7-inches as a
minimum catch size for key edible freshwater fish species (Pennsylvania Fish and Boating
Commission, 2011),57 as an example to support a specific fish size assumption in the Hg Risk
TSD, not as an argument (as suggested by Petitioner) that subsistence fishers necessarily follow
fishing regulations. In addition, Petitioner neglects to mention that minimum catch sizes are
generally codified in state laws or regulations, which can have legal consequences for
noncompliance, whereas disobeying voluntary fish advisories has no legal consequences. In
addition, the SAB supported the Agency's approach in the Hg Risk TSD.
Because Petitioners have not demonstrated that it was impracticable to comment on this issue
during the public comment period on the proposed MATS rule, the EPA is denying the petitions
for reconsideration of this issue. In any case, the issue is not of central relevance for the reasons
set forth above.
1.13 Rounding of EGU contribution thresholds in Hg Risk TSD
Issue 28: Petitioner 20183 (att. 2) and Petitioner 20180 (att. 3) allege that the EPA changed the
rounding convention for determining the Hg contribution from U.S. EGUs between the draft and
revised Hg Risk TSDs, thus this change was not available for public comment.
Response to Issue 28: In the review of the draft Hg Risk TSD, the SAB asked EPA to "discuss
the basis for selecting a HQ at or above 1.5 as the criteria for selecting potentially impacted
watersheds" (U.S. EPA-SAB, 201 la). In preparing this discussion, the EPA confirmed that the
appropriate level of precision for the HQ should be determined by the level of precision in the
input data, specifically, that of the RfD, which is reported at one significant figure. In Section
1.4.5 of the revised Hg Risk TSD, the EPA explains the use of an HQ of 2 or greater as an
indication that a watershed has populations potentially at risk. Specifically, in the risk
calculations, an HQ of 2 is equivalent to an HQ of 1.5 or greater, because any HQ of 1.5 or
greater is rounded to an HQ of 2. For consistency throughout the risk calculations, the EPA
applied the same rounding convention throughout the remainder of the revised Hg Risk TSD
(i.e., one significant figure) in the calculation of the percent contribution from U.S. EGU. For
example, any watershed with 4.5 percent or greater would round up to 5 percent. Further, in the
preamble to the proposed rule (76 FR 25010), the EPA stated, "Any contribution of Hg
emissions from EGUs to watersheds where potential exposures from total Hg deposition exceed
the RfD is a hazard to public health, but for purposes of our analyses we evaluated only those
watersheds where we determined EGUs contributed 5 percent or more to deposition to the
watershed. The EPA believes this is a conservative approach given the increasing risks
associated with incremental exposures above the RfD."
Although the EPA slightly revised the specific rounding convention applied in the contribution
thresholds, it did not change the overall approach used to calculate the EGU attributable
57 Pennsylvania Fish and Boat Commission. 2011. Summary Book: 2001 Pennsylvania Fishing
Laws and Regulations. Available on the Internet at
http: //fi shandb oat. com/fi shpub/summary/inl and. html.
39
-------
contribution to each watershed. Specifically, the draft Hg Risk TSD applied a rounding
convention that the EPA documented in the Hg Risk spreadsheets in the docket (EPA-HQ-OAR-
2009-0234-3079), and the EPA received comments on other rounding conventions applied in the
draft Hg Risk TSD from the SAB. Therefore, this revision was a logical outgrowth of the
comments received from the SAB during the public comment period. See NRDC v. Thomas, 838
F.2d 1224, 1242 (D.C. Cir. 1988) and Small Refiner Lead Phase-Down Task Force v. EPA, 705
F.2d at 547 (agency may make changes to proposed rule without triggering new round of
comments, where changes are logical outgrowth of proposal and comments). Because the
revision did not change the manner of calculating the contribution or the conclusions of the Hg
Risk TSD, the EPA does not believe it was necessary to provide additional opportunity to
comment on this issue.
Although this change had a small effect on the estimated percent of watersheds with populations
potentially at risk where the RfD is exceeded from total Hg deposition, the EPA determined that
the percent of watersheds would remain unacceptable under either rounding convention because
it would not substantially change the results. In addition, the rounding convention for the percent
of contribution from U.S. EGU emissions has no impact on the other risk metric: the percent of
watersheds where potential exposures exceed the RfD from U.S. EGUs alone. Specifically, the
revised Hg Risk TSD found that up to 10 percent of the watersheds exceed the RfD due to U.S.
EGU emissions alone even before considering other sources of Hg deposition. Because this
estimate is not affected by the percent contribution from U.S. EGU emissions, it also is not
affected by the rounding convention. As stated in the preamble to the final MATS rule (76 FR
93111), this result "independently supports our conclusion that Hg emissions from U.S. EGUs
pose hazards to public health." Therefore, for either risk metric, the rounding convention is not
of central relevance to the "Appropriate and Necessary" finding.
Because the EPA received comments on rounding conventions used in the draft Hg Risk TSD
and Petitioners have not demonstrated that it was impractical for them to comment on this
rounding convention, we are denying the petition for reconsideration on this issue. In addition,
the change in rounding convention was a logical outgrowth of the proposed rule and the
comments received on the draft Hg Risk TSD and the Petitioner has provided no new data or
information that calls into question the underlying determination. The EPA is also denying
reconsideration because the issue is not of central relevance to the "Appropriate and Necessary"
finding as an unacceptable level of risk exists under either rounding convention.
Issue 29: Petitioner 20183 (att. 2) and Petitioner 20180 (att. 3) allege that the change in the
rounding convention for determining the Hg contribution from U.S. EGUs would reduce the
number of watersheds that exceed 5 percent Hg contribution from U.S. EGUs. Specifically,
Petitioner provides an analysis that shows that retaining the rounding convention used in the
draft Hg Risk TSD would decrease the percentage of watersheds with populations that are
potentially at-risk due to total deposition from 22 percent in the draft Hg Risk TSD to 20 percent
instead of increasing to 24 percent as concluded by the EPA. In addition, Petitioner claims that
retaining the rounding convention used in the draft Hg Risk TSD would decrease the percent of
at-risk-watersheds from 28 percent in the draft Hg Risk TSD to 26 percent instead of increasing
to 29 percent as concluded by the EPA.
40
-------
Response to Issue 29: As fully explained in the response to Issue 28, the SAB asked EPA to
"discuss the basis for selecting a HQ at or above 1.5 as the criteria for selecting potentially
impacted watersheds" (U.S. EPA-SAB, 201 la). For consistency throughout the risk calculations,
at final the EPA applied the same rounding convention throughout the remainder of the revised
Hg Risk TSD (i.e., one significant figure) in the calculation of the percent contribution from U.S.
EGU. For all the reasons stated in response to Issue 28, we are denying the petitions for
reconsideration of this issue.
In addition, although the change in the rounding convention had a small effect on the estimated
percent of watersheds with populations potentially at risk where the RfD is exceeded from total
Hg deposition, the EPA determined that the percent of watersheds would remain unacceptable
under either rounding convention. Further, in the preamble to the proposed rule (76 FR 25010),
the EPA stated, "Any contribution of Hg emissions from EGUs to watersheds where potential
exposures from total Hg deposition exceed the RfD is a hazard to public health, but for purposes
of our analyses we evaluated only those watersheds where we determined EGUs contributed 5
percent or more to deposition to the watershed. The EPA believes this is a conservative approach
given the increasing risks associated with incremental exposures above the RfD." In addition, the
rounding convention for the percent of contribution from U.S. EGU emissions has no impact on
the other risk metric: the percent of watersheds where potential exposures exceed the RfD from
U.S. EGUs alone. Because this estimate is not affected by the percent contribution from U.S.
EGU emissions, it also it also is not affected by the rounding convention. Therefore, for either
risk metric, the rounding convention is not of central relevance to the "Appropriate and
Necessary" finding. For all these reasons, the EPA is denying the petitions for reconsideration of
this issue.
Issue 30: Petitioner 20183 (att. 2) and Petitioner 20180 (att. 3) allege that the rounding
convention that the EPA applied in the Hg Risk TSD is arbitrary because the EPA did not
discuss the reason for applying a rounding convention, the rounding convention itself, or the
reason for changing the rounding convention between the proposed and final MATS rules.
Response to Issue 30: As fully explained in the response to Issues 28 and 29, the SAB (p. 8)
asked EPA to "discuss the basis for selecting a HQ at or above 1.5 as the criteria for selecting
potentially impacted watersheds" (U.S. EPA-SAB, 201 la). For consistency throughout the risk
calculations, at final the EPA applied the same rounding convention throughout the remainder of
the revised Hg Risk TSD (i.e., one significant figure) in the calculation of the percent
contribution from U.S. EGU. For all the reasons set forth in responses to Issues 28 and 29, we
are denying the petitions for reconsideration on this issue.
1.14 Modeling Hg emissions from final CSAPR in Hg Risk TSD
Issue 31: Petitioner 20183 (att. 3) and Petitioner 20180 (att. 4) claim that the Hg Risk TSD
overestimates risk because the EPA did not re-estimate Hg deposition based on updated Hg
emissions from the final Cross-State Air Pollution Rule (CSAPR) (formerly called the
"Transport Rule") and issued under CAA section 110.
41
-------
Response to Issue 31: The EPA Administrator signed the final CSAPR on July 6, 2011, which
was during the public comment period for MATS. The EPA included emission estimates
reflecting the proposed CSAPR in the Hg deposition modeling for the Hg Risk TSD. Issues
regarding the inclusion of the CSAPR emissions characterizations in the Hg deposition modeling
were raised in public comments submitted in response to the proposed MATS rule, and the EPA
responded to those comments in Section IB (pp. 43 - 48) of the RTC.
Because Petitioners have not demonstrated that it was impracticable to comment on this issue
during the public comment period on the proposed MATS rule, the EPA is denying the petitions
for reconsideration of this issue. Furthermore, in the White Stallion decision, the Court found that
"EPA's 'appropriate and necessary' determination in 2000, and its reaffirmation of that
determination in 2012, are amply supported by EPA's findings regarding the health effects of
mercury exposure." White Stallion, 748 F.3d at 1245-46. As the Court may only consider issues
raised during the period for public comment, issues raised in the litigation and addressed by the
Court clearly do not meet the criteria for reconsideration in CAA section 307(d)(7)(B).
Moreover, parties may not use this final action denying reconsideration as a basis to litigate
issues that could have been raised in the initial litigation.
As stated in the preamble to the final MATS rule (77 FR 9316), although CSAPR would likely
achieve some Hg co-benefits from controls on sulfur dioxide (SO2) emissions, the revised Hg
Risk TSD indicates that an unacceptable percentage of modeled watersheds with populations
potentially at-risk from U.S. EGU-attributable Hg deposition would remain after implementation
of CSAPR. As stated in the preamble to the final MATS rule (77 FR 9316), modeling emissions
based on the final CSAPR (in conjunction with additional updates to the emissions modeling)
would not result in a substantial change in risk, because the final CSAPR would reduce Hg
emissions by 2 tons nationally - a reduction of less than 10 percent - compared to the 29 tons of
Hg emissions used in the Hg risk modeling. The EPA concluded that this would not result in a
substantial change in risk. Further, as noted in the preambles to the proposed and final MATS
rules (76 FR 25017, 77 FR 9311), owners and operators of EGUs can change their criteria
pollutant compliance strategies and use methodologies that do not achieve the same level of Hg
or other HAP co-benefit, and the only way to ensure permanent reductions in HAP emissions
from U.S. EGUs and the associated risks to public health and the environment is through
standards set under CAA section 112, such as MATS.
In addition, Petitioners also assert that the impact from CSAPR is likely much greater when
considering individual EGUs. However, Petitioner provides no specific information to support
this assertion and, as such, the assertion is speculation. As discussed in the revised Hg Risk TSD
(pp. 5 - 7), the EPA evaluated the nature of risks likely to result from Hg emissions from U.S.
EGUs and determined that a national-scale risk assessment was the appropriate scale of analysis
for the "Appropriate and Necessary" finding, in part because of the regional nature of Hg
transport and deposition. The specific impacts on individual EGUs of small changes in Hg
emissions are not likely to be significant determinants of national risk estimates.
Because the EPA received comments on the CSAPR modeling issues, and the EPA responded to
those comments, we are denying reconsideration on this issue. The EPA is also denying
reconsideration because the issue is not of central relevance for the reasons set forth above.
42
-------
1.15 Boundary conditions in Hg deposition modeling
Issue 32: Petitioner 20180 states that a new study (Baker and Bash, 2012),58 which was
published after the public comment period closed, suggests that non-U.S. emissions may
contribute to overestimation of Hg deposition in some locations in the MATS modeling. Further,
Petitioner 20180 (att. 6) also notes that this study hypothesized that too much Hg may be flowing
into the model from outside the model boundaries in certain areas of the model domain (i.e., the
continental U.S. including northern Mexico and southern Canada), which Petitioner contends
would lead to a mischaracterization of Hg from inter-hemispheric transport into the U.S.
Response to Issue 32: The Baker and Bash study does not present any new information because
the Hg deposition modeling presented in that study is the same as the Hg deposition modeling
conducted for the Hg Risk TSD. In fact, both authors of the study are EPA staff. Although the
publication of the study describing the modeling was published after the public comment period,
all the Hg deposition modeling inputs and outputs for the Hg modeling assessment in the study
were available in the docket (EPA-HQ-OAR-2009-0234-2989) as part of the proposed MATS
rule. The EPA did not generate any new Hg deposition modeling using the Community Multi-
scale Air Quality (CMAQ) model for the Baker and Bash study. Contrary to Petitioner's
assertions, the Baker and Bash study does not quantitatively evaluate mercury concentrations at
the boundary of the model domain. Further, lowering the amount of Hg flowing in from the
boundaries would increase the percentage of Hg deposition from U.S. EGUs, which would
strengthen the conclusions of the Hg Risk TSD. Because the information in the study is the same
as the Hg deposition modeling for MATS and considering the study would not lead to any
changes in the Hg deposition modeling even if it had been published prior to the end of the
public comment period, this issue is not of central relevance to the "Appropriate and Necessary"
finding.
The public had ample opportunity to review the Hg deposition modeling inputs and outputs in
the docket for the proposed MATS rule, and the EPA received comments on the modeling during
the public comment period. Because Petitioner has not demonstrated that it was impracticable to
comment on the modeling, including the boundary conditions, during the public comment period
on the proposed MATS rule, the EPA is denying reconsideration of this issue. Furthermore, for
the reasons discussed above, the issues raised in the study cited by the Petitioner are not of
central relevance to the "Appropriate and Necessary" finding, and we are denying the petition for
reconsideration for that reason as well.
Issue 33: Petitioner 20180 claims that the EPA inadequately addressed Hg emissions from China
in the boundary conditions for modeling Hg deposition. Petitioner 20180 (att. 6) asserts that the
EPA should conduct an additional sensitivity analysis for Hg deposition modeling for variable
boundary conditions due to changes in global Hg emissions.
58 Baker, K.R., Bash, J.O., 2012. "Regional scale photochemical model evaluation of total
mercury wet deposition and speciated ambient mercury." Atmospheric Environment 49: 151-162.
43
-------
Response to Issue 33: Issues regarding the EPA's treatment of non-U.S. Hg emissions were
raised in public comments submitted in response to the proposed MATS rule. The EPA
responded to these comments in Section ID (pp. 93 - 100) and in Section IE (pp. 101 - 108) of
the RTC.
Furthermore, in the While Stallion decision, the Court found that "EPA's 'appropriate and
necessary' determination in 2000, and its reaffirmation of that determination in 2012, are amply
supported by EPA's findings regarding the health effects of mercury exposure." White Stallion,
748 F.3d at 1245-46. As the Court may only consider issues raised during the period for public
comment, issues raised in the litigation and addressed by the Court clearly do not meet the
criteria for reconsideration in CAA section 307(d)(7)(B). Moreover, parties may not use this
final action denying reconsideration as a basis to litigate issues that could have been raised in the
initial litigation. For these reasons, we are denying reconsideration.
During their peer review, the SAB raised the issue of uncertainties in the global Hg emissions
and supported a qualitative discussion of these uncertainties (U.S. EPA-SAB 2011, pp. 22 and
26). The SAB did not request a sensitivity analysis for global Hg emissions. The EPA fully
responded to each of the SAB recommendations, including describing the boundary conditions
used in the Hg deposition modeling and the associated uncertainties.
The public had ample opportunity to comment on this issue and EPA did, in fact, receive such
comments and the EPA responded to those comments. Because Petitioner has not demonstrated
that it was impracticable to comment on this issue during the public comment period on the
proposed MATS rule, the EPA is denying the petition for reconsideration of this issue.
Issue 34: Petitioner 20180 (att. 6) asserts that the RIA for MATS acknowledges that
international Hg emissions have been increasing.
Response to Issue 34: Issues regarding the EPA's treatment of non-U.S. Hg emissions were
raised in public comments submitted in response to the proposed MATS rule. The EPA
responded to these comments in Section ID (pp. 93 - 100) and in Section IE (pp. 101 - 108) of
the RTC.
The comments that the EPA received on these issues demonstrate that the public had ample
opportunity to comment and the EPA responded to those comments. Furthermore, in the White
Stallion decision, the Court found that "EPA's 'appropriate and necessary' determination in
2000, and its reaffirmation of that determination in 2012, are amply supported by EPA's findings
regarding the health effects of mercury exposure." White Stallion, 748 F.3d at 1245-46. As the
Court may only consider issues raised during the period for public comment, issues raised in the
litigation and addressed by the Court clearly do not meet the criteria for reconsideration in CAA
section 307(d)(7)(B). Moreover, parties may not use this final action denying reconsideration as
a basis to litigate issues that could have been raised in the initial litigation. The EPA is denying
the petition for reconsideration for these reasons.
Further, Petitioner misinterprets the context of the EPA's statement in the RIA that "international
emissions have increased." This statement refers to the general trend of increasing global Hg
44
-------
emissions during previous years, not a potential change in international emissions from the
current year towards the future. In addition, Petitioner has only provided information about
changes in coal use in China, but the Petitioner has not provided additional information to
support an assertion that international Hg emissions will increase into the future. For this reason,
we are also denying reconsideration because the issue as presented is not of central relevance.
Because Petitioner has not demonstrated that it was impracticable to comment on this issue
during the public comment period on the proposed MATS rule and because the issue is not of
central relevance, the EPA is denying the petition for reconsideration of this issue.
Issue 35: Petitioner 20183 (att. 3) and Petitioner 20180 (att. 4) contend that the Hg boundary
conditions used in the Hg deposition modeling for the Hg Risk TSD do not correctly account for
Hg emissions from China, which underestimates the Hg boundary conditions. Petitioners also
claim that a study (Jaffe, et al., 2005)59 shows that Chinese emissions have a large impact on Hg
deposition in the U.S.
Response to Issue 35: Issues regarding the EPA's treatment of non-U.S. Hg emissions were
raised in public comments submitted in response to the proposed MATS rule. The EPA
responded to these comments in Section ID (pp. 93 - 100) and in Section IE (pp. 101 - 108). The
comments that the EPA received on these issues demonstrate that the public had ample
opportunity to comment and the EPA responded to those comments. Furthermore, in the White
Stallion decision, the Court found that "EPA's 'appropriate and necessary' determination in
2000, and its reaffirmation of that determination in 2012, are amply supported by EPA's findings
regarding the health effects of mercury exposure." White Stallion, 748 F.3d at 1245-46. As the
Court may only consider issues raised during the period for public comment, issues raised in the
litigation and addressed by the Court clearly do not meet the criteria for reconsideration in CAA
section 307(d)(7)(B). Moreover, parties may not use this final action denying reconsideration as
a basis to litigate issues that could have been raised in the initial litigation. The EPA is denying
the petition for reconsideration for these reasons.
In addition, the EPA disagrees with Petitioners' assertion because Petitioners have provided no
data to support it. Specifically, Petitioners do not provide any quantitative analysis of the EPA's
boundary conditions that compares CMAQ-modeled Hg concentration and deposition near the
boundary when airflow is coming from outside the model domain with Hg measurements.
Petitioners' claim that boundary conditions do not have high enough Hg concentrations is
speculative and not based on any data.
Further, Petitioner misinterprets the results of the Jaffe, et al., study, the methodology employed,
and the relevance to the entire continental U.S. The Jaffe, et al., study found a connection
between Hg measured at one place in the western United States at an extreme elevation (Mt.
Bachelor, Oregon) and the continent of Asia, not specifically China. The authors of the Jaffe, et
59 Jaffe, D., Prestbo, E., Swartzendruber, P., Weiss-Penzias, P., Kato, S., Takami, A.,
Hatakeyama, S., Kajii, Y., 2005. "Export of atmospheric mercury from Asia ."Atmospheric
Environment 39, 3029-3038.
45
-------
al., study acknowledge their techniques did not have the precision to differentiate the origin of
Hg from specific countries.
The EPA has concluded that given the lack of new information provided by Petitioner and, as
discussed in the RTC (pp. 94 - 95), the similarity in continental emissions totals between 2000
and 2006 in Streets, et al., (2009),60 the boundary conditions used in the Hg deposition modeling
are appropriate. Because Petitioner has not provided any new information that would, if accepted
by EPA, change the results of the analysis, this objection is also not of central relevance.
Because Petitioners have not demonstrated that it was impracticable to comment on this issue
during the public comment period on the proposed MATS rule, and because the issue is not of
central relevance for the reasons set forth above, the EPA is denying the petition for
reconsideration of this issue.
Issue 36: Petitioner 20180 (att. 6) asserts the EPA did not account for increasing global Hg
emissions in the future associated with higher coal use in China and provides several citations
intended to support this assertion (Shen, et al., 2012;61 Wang, et al., 2010;62 Li and Leung,
201263).
Response to Issue 36: Issues regarding the EPA's treatment of non-U.S. Hg emissions were
raised in public comments submitted in response to the proposed MATS rule. The EPA
responded to these comments in Section ID (pp. 93 - 100) and in Section IE (pp. 101 - 108) of
the RTC. Furthermore, in the White Stallion decision, the Court found that "EPA's 'appropriate
and necessary' determination in 2000, and its reaffirmation of that determination in 2012, are
amply supported by EPA's findings regarding the health effects of mercury exposure." White
Stallion, 748 F.3d at 1245-46. As the Court may only consider issues raised during the period for
public comment, issues raised in the litigation and addressed by the Court clearly do not meet the
criteria for reconsideration in CAA section 307(d)(7)(B). Moreover, parties may not use this
final action denying reconsideration as a basis to litigate issues that could have been raised in the
initial litigation.
The citations that Petitioner provides to support their assertion omit important concepts and
information. Specifically, long-term measurements taken in the northern hemisphere since 1995
show, at the majority of sites measured, steadily decreasing total Hg concentrations (Slemr, et
60 Streets, D.G., Zhang, Q., Wu, Y., 2009. "Projections of Global Mercury Emissions in 2050."
Environmental Science & Technology 43, 2983-2988.
61 Shen L, Gao T-m, Cheng X. 2012. "China's coal policy since 1979: A brief overview." Energy
Policy 40: 274-281.
62 Wang SX, Zhang L, Li GH, Wu Y, Hao JM, Pirrone N, Sprovieri F, Ancora MP. 2010,
"Mercury emission and speciation of coal-fired power plants in China." Atmos. Chem. Phys. 10,
1183-1192, 2010.
63 Li R, Leung GCK. 2012. "Coal consumption and economic growth in China." Energy Policy
40 (2012) 438-443.
46
-------
al., 2011),64 results that are as yet unexplained in the scientific literature. Moreover, monitoring
data taken in the U.S. since 1996 show no trend in increased Hg wet deposition (Presbo and Gay,
2009).65 Additionally, there are uncertainties in estimates of Chinese emissions and only
considering coal use without accounting for emissions control implementation would lead to an
erroneous conclusion about the magnitude of potential growth in emissions. Improvements are
being made in efficiency and in application of technology that are offsetting at least part of the
associated increase in emissions. For example, application of emission controls for particulate
matter and SO2 in China has reportedly resulted in co-benefit emissions reduction of power plant
Hg emissions in recent years despite an increase in coal use (Tian, et al., 2012;66 Wang, et al.,
201267).
Although Petitioner cites two new studies (Shen, et al., 2012; Li and Leung, 2012), these studies
still do not provide information regarding how changes in coal use in China would necessarily
increase Hg deposition in the continental U.S. for the scenario that the EPA modeled. For these
reasons, this issue is not of central relevance to the "Appropriate and Necessary" finding.
The comments that the EPA received on these issues demonstrate that the public had ample
opportunity to comment and the EPA responded to those comments. Petitioner has not
demonstrated that it was impracticable to comment on this issue during the public comment
period on the proposed MATS rule. Further, for the reasons set forth above, this issue is not of
central relevance to the "Appropriate and Necessary" finding. For all these reasons, the EPA is
denying the petition for reconsideration of this issue.
1.16 Chemical reactions in the Hg deposition modeling
Issue 37: Petitioner 20183 (att. 3), Petitioner 20180 (att. 4), and Petitioner 20180 (att. 6) contend
that chemical reactions take place in plumes downwind from coal-fired EGUs that convert the
fast depositing form of Hg (i.e., divalent Hg or Hg(II)) to the slower depositing form of Hg (i.e.,
elemental Hg or Hg(0)) and that the EPA should have used these reactions in the Hg deposition
modeling. Petitioners assert that including these reactions in the model would convert some of
the local Hg deposition shown in EPA's modeling into regional or continental Hg deposition.
64 Slemr, F., Brunke, E.G., Ebinghaus, R., Kuss, J., 2011. "Worldwide trend of atmospheric
mercury since 1995 ''Atmospheric Chemistry and Physics 11, May, 4779-4787.
65 Prestbo, E.M. and Gay, D.A. 2009. "Wet deposition of mercury in the US and Canada 1996-
2005: Results and Analysis of the NADP mercury deposition network (MDN)," Atmos. Environ.
43, 4332-4233.
66 Tian, Hezhong, Yan Wang, Ke Cheng, Yiping Qu, Jiming Hao, Zhigang Xue, and Fahe Chai.
2012. "Control strategies of atmospheric mercury emissions from coal-fired power plants in
China." Journal of the Air & Waste Management Association 62(5):576-586. DOI:
10.1080/10962247.2012.663733
67 Wang, Shuxiao, Lei Zhang, Bin Zhao, Yang Meng, and Jiming Hao. 2012. "Mitigation
Potential of Mercury Emissions from Coal-Fired Power Plants in China." Energy Fuels 26,
4635-4642. dx.doi.org/10.1021/ef201990x. Publication Date (Web): February 21, 2012.
47
-------
Petitioner provides several citations to support this assertion, including Kolker, et al., (2010),68
Lohman, et al., (2006),69 Amos, et al., (2012), 70 Landis, et al., (2009),71 and ter Schure, et al.,
(2011).72 In addition, Petitioner 20180 (att. 6) contends that the EPA's Hg deposition modeling
should have used "Advanced Plume Treatment," which would have included these reactions.
Response to Issue 37: Issues regarding the EPA's treatment of in-plume Hg chemistry in the Hg
deposition modeling were raised in public comments submitted in response to the proposed
MATS rule. The EPA responded to these comments in Section IE (pp. 107 - 111) of the RTC,
which include discussions of Petitioner's citations including Kolker, et al., (2010), Lohman, et
al., (2006), and Landis, et al., (2009). Furthermore, in the White Stallion decision, the Court
found that "EPA's 'appropriate and necessary' determination in 2000, and its reaffirmation of
that determination in 2012, are amply supported by EPA's findings regarding the health effects
of mercury exposure." White Stallion, 748 F.3d at 1245-46. As the Court may only consider
issues raised during the period for public comment, issues raised in the litigation and addressed
by the Court clearly do not meet the criteria for reconsideration in CAA section 307(d)(7)(B).
Moreover, parties may not use this final action denying reconsideration as a basis to litigate
issues that could have been raised in the initial litigation.
Although Petitioners provided several citations to support their assertions that in-plume chemical
reactions would change divalent Hg to elemental Hg and, thus, alter the Hg deposition,
Petitioners are incorrect in stating that all of the references cited in the petition have been
published in peer-reviewed literature. Specifically, the EPA does not agree with Petitioners that
the conference presentations cited by Petitioners in support of these reactions constitute peer-
reviewed literature. Petitioners are incorrect to contend that a conference presentation such as
Landis, et al., (2009) or a conference presentation cited in a peer-reviewed journal article such as
ter Schure, et al., (2011) is a peer-reviewed study itself. Further, the field studies at Plant Crist in
68 Kolker, A., Olson, M.L., Krabbenhoft, D.P., Tate, M.T., Engle, M.A., 2010. "Patterns of
mercury dispersion from local and regional emission sources, rural Central Wisconsin, USA."
Atmospheric Chemistry and Physics 10, 4467-4476.
69 Lohman, K., Seigneur, C., Edgerton, E., Jansen, J., 2006. "Modeling mercury in power plant
plumes." Environmental Science & Technology 40, 3848-3854.
70 Amos, H.M., Jacob, D.J., Holmes, C.D., Fisher, J.A., Wang, Q., Yantosca, R.M., Corbitt, E.S.,
Galarneau, E., Rutter, A.P., Gustin, M.S., Steffen, A., Schauer, J.J., Graydon, J.A., St Louis,
V.L., Talbot, R.W., Edgerton, E.S., Zhang, Y., Sunderland, E.M. 2012. "Gas-particle partitioning
of atmospheric Hg(II) and its effect on global mercury deposition." Atmospheric Chemistry and
Physics 12, 591-603.
71 Landis, M, Ryan, J, Oswald, E, Jansen J., Monroe, L, Walters, J, Levin, L, ter Schure, AFH,
Laudal, D, Edgerton, E. 2009. "Plant Crist Mercury Plume Study." Presented at Air Quality VII
Conference, Arlington, VA, October 27.
72 ter Schure A, Caffrey, J, Gustin,M, Holmes, C, Hynes, A, Landing, W, Landis, M, Laudal, D,
Levin L, Nair, U, Jansen, J, Ryan, J, Walters, J, Shauer, J, Volkamer, R, Waters, D, Weiss, P.
2011. "An Integrated Approach to Assess Elevated Mercury Wet Deposition and Concentrations
in the South Eastern United States." Presented at the 10th International Conference on Mercury as
a Global Pollutant, Halifax, Nova Scotia, Canada, July 27, 2011.
48
-------
west Florida used in the Landis, et al., presentation were sponsored by the authors of Petitioner
20180's att. 6.
Petitioners are also incorrect in their interpretation of two of the peer-reviewed studies related to
this issue (Kolker, et al., 2010; Lohman, et al., 2006). The Kolker, et al., (2010) study, which was
cited by both Petitioners and the EPA in response to comment on the proposed MATS rule (pp.
102 - 112), shows that when wind blows from the direction of a coal-fired EGU in central
Wisconsin that divalent Hg increases as the plume travels away from that source. This
contradicts Petitioner's assertion that in-plume chemistry converts divalent Hg, the form that
deposits quickly near sources, to elemental Hg, the form that has local to global impacts.
Petitioners also mistakenly contend that the Kolker, et al., study is not relevant to an analysis of
coal-fired EGUs, but Petitioners acknowledge that this study evaluated dispersion from a variety
of sources that included coal-fired EGUs.
Contrary to the assertion by Petitioners, the Lohman, et al., study, which was co-authored by
Petitioner 20183, does not observe in-plume oxidation. Instead, the study merely suggests one
possible explanation for the authors' observations and then finds that in-plume oxidation could
explain some of their observed SO2 to speciated Hg ratios. Specifically, the Lohman, et al., study
used ambient measurements to infer two possible reactions and concluded that "those potential
Hg(II) to Hg(0) reactions need to be studied in the laboratory to further investigate this
hypothesis." In addition, the Lohman, et al., study fails to appropriately characterize model
performance evaluation, which raises questions regarding whether the performance of the
dispersion model used in the Lohman, et al., study is appropriate for local and regional scale
regulatory modeling.
Although the Amos, et al., (2012) study is a new study, it provides no new information on in-
plume Hg reactions. Instead, this study simply uses the inferred reactions presented by the
Lohman, et al., study. Further, the Amos, et al., study clearly states that "[t]he implicit inclusion
of Hg(II) in-plume reduction in the model comes with the important caveats that a chemical
mechanism has not been identified (Lohman, et al., 2006)." Until in-plume reactions are clearly
identified in the peer-reviewed literature, it would be inappropriate for the EPA to employ those
reactions for tools such as the "Advanced Plume Treatment" that include inferred reactions based
on the Lohman, et al., study. For these reasons, this issue is not of central relevance to the
"Appropriate and Necessary" finding.
The comments that the EPA received on in-plume chemistry issues demonstrate that the public
had ample opportunity to comment on this issue, and the EPA responded to those comments.
Petitioners have not demonstrated that it was impracticable to comment on this issue during the
public comment period on the proposed MATS rule. Further, because the new study provided by
the Petitioners does not provide new data or information relevant to the EPA analysis, as
explained above, this issue is not of central relevance to the "Appropriate and Necessary"
finding. For all these reasons, the EPA is denying the petitions for reconsideration of this issue.
Issue 38: Petitioner 20183 (att. 3) and Petitioner 20180 (att. 4) suggest the EPA should have
included Hg reactions with bromine, an ocean-emitted halogen, in the CMAQ modeling
49
-------
assessment. Petitioners cite a study by Holmes, et al., (2010)73 as a potential source of
information about Hg chemical reactions with halogens emitted by oceans.
Response to Issue 38: Issues regarding potential Hg reactions with bromine were not raised
during the public comment period for this rule, and the Holmes, et al., study was published prior
to the close of the comment period for the proposed MATS rule. Petitioners did not raise the
issue regarding bromine reactions during the public comment period on the proposed MATS rule
despite providing extensive comments on other aspects of the Hg deposition modeling, and they
have not provided a reason why they were unable to do so. Because Petitioners have not
demonstrated that it was impracticable to comment on this issue during the public comment
period on the proposed MATS rule, the EPA is denying reconsideration of this issue.
In addition, inclusion of these reactions would not be appropriate for this analysis. The EPA's
Hg deposition modeling focuses on point source impacts on local and regional Hg deposition in
the continental U.S., and including bromine as the sole oxidant of elemental Hg would likely
degrade model performance due to the distance of many U.S. EGUs to open oceans. The EPA
notes that the CMAQ model used for the Hg deposition modeling includes Hg reactions with
other ocean-emitted halogens (e.g., chlorine, CI and Ch) to appropriately represent the fate of Hg
in marine environments. Because the EPA already includes reactions to represent ocean-emitted
halogens in the modeling, the inclusion of bromine is not necessary and, therefore, this issue is
not of central relevance to the "Appropriate and Necessary" finding. For all of these reasons, the
EPA is denying the petitions for reconsideration.
1.17 Uncertainties in local and regional Hg deposition
Issue 39: Petitioner 20180 (att. 6) disagrees with EPA's characterization of U.S. coal-fired
power plant impacts on local and regional Hg deposition in the excess local deposition analysis.
Specifically, Petitioner states in att. 6 that because the Kolker, et al., (2010) study cited in
comments submitted during the public comment period (EPA-HQ-OAR-2009-0234-17621)
evaluated ambient Hg concentrations rather than deposition, the EPA incorrectly discounted the
results of that study in its response that "the focus was on excess local deposition due solely to
EGU emissions, rather than all local and regional deposition." Petitioner states that Kolker, et al.,
did not study (local) deposition of Hg, but instead measured ambient atmospheric Hg
concentrations.
Response to Issue 39: Issues regarding the EPA's characterization of elevated Hg deposition in
the excess local deposition analysis were raised in public comments submitted in response to the
proposed MATS rule. The EPA responded to these comments in Section IE (pp. 114 - 119) of
the RTC. Furthermore, in the White Stallion decision, the Court found that "EPA's 'appropriate
and necessary' determination in 2000, and its reaffirmation of that determination in 2012, are
amply supported by EPA's findings regarding the health effects of mercury exposure." White
Stallion, 748 F.3d at 1245-46. As the Court may only consider issues raised during the period for
73 Holmes, C.D., Jacob, D.J., Corbitt, E.S., Mao, J., Yang, X., Talbot, R., Slemr, F., 2010.
"Global atmospheric model for mercury including oxidation by bromine atoms." Atmospheric
Chemistry and Physics 10, 12037-12057.
50
-------
public comment, issues raised in the litigation and addressed by the Court clearly do not meet the
criteria for reconsideration in CAA section 307(d)(7)(B). Moreover, parties may not use this
final action denying reconsideration as a basis to litigate issues that could have been raised in the
initial litigation.
The EPA has concluded that the Hg deposition modeling conducted for the Hg Risk TSD
appropriately reflects the impact of U.S. EGUs on Hg deposition. This modeling assesses local
and regional changes in Hg deposition that result from the removal of Hg from U.S. EGUs.
Because the Hg risk estimates are based on total and EGU-attributable CMAQ-modeled Hg
deposition, delineation about whether Hg impacts from U.S. EGUs would be considered local or
regional is not of central relevance to the "Appropriate and Necessary" finding.
Additionally, the EPA disagrees with Petitioner's interpretation of the EPA's responses to
comments on local Hg deposition. The EPA believes Petitioner likely misinterpreted "the
analysis" in the RTC (p. 117) to refer to the Kolker, et al., (2010) study when the EPA was
instead referring to the "Potential for Excess Local Deposition of Hg in Areas near U.S. EGUs"
analysis. The EPA agrees that the Kolker, et al., (2010) study does not provide information about
Hg deposition. Therefore consistent with the RTC (pp. 117 - 118), the EPA maintains that this
study does not provide information about the spatial extent of Hg deposition near U.S. EGUs.
Because Petitioner has not demonstrated that it was impracticable to comment on this issue
during the public comment period on the proposed MATS rule, as evidenced by the comments
received on this issue, the EPA is denying the petitions for reconsideration of this issue. The
EPA also stands by the responses to the comments received and the Agency does not believe
reconsideration is warranted based on the Petitioner's misinterpretation of the EPA response.
Finally, as stated above, the EPA would not have changed the Hg risk assessment or the excess
local deposition analysis based on the issues raised in the petition because the Kolker, et al.,
(2010) study does not provide information on Hg deposition, so the issue is not of central
relevance and the Agency is denying the petitions for reconsideration for that reason as well.
Issue 40: Petitioner 20183 (att. 3) and Petitioner 20180 (att. 4) claim that the Caffrey, et al.,
(2010) study,74 which was cited by the EPA in the RTC, does not support the EPA's argument
regarding local impacts of U.S. EGUs.
Response to Issue 40: Petitioner disagrees with the EPA's interpretation of a scientific study
cited in response to comments regarding local and regional Hg deposition from U.S. EGUs.
Specifically, the EPA cited the Caffrey, et al., study and two additional studies in Section IE (pp.
96 and 100) of the RTC as evidence that U.S. coal-fired power plants can significantly enhance
local and regional Hg deposition. The Caffrey, et al., study shows that Hg deposition
measurements highly correlate to sulfate deposition, which is similar to findings in other studies
74 Caffrey, J.M., Landing, W.M., Nolek, S.D., Gosnell, K.J., Bagui, S.S., Bagui, S.C., 2010.
"Atmospheric deposition of mercury and major ions to the Pensacola (Florida) watershed:
spatial, seasonal, and inter-annual variability ''Atmospheric Chemistry and Physics 10, 5425-
5434.
51
-------
in Michigan and the Ohio Valley (Keeleret al, 2006;75 Landis and Keeler, 200276). The
relationship between increased Hg and sulfate deposition in the Caffrey, et al., study is consistent
with the conclusion that coal-fired power plants contribute to wet Hg deposition in that area. The
EPA maintains that citing the Caffrey, et al., study is appropriate and supports the EPA's
conclusions regarding local Hg deposition, and Petitioner has provided no data or evidence to
support a contrary position. Therefore, this issue is not of central relevance to the "Appropriate
and Necessary" finding.
The comments that the EPA received on these issues demonstrate that the public had ample
opportunity to comment on the issue, and the EPA responded to those comments. Because
Petitioners have not demonstrated that it was impracticable to comment on this issue during the
public comment period on the proposed MATS rule, and because the issue is not of central
relevance to the "Appropriate and Necessary" finding, the EPA is denying the petitions for
reconsideration of this issue.
Issue 41: Petitioner 20180 (att. 6), disagrees with the EPA's conclusion that grid models such as
CMAQ do not overestimate local deposition, and Petitioner alleges that CMAQ overestimates
local deposition by up to 30 percent based on Karamchandani, et al., (2006),77 which simulated
deposition of sulfur and nitrogen oxides. Petitioner contends that the EPA's Hg deposition
modeling should have included sub-grid plume treatment, such as the "Advanced Plume
Treatment." Petitioner also questions the use of 12 km grid cells and cites concerns about the
CMAQ-Hg model, including ratio of wet to dry Hg deposition.
Response to Issue 41: Issues regarding the EPA's treatment of regional Hg deposition from
point sources with the CMAQ model were raised in public comments submitted in response to
the proposed MATS rule. The EPA responded to these comments in Section IE (pp. 101 - 106)
of the RTC. Furthermore, in the White Stallion decision, the Court found that "EPA's
'appropriate and necessary' determination in 2000, and its reaffirmation of that determination in
2012, are amply supported by EPA's findings regarding the health effects of mercury exposure."
White Stallion, 748 F.3d at 1245-46. As the Court may only consider issues raised during the
period for public comment, issues raised in the litigation and addressed by the Court clearly do
not meet the criteria for reconsideration in CAA section 307(d)(7)(B). Moreover, parties may not
use this final action denying reconsideration as a basis to litigate issues that could have been
raised in the initial litigation. The EPA is denying the petition for reconsideration for these
reasons.
As the EPA notes in the RTC (p. 102), the "Advanced Plume Treatment" was not an option in
the most recent versions of the CMAQ model. The EPA disagrees with Petitioner's assertion that
75 Keeler, G.J., Landis, M.S., Norris, G.A., Christianson, E.M., Dvonch, J.T., 2006. "Sources of
mercury wet deposition in Eastern Ohio, USA." Environmental Science & Technology 40, 5874-
5881.
76 Landis, M.S., Keeler, G.J. 2002. "Atmospheric mercury deposition to Lake Michigan during
the Lake Michigan Mass Balance Study." Environmental Science & Technology 36, 4518-4524.
77 Karamchandani, P., Vijayaraghavan, K., Chen, S.-Y., Seigneur, C., Edgerton, E.S., 2006.
"Plume-in-grid modeling for particulate matter." Atmospheric Environment 40, 7280-7297.
52
-------
the regional fate of nitrogen oxides (NOx), SO2, and Hg emissions is likely to be similar for all
three pollutants from point sources. While there may be correlations between ambient
concentrations and deposition of Hg, SO2, and NOx because some sources such as EGUs may
emit all 3 of these species, the atmospheric chemistry is different for each and the amount
deposited locally and regionally may be different. The Karamchandani, et al., (2006) study
provides no evidence about near-source Hg plume dispersion and chemistry. The "Advanced
Plume Treatment" in versions of CMAQ that are older than the version used for the Hg
deposition analysis include in-plume Hg reactions that have not been demonstrated to exist,
which means its use would be inappropriate. Petitioner does not support its assertions regarding
potential bias when using grid cells to represent impacts from single sources with any
quantitative evidence. In addition, the EPA notes that the Hg deposition modeling conducted for
MATS represents the most advanced modeling platform used for a national Hg deposition
assessment and the application of a photochemical grid model at 12 km resolution for the entire
continental U.S. provides the most robust grid resolution and scale in the published literature at
the time of the modeling assessment. The EPA has concluded that the modeling analysis
appropriately captures the local and regional impacts of coal-fired power plants, and Petitioners
show no compelling evidence to the contrary.
The EPA disagrees with Petitioner that the ratio between wet and dry Hg deposition predicted by
the CMAQ model does not match Hg deposition measurements. Dry deposition is not routinely
measured, so it is not possible to evaluate this ratio. Petitioner claims that the EPA's ratio has a
"high bias of roughly 30%," but the cited Karamchandani, et al., study provides no evidence
related to Hg deposition to support this claim. Petitioner provides no new quantitative
information about the EPA's Hg deposition modeling. In addition, Petitioner could have raised
this issue during the public comment period. The EPA provided all CMAQ model inputs and
outputs in the docket (EPA-HQ-OAR-2009-0234-2989), which Petitioners could have used to
generate quantitative information about the EPA's Hg deposition analysis to support this alleged
issue.
Because Petitioner has not demonstrated that it was impracticable to comment on this issue
during the public comment period on the proposed MATS rule, as evidenced by the comments
EPA received on these issues, the EPA is denying the petition for reconsideration of this issue.
The EPA is also denying reconsideration because the issue is not of central relevance to the rule
for the reasons cited above.
Issue 42: Petitioner 20180 (att. 6) claims that the uncertainty analysis for Hg deposition "appears
cursory" and that the EPA could have used dry deposition data from the Atmospheric Mercury
Network (AMNet) stations to better evaluate the model results.
Response to Issue 42: Issues regarding the use of AMNet were raised in public comments
submitted in response to the proposed MATS rule. The EPA responded to this comment in
Section IE (pp. 114) of the RTC. Furthermore, in the White Stallion decision, the Court found
that "EPA's 'appropriate and necessary' determination in 2000, and its reaffirmation of that
determination in 2012, are amply supported by EPA's findings regarding the health effects of
mercury exposure." White Stallion, 748 F.3d at 1245-46. As the Court may only consider issues
raised during the period for public comment, issues raised in the litigation and addressed by the
53
-------
Court clearly do not meet the criteria for reconsideration in CAA section 307(d)(7)(B).
Moreover, parties may not use this final action denying reconsideration as a basis to litigate
issues that could have been raised in the initial litigation. The EPA is denying the petition for
reconsideration for these reasons.
Contrary to Petitioner's assertion, AMNet does not routinely measure dry deposition and was not
operational in 2005, which is the base year of the Hg deposition modeling analysis, which the
EPA noted in the RTC. It would be impossible for the EPA to compare CMAQ-modeled
estimates of Hg dry deposition to AMNet monitors because those monitors did not collect that
data in 2005. As such, Petitioner provides no new information or data that could have been
incorporated into the Hg deposition modeling. The public had ample opportunity to comment on
this issue, and the EPA responded to those comments.
Regarding the uncertainty analysis, the SAB recommended (U.S. EPA-SAB, 201 la), and the
EPA added, additional discussion on deposition uncertainty to the revised Hg Risk TSD. The
EPA fully responded to each of the SAB's recommendations including adding a more detailed
description of uncertainty in CMAQ and referencing existing evaluations of the model. Because
uncertainty assessments are generally ancillary to the core modeling, expanding the uncertainty
analysis would not have changed the results of the Hg deposition modeling. Therefore, this issue
is not of central relevance to the "Appropriate and Necessary" finding.
Because Petitioner has not demonstrated that it was impracticable to comment on AMNet during
the public comment period on the proposed MATS rule, as evidenced by comments on this issue,
and because the issue is not of central relevance to the "Appropriate and Necessary" finding, the
EPA is denying the petition for reconsideration of this issue.
Issue 43: Petitioner 20183 (att. 3) and Petitioner 20180 (att. 4) assert that the EPA did not
address critical comments from the SAB regarding uncertainty in the Hg emissions and Hg
deposition modeling. Specifically, Petitioners state that the EPA did not perform a quantitative
CMAQ modeling assessment of Hg deposition uncertainty as directed by the SAB.
Response to Issue 43: Petitioners are incorrect in their interpretation of the SAB's
recommendations to the EPA. The SAB stated that "[t]he qualitative nature of the [uncertainty]
discussion is appropriate because this is a conditional analysis. However, the SAB recommends
an expanded discussion in Appendix F of variability and uncertainty to make explicit the
uncertainties associated with the Agency's key analytical choices, which the SAB supports."
(U.S. EPA-SAB, 201 la, p. 22). The SAB did not request a quantitative evaluation of uncertainty.
The EPA fully responded to each of the SAB recommendations, including adding a more
detailed description of uncertainty in CMAQ and referencing existing evaluations of the model.
Because uncertainty assessments are generally ancillary to the core modeling, expansion of the
uncertainty analysis would not cause EPA to revise its conclusions concerning the Hg deposition
modeling. Therefore, this issue is not of central relevance to the "Appropriate and Necessary"
finding, and the EPA is denying the petition for reconsideration of this issue.
1.18 Alternate analysis of Hg deposition
54
-------
Issue 44: Petitioner 20183 (att. 3) and Petitioner 20180 (att. 4) claims that the EPA's RTC
regarding their alternative analysis of Hg deposition is inadequate.
Response to Issue 44: The EPA's RTC regarding the alternative analysis performed by
Petitioner 20183 (att. 3) and Petitioner 20180 (att. 4) is in Section ID (p. 99) of the RTC.
The Petitioners' assertion demonstrates that it was not impractical to comment on this issue
during the public comment period and that the EPA responded to the comments. In the White
Stallion decision, the Court found that "EPA's 'appropriate and necessary' determination in
2000, and its reaffirmation of that determination in 2012, are amply supported by EPA's findings
regarding the health effects of mercury exposure." White Stallion, 748 F.3d at 1245-46. As the
Court may only consider issues raised during the period for public comment, issues raised in the
litigation and addressed by the Court clearly do not meet the criteria for reconsideration in CAA
section 307(d)(7)(B). Moreover, parties may not use this final action denying reconsideration as
a basis to litigate issues that could have been raised in the initial litigation.
In addition, the EPA disagrees that the response was inadequate and repeating the comment does
not provide any additional data or information to change the EPA's opinion. The analysis
conducted by Petitioner 20183 (att. 3) and Petitioner 20180 (att. 4) adjusts CMAQ Hg deposition
estimates using incorrect assumptions about boundary inflow, Hg chemical reactions, and
incommensurate models. As stated in more detail the EPA's RTC (p. 99), the EPA finds the
methodology used by Petitioner 20183 (att. 3) and Petitioner 20180 (att. 4) to be inappropriate,
and the EPA maintains that conclusion. For these reasons, this issue is not of central relevance to
the "Appropriate and Necessary" finding.
Because Petitioners have not demonstrated that it was impracticable to comment on this issue
during the public comment period on the proposed MATS rule, and because this issue is not of
central relevance, the EPA is denying the petition for reconsideration of this issue.
55
-------
1.19 Emissions in non-Hg risk case studies
Issue 45: Petitioners 20180 and 20183 request the EPA reconsider the risk posed by chromium
(Cr) in light of new Cr emissions data from five coal-burning EGUs.78 Petitioner 20180 repeats
the allegation made during the public comment period that the Cr and nickel (Ni) measurements
collected from these EGUs as part of the 2010 Information Collection Request (ICR) are
anomalous, were most likely caused by sample contamination - perhaps by the use of metallic
fittings, as purportedly observed at one EGU during subsequent testing, and do not represent
actual emissions. Petitioner 20180 identifies these new data, suggests these new data confirm
sample contamination of previous data, and asserts the older data should not have been used in
the risk assessment. Further, Petitioner 20180 (att. 1) asserts that these data issues were raised
during the public comment period, to which the EPA provided a condescending response.
78 The Petitioner identified the following EGUs as having excessive Cr emissions obtained as
part of the ICR data collection process:
• City Utilities' James River Power Plant Unit 4 (The ICR data submittal code for this
EGU is 2135);
• City Utilities' James River Power Plant Unit 5 (The ICR data submittal code for this
EGUs is 21351);
• We Energies' Valley Power Plant Boiler 3 (The ICR data submittal code for this EGU is
735);
• We Energies' Valley Power Plant Boiler 4 (The ICR data submittal code for this EGU is
736);
• Tennessee Valley Authority's Gallatin Plant Unit 2; and
• Tennessee Valley Authority's Colbert Plant Units 3 and 4.
Petitioners provided a summary sheet listing non-Hg metals results identified as being from Unit
3 of American Electric Power Company's Conesville Power Plant, as well as copies of non-Hg
metals test reports identified as being from Units 4 and 5 of City Utilities' James River Power
Plant and from Boilers 3 and 4 of We Energies' Valley Power Plant. These tests were conducted
in February and March 2012, at least 6 months after the close of the public comment period and
at least 17 months after the end of the ICR data collection period. In addition, Petitioners
submitted these data without the EGU owner or operator certifications (of truth, accuracy, and
completeness) required by the ICR, without an explanation as to why the testing had to wait so
long given that the proposed rule had been available for almost a year, without an indication that
these were the only data collected from these EGUs during the retesting period, and without the
testing representative certification (that the test methods were followed) required by the ICR.
Despite these limitations, the EPA accepted and reviewed the deficient data summary and
deficient test reports. The Petitioners provided no TVA data, but TVA provided test reports for
Gallatin Unit 2 and Colbert Unit 4 on November 6, 2010; these data, which suffer from the same
limitations as the data from the other 5 EGUs, were tested after the deadline for submission of
petitions for judicial review for the MATS rule, unlike the other EGUs, which were tested prior
to the deadline for submission of petitions for judicial review for the MATS rule. As a result, the
TVA test reports were not reviewed by EPA.
56
-------
Response to Issue 45: As a general matter, the EPA disagrees that the Cr and Ni measurements
collected as part of the ICR are anomalous, that the Cr and Ni measurements represent values
caused by sample contamination, that the Cr and Ni measurements do not represent actual
emissions, and that the original Cr and Ni measurements should be excluded from emissions
factor development used in the risk assessment. Moreover, these allegations are not new; they
were raised in public comments submitted in response to the proposed MATS rule. The EPA's
responses to the comments are in section 1G (pp. 187 - 188) of the RTC. The EPA disagrees that
the responses were condescending.
The Cr and Ni measurements are not anomalous. In describing individual test run values as
anomalous, Petitioners appear to lack understanding of EPA's emissions factor development
process, because the average of individual test run values - not the individual test run values - is
used in determining emissions factors. In addition, the emission factor development process
includes a step for conducting outlier analysis; this analysis allows the EPA to identify and not
use those test averages from facilities with similar emissions control devices and operational
characteristics but with unrepresentative emissions when calculating emissions factors. As
emissions are assumed to follow a lognormal distribution, the outlier analysis is conducted on the
logarithm of test averages, and there must be at least 3 discrete test averages before an outlier
analysis can be conducted.
With respect to the EGU ICR data, outlier analyses were not conducted for specific EGUs,
because the ICR required just one test per pollutant from each EGU. Consistent with the
emission factor development process, at least 3 tests per pollutant would have been necessary for
an outlier analysis to have been conducted for a specific pollutant for a specific EGU. While
nothing in the ICR precluded an EGU owner or operator from conducting 3 or more tests for a
specific pollutant at a specific EGU during the specified data collection period, the EPA is
unaware of any additional testing from any EGU during the specified data collection period. On
the other hand, the EPA conducted outlier analyses for groups (or "bins") of at least 3 EGUs with
similar emissions control devices and operational characteristics as a part of developing
emissions factors for those bins.
As there is just one test value per pollutant per specific EGU, no outlier can exist for that specific
EGU. Because no outlier can exist for a site-specific EGU with just one test value, the EPA has
no reason - and the Petitioners have provided no evidence - to exclude these test values either
from site-specific emissions factors or from the emissions factor bins for EGUs with similar
control device and operational characteristics. While other situations, including sample
contamination, could cause test values to be excluded from emissions factor calculations, sample
contamination as alleged by the Petitioners appears unlikely and Petitioners have provided no
data or evidence to suggest contamination may have occurred.
The record does not support the assertion that the Cr and Ni measurements were contaminated.
The ICR placed responsibility for emissions test quality assurance and control on EGU owners or
operators. Efforts to identify and prevent sample contamination would be a part of emissions test
quality assurance and control, as would development of a test plan that would specify means of
57
-------
corrective action should sample contamination occur. Moreover, adherence to EPA Method 2979,
which contains numerous warnings not to contaminate samples80, should have caused emissions
testing representatives, as well as EGU owners or operators, to exercise great caution with
respect to ensuring contamination did not occur. As part of the ICR data submission, emissions
testing representatives from these EGUs provided certifications that the testing details and results
had been reviewed and that the test reports were authentic and accurate, and these EGU owners
or operators provided certifications that the statements and information in the test report were
true, accurate, and complete, based on information and belief formed after reasonable inquiry.
If sample contamination had occurred, as the Petitioners allege, the EGU owners or operators
would have neglected their quality assurance and control activities and the testing
representatives, along with the EGU owners or operators, would have falsely certified their ICR
data submissions, because the test method directs users to prevent possible sample contamination
problems. Petitioners cannot have it both ways. Either testing representatives and EGU owners
or operators willfully ignored their quality assurance and control obligations, as well as test
method requirements, and knowingly provided false certifications, or testing representatives and
EGU owners or operators conducted the ICR emissions testing appropriately. Given that it is the
Petitioners who, as third parties neither participated in the operation, emissions sampling,
emissions analysis, certification, or submission of ICR data, assert the claim of sample
contamination, rather than the EGU owners or operators, and that the EGU owners or operators
have not withdrawn the original certifications and provided revised certifications or requested
that the ICR data be withdrawn, the EPA may rely on the certifications of actual participants in
the ICR data collection and submissions. See White Stallion, 1222 F.3d at 1249 (finding that the
EPA acted reasonably in relying on certified data).
The Cr and Ni measurements represent actual emissions.
The ICR also directed owners or operators of EGUs to conduct emissions testing at a load and
with a fuel that reflects typical operations at the facility. Moreover, the ICR required the EGU
owners or operators to certify that the fuel fired during emissions testing was representative of
the fuel that is burned routinely at their EGUs; that all of the pollution control equipment was
operated in accordance with manufacturer's specifications and requirements for proper
operation; and that pollution control equipment was operated to optimize reduction of the
pollutants for which the equipment was designed. Given these requirements and the EGU owners
or operators' certifications, the EPA has no reason to believe EGU operation during emissions
testing was anything other than representative. Moreover, as the emissions testing was conducted
during representative EGU operation, the EPA has no reason - and Petitioner has provided no
79 See Method 29, Determination of Metals Emissions From Stationary Sources, in Appendix A-
8 to 40 CFR part 60
80 See section 6.1.1 concerning connectors to the sampling train for the probe nozzle: "... probe
fittings of plastic such as Teflon, polypropylene, etc. are recommended instead of metal fittings
to prevent contamination..." and section 8.1.3.4 concerning ensuring the sampling train is sealed
so as not to introduce outside contamination: ".. .to prevent possible sample contamination
problems, use Teflon tape or other non-contaminating material..."
58
-------
relevant evidence - to assume that the Cr or Ni emissions obtained from the EGUs in question
during the period of ICR data collection are unrepresentative of normal operation.
Moreover, as discussed in footnote 77, when the EPA, for its own purposes, combined the new
data with the ICR data, conducted an outlier analysis, generated emissions factors from the
combination of new and ICR data, and used those emissions factors in risk assessments, it found
the resulting cancer risk estimates for three facilities (Chesapeake, James River, and Yorktown)
were greater than 1-in-l million, which is consistent with the conclusions in the final rule.
Because the new data would not have affected the conclusions of the non-Hg case studies, this
issue is not of central relevance to the "Appropriate and Necessary" finding.
Given that the Petitioners' issues regarding the ICR data were raised in public comments on the
proposed MATS rule, and the EPA responded to those comments and that the Petitioners have
not demonstrated that the Cr and Ni measurements are anomalous, or were caused by sample
contamination, or were not representative of actual emissions, the EPA denies the Petition for
Reconsideration on these issues.
In addition, in the While Stallion decision, the Court affirmed the EPA's assertion that a finding
that it is appropriate and necessary to regulate any one HAP is sufficient to list EGUs and
regulate all HAP from those sources, and the Court found that the appropriate and necessary
finding is "amply supported by EPA's findings regarding the health effects of mercury
exposure." 748 F.3d at 1244-45. Because of the Court's decision, the non-Hg metals findings are
not of central relevance and are moot because the appropriate and necessary finding can be
sustained on the Hg findings alone.
Issue 46: Petitioners 20180 and 20183 state that new emissions data available since the close of
the public comment period voids the EPA's conclusions in the non-Hg chronic inhalation risk
assessment and that, using retested Cr emissions, no coal-fired facility considered in that
assessment exceeds 1-in-l million cancer risk. Petitioners state that because the EPA cannot
show a coal-fired facility that exceeds 1-in-l million cancer risk, the EPA has no factual basis for
concluding that it is appropriate to regulate non-Hg HAP, and request that the EPA reconsider
the risks posed by Cr.
Response to Issue 46: As stated above, the Court in White Stallion affirmed EPA's conclusion
that a finding that it is appropriate and necessary to regulate any one HAP is sufficient to list
EGUs and regulate all HAP from those sources. 748 F.3d at 1244-45. The Court also found that
the appropriate and necessary finding is "amply supported by EPA's findings regarding the
health effects of mercury exposure." Id. at 1244-45. Because of the Court's decision, the
objections relating to the non-Hg metals findings are not of central relevance and are moot
because the appropriate and necessary finding can be sustained on the Hg findings alone.
Furthermore, the EPA is under no obligation to consider any of the additional test data provided
by Petitioner because, as noted above, the additional data are not comparable to those data
submitted under auspices of the ICR. Specifically, these additional data were not submitted using
the ERT; no certification as specified in the ICR regarding representative operation, fuel use,
control device operation, and control device efficiency during the emissions testing period was
59
-------
provided; and the test results were not certified as specified in the ICR by either the testing
representatives or the EGU owners or operators. 81 Moreover, the EPA does not know the testing
conditions, number of additional test runs performed, or if the Petitioner provided all available
data. For these reasons, the data are not of central relevance to the final MATS rule and the EPA
denies the petition for reconsideration on this issue.
Issue 47: Petitioners 20180 and 20183 suggest that the EPA used a Ni emission factor
approximately three times greater than the ICR test value to estimate annual Ni emissions for
Yorktown Unit 3.
Response to Issue 47: Issues regarding the emission rates used in the non-Hg risk case studies
were raised in public comments submitted in response to the proposed MATS rule. The EPA
responded to these comments in Section 1G (p. 185) of the RTC.
81 Nevertheless, the EPA decided for its own purposes to determine what, if any, impact these
new uncertified data could have on the non-Hg risk assessment. Although Petitioners seek
review of just the Cr emissions factors obtained during the ICR data collection period, the EPA
reviewed not only the Cr emissions data, but also Ni and As emissions data when conducting this
voluntary analysis.
The EPA recalculated the emissions factors in the same manner as was done for the final MATS
rule, except that for this situation the combined set of individual test run values (not test
averages) were subjected to an outlier analysis. ICR and retest data were combined and used,
rather than using just the retest data as suggested by Petitioners. Using emissions factors
developed from the combined dataset, the EPA conducted dispersion modeling and calculated
risks for the facilities noted by Petitioners: Dominion Chesapeake VA, Conesville, OH, City
Utilities of Springfield - James River, MO, AmerenUE - Labadie, MO, OG & E - Muskogee, and
Yorktown, VA. The dispersion modeling and risk calculations followed the same methodologies
described in the TSDs for the non-Hg case study assessments.
Even after incorporating this combined emissions dataset, the resulting cancer risk estimates for
three facilities (Chesapeake, James River, and Yorktown) were greater than 1-in-l million,
which is consistent with the conclusions in the final rule. Because the new data would not have
affected the conclusions of the non-Hg case studies, this issue is not of central relevance to the
"Appropriate and Necessary" finding.
Further, the EPA disagrees with the premise that a non-Hg finding was required to support the
"Appropriate and Necessary" finding. As stated in the preamble to the final MATS rule (77 FR
9326), "[t]he EPA reasonably concluded that we must find it "appropriate" to regulate EGUs
under CAA section 112 if we determine that a single HAP emitted from EGUs poses a hazard to
public health or the environment." In addition, the preamble to the final MATS rule (77 FR
9307) also stated, "[o]nce the EPA lists a source category pursuant to CAA section 112(c), the
EPA must then establish technology-based emission standards under CAA section 112(d)". See
White Stallion, 748 F.3d at 1244-45 (affirming EPA's interpretation). Therefore, this issue is not
of central relevance to the "Appropriate and Necessary" finding.
60
-------
The EPA disagrees with Petitioners regarding the use of the Ni emissions factor from the ICR, as
Petitioners incorrectly seek to use an emissions factor developed for MACT floor purposes
(2.43E-03 pounds per million British thermal units (lb/MMBtu))82 for case study purposes. As
stated in the RTC (p. 185), the purposes of the emissions factors developed for the case study
emissions is quite different from the MACT program, which establishes emissions limits using
emissions data from the best performing sources in a specific category.
As shown in the Part III ICR data, Ni emissions run data obtained during ICR testing from
Yorktown Unit 3, in lb/MMBtu, were 1.73E-03, 1.96E-02, and 1.40E-03.83 The average of these
three runs is 7.58E-03 lb/MMBtu, which is the value the EPA used in the non-Hg risk case
studies for Yorktown Unit 3's site-specific emissions factor for Ni.84
These comments demonstrate that the public had ample opportunity to comment on the issue,
and EPA responded to those comments. Because Petitioners have not demonstrated that it was
impracticable to comment on this issue during the public comment period on the proposed
MATS rule, the EPA is denying the petition for reconsideration of this issue.
In addition, as stated above, the Court in While Stallion affirmed EPA's conclusion that a finding
that it is appropriate and necessary to regulate any one HAP is sufficient to list EGUs and
regulate all HAP from those sources. 748 F.3d at 1244-45. The Court also found that the
appropriate and necessary finding is "amply supported by EPA's findings regarding the health
effects of mercury exposure." Id. at 1244-45. Because of the Court's decision, the objections
relating to the non-Hg metals findings are not of central relevance and are moot because the
appropriate and necessary finding can be sustained on the Hg findings alone.
Issue 48: Petitioners 20180 and 20183 state that hexavalent Cr has not been detected in any coal-
fired power plant stack measurements.
Response to Issue 48: Issues regarding hexavalent Cr emissions were raised in public comments
submitted in response to the proposed MATS rule. The EPA's responses to the comments are in
section 1G (pp. 192 - 193) of the RTC.
Although the EPA does not have a means of selectively identifying hexavalent Cr within the Cr
measurements obtained by employing EPA Method 29, and the ICR did not request such data,
82 See the "Metallic_oil_Metals US" tab in the "a5_oil_mact_floor_analysis_12161 l.xlsx" file
labeled as the "MACT Floor analysis -oil" in the MATS ICR Data section of the Air Toxics
Standards for Utilities website, which is located at the following internet address:
.
83 See the Microsoft database file contained in the "EGU ICR PartHI" portion of the MATS ICR
Data section of the Air Toxics Standards for Utilities website, which is located at the following
internet address: .
84 See the "mats_final_current_base_hap.inven.xlsx" file labeled as the "Emissions Spreadsheet"
in the Technical Support Documents section of the Air Toxics Standards for Utilities website,
which is available on the Internet at .
61
-------
the singular Cr measurement supplied by Method 29 includes any hexavalent Cr that may be
present. A portion of the measured total Cr compounds (12 percent) was assigned as hexavalent
Cr, which the EPA had peer-reviewed as part of the methods to assess Ni and Cr risk. As stated
in the final rule (77 FR 9317), "[a]ll three authors also considered EPA's use of the average of
the range of the available speciation data (i.e., 12 percent and 18 percent Cr(VI) contained in
coal- and oil-fired EGUs, respectively) as a reasonable approach for the derivation of default
speciation profiles to be used when there is no speciation data available. All reviewers agreed
that there is high uncertainty associated with the variability in the speciation data available for Cr
(e.g., range of approximately 4 to 23 percent Cr(VI) from coal-fired units)." All materials related
to this peer review are available in the docket at EPA-HQ-OAR-2009-0234-19744.
These comments demonstrate that the public had ample opportunity to comment on this issue,
and EPA responded to those comments. Because Petitioners have not demonstrated that it was
impracticable to comment on this issue during the public comment period on the proposed
MATS rule, the EPA is denying the petition for reconsideration of this issue.
In addition, as stated above, the Court in While Stallion affirmed EPA's conclusion that a finding
that it is appropriate and necessary to regulate any one HAP is sufficient to list EGUs and
regulate all HAP from those sources. 748 F.3d at 1244-45. The Court also found that the
appropriate and necessary finding is "amply supported by EPA's findings regarding the health
effects of mercury exposure." Id. at 1244-45. Because of the Court's decision, the non-Hg metals
findings are not of central relevance and are moot because the appropriate and necessary finding
can be sustained on the Hg findings alone.
Issue 49: Petitioner 20180 (att. 1) commented that during the ICR review and OMB approval
process the EPA should have been better prepared to address data quality given the amount of
testing conducted under the EPA's compressed schedule.
Response to Issue 49: The EPA disagrees with Petitioner. As mentioned earlier, the ICR placed
responsibility for emissions test quality assurance and control on EGU owners or operators, and
it advised them to prepare Quality Assurance Test Plans specific for each facility required to
perform emissions testing (See ICR Part A - Page 10 of 22). Further, as part of the data
submission, the ERT required testing company representatives to certify that the testing details
and results had been reviewed and that the test report was authentic and accurate. See White
Stallion, 1222 F.3d at 1249 (finding that the EPA acted reasonably in relying on certified data).
Also as part of the data submission, the ERT required all EGU owners and operators to certify
that the statements and information contained in the test report were true, accurate, and complete,
based on belief formed after reasonable inquiry. Moreover, the ICR Part B (page 35) required
EGU owners or operators to certify process and pollution control device operational
characteristics. Given these requirements and certifications, the EPA has no reason - and
Petitioner has provided no relevant evidence - to assume that the data quality is compromised.
For this reason, this issue is not of central relevance and we are denying reconsideration for this
reason. In addition, the Petitioners have not demonstrated that it was impractical to comment on
this issue during the public comment period, and we are denying the petition for reconsideration
for this reason as well.
62
-------
In addition, as stated above, the Court in White Stallion affirmed EPA's conclusion that a finding
that it is appropriate and necessary to regulate any one HAP is sufficient to list EGUs and
regulate all HAP from those sources. 748 F.3d at 1244-45. The Court also found that the
appropriate and necessary finding is "amply supported by EPA's findings regarding the health
effects of mercury exposure." Id. at 1244-45. Because of the Court's decision, the objections
relating to the non-Hg metals findings are not of central relevance as the appropriate and
necessary finding can be sustained on the Hg findings alone.
Issue 50: Petitioners 20180 and 20183 suggest the emission rates used in the MACT analysis are
more appropriate than the emission rates used in the non-Hg metals risk case studies.
Response to Issue 50: The EPA disagrees with Petitioners. This issue was raised in public
comments submitted in response to the proposed MATS rule. As mentioned earlier, the purpose
of the emissions factors developed for the case study emissions is quite different from the MACT
program, which establishes emissions limits using emissions data from the best performing
sources in a specific category. The EPA's responses to the comments on this issue are in section
1G (pp. 184 - 188) of the RTC.
These comments demonstrate that the public, including Petitioner 20180, had ample opportunity
to comment on this issue, and EPA responded to those comments. Because Petitioner has not
demonstrated that it was impracticable to comment on this issue during the public comment
period on the proposed MATS rule, the EPA is denying the petition for reconsideration of this
issue.
In addition, as stated above, the Court in White Stallion affirmed EPA's conclusion that a finding
that it is appropriate and necessary to regulate any one HAP is sufficient to list EGUs and
regulate all HAP from those sources. 748 F.3d at 1244-45. The Court also found that the
appropriate and necessary finding is "amply supported by EPA's findings regarding the health
effects of mercury exposure." Id. at 1244-45. Because of the Court's decision, objections relating
to the non-Hg metals findings are not of central relevance and are moot as the appropriate and
necessary finding can be sustained on the Hg findings alone.
Issue 51: Petitioners 20180 and 20183 say the EPA used the arithmetic mean rather than the
geometric mean to develop emission factors for the different groupings of units, and Petitioners
maintain that this approach is neither statistically sound nor defensible, as shown by the EPA's
own analysis that the distribution of emission rates from the ICR units has a log-normal
distribution.
Response to Issue 51: The EPA disagrees with Petitioners. This issue was raised in public
comments submitted in response to the proposed MATS rule. The EPA's responses to the
comments are in section 1G (pp. 184 - 187) of the RTC.
These comments demonstrate that the public had ample opportunity to comment on this issue,
and EPA responded to those comments. Because Petitioner has not demonstrated that it was
impracticable to comment on this issue during the public comment period on the proposed
MATS rule, the EPA is denying the petitions for reconsideration of this issue.
63
-------
In addition, as stated above, the Court in White Stallion affirmed EPA's conclusion that a finding
that it is appropriate and necessary to regulate any one HAP is sufficient to list EGUs and
regulate all HAP from those sources. The Court also found that the appropriate and necessary
finding is "amply supported by EPA's findings regarding the health effects of mercury
exposure." 748 F.3d at 1244-45. Because of the Court's decision, objections relating to the non-
Hg metals findings are not of central relevance and are moot as the appropriate and necessary
finding can be sustained on the Hg findings alone.
Issue 52: Petitioners 20180 and 20183 say using a statistically robust methodology for
calculating the emission factor for the arsenic (As) bin would have substantially lowered the risk
at Chesapeake.
Response to Issue 52: The EPA disagrees with Petitioners. Issues related to the non-Hg case
studies were raised in public comments submitted in response to the proposed MATS rule. The
EPA's responses to the comments are in section 1G (pp. 184 - 193) of the RTC.
In addition, As was not the cancer driver for any of the facilities estimated to have cancer risk
above 1-in-l million in the final MATS rule. Even if the EPA were to assume that As emissions
at Chesapeake were more than 90 percent lower, the cancer risk at Chesapeake would still
exceed 1-in-l million. Therefore, this issue is not of central relevance to the "Appropriate and
Necessary" finding.
The comments received during the comment period demonstrate that the public had ample
opportunity to comment on these issues, and EPA responded to those comments. Because
Petitioners have not demonstrated that it was impracticable to comment on this issue during the
public comment period on the proposed MATS rule, the EPA is denying the petitions for
reconsideration of this issue.
In addition, as stated above, the Court in White Stallion affirmed EPA's conclusion that a finding
that it is appropriate and necessary to regulate any one HAP is sufficient to list EGUs and
regulate all HAP from those sources. 748 F.3d at 1244-45. The Court also found that the
appropriate and necessary finding is "amply supported by EPA's findings regarding the health
effects of mercury exposure." Id. at 1244-45. Because of the Court's decision, objections relating
to the non-Hg metals findings are not of central relevance and are moot as the appropriate and
necessary finding can be sustained on the Hg findings alone.
Issue 53: Petitioners 20180 and 20183 suggest that the EPA erroneously employed an arithmetic
mean, as compared with a geometric mean, to develop plant-specific emission estimates.
Response to Issue 53: The EPA disagrees with Petitioners. This issue was raised in public
comments submitted in response to the proposed MATS rule. The EPA's responses to the
comments are in section 1G (pp. 184 - 187) of the RTC.
The comments received on this issue during the comment period demonstrate that the public had
ample opportunity to comment on this issue, and EPA responded to those comments. Because
64
-------
petitioners have not demonstrated that it was impracticable to comment on this issue during the
public comment period on the proposed MATS rule, the EPA is denying the petition for
reconsideration of this issue.
In addition, as stated above, the Court in While Stallion affirmed EPA's conclusion that a finding
that it is appropriate and necessary to regulate any one HAP is sufficient to list EGUs and
regulate all HAP from those sources. 748 F.3d at 1244-45. The Court also found that the
appropriate and necessary finding is "amply supported by EPA's findings regarding the health
effects of mercury exposure." Id. at 1244-45. Because of the Court's decision, objections relating
to the non-Hg metals findings are not of central relevance and are moot as the appropriate and
necessary finding can be sustained on the Hg findings alone.
Issue 54: Petitioners 20180 and 20183 suggest that the geometric mean provides a more robust
estimate for an individual plant's emission rate (i.e., for individual plant data where the sample
population dataset is not normally distributed).
Response to Issue 54: The EPA disagrees with Petitioners. As mentioned earlier, this issue was
raised in public comments submitted in response to the proposed MATS rule. The EPA's
responses to the comments are in section 1G (pp. 184 - 187) of the RTC.
These comments demonstrate that the public had ample opportunity to comment on this issue,
and EPA responded to those comments. Because Petitioner has not demonstrated that it was
impracticable to comment on this issue during the public comment period on the proposed
MATS rule, the EPA is denying the petitions for reconsideration of this issue.
In addition, as stated above, the Court in White Stallion affirmed EPA's conclusion that a finding
that it is appropriate and necessary to regulate any one HAP is sufficient to list EGUs and
regulate all HAP from those sources. 748 F.3d at 1244-45. The Court also found that the
appropriate and necessary finding is "amply supported by EPA's findings regarding the health
effects of mercury exposure." Id. at 1244-45. Because of the Court's decision, objections relating
to the non-Hg metals findings are not of central relevance and are moot as the appropriate and
necessary finding can be sustained on the Hg findings alone.
Issue 55: Petitioners 20180 and 20183 reiterate their claim on the proposed rule that some of the
ICR chromium measurements from the 2010 ICR testing were contaminated and suggest that the
EPA must grant reconsideration to allow the new chromium emissions data from the retested
EGUs - James River 4, James River 5, Valley 3, and Valley 4 - as well as the summary data
from Conesville 3, to be fully and fairly considered. Petitioners claim use of those data as
replacements for the ICR data would lead to revised emissions factors and risk assessments less
than 1 -in-1 million for Cr for the Dominion Chesapeake Energy Center, Conesville, TVA
Gallatin, and City Utilities of Springfield - James River coal-fired EGUs and for Ni for the
Dominion - Yorktown oil-fired EGU.
Response to Issue 55: The EPA disagrees with Petitioners. The sample contamination claim for
Cr emissions from James River 4, James River 5, Valley 3, Valley 4, and Conesville 3 (and from
12 other EGUs) was raised in public comments submitted in response to the proposed MATS
65
-------
rule. The EPA's response to the comments are in section 1G (pp. 187) of the RTC. Moreover, as
discussed earlier, the Petitioners have not demonstrated that the Cr and Ni were caused by
sample contamination. The EPA responded to these comments in Section 1G (p. 185) of the
RTC.
These comments demonstrate that the public had ample opportunity to comment on this issue,
and EPA responded to those comments. Because the Petitioners have not demonstrated that it
was impracticable to comment on these issues during the public comment period on the proposed
MATS rule, the EPA is denying the petitions for reconsideration of these issues.
In addition, as stated above, the Court in While Stallion affirmed EPA's conclusion that a finding
that it is appropriate and necessary to regulate any one HAP is sufficient to list EGUs and
regulate all HAP from those sources. 748 F.3d at 1244-45. The Court also found that the
appropriate and necessary finding is "amply supported by EPA's findings regarding the health
effects of mercury exposure." Id. at 1244-45. Because of the Court's decision, objections relating
to the non-Hg metals findings are not of central relevance and are moot as the appropriate and
necessary finding can be sustained on the Hg findings alone.
Issue 56: Petitioners 20180 and 20183 allege that EPA ignored comments in the ERT Comments
field of the 2010 ICR Cr and Ni test results that noted the likely presence of metallic
contamination in the sample.
Response to Issue 56: The EPA disagrees with Petitioners. As mentioned earlier, the ICR placed
responsibility for emissions test quality assurance and control on EGU owners or operators.
Therefore, EGU owners or operators should have developed specific procedures to detect sample
contamination and to correct it and provide replacement data, if necessary. No specific
procedures for sample contamination detection, correction, or supplemental data appear in the
test reports for the EGUs in question, nor are any such measures or data described. However, as
mentioned earlier, each test report contains a certification from the testing representative that the
testing details and results had been reviewed and that the test reports were authentic and
accurate, and these EGU owners or operators provided certifications that the statements and
information in the test report were true, accurate, and complete, based on information and belief
formed after reasonable inquiry. See White Stallion, 1222 F.3d at 1249 (stating that the EPA may
rely on certified data).
If sample contamination had occurred, as the Petitioners allege, the EGU owners or operators
would have neglected their quality assurance and control activities and the testing
representatives, along with the EGU owners or operators, would have falsely certified their ICR
data submissions, because the test method directs users to prevent possible sample contamination
problems. Petitioners cannot have it both ways. Either testing representatives and EGU owners
or operators willfully ignored their quality assurance and control obligations, as well as test
method requirements, and/or knowingly provided false certifications, or testing representatives
and EGU owners or operators conducted the ICR emissions testing appropriately. Given that it is
the Petitioners who, as third parties neither participated in the operation, emissions sampling,
emissions analysis, certification, or submission of ICR data, assert the claim of sample
contamination, rather than the EGU owners or operators, the EPA reasonably relies on the
66
-------
certifications of actual participants in the ICR data collection and submissions. Rather than
ignore the comments, the EPA reviewed the Petitioner's concern during the public comment
period on the proposed MATS rule and upon consideration of all factors- such as the warnings to
preclude sample contamination contained in the test method, no specific objection related to Cr
testing, no additional emissions testing runs to offset or replace deficient runs conducted during
the ICR data collection period or the rule comment period, the inability to determine the Cr
emissions as outliers, and the certifications by owners or operators of the EGUs in question - the
EPA found the comments suggesting potential sample contamination not to be reasonably
supported. See Responses 3 and 4 in section 1G (pp. 187 - 188) of the RTC.
Because the Petitioners have not demonstrated that it was impractical to comment on this issue
during the comment period, and because the issue is not of central relevance for the reasons set
forth above, the EPA denies the Petition for Reconsideration on this issue.
In addition, as stated above, the Court in While Stallion affirmed EPA's conclusion that a finding
that it is appropriate and necessary to regulate any one HAP is sufficient to list EGUs and
regulate all HAP from those sources. 748 F.3d at 1244-45. The Court also found that the
appropriate and necessary finding is "amply supported by EPA's findings regarding the health
effects of mercury exposure." Id. at 1244-45. Because of the Court's decision, objections relating
to the non-Hg metals findings are not of central relevance and are moot as the appropriate and
necessary finding can be sustained on the Hg findings alone.
Issue 57: Petitioners 20180 and 20183 recommend that the EPA perform a detailed evaluation of
all Method 29 ICR results because other test runs may have data quality issues.
Response to Issue 57: The EPA disagrees with Petitioners. As mentioned earlier, the ICR placed
responsibility for emissions test quality assurance and control on EGU owners or operators, and
it advised them to prepare Quality Assurance Test Plans specific for each facility required to
perform emissions testing. Further, as part of the data submission, the ERT required testing
company representatives to certify that the testing details and results had been reviewed and that
the test reports were authentic and accurate. Also as part of the data submission, the ERT
required all EGU owners and operators to certify that the statements and information contained
in the test report were true, accurate, and complete, based on belief formed after reasonable
inquiry. See White Stallion, 1222 F.3d at 1249 (finding that the EPA acted reasonably in relying
on certified data). Moreover, the ICR required EGU owners or operators to certify process and
pollution control device operational characteristics. Given these requirements and certifications,
the EPA has no reason - and Petitioner has provided no relevant evidence - to assume that the
data quality is compromised. For these reasons, this issue is not of central relevance.
Because the Petitioners have not demonstrated that it was impractical to comment on this issue
during the comment period, and because the issue is not of central relevance, we are denying the
petitions for reconsideration of this issue.
In addition, as stated above, the Court in White Stallion affirmed EPA's conclusion that a finding
that it is appropriate and necessary to regulate any one HAP is sufficient to list EGUs and
regulate all HAP from those sources. 748 F.3d at 1244-45. The Court also found that the
67
-------
appropriate and necessary finding is "amply supported by EPA's findings regarding the health
effects of mercury exposure." Id. at 1244-45. Because of the Court's decision, objections relating
to the non-Hg metals findings are not of central relevance and are moot as the appropriate and
necessary finding can be sustained on the Hg findings alone.
Issue 58: Petitioner 20180 (att. 6) claims that it could not replicate EPA's analysis of EPRI
equations and requested the underlying spreadsheet and supporting information.
Response to Issue 58: All components of the analysis are available to the public, and the
Petitioners could use these components to replicate EPA's analysis. Given that the components
are contained in the rule docket, the ICR Part III database, and AP-42 all of the relevant
information is available to the public, and Petitioners have not shown they were unable to access
and use this information. For this reason, the issue is not of central relevance and the EPA denies
the petition for reconsideration on this issue
In addition, as stated above, the Court in While Stallion affirmed EPA's conclusion that a finding
that it is appropriate and necessary to regulate any one HAP is sufficient to list EGUs and
regulate all HAP from those sources. 748 F.3d at 1244-45. The Court also found that the
appropriate and necessary finding is "amply supported by EPA's findings regarding the health
effects of mercury exposure." Id. at 1244-45. Because of the Court's decision, objections relating
to the non-Hg metals findings are not of central relevance and are moot as the appropriate and
necessary finding can be sustained on the Hg findings alone.
1.20 Nickel risk
Issue 59: Petitioner 20180 (att. 6) alleges that the conclusions regarding Ni risk were based on a
peer review sponsored by the EPA that was limited to two charge questions, was not open to
external review or comment, and was not clearly documented.
Response to Issue 59: The EPA followed the process that is described in the EPA's Peer Review
Handbook (U.S. EPA, 2012). The Handbook provides several examples of peer review
mechanisms, and describes factors to consider in selecting a particular mechanism (pp. 44 - 47).
For the Ni methods review, the EPA selected a formal external letter review by independent
experts from outside the agency. The EPA chose this mechanism because the issues associated
with the Ni and Cr methods review were relatively narrow in scope and were not scientifically
complex. The EPA also followed the method in the handbook for selecting the peer reviewers,
and per the handbook, considered the following factors: expertise of the peer reviewers, balance,
independence and potential conflicts of interest (peer review handbook pp. B11 - 14). The
Handbook states that public participation can be important (p. B14), but it does not require the
EPA to include public comment as part of every peer review. In this case, the agency determined
that the Ni and Cr methods review was not a highly influential assessment, and therefore the
agency was not required to provide an opportunity for public comment during the peer review.
Regarding documentation of the peer review, the EPA followed the recommendations in the
Handbook (pp. 49 - 50). The EPA developed a "Peer Review Record" that contains sufficient
documentation for an uninvolved person to understand the decision to conduct the peer review,
68
-------
the type of peer review conducted, and how the peer review comments were addressed. In
accordance with the Handbook recommendations, this Peer Review Record includes the
following materials: (1) draft work submitted for the peer review, (2) charge questions and all
additional information provided to the peer reviewers, (3) the peer review report, which contains
a summary of the review findings and detailed information about the reviewers, (4) logistical
information about the review (e.g., when it was conducted, time allotted for the review of the
draft document), (5) a written record of the response to peer review comments with specifics on
acceptance or, where appropriate, rebuttal and non-acceptance, and (6) the final work product.
The final product of this review is available in the docket (EPA-HQ-OAR-2009-0234-1999), as
well as a memorandum containing all documents developed in connection with the peer review
(i.e., all materials considered by the individual peer reviewers, the peer review report, and other
input) (EPA-HQ-OAR-2009-0234-19744). Because the EPA appropriately conducted the
external review and clearly documented all aspects of this review according to the
recommendations of EPA's Handbook, and parties were able to comment on this issue during the
comment period, the issue is not of central relevance and the EPA is denying the petition for
reconsideration of these issues.
The EPA is also denying reconsideration because the Petitioners have not demonstrated that it
was impractical to comment on this issue during the public comment period.
In addition, as stated above, the Court in While Stallion affirmed EPA's conclusion that a finding
that it is appropriate and necessary to regulate any one HAP is sufficient to list EGUs and
regulate all HAP from those sources. 748 F.3d at 1244-45. The Court also found that the
appropriate and necessary finding is "amply supported by EPA's findings regarding the health
effects of mercury exposure." Id. at 1244-45. Because of the Court's decision, objections relating
to the non-Hg metals findings are not of central relevance and are moot as the appropriate and
necessary finding can be sustained on the Hg findings alone.
Issue 60: Petitioner 20180 (att. 6) disagrees with EPA's conclusions that all Ni compounds are
of equal cancer potency, and thus the EPA should continue to apply unit risk estimates based
solely on nickel subsulfide.
Response to Issue 60: Issues regarding Ni risk were raised in public comments submitted in
response to the proposed MATS rule. The EPA's responses to the comments are in Section 1G
(pp. 210-212) of the RTC.
In the response, the EPA states that based on the available scientific evidence and in agreement
with major scientific bodies (the International Agency for Research on Cancer, the National
Toxicology Program, the World Health Organization), the EPA considers all Ni compounds to
be potentially carcinogenic to humans. The EPA also acknowledged that there are potential
differences in toxicity and/or carcinogenic potential across the different Ni compounds and
considers it reasonable to use values other than that existing for Ni subsulfide (0.00048 per
|ig/mJ), which has the highest existing Unit Risk Estimate (URE) value and therefore it is
69
-------
potentially the most potent carcinogen among Ni compounds.85 The EPA also stated that there
are three existing Ni UREs other than that derived for nickel subsulfide including one derived by
EPA's Integrated Risk Information System for Ni refinery dusts (0.00024 |ig/m3), and two others
that consider Ni compounds as a group derived by the California Department of Health Services
(CDHS, 1991)86 and the Texas Commission on Environmental Quality (TCEQ, 2011),87 with
values of 0.00026 per [j,g/m3 and 0.00017 [^g/m3, respectively. In the response to comments, the
EPA also noted that the existing UREs for Ni only vary quantitatively 2-3 fold, and that the use
of any of the existing UREs would yield roughly similar risk estimates. In addition, the UREs
derived by CDHS and TCEQ have considered all Ni compounds as a group which indicates that
these values can be applied across all Ni compounds. In the case studies for MATS, the EPA
applied 100 percent of the current IRIS URE for nickle subsulfide because IRIS is at the top of
the list of the preferred sources of dose response values used in the EPA's risk characterizations,
and because of the concerns about the potential carcinogenicity of all forms of Ni raised by the
major national and international scientific bodies. The EPA also concluded that in certain
instances (e.g., when high quality data is available on the composition of Ni emissions from a
specific source), it is reasonable to consider a value that is 50 percent of the IRIS URE for nickle
subsulfide for providing an estimate of the lower end of a plausible range of cancer potency
values for different mixtures of Ni compounds. Even if the EPA applied a value 50 percent of the
URE for nickel subsulfide in the case studies, several coal-fired facilities and one oil-fired
facility would still exceed 1-in-l million cancer risk. Thus this issue is not of central relevance to
the "Appropriate and Necessary" finding.
The comments submitted on this issue during the comment period demonstrate that the public
had ample opportunity to comment on the issue and the EPA responded to those comments.
Because Petitioner did not demonstrate that it was impracticable to comment on this issue during
the public comment period on the proposed MATS rule, the EPA is denying reconsideration of
this issue. In addition, because we determined that there would still be EGU facilities with a risk
greater than 1 in a million even if we took the Petitioner's approach, the issue is not of central
relevance to the "Appropriate and Necessary" finding.
In addition, as stated above, the Court in While Stallion affirmed EPA's conclusion that a finding
that it is appropriate and necessary to regulate any one HAP is sufficient to list EGUs and
regulate all HAP from those sources. 748 F.3d at 1244-45. The Court also found that the
85 U. S. EPA. 201 lg. Methods to Develop Inhalation Cancer Risk Estimates for Chromium and
Nickel Compounds. Available online at
http://cfpub.epa.gov/si/si_public_record_report.cfm?dirEntryId=238881. EPA-HQ-OAR-2009-
0234-19909.
86 California Department of Health Services (CDHS), 1991. Health Risk Assessment for Nickel.
Initial Statement for Rulemaking, Proposed Identification of Nickel as a Toxic Air Contaminant -
Technical Support Document, partB. Available online at
http://www.arb. ca.gov/toxics/id/ summary/nickel_tech_b .pdf.
87 Texas Commission on Environmental Quality (TCEQ), 2011. Development Support Document
for nickel and inorganic nickel compounds. Available online at
http://www.tceq.state.tx.us/assets/public/implementation/tox/dsd/final/junel l/nickel_&_compou
nds.pdf
70
-------
appropriate and necessary finding is "amply supported by EPA's findings regarding the health
effects of mercury exposure." Id. at 1244-45. Because of the Court's decision, objections relating
to the non-Hg metals findings are not of central relevance and are moot as the appropriate and
necessary finding can be sustained on the Hg findings alone.
Issue 61: Petitioner 20180 (att. 6) disagrees with EPA in that the relative speciation/component
data for Ni from EGUs does not need to be considered. According to Petitioner, the EPA
inappropriately equates the statements that all Ni compounds should be considered as a group for
cancer classification with the statement that all Ni should be considered as a group for
quantitative risk characterization, regardless of the toxicological and epidemiologic evidence.
Response to Issue 61: Issues regarding Ni compounds being considered as a group were raised
in public comments submitted in response to the proposed MATS rule. The EPA's responses to
the comments are in Section 1G (pp. 210 - 212) of the RTC.
In the RTC, the EPA stated that, in agreement with the peer reviewers of the Ni and Cr methods
document, it considers all Ni compounds to be potentially carcinogenic and did not focus on Ni
speciation or Ni solubility as strong determinants of Ni carcinogenicity. The EPA noted that
these views are based on reviews of the carcinogenic potential of Ni compounds by major
scientific bodies (i.e., the International Agency for Research on Cancer, the National Toxicology
Program, the World Health Organization) which conclude that the integrated evidence from
epidemiological studies, mechanistic studies and carcinogenesis studies in rodents support the
concept that Ni compounds should be considered carcinogenic, as a group. The EPA also noted
that there may be differences in toxicity and/or carcinogenic potential across the different Ni
compounds and the available Ni UREs are only 2-3 fold different. In the case studies for
MATS, the EPA applied 100 percent of the IRIS URE for nickle subsulfide, but it is reasonable
to consider a value that is 50 percent of the IRIS URE for nickel subsulfide for providing an
estimate of the lower end of plausible range of cancer potency values for different mixtures of Ni
compounds. Even if the EPA applied a value 50 percent of the URE for nickel subsulfide in the
case studies, several facilities would still exceed 1-in-l million cancer risk. Thus, this issue is not
of central relevance to the "Appropriate and Necessary" finding.
The comments received on this issue demonstrate that the public had ample opportunity to
comment on the issue and the EPA responded to those comments. Because Petitioner did not
demonstrate that it was impracticable to comment on this issue during the public comment period
on the proposed MATS rule, the EPA is denying the petition for reconsideration of this issue. In
addition, the issue is not of central relevance to the "Appropriate and Necessary" finding for the
reasons set forth above.
In addition, as stated above, the Court in While Stallion affirmed EPA's conclusion that a finding
that it is appropriate and necessary to regulate any one HAP is sufficient to list EGUs and
regulate all HAP from those sources. 748 F.3d at 1244-45. The Court also found that the
appropriate and necessary finding is "amply supported by EPA's findings regarding the health
effects of mercury exposure." Id. at 1244-45. Because of the Court's decision, objections relating
to the non-Hg metals findings are not of central relevance and are moot as the appropriate and
necessary finding can be sustained on the Hg findings alone.
71
-------
Issue 62: Petitioner 20180 (att. 6) alleges that the EPA continues to be overly conservative by
relying on dose-response values derived from exposures to mixed Ni compounds or solely to
nickel subsulfide.
Response to Issue 62: Issues regarding the Ni risk analyses based solely on nickel subsulfide
were raised in public comments submitted in response to the proposed MATS rule. The EPA's
responses to the comments are in Section 1G (pp. 210 - 212) of the RTC.
In the response to this previously raised issue, the EPA stated that after consideration of all the
available scientific evidence, and the recommendations from the Ni methods peer reviewers
(EPA-HQ-OAR-2009-0234-19744), the EPA decided to continue using 100 percent of the
current IRIS URE for Ni subsulfide because IRIS is at the top of preferred sources of dose
response information used in the EPA's risk characterizations, and because of the concerns about
the potential carcinogenicity of all forms of Ni raised by the major national and international
scientific bodies. The EPA also concluded that it is reasonable, in certain instances (e.g., when
high quality data are available on the composition of Ni emissions from a specific source), to
consider a value that is 50 percent of the IRIS URE for Ni subsulfide for providing an estimate of
the lower end of a plausible range of cancer potency values for different mixtures of Ni
compounds. Even if the EPA applied a value 50 percent of the URE for nickel subsulfide in the
case studies, several facilities would still exceed 1-in-l million cancer risk. Thus this issue is not
of central relevance to the "Appropriate and Necessary" finding.
The comments raised on this issue during the public comment period demonstrate that the public
had ample opportunity to comment on the issue and the EPA responded to those comments.
Because Petitioner did not demonstrate that it was impracticable to comment on this issue during
the public comment period on the proposed MATS rule, and because the issue is not of central
relevance, the EPA is denying the petition for reconsideration of this issue.
In addition, as stated above, the Court in While Stallion affirmed EPA's conclusion that a finding
that it is appropriate and necessary to regulate any one HAP is sufficient to list EGUs and
regulate all HAP from those sources. 748 F.3d at 1244-45. The Court also found that the
appropriate and necessary finding is "amply supported by EPA's findings regarding the health
effects of mercury exposure." Id. at 1244-45. Because of the Court's decision, objections relating
to the non-Hg metals findings are not of central relevance and are moot as the appropriate and
necessary finding can be sustained on the Hg findings alone.
1.21 EPRI's Multipathway Risk Assessment
Issue 63: Petitioner 20180 (att. 6) disagrees with the EPA's conclusion that EPRI's
multipathway risk modeling88 (submitted during the public comment period) does not support
delisting coal-fired EGUs and the report provides additional documentation supporting the
88 Electric Power Research Institute, 2011. Multi-Pathway Human Health and Ecological Risk
Assessment for a Model Coal-Fired Power Plant. August 2011. EPA-HQ-OAR-2009-0234-
17773.
72
-------
previous EPRI modeling. Further, Petitioner 20180 (att. 6) claims that the EPA's assertion that
EPRI's modeling does not represent a suitably worst-case scenario is unfounded. Petitioner
20180 (att. 6) asserts that EPRI's modeling, although indicating cancer risks up to 4-in-l million,
is overly conservative because of the modeling scenario and parameters for As, which is the
pollutant driving the cancer risk in the multipathway risk modeling. Petitioner 20180 (att. 6)
claims that if the As assumptions in EPRI's modeling were refined, the lifetime cancer risk for
all receptors and pathways would be well below 1-in-l million.
Further, Petitioner 20180 disagrees with EPA's conclusion that EPRI multipathway screening
risk modeling does not support delisting coal-fired EGUs, and the Petition states an intention to
".. .renew its petition to delist once non-Hg metals testing has been completed..." for certain
EGUs.
Response to Issue 63: Because the EPA did not rely on the EPRI multipathway assessment in
the "Appropriate and Necessary" finding, the EPA does not consider the additional
documentation on the EPRI multipathway risk assessment to be relevant to the "Appropriate and
Necessary" finding. Although the EPA did rely on EPRI's multipathway assessment in denying
the petition to delist coal-fired EGUs, Petitioner did not request reconsideration of the EPA's
decision to not delist coal-fired EGUs. Even if the As assumptions in EPRI's multipathway
assessment were overly conservative, as Petitioner claims, this issue is not of central relevance to
the decision to regulation EGUs under section 112. For these reasons, the EPA is denying the
petition for reconsideration of this issue.
In the proposed MATS rule (76 FR 24998), the EPA stated that EGUs do not meet the delisting
criteria in CAA section 112(c)(9) and, consequently, that they may not be delisted. Specifically,
the EPA stated that information in the Utility Study Report to Congress (U.S. EPA, 1998)
indicated that HAP emissions from a number of EGUs caused a lifetime cancer risk greater than
1-in-l million, and the EPA's non-Hg inhalation risk case studies indicated risks greater than 1-
in-1 million. (EPA-HQ-OAR-2009-0234-2939).89 During the public comment period for the
proposed MATS rule, the petition to delist coal-fired EGUs and the EPRI multipathway risk
assessment were both submitted. In the preamble to the final MATS rule (77 FR 9364), the EPA
stated that the petition to delist "improperly seeks to delist a portion of a CAA section 112(c)
listed source category", and that because the "request to delist is contrary to the plain language of
CAA section 112(c)(9)(B) and NRDC, we are denying the delisting petition." In addition to this
primary basis for denying the delisting petition, the EPA also concluded that estimated cancer
risks were still greater than 1-in-l million for several facilities in the non-Hg inhalation risk case
studies even after addressing public comments and peer review comments (EPA-HQ-OAR-2009-
0234-19912).90 These results alone are sufficient to deny the delisting petition. Lastly, EPRI's
multipathway assessment concluded that there were potential cancer risks greater than 1-in-l
89 U.S. EPA. 201 lh. Non-Hg Case Study Chronic Inhalation Risk Assessment for the Utility
MACT Appropriate and Necessary Analysis. Office of Air Quality Planning and Standards.
March 2011. EPA-HQ-OAR-2009-0234-2939.
90 U.S. EPA. 201 li. Supplement to Non-mercury Case Study Chronic Inhalation Risk Assessment
for the Utility MACT Appropriate and Necessary Analysis. Office of Air Quality Planning and
Standards. November 2011. EPA-HQ-OAR-2009-0234-19912.
73
-------
million, even though EPRI's multipathway assessment was not the primary basis for denying the
petition.
In addition, the EPA's denial of the delisting petition was challenged in the White Stallion case.
The Court affirmed the EPA's denial of the petition to delist. As the Court may only consider
issues raised during the period for public comment, issues raised in the litigation and addressed
by the Court clearly do not meet the criteria for reconsideration in CAA section 307(d)(7)(B)..
We are denying the petitions for reconsideration of this issue for all of these reasons.
74
-------
2.0 Affirmative Defense
Issue 64: Petitioner 20187 states that in an attempt to defend its "affirmative defense" for
violations of the CAA that occur during malfunctions, the EPA has advanced new arguments to
which Petitioners did not have an opportunity to object. Petitioner maintains that the affirmative
defense functions as a de facto exemption for facilities with a chronic history of violating CAA
limits.
Petitioner notes that during the comment period, the EPA received comments pointing out that
its proposed affirmative defense to penalties for violations of emission standards during periods
of malfunction ("affirmative defense") was unlawful and arbitrary because, among other things,
it purports to limit federal district courts' statutory authority to determine the appropriate penalty
for such violations. In its final rule, Petitioner states that the EPA revised the affirmative defense
and advanced new arguments in attempt to explain why it is not unlawful. Petitioner maintains
that because the EPA did not include these revisions and arguments in the proposed rule, it was
impracticable to raise objections to them and that because the EPA's affirmative defense
purports to limit the statutory authority of federal courts and undermines the enforceability of its
standards, Petitioner's objection is of central relevance and the final rule must be reconsidered on
this point.
Petitioner alleges that the affirmative defense in the final rule does nothing to meet the
continuous compliance requirement of the CAA. Petitioner maintains that limiting the
enforceability of these standards to obtaining injunctive relief for violations that meet certain
agency requirements does not make such standards enforceable. To the contrary, Petitioner states
that it makes them less enforceable by purporting to prevent district courts from imposing
penalties where the CAA unambiguously gives them authority to do so.
Petitioner also states that the EPA expresses a new "view" "that the affirmative defense is part of
the emission standard and defines two categories of violation," one category of violation being
when all the elements of the affirmative defense are met and only injunctive relief is available
and the other category including all other violations, for which both injunctive relief and
penalties are available. Petitioner states that the EPA further argues that because the CAA allows
it to establish enforceable emission limitations, it also allows the agency to establish some
limitations that are enforceable only through injunctive relief and that from that premise,
apparently the Petitioner infers that the EPA claims that its affirmative defense is "part of the
emission standard." Petitioner maintains that Congress did not intend to authorize citizen suits
for past violations where there was no "evidence that the violation has been repeated" but, other
than that, intended all emission limitations to be enforceable without limitation by "any person"
against "any person" and that once a violation has been established in a citizen suit, it is up to the
district court, not EPA, to determine what penalty if any should be applied. Petitioner states that,
contrary to EPA's claim, the agency may not establish any "category" of emission limitation that
is enforceable only by injunction and, thus, the affirmative defense is not, in any sense, a part of
any emission standard but rather an unlawful attempt by EPA to infringe on district courts'
jurisdiction to set penalties.
75
-------
Petitioner notes that the EPA argues that the affirmative defense "appropriately balances
competing concerns:" citizens' ability to enforce the law on the one hand and violators' concerns
about being penalized for penalties they regard as unavoidable and appropriately handled on the
other. Petitioner alleges that Congress struck the balance it viewed as appropriate in the text of
the CAA and that the EPA lacks authority to strike a different balance regardless of whether it
thinks that balance is more appropriate than the one Congress chose.
Petitioner states that the EPA also argues that the affirmative defense will not hamper citizen
enforcement by claiming that penalties will not deter violations where the criteria for the
affirmative defense are met and further claiming that citizens will have to litigate some of the
issues relevant to the affirmative defense in the penalty phase of litigation. Petitioner maintains
that the EPA's arguments are without merit - citizens may need to litigate some issues that might
be criteria for the affirmative defense at the penalty phase, such as whether the violation was
avoidable and whether the violator took adequate steps to prevent it and to minimize emissions.
Petitioner alleges that the EPA's affirmative defense creates factors that citizens must litigate
that they otherwise would not have to litigate - thus necessarily adding the expense and
complexity of enforcing emission standards - and prevents the consideration of factors that
might lead to the imposition of penalties.
Petitioner states that the EPA attempts to dismiss the significant differences between the penalty
criteria established by Congress and the criteria in its affirmative defense. Petitioner maintains
that the EPA's argument merely seeks to elevate EPA's policy preferences over the law and that
it is Congress and not the EPA that writes the law and Congress' decisions must be respected.
Further, Petitioner alleges that the EPA conveniently ignores other differences between CAA
section 113(e)(1) and the EPA's affirmative defense, including that CAA section 113(e)(1) does
not impose an absolute bar to penalties under any circumstances and directs courts to consider
"the violator's full compliance history."
Response to Issue 64: The EPA is denying this petition because the issue is moot. An
affirmative defense provision substantively the same as the MATS affirmative defense was
recently struck down in NRDC v. EPA, 749 F.3d 1055, 1063 (D.C. Cir. 2014). For this reason,
the EPA recently published proposed corrections to the MATS rule that include a proposal to
remove the affirmative defense provisions from the rule. See 80 FR 8442, 8448-9 (February 17,
2015)
Moreover, the EPA received comments on the affirmative defense to civil penalties in the
proposed MATS rule. As Petitioner notes, the EPA responded to the comments raised and made
revisions to the affirmative defense when issuing the final rule. See 77 FR 9382 - 9383; RTC,
Vol. 2, pp. 429 - 437. Petitioner maintains that the EPA raised "new arguments" in support of the
affirmative defense developed in response to the comments and the revisions to the affirmative
defense require the EPA to reconsider the affirmative defense so that it has an opportunity to
comment on the new rationale. The EPA does not agree that its response to comments requires
reconsideration of the affirmative defense under CAA section 307(d)(7)(B).
It is well established that "an agency must be able to respond flexibly to comments and need not
provide a new round of notice and comment every time it modifies a proposed rule." Fertilizer
76
-------
Institute v. EPA, 935 F.2d 1303, 1311 (D.C. Cir 1991); see also, InternationalFabricare
Institute v. EPA, 972 F.2d 384, 399 (D.C. 1992) (notice and comment is not intended to result in
"interminable back-and-forth"). Courts will generally not find a procedural error unless "the
agency's final rule so departs from its proposed rule as to constitute more surprise than notice."
Air Transport Ass 'n v. FAA, 169 F.3d 1, 7 (D.C. Cir 1999). The petition for reconsideration
generally argues that the affirmative defense is not consistent with the CAA, but suggests that the
new rationale put forth by the Agency is what justifies reconsideration. But the EPA responded
to arguments questioning the legality of the affirmative defense, and providing another round of
notice and comment is unnecessary.
77
-------
3.0 Averaging
Issue 65: Petitioner 20180 objects to the requirement in the final rule that facilities using an
alternative emissions averaging period (i.e., 90-day group boiler operating day) must comply
with a more stringent numerical Hg emissions limit than other facilities, even if the compliance
averaging period is longer, and claims that it is not clearly articulated as an alternative. Petitioner
also claims that because the EPA did not propose this approach, Petitioner was unable to
comment on it.
Response to Issue 65: The EPA proposed a 30-boiler operating day averaging period for all
standards and requested comments on all aspects of the proposed averaging approach. The EPA
specifically requested comment on whether the Agency should include a discount factor for
EGUs that comply with MATS by averaging emissions from multiple units. See 76 FR 25053-
54. The EPA received comments on the proposed averaging, and the Agency responded to the
comments received on the issue. RTC, Vol, 2, pp. 342-365. The EPA received comments both in
favor and opposed to a discount factor, and the Agency also received comments that suggested a
longer averaging period should be employed. See, e.g., RTC, Vol. 2, p. 354. In the final MATS
rule, the EPA determined that a discount factor was not appropriate for sources that average on a
30-day basis, but the Agency concluded that with the increase in averaging time to 90 days a
discount factor was appropriate, in part, because there is less variability in Hg emissions over a
90-day period. 77 FR 9385.
Because the commenter has not demonstrated that it was impracticable to comment on this issue
during the public comment period, and because the revisions to the averaging provisions for Hg
constitute a logical outgrowth of the proposed rule, the Agency is denying reconsideration on
this issue.
In addition, the averaging provisions were challenged in the White Stallion case, and the Court
affirmed the EPA's action. As the Court may only consider issues raised during the period for
public comment, issues raised in the litigation and addressed by the Court clearly do not meet the
criteria for reconsideration in CAA section 307(d)(7)(B).
Issue 66: Petitioner 20180 claims that it was unable to comment on the definition of the term
"30- (or 90-) group boiler operating day," which was included in the final rule but not the
proposal, though Petitioner supports the use of an operating day period. Petitioner also claims
that it was unable to comment on the equations used to calculate weighted rolling averages, as
they were included in the final rule but not the proposal. Petitioner also suggests corrections to
the equations.
Response to Issue 66: The EPA proposed a 30-boiler operating day averaging period be used for
those EGUs contained in an emissions averaging group and requested comment on all aspects of
the proposed emissions averaging approach.91 As mentioned above, the EPA also received
comments that suggested a longer averaging period should be employed. See, e.g., RTC, Vol. 2,
p. 354. In the final MATS rule, the EPA determined that an increase in averaging time to 90
91 See 76 FR 25078, May 3, 2011.
78
-------
days, along with use of a discount factor, to be appropriate. 77 FR 9385. Although the final rule
maintained a 30-group boiler operating day averaging period for those EGUs contained in an
emissions averaging group, it also included a 90-group boiler operating day averaging period as
an alternative standard for Hg emissions from existing coal-fired EGUs that do not use low rank
virgin coal. Just as adding the alternative averaging period of 90 days was a logical outgrowth
based on comments, so is including this 90-day averaging period in the definition of group boiler
operating day. This affords EGU owners or operators the ability to choose between a 30- or a 90-
group boiler operating day emissions standard for Hg emissions from existing coal-fired EGUs
that do not use low rank virgin coal and reduces the number of steps necessary to determine
compliance. For these reasons, we maintain that the final rule is a logical outgrowth of the
proposed rule.
The technical corrections identified in Table 4 and made to Equations 2a and 3a in 40 CFR
63.10009(b) in the proposed MATS reconsideration notice (see 77 FR 71332; November 13,
2012), proposed clarifications that address the concerns identified by the Petitioner relating to the
equations. These clarifications were finalized on April 24, 2013 (78 FR 24073). This issue was
further addressed in the Technical Corrections notice proposed on February 17, 2015 (see 80 FR
8445) by proposed revisions to equations la and lb in 40 CFR 63.10009(b)(1). As a result, the
Petitioner's comments are moot and we are denying reconsideration on this issue for this reason
as well.
79
-------
4.0 Beyond-the-floor - Low Rank Virgin Coal Hg Limit
Issue 67: Petitioner 20174 objects to the EPA's response to its and others' comments on the
proposed rule regarding the beyond-the-floor (BTF) analysis for the Hg emission limit for
existing sources designed to burn low rank virgin coal. Petitioner claims that the EPA's response
is "both unresponsive and misrepresentative of the comment and the information that EPA
provided at proposal."
Petitioner claims that at proposal the BTF TSD analysis for low rank virgin coal EGUs did not
specifically explain the technological feasibility of the BTF limit. Petitioner acknowledges that
the TSD did provide cost estimates associated with fuel switching to non-low rank virgin coal
coals or activated carbon injection (ACI) control retrofits. Petitioner also states that the TSD
specifically indicated that EPA's IPM modeling projected 4.7 GW of capacity would switch
from low rank virgin coal to subbituminous coal to comply with the BTF Hg limit.
Petitioner states that though the EPA indicated in the proposed rule's preamble that ACI was not
being used to its "fullest extent," on the low rank virgin coal EGUs not meeting the proposed
limit, the EPA provided no technical analysis to justify the BTF limit in the rule. Petitioner states
that its comment on the proposed rule challenged the EPA's BTF determination and questioned
the appropriateness of setting a BTF standard for a subcategory that resulted in EGUs switching
out of that fuel category. Petitioner maintains that in response to its comment, the EPA denied
knowledge of any EGUs that switch fuel to comply with the BTF MACT limit. Petitioner
contends that the EPA also removed all discussion and cost information associated with fuel
switching from low rank virgin coal to subbituminous coal in the final version of the BTF TSD.
Petitioner maintains that the EPA "failed to address the specific question raised by its own initial
IPM modeling results and questioned in comments" by Petitioner. Petitioner further maintains
that the EPA's statement that it knows of no EGUs that utilize fuel switching to comply with the
Hg MACT limit does not answer the question that Petitioner specifically asked in its comments
on the proposed rule. Petitioner acknowledges that the EPA did supplement the final BTF TSD
with expected EGU-specific Hg reductions necessary to meet the BTF Hg limit based on IPM
projections and that the EPA states in the final rule's BTF TSD that the facilities would need less
than 90 percent Hg reduction to meet the standard and that use of ACI could achieve greater than
90 percent Hg reduction. Petitioner contends that this technical information, however, was not
provided in the TSD with the proposed rule for comment.
Response to Issue 67: The EPA received comments on the proposed MACT standard for Hg
emissions from low rank virgin coal-fired EGUs that was based on a BTF level of control, and
the EPA responded to the comments received (see, e.g., RTC, Vol. 1, section 4C01). The EPA
disagrees with Petitioner's claims that the response to the comments was inadequate and the
Petitioner does not provide any new information that could not have been submitted during the
comment period. For the reasons discussed below, the EPA is denying the Petition for
Reconsideration on the issues raised.
At proposal, the EPA provided two documents supporting the BTF analysis for the Hg emission
limit for existing low rank virgin coal EGUs. One is titled "National Emission Standards for
80
-------
Hazardous Air Pollutants (NESHAP) Beyond the Maximum Achievable Control Technology
(MACT) Floor Analysis for Coal- and Oil-fired Electric Steam Generating EGUs" (EPA-HQ-
OAR-2009-0234-2924), and the other is titled "Emission Reduction Costs for the Beyond-the-
Floor Mercury Rate in the Toxics Rule" (EPA-HQ-OAR-2009-0234-2925). These documents
provide details on the data, assumptions and methodologies used to set the proposed BTF
emissions standard and to evaluate the cost-effectiveness and other impacts from achieving the
stricter standard.
The analysis in the first TSD noted above (EPA-HQ-OAR-2009-0234-2924) analyzes facilities'
current control configurations and the controls needed to meet other MATS emissions standards
(especially the particulate matter (PM) standard). The document explains that the emission limit
of 4.0 lb/TBtu was based on the performance of the best controlled similar source which "is
equipped with ACI and a fabric filter (FF)." The document then explains that all EGUs without
ACI or a FF would add the technology - with the FF being added primarily to meet the PM limit.
Clearly the expectation was that existing EGUs would be able to meet the BTF limit using the
ACI plus FF technology configuration. Table 5 in the TSD lists specific EGUs with expected
BTF controls added.
Further, in "Documentation Supplement for EPA Base Case v4. lO PTox - Updates for Proposed
Toxics Rule" (March 2011, available at http://www.epa.gov/airmarkt/progsregs/epa-
ipm/docs/suppdoc.pdf and at EPA-HQ-OAR-2009-0234-3048), the EPA explains specific
assumptions in IPM regarding the technical performance of ACI:
The technology specifically designated for mercury control is Activated Carbon
Injection (ACI) downstream of the combustion process in coal fired units. In
preparation for performing modeling of air toxics, a comprehensive update of ACI
cost and performance assumptions was undertaken by Sargent & Lundy, the same
engineering firm that developed the SO2 and NOx control assumptions used in
EPA Base Case v4.10. The ACI update, whose elements are described below,
incorporates the latest field experience through 2010.
Assuming a target of 90% removal from the level of mercury in the coal, three
alternative ACI options were identified as providing the required rate of removal
for all possible configurations of boiler, emission controls, and coal types used in
the U.S. electric power sector. The three ACI options differed based on the type
of particulate control device - electrostatic precipitator (ESP) or pre-existing or
new fabric filter (also called a "baghouse"), i.e.,
• ACI with Existing ESP
• ACI with Existing Baghouse
• ACI with an Additional Full Baghouse (also referred to as Toxecon)
All three configurations assume the use of brominated ACI, where a small amount
of bromine is chemically bonded to the powdered carbon which is injected into
the flue gas stream. The use of brominated ACI exploits the discovery that by
converting elemental mercury to oxidized mercury, halogens (like chlorine,
iodine, and bromine) can make activated carbon more effective in capturing the
mercury at the high temperatures found in industrial processes like power
81
-------
generation. The ionic mercury adheres to the activated carbon (and to fly ash and
unburned carbon in the fuel gas) which can be removed efficiently from the flue
gas by the particulate control device (ESP or fabric filter). In the third option
listed above the additional baghouse is installed downstream of the pre-existing
particulate matter device and the activated carbon is injected after the existing
controls. This configuration allows the fly ash to be removed before the mercury
controls to preserve its marketability.
The applicable ACI option depends on the coal type burned, its SO2 content, the
boiler and particulate control type and, in some instances, consideration of
whether an SO2 scrubber (FGD) system and SCRNOx post-combustion control
are present. Table 5-16 shows the ACI assignment scheme used in EPA Base
Case v4.10_PTox to achieve 90% mercury removal.
Table 5-16 of the document details control configurations and assumed sorbent injection rates to
achieve 90 percent Hg removal. This includes configurations for EGUs burning low rank virgin
coal (listed as lignite in the table).
The second TSD noted above (EPA-HQ-OAR-2009-0234-2925) details IPM modeling results
for the proposed MATS. In the document, the EPA explains that it projects two methods for
compliance: fuel switching from lignite to subbituminous coal and retrofitting ACI systems.
And, as Petitioner claims, the TSD specifically indicates that the EPA's IPM modeling projected
that 4.7 GW of capacity would switch from low rank virgin coal to subbituminous coal to
comply with the MATS standards (though it is not certain that the BTF Hg limit would be the
basis for the fuel switching decision). Petitioner claims that "in response to [Petitioner's]
comment, the EPA denied having any knowledge of any EGUs switching fuel to comply with the
BTF MACT limit." The EPA disagrees with Petitioner's claim. In the RTC (Vol. 1, p. 578) the
EPA explained as follows:
The EPA did not state that some units in this subcategory would have to switch
fuels in order to comply with the beyond-the-floor Hg standard as the commenter
suggests, but even that conclusion would not have made the beyond-the-floor
standard invalid, [emphasis added]
The TSDs make clear that the EPA believed that ACI plus FF was an adequate control
technology. The EPA's analysis in the first TSD was based entirely on the use of ACI plus FF.
The EPA did not state that some EGUs would have to switch fuels in order to comply with the
BTF Hg standard. Yet, EPA also did not claim that no EGUs would decide to switch fuels in
order to comply. The IPM considers fuel switching as an operational decision, and EGUs switch
fuels for a variety of reasons, such as compliance with regulatory requirements or economic
factors (i.e., burning an alternative fuel may cost less). The EPA is aware of several EGUs
designed to burn low rank coal that switched to burning subbituminous coal - a switch done
before MATS was proposed. In the RTC (Vol. 1, p. 578), as noted above, the EPA stated:
As with all MACT standards, sources are required to comply with the emissions
limits and the EPA does not dictate how compliance must be attained. Fuel
82
-------
switching is certainly among the available options for all EGUs subject to this
final rule, including EGUs in the low rank, virgin coal subcategory. We disagree
with the commenter that the beyond the floor standard assumes sources will have
to switch fuels to comply and we have no data to suggest that is the case.
[emphasis added]
In the final MATS rule, the EPA provided a TSD to support the BTF Hg emission limit. This
document is titled "Emission Reduction Costs for Beyond-the-floor Mercury Rate for Existing
Units Designed to Burn Low Rank Virgin Coal" (EPA-HQ-OAR-2009-0234-20130). Petitioner
is correct in its claim that the final BTF TSD does not discuss any EGU having to switch fuels in
order to comply with the BTF Hg limit or costs associated with fuel switching. Rather, the
document discusses the use of brominated ACI as the expected compliance option. The IPM
modeling for the final MATS rule predicted that only one low rank coal-fired EGU would
choose to switch to subbituminous coal. However, that EGU was also predicted to install an ACI
system, and it is not clear if the fuel switch would be for compliance or economic reasons.
Petitioner states that the EPA, in the MATS proposal preamble, indicated that ACI on the low
rank virgin coal EGUs not meeting the proposed BTF limit was not being used to its "fullest
extent" but provided no technical analysis justifying the BTF limit. In the preamble the EPA
wrote:
EPA has learned that the units of this design that were using ACI during the
testing were using ACI to meet their permitted Hg emission levels. However,
EPA believes that the control level being achieved is still not that which could be
achieved if ACI were used to its fullest extent. (76 FR 25046)
The EPA clearly stated that the EGUs using the ACI during the testing were using the
technology to meet permitted emission levels, which are much higher than the Hg BTF value.
ACI is a "tunable" technology - i.e., the performance can be "tuned" to a certain degree by
controlling the amount of injected sorbent. Increased sorbent injection rates result in increased
Hg capture up to a certain point. The primary cost of ACI operation is the cost of sorbent, and
the EPA maintains that EGUs striving to meet permit levels less stringent than the MATS
emission limit would not be motivated or obligated to "over control" emissions, because it would
require increased sorbent use beyond that which is required for compliance and hence incur
unnecessary cost. So, if an EGU is using ACI to meet its "permitted Hg emission levels," and
only to those levels, then the EPA believes that ACI is not being used "to its fullest extent." In
fact, the only EGU (Monticello, ORIS ID 6147) using ACI that did not meet the BTF Hg limit
had the lowest ACI rate of all of the EGUs that were using ACI during ICR testing. Commenters
did not provide data to rebut the EPA's statements concerning ACI utilization by EGUs in this
subcategory.
Because the Petitioner has not demonstrated that it was impracticable to comment on this issue
during the comment period, we are denying reconsideration of this issue.
In addition, the BTF limit was challenged in the White Stallion case. The Court upheld the BTF
standard, noting that the issue ultimately amounted to a factual dispute, and found that "because
83
-------
the record contains no data inconsistent with EPA's position on the efficacy of activated carbon
injection, we defer to the agency's determination that the beyond-the-floor emission standard for
lignite-fired EGU's is achievable." White Stallion, 1222 F.3d at 1251. As the Court may only
consider issues raised during the period for public comment, issues raised in the litigation and
addressed by the Court clearly do not meet the criteria for reconsideration in CAA section
307(d)(7)(B). The Petitioners here have not provided any new data calling into question the
EPA's factual conclusion, therefore, we are also denying this petition as moot.
84
-------
5.0 Bias Against New Cogeneration EGUs
Issue 68: Petitioner 20174 states that all new EGUs subject to the rule must comply with output-
based only emission standards and, although the definition of "gross output" in the rule considers
cogeneration EGUs, the output-based emission calculations in the rule's appendices are based
solely on "gross electrical output." Petitioner claims that because cogeneration EGUs only use a
portion of their steam to produce electric power, the rule is inherently biased against new
cogeneration EGUs. Petitioner claims that the rule's requirement that new EGUs comply with
output-based emission limits, although only considering gross electrical output in the emission
calculations for those limits, results in inflated lb/MW or lb/GW emissions for cogeneration
EGUs, because these EGUs' gross electric output can be substantially lower than their true total
output. Petitioner suggests that the rule should be revised to allow new cogeneration EGUs to use
an alternate output-based emission calculation.
Response to Issue 68: The issue of addressing thermal energy production in MATS was raised
in public comments submitted in response to the May 3, 2011, proposed MATS rule (76 FR
24976). The EPA responded to the comments submitted on this issue and revised the final rule to
address this issue. (RTC, Vol. 2, p. 482).
As noted in the RTC, the final rule provides 75 percent credit for thermal output in the definition
of "gross output." We believe that this is a solution to the Petitioner's issue, and the Petitioner
does not provide any new information or data to support its argument to the contrary (e.g., what
specific "alternate output-based emission calculations" should be allowed). The EPA believes
that the final definition is consistent with the EPA's stated intent of accommodating cogeneration
EGUs. The EPA believes that applicability determinations in which the case-by-case specifics
may be analyzed and addressed are the more appropriate forum to address the other issues raised
by Petitioner, rather than a rule.
Because Petitioner has not demonstrated that it was impracticable to comment on this issue
during the comment period on the proposed rule, the EPA is denying the petition of this issue.
The EPA is also denying this issue because it is not of central relevance for the reasons set forth
above.
85
-------
6.0 Common Stack
Issue 69: Petitioner 20180 claims that the rule does not resolve how the startup and shutdown
definitions and work practice provisions apply to EGUs that share a common stack and
acknowledges that the work practice provisions can be applied separately on each EGU.
Petitioner requests that the rule establish that emissions limits do not apply unless both EGUs
sharing a common stack are not starting up or shutting down. Petitioner also claims that it was
unable to comment on a new option that allows owners or operators of EGUs that share a
common control device or devices that discharge to the atmosphere through multiple stacks the
ability to monitor each stack or duct or to monitor with a CMS or sorbent trap monitoring system
for one stack and flow and dilution rates in all stacks or ducts.
Response to Issue 69: The EPA received comments requesting a work practice standard for
periods of startup and shutdown and also on the optional common stack monitoring approach
included in the final rule. See, e.g., RTC, Vol.1, p. 563 and RTC, Vol. 2 pp. 138 and 354,
respectively. The Agency responded to these comments.
First, concerning startup and shutdown, the EPA finalized a work practice standard in lieu of
requiring compliance with the numerical standards during periods of startup and shutdown. The
EPA reconsidered the startup and shutdown work practice in the MATS reconsideration and
provided an opportunity for comment on all aspects of that work practice (see 77 FR 71323;
November 30, 2012). The EPA received comments on common stack issues related to startup
and shutdown periods, and the Agency responded to those comments in the final startup and
shutdown reconsideration action that published in the Federal Register on November 19, 2014.
79 FR 68777. Because the Agency has already reconsidered these issues, the request for
reconsideration of these issues is now moot.
As to Petitioner's second claim concerning being unable to comment on the optional monitoring
approach, the EPA disagrees. As noted above, the EPA received comments relevant to this
option (see, e.g., RTC, Vol. 2, p. 138), and the change in the final rule was based on those
comments. Thus, the Petitioner has not demonstrated it was impracticable to comment during the
comment period. See NRDC v. Thomas, 838 F.2d 1224, 1242 (D.C. Cir 1988) and Small Refiner
Lead Phase-Down Task Force v. EPA, 705 F.2d 506, 547 (D.C. Cir. 1983) (Agency may make
changes to proposed rule without triggering new round of comments, where the changes are
logical outgrowth of the proposal and comments). Where, as here, the EPA framed the subjects
for discussion, interested parties commented on the subjects, and EPA responded directly to
these comments by modifying the proposal, the purpose of the notice requirement is satisfied.
See Fertilizer Inst. V. EPA, 935 F.2d 1303 (D.C. Cir. 1991) (citing Small Refiner, 705 F.2d at
547) ("an agency must be able to respond flexibly to comments and need not provide a new
round of notice and comment every time it modifies a proposed rule").
This issue was further addressed in the final action on reconsideration of the startup and
shutdown provisions in the MATS and Utility NSPS on November 19, 2014 (see 79 FR 68784).
Consistent with the monitoring provisions in the final rule, owners or operators of EGUs with
common stacks are required to monitor and report emissions for compliance purposes at all times
when any EGU using a common stack is operating in a non-startup/shutdown mode, even if
86
-------
another EGU using that common stack is in startup/shutdown mode. 40 CFR 63.10005(a)(2) was
amended by adding paragraph (iii) which reinforces and clarifies this requirement.
Moreover, as the monitoring approach is optional, the Petitioner's comment is not of central
relevance. Since the Petitioner is able to choose another approach for stack measurement, even
acceptance of Petitioner's claim would not necessitate a change in the final rule. Because the
commenter has not demonstrated that it was impracticable to comment on the optional
monitoring approach, as evidenced by the comments the EPA received on the issue, we are
denying reconsideration on this issue. Further, because the alternative monitoring approach was
included in response to comments and is optional, we are denying reconsideration of the issue for
that reason as well. The petition issues concerning startup and shutdown are moot as the EPA has
reconsidered those provisions and the commenter had an opportunity to comment on that rule.
87
-------
7.0 Deadlines
7.1 Deadlines for "Newly Applicable Monitoring Requirements"
Issue 70: Petitioner 20180 states that the EPA added a new provision establishing a deadline for
monitoring system installation, operation and testing for EGUs that become subject to "newly
applicable monitoring requirements" as a result of becoming subject to 40 CFR Part 63, subpart
UUUUU. Petitioner states that because EPA did not propose this provision, Petitioner was not
able to comment on it.
Petitioner maintains that the provision makes no sense as promulgated. According to Petitioner,
in addition to addressing EGUs that become "subject to" 40 CFR Part 63, subpart UUUUU, the
provision requires installation of monitoring systems and completion of some tests "as of the
date your source ceases to be" subject to 40 CFR Part 63, subpart UUUUU, and that the
"[rjelative accuracy test must be performed as of the performance test deadline for PM CEMS, if
applicable." Petitioner maintains that the reference to EGUs that "cease to be" subject to the rule
is incongruent. According to Petitioner, if the provision is intended to address EGUs that are
newly subject to 40 CFR Part 63, subpart UUUUU, a deadline cannot be established based on the
date the EGU ceases to be subject to the subpart. Petitioner asserts that establishment of a
relative accuracy test audit (RATA) deadline based on the "performance test" deadline for PM
continuous emission monitoring systems (CEMS) also is nonsensical, as PM CEMS are not
subject to relative accuracy testing, and not all EGUs that must perform RAT As under the rule
will use PM CEMS. Petitioner asserts that the EPA must issue and solicit comment on a proposal
that articulates clearly what the Agency intended.
Response to Issue 70: In the proposed MATS rule, the EPA discussed the fact that certain units
could at times meet the definition of an EGU subject to the MATS rule and at other times not
meet the definition of an EGU and instead meet a definition that subjects the unit to another
CAA rule. In the proposal, the Agency specifically discussed co-generation units because the
CAA section 112 definition of such units includes thresholds that could be met at some times but
not others. 76 FR 25025-26. The same is also true for an EGU that combusts solid waste such
that it becomes as solid waste incineration unit. The proposed rule stated, in part:
EPA solicits comment on the extent to which [sources changing applicability] might
occur and whether the 6-month period is appropriate. Given the difference between the
rules, should EPA address the reclassification of the sources between the rule,
particularly with regard to initial and ongoing compliance requirements and
schedules^....We specifically solicit comment on whether we should include provisions
similar to those included in the final CISWI rule to address units that combust different
fuels at different times. See Final CISWI Rule, 40 CFR 60.2145.
76 FR 25026 {emphasis added).
As can be seen from the proposed rule, the Agency did provide notice that it was considering the
inclusion of additional provisions to address EGUs that might at times not meet the definition of
an EGU and potentially be subject to another CAA rule. We specifically requested comment on
88
-------
the need for additional compliance requirements, such as the monitoring provisions that the
Petitioner identifies. Further, the CISWI provisions that are referenced in the rule address units
that cease to be subject to that rule and provide for the compliance assurance mechanisms to
ensure that the source is in continual compliance with the alternately applicable requirements.
The EPA received comments suggesting that EPA include provisions similar to those in the
CISWI rule. RTC, Vol. 1, pp. 272 - 74. The EPA responded that it was including such provisions
in the final MATS rule.
The EPA finds the Petitioner's claims concerning relative accuracy testing and PM CEMS
mistaken. Petitioner misreads 63.10000(k) - ".. .Relative accuracy tests must be performed as of
the performance test deadline for PM CEMS, if applicable..." - as mandating relative accuracy
test audits (RAT As), despite that language having no mention of RAT As, and the Petitioner
chooses to ignore the conditional phrase "if applicable."
While RATAs are one means of providing relative accuracy checks, they are not the sole means,
and, in any event, other procedures for ensuring the quality of data (i.e., its accuracy) produced
by PM CEMS exist. These other procedures include relative response audits (RRAs) and
response correlation audits (RCAs) and their use is specified in Procedure 2 in Appendix F to 40
CFR part 60 for use by owners or operators of PM CEMS. Moreover, the final rule in
63.10010(i)(2) and Table 5, as well as the preamble at 77 FR 9371 and 77 FR 9420, specifies use
of Procedure 2 in Appendix F to 40 CFR 60. Furthermore, the instruments require ongoing
quality assurance checks and that one instance of an overly generic term (relative accuracy) to
refer to ongoing quality assurance checks does not negate the need for such checks or the many
other instances in which the specific terms for ongoing quality assurance checks for PM CEMS
(RRAs and RCAs, as well as the use of QA Procedure 2) were used in the rule. In addition, the
petitioner's comment regarding timing being 'incongruent' makes no sense. Every instrument,
once installed, must meet the ongoing accuracy provisions. In this case, the installed instrument
(PM CEMS) must be current with its ongoing accuracy requirements prior to its use, meaning
that if the instrument has already performed its accuracy checks, no further activity need occur
before beginning to collect data. Such a condition is unlikely for an instrument connected to a
source that has just become applicable to subpart UUUUU. The more likely case is as the rule
describes: A source becomes an EGU by switching fuels and must conduct the relative accuracy
checks upon switching, so that the data collected after the switch are valid. For the reasons
above, we find this issue is not of central relevance and we are denying reconsideration for this
reason.
Because EPA specifically identified in the proposed MATS rule the provisions identified in the
petition for reconsideration, and some commenters suggested EPA include provisions similar to
the CISWI rule, the Petitioner was on notice that additional monitoring and other requirements
might be included in the final rule to address sources that might at times meet and not meet the
definition of an EGU subject to the MATS rule. Because the Petitioner has not demonstrated that
is was impractical to comment on this issue during the comment period, the EPA is denying
reconsideration of this issue. See NRDC v. Thomas, 838 F.2d 1224, 1242 (D.C. Cir 1988) and
Small Refiner Lead Phase-Down Task Force v. EPA, 705 F.2d 506, 547 (D.C. Cir. 1983)
(Agency may make changes to proposed rule without triggering new round of comments, where
the changes are logical outgrowth of the proposal and comments).
89
-------
7.2 January 1, 2012, start date for reporting of performance test results
Issue 71: Petitioner 20180 states that the EPA proposed to require, "[a]s of January 1, 2012,"
electronic reporting of any performance test data recorded under the rule within 60 days of
completing the test. Petitioner states that it commented on the reporting requirement but did not
object to the start date because it was not clear at the time of proposal that the rule would not be
effective on the first possible date a report could be due under the rule, and Petitioner assumed
the EPA would adjust the date in the final rule to account for any change in timing. Petitioner
asserts that the EPA did not finalize the rule on the schedule it had initially announced and also
did not adjust the reporting start date. According to Petitioner, the rule requires EGUs to report
the results of performance tests required under subpart UUUUU starting on January 1, 2012, and
Petitioner objects to this requirement. Petitioner states that although existing EGUs are not
required to conduct performance tests until 180 days after April 16, 2015, at the earliest, new and
reconstructed EGUs must comply by 180 days after April 16, 2012, (or, if later, startup) and may
rely on tests conducted before that date. As a result, according to Petitioner, a new or
reconstructed EGU that conducts a stack test to show compliance with 40 CFR Part 63, subpart
UUUUU, after January 1, 2012, could be required under the rule to report the results "before the
rule becomes effective on April 16, 2012." Petitioner maintains that the EPA could not have
intended that result. Petitioner states that if it had known at the time of proposal that finalization
of the rule would be delayed, Petitioner would have objected to the January 1, 2012, date.
Response to Issue 71: The EPA disagrees with Petitioner because the Petitioner misunderstands
the basis for establishing a starting date for electronic reporting of performance test data.
Although no EGU owner or operator is required to begin electronic reporting of performance test
data prior to the applicability date in the rule (no later than April 2015 for many existing EGUs),
some EGU owners or operators may choose to seek LEE status for their EGUs as early as the
promulgation date of the rule. As 3 years of performance test results are needed to qualify for
LEE status for most non-Hg emissions, it is important to let EGU owners or operators know that
electronic reporting of performance test data is available in 2012. Further, as Petitioner admits,
the proposal contained the starting date for electronic reporting, Petitioner provided comments
but did not object to the starting date, and the rule was finalized with the proposed starting date.
In addition, the Petitioner's claim that it did not comment because initially the final MATS
signature date was supposed to be one month earlier is not credible. Major rules often take well
over a month to be published in the Federal Register (the proposed MATS rule took over a
month and a half), and major rules are not effective until 60 days after publication in the Federal
Register pursuant to the Congressional Review Act. Thus, even if the EPA had signed the rule on
November 16, 2011, as initially required under Consent Decree, and the rule published in the
Federal Register the next day, the January 1, 2012, date would still have been prior to the
effective date of the final rule. In any case, the EPA received comments on this issue, which
demonstrates that Petitioner had ample opportunity to comment on various aspects of the January
1, 2012, issue, and the EPA responded to those comments (see, e.g., RTC, Vol. 2, pp. 238, 278,
279 - 280, 283). Because Petitioner did not demonstrate that it was impracticable to comment on
this issue during the comment period on the proposed rule, the EPA is denying reconsideration of
this issue.
90
-------
In addition, the EPA is denying the Petition for Reconsideration on this issue because the issue is
moot. The EPA is not aware of any EGU that faced reporting requirements before the effective
date of the rule. The EPA maintained the January 1, 2012, date in order to allow owners or
operators to be able to provide emissions data in advance of their compliance date, which is
necessary if a source owner or operator wants to qualify for low emitting EGU (LEE) status on
the compliance date for existing EGU. For all of these reasons, we are denying reconsideration
of this issue.
7.3 Deadline for certification of Hg CEMS and sorbent trap monitoring systems
Issue 72: Petitioner 20180 states that the EPA proposed to require that all Hg CEMS and sorbent
trap monitoring systems be "certified" prior to "the applicable compliance date." According to
Petitioner, under the proposed rule, new and reconstructed EGUs were required to comply by the
later of publication of the final rule or startup, and existing EGUs had 3 years from publication of
the final rule to comply.
Petitioner asserts that the final rule differs in several respects from the proposed rule. First, rather
than establish a fixed deadline, the final rule requires "certification" of monitoring systems "in a
timely manner, such that the initial compliance demonstration is completed" by the applicable
deadline. Petitioner supports this change because the change allows EGUs to certify their
monitoring systems during the 180-day period provided for the initial compliance demonstration.
Petitioner notes that rather than cite final 40 CFR 63.9984(f), which establishes the deadlines for
initial compliance demonstrations for all EGUs, the final rule refers to the applicable date in 40
CFR 63.10005(g), which establishes the deadline for initial compliance demonstrations for new
and reconstructed units only. Petitioner states that the rule references no deadline for existing
units and that other provisions discussing the deadlines for initial compliance demonstrations
include references to both 40 CFR 63.9984(f) and 40 CFR 63.10005(g).
Petitioner maintains that the EPA should revise Appendix A, section 4.1 to similarly reference
both provisions. Petitioner also states that alternatively the EPA could remove the deadline.
According to Petitioner, because the rules already require performance tests to be conducted with
a "certified CEMS," a deadline for CEMS certification is not necessary; the deadline for the
initial compliance demonstration achieves the same result. Petitioner notes that the rules do not
specify a separate "certification" deadline for EGUs opting to use SO2, PM, hydrogen chloride
(HC1) or hydrogen fluoride (HF) CEMS.
Response to Issue 72: As explained in the November 2012 proposed MATS reconsideration
notice, the reference to 40 CFR 63.100005(g) in Appendix A, section 4.1, was incorrect; this was
corrected (with the correct cite to 40 CFR 63.9984(f) as the Petitioner suggests) in the final new
source reconsideration notice (see 78 FR 24073; April 24, 2013). Because the issue has been
addressed in the MATS new source reconsideration, the EPA is denying reconsideration of this
issue as moot.
91
-------
8.0 Definitions - Applicability
Issue 73: Petitioner 20180 states that proposed Subpart UUUUU would have applied to each
coal- or oil-fired EGU, as defined in 40 CFR 63.10042. In May 2011, the EPA proposed
definitions of "fossil fuel-fired," "coal-fired electric utility steam generating unit," and "oil-fired
electric utility steam generating unit" that would have covered any unit that combusted a fossil
fuel (e.g., natural gas, oil, or coal) for more than a 10 percent of its average annual heat input in
the "previous 3 calendar years" or 15 percent in "any one of those calendar years," or that
currently combusts oil "alternately with" other fuels. Petitioner notes that where the EPA
proposed to exclude EGUs that did not actually combust coal or oil in significant amounts (e.g.,
gas-fired EGUs and biomass-fired EGUs), the EPA also referred to fuel combustion in
"previous" calendar years. Commenters noted that under those definitions and provisions, the
rule would apply to coal- and oil-fired EGUs that convert to combustion of alternative fuels (e.g.,
natural gas or biomass) between 2012 and 2015, because they either would have combusted
fossil-fuel in a previous calendar year (or years) or because they still combust some oil (e.g., for
startup or emergencies).
Petitioner notes that in the preamble to the final rule, the EPA explains which EGUs it intended
to cover and acknowledges that the definitions "as proposed, were not sufficiently descriptive."
Petitioner states that to address concerns regarding fuel conversion, the EPA says it revised the
definition of "fossil fuel fired" so that the fossil fuel combustion thresholds "are evaluated after
the applicable compliance date of the final rule on a rolling basis." Petitioner states that the EPA
similarly states in the response to comments that the definition was revised "to make clear that
units must only look at the present capability and utilization of the unit to determine if it is an
affected source on the compliance date."
Petitioner also states that the EPA goes on to state that an EGU "that spends the 3 year (or 4
years if necessary for the installation of controls) compliance period converting the unit to
natural gas or biomass will not be subject to this rule based on our revised definition of fossil
fuel-fired."
However, Petitioner alleges that the only change the EPA made to the final definition of "fossil
fuel-fired" was to change the reference to "previous" year(s) to "any" year(s) and add the "any"
calendar year thresholds to the final definitions of "coal-fired electric utility steam generating
unit" and "oil-fired electric utility steam generating unit," and included similar thresholds in a
new definition of "natural gas-fired electric utility steam generating unit."
Petitioner alleges that there is no language in any of the relevant definitions, or applicability
provisions, to implement the Agency's stated intent to look only at present capability and
utilization when determining how to categorize an EGU with respect to fuel combustion and
applicability. Moreover, Petitioner notes that because "any" year can include not only the
"previous 3 calendar years," but also "any" year in the past, the final definitions would actually
expand the number of EGUs covered by the rule to include EGUs that converted to an alternative
fuel more than three years ago. Because EPA did not propose these final definitions or
applicability provisions, Petitioner states that it was not able to comment on them.
92
-------
Petitioner indicates that it supports EPA's stated intent, but objects to the final definitions of
"fossil fuel fired," "coal-fired electric utility steam generating unit," "oil-fired electric utility
steam generating unit," and "natural gas-fired electric utility steam generating unit" because they
are not consistent with that intent and will result in improper application of the rule to EGUs that
do not combust coal or oil in significant quantities
Petitioner similarly objects to the final exclusions in 40 CFR 63.9983(b) and (d). Petitioner states
that to implement its stated intent and avoid expanding application of the EGU MACT standards,
the EPA must revise the definitions to allow EGUs to be categorized based on actual (and
projected) fuel combustion on the applicable compliance deadline, even if the EGU may have
previously combusted other fuel (and may still be capable of combusting other fuels).
Response to Issue 73: The EPA provided an opportunity to comment on the definitions included
in the proposed MATS rule. In response, the EPA received comments raising the concern that
EGUs that converted to fuel other than coal or oil close to the compliance date would potentially
be subject to MATS, which was not the EPA's intent. (See, e.g., RTC, Vol. 2, pp. 705 - 709).
Specifically, commenters were concerned that the definitions would require units that convert to
natural gas or other non-fossil fuels prior to the compliance date would still be subject to the rule
for up to 3 years after converting based on the proposed definitions. The EPA responded to the
comments at length in the preamble to the final rule. (See 77 FR 9376 - 9378.) Because
Petitioners have not shown they were incapable of raising these issues during the comment
period, EPA is denying the petition for reconsideration on this issue.
The EPA stated that it revised the definitions so that sources that convert to a non-regulated fuel
prior to the compliance date will not be subject to the final rule. The Agency accomplished this
change by including in the definition of "coal-fired electric utility steam generating unit," "oil-
fired electric utility steam generating unit," and "natural gas-fired electric utility steam
generating unit" a requirement that the units be "fossil fuel-fired." The definition of "fossil fuel
fired" states, in part, that "fossil fuel-fired means any EGU that fired fossil fuels for more than
10.0 percent of the average annual heat input during any 3 consecutive years or for more than
15.0 percent of the annual heat input during any 1 calendar year after the applicable compliance
date." 40 CFR 63.10042 (emphasis added). As can be seen from that definition, sources are to
consider their fossil fuel usage after the applicable compliance date. This means that sources that
do not convert prior to the compliance date must project their fossil-fuel usage for the first year
after the compliance date and comply with the MATS rule if their coal or oil usage will exceed
the definitional thresholds in the rule. The Petitioner's interpretation of the MATS rule is not
correct and sources are not required to comply with the MATS rule if they convert to a non-
regulated fuel prior to the compliance date.. Because Petitioner's objection is based on a
misunderstanding of the MATS rule, this issue is not of central relevance.
The provision at 40 CFR 63.9983(b) was included in the proposed rule and merely excludes
natural gas-fired EGUs from applicability under MATS. This exclusion is appropriate because
the EPA did not find regulation of natural gas-fired EGUs appropriate and necessary. The
provision at 40 CFR 63.9983(d) was added in the final rule because EGUs meeting the definition
of a solid waste incineration unit may not be subject to standards under CAA section 112
93
-------
pursuant to CAA section 129(h)(2). For these reasons, these issues are not of central relevance
and the EPA is denying reconsideration.
In addition, on February 17, 2015, EPA proposed further revisions to the definitions identified
above. 80 FR 8442, 8447. For this reason, EPA is also denying the petition for reconsideration
as moot.
9.0 Definitions - FGD Definition Should Include CFB
Issue 74: Petitioner 20175 states that MATS would authorize the measurement of SO2 as a
surrogate for HC1 emissions under certain circumstances, including for units equipped with FGD.
In addition, Petitioner notes that the definition of "FGD" in MATS appears to properly
acknowledge that such technology should qualify as FGD. Specifically, the second sentence of
the relevant definition expressly states that "[ajkaline sorbent injection systems in fluidized bed
combustors (FBC) or circulating fluidized bed (CFB) boilers are included in this definition."
However, Petitioner alleges that the definition otherwise creates confusion because the first
sentence of the definition appears to state that any dry FGD technology shall require an "add-on
air pollution control system located downstream of the steam generating unit." In order to avoid
the potential inconsistency between these two sentences comprising the relevant definition,
Petitioner maintains that the EPA should revise the definition of "FGD" to expressly state in the
second sentence that the sorbent injection system in a CFB boiler need not be a separate add-on
pollution control system located downstream of the steam generating unit. Petitioner asserts that
revising the definition of "FGD" in this manner would be consistent with the EPA's apparent
intent to classify sorbent injection systems in FBCs as FGD for purposes of MATS,
notwithstanding that such systems accomplish desulfurization without the use of an add-on,
back-end control device. For these reasons, Petitioner requests that EPA revise the definition of
"FGD" by either eliminating the reference to a required back-end pollution control device, or
clarifying that such reference does not apply to sorbent injection systems associated with CFB
boilers.
Petitioner 20174 alleges that for the coal-fired and petroleum coke-fired new unit SO2 surrogate
emission standards, the EPA used units that do not actually qualify for the alternate standard.
Petitioner maintains that neither unit is equipped with FGD controls, which the rule requires to
qualify for the surrogate limit, and that both facilities are fluidized bed units and are equipped
with fluidized bed lime injection for SO2 control. Petitioner asserts, however, that fluidized bed
lime injection does not meet the definition of "FGD" specified in 40 CFR 63.10042. Therefore,
Petitioner maintains that the facilities should not have been used to set a new unit surrogate
standards. Petitioner notes that it commented on this discrepancy in the proposed rule.
Response to Issue 74: The issue of clarifying the definitions of FGD was raised in public
comments submitted in response to the May 3, 2011, proposed MATS rule (76 FR 24976). The
EPA's responses to the comments are in the RTC (see, e.g., Vol. 2, p. 715). Because the
Petitioners have not demonstrated that it was impractical to comment on these issues during the
94
-------
comment period, and in fact acknowledge that they did comment on it, we are denying
reconsideration of this issue.
Further, as noted in the RTC (and by Petitioner 20175), the definition in 40 CFR 63.10042 for
"Dry flue gas desulfurization technology, or dry FGD, or spray dryer, or dry scrubber" indicates
that CFB technology is considered to be an "FGD system" ("[ajkaline sorbent injection systems
in fluidized bed combustors (FBC) or circulating fluidized bed (CFB) boilers are included in this
definition"). The EPA believes that the definition, when taken as a whole, adequately addresses
Petitioners' concerns. For these reasons, the issue is not of central relevance and the EPA is
denying the petition for reconsideration.
The EPA disagrees with Petitioner's allegation that the EGUs selected as the basis for the solid
oil-derived fuel-fired new source SO2 alternate limits do not qualify for the alternate standard
because they are FBC EGUs because, as noted above, the EPA believes that the definition of
"Dry flue gas de sulfur ization technology, or dry FGD, or spray dryer, or dry scrubber" in 40
CFR 63.10042 includes FBC systems. The EPA, therefore, is denying reconsideration on this
issue because it is not of central relevance.
95
-------
10.0 Definitions - "Modification" in Definition of New or Reconstructed
Issue 75: Petitioners 20174 and 20185 state that the final rule defines a new EGU as either (1) an
EGU that commenced construction after May 3, 2011, or (2) an EGU that commenced
reconstruction or modification after May 3, 2011, and that the term "modification" is not defined
under 40 CFR Part 63. Petitioners state that because this was a change in the final rule from
proposal, they were unable to comment on the error previously. Petitioners assert that the
preamble of the final rule does not explain why modification was included. According to
Petitioners, if including modification in the trigger classification for new sources is intentional,
the EPA must provide an explanation and legal basis for inclusion and a definition of
modification for 40 CFR Part 63.
Response to Issue 75: The issue noted by Petitioners was a typographical error in the final rule.
The error was corrected in the technical corrections notice published on April 19, 2012 (77 FR
23399) and thus the petition for reconsideration of this issue is moot.
96
-------
11.0 Definitions - Natural Gas
Issue 76: Petitioner 20192 requests the EPA to reconsider the definition of "natural gas" in the
final rule to include synthetic natural gas derived from coal that otherwise satisfies the chemical
definition of natural gas and meets a pipeline-tariff-approved quality specification for natural
gas, or to otherwise clarify that the use of synthetic natural gas is acceptable during startup and
shutdown. Petitioner notes that the final MATS rule added the definition for natural gas, which
excludes "coal-derived gas" from the definition of natural gas and that the work practice
standards in the final rule require sources to use either natural gas or distillate oil for ignition
during startup.
Petitioner notes that it would like the option of using any natural gas that meets the specification
necessary for that gas to be transported on a Federal Energy Regulatory Commission (FERC)
regulated pipeline, during startup and shutdown, including synthetic natural gas. Petitioner
alleges that even though the synthetic natural gas it would like to use chemically satisfies the
definition of "natural gas" contained in the first sentence of the new rule, the phrase "coal-
derived gas" in the last sentence of the new definition may technically make the synthetic natural
gas fall outside of the requirement for "sources to operate using either natural gas or distillate oil
for ignition during startup." Petitioner believes that the exclusion of synthetic natural gas from
the definition of "natural gas" was an unintended oversight by the EPA.
Petitioner asks that the EPA reconsider this oversight, and allow synthetic natural gas to be used
as a startup fuel. Petitioner alleges that it was impractical to raise this issue during the public
comment period because synthetic natural gas was included in the definition for natural gas in
the proposed rule.
Response to Issue 76: EPA provided an opportunity for comment on the proposed definition of
"natural gas" (76 FR 25123), which did not include synthetic natural gas or "coal-derived gas,"
as the Petitioner implies, because the definition specifically referred to a "naturally occurring"
material or liquid petroleum gas:
Natural gas means:
(1) A naturally occurring mixture of hydrocarbon and nonhydrocarbon gases
found in geologic formations beneath the earth's surface, of which the principal
constituent is methane; or
(2) Liquid petroleum gas, as defined by ASTM Method D183 5-03a (incorporated
by reference, see § 63.14(b)(41)).
Moreover, numerous comments on the definition of "natural gas" (see, e.g. RTC Vol.2 p.700)
were submitted during the period for public comment, and revisions to the definition were made
to ensure consistency with other EPA rules. The issue of including synthetic natural gas or "coal-
derived gas" in the definition of "natural gas" under MATS was not raised during the public
comment period. Because Petitioner has not demonstrated that it was impracticable to comment
on this issue during the comment period on the proposed rule, the EPA is denying the petition for
reconsideration of the issue. However, we note that Petitioner's general issue relates more to the
use of syngas as a clean fuel that can be used during startup rather than to its exclusion from the
97
-------
definition of "natural gas." The definition of clean fuels that may be used during startup was
reconsidered in the MATS reconsideration rulemaking (see 79 FR 68777; November 19, 2014),
and, thus, the petition is moot.
98
-------
12.0 Differentiation Between "Coal-fired" and "Solid Oil-derived Fuel-fired" EGUs
Issue 77: Petitioner 20184 alleges that the proposed rule did not provide clear notice that the
EPA was contemplating classifying any "unit designed to burn solid oil derived fuel" as a "coal-
fired EGU" if it co-fired coal above nominal levels. According to Petitioner, the record appears
to lack a reasoned explanation, relevant to establishing emission limitations under CAA section
112(d)(3), for classifying as "coal-fired EGUs" EGUs otherwise clearly designed for combusting
petroleum coke if they co-fire coal at any rate minimally greater than the specified 10/15 percent
threshold. Petitioner maintains that the EPA first needs to reconsider and subsequently explain
the criteria it uses to categorize and differentiate "solid oil derived fuel units" from "coal-fired"
EGUs. Petitioner states that the Rule's revisions are far more than clarifications of the
subcategory definitions. Petitioner maintains that the changes are "a significant and a material
redefinition of the sub-categories, moving new and existing EGUs operating with or intended to
operate with flexibility in the blend of coals, petroleum coke and other fuels from the solid oil-
derived subcategory to a coal-fired subcategory whose floor units are completely dissimilar in
boiler design, applicable controls and fuels." Petitioner also suggests that the EPA may need to
reconsider the MACT "floors" for the "solid oil derived fuel" EGUs based upon the EGUs that
will remain within that subcategory.
Petitioner further asserts that adding the phrase "except solid waste" in the discussion of the
changes made since proposal to the definitions (77 FR 9376) "changes the meaning of the quoted
section" as discussed at proposal (76 FR 25020).
Response to Issue 77: EPA provided an opportunity to comment on issues related to how "coal
fired" and "solid oil-derived fuel-fired" should be defined. The issue of clarifying the proposed
definitions of "coal fired" and "solid oil-derived fuel-fired" was raised in public comments
submitted in response to the May 3, 2011, proposed MATS rule (76 FR 24976). The EPA's
responses to the comments are in the RTC (see, e.g., Vol. 1, pp. 303 and 354 - 355, Vol. 2, pp.
392 - 393 and 708 - 709). Based on the comments the EPA received, the EPA recognized that
there was a conflict between some definitions that may cause confusion in applying those
definitions to a specific EGU. The EPA explained at length in the preamble to the final rule (77
FR 9376 - 9378) the reasons the Agency was revising the definitions and why the revised
definitions were a logical outgrowth of the proposed rule. Because Petitioners have not
demonstrated that it was impractical to comment on this issue during the comment period, we are
denying the petition for reconsideration on this issue.
The EPA does not agree with Petitioner's statement that the rule's revisions are far more than
clarifications of the sub-category definitions. While the Petitioner alleges that the revised
definitions may or will cause units to shift subcategories, the Petitioner did not provide data or
other information to support the claim. The EPA did not identify any EGUs that changed
subcategories as a result of the revisions to the definitions, and the Agency would not have
expected to have EGUs shift subcategories because the definitions in the final rule reflected how
the Agency approached its evaluation of sources when determining the appropriate subcategory
of particular EGUs in developing the proposed rule.
99
-------
The EPA also disagrees that addition of the phrase "except solid waste" in the discussion of the
changes made since proposal to the definition of solid oil-derived fuel "changes the meaning of
the quoted section," as alleged by Petitioner, because solid oil-derived fuel is not solid waste, and
thus the phrase is not relevant to the discussion. For these reason, the objection is not of central
relevance.
100
-------
13.0 Emission Standards - Based on Non-detectable Test Results
Issue 78: Petitioner 20182 claims that reliance on measurements that are below method detection
levels (MDLs) to establish emissions standards is problematic, because low emissions
concentration levels have large potential measurement errors or uncertainty. Petitioner also
claims that the 3xRDL approach, in which the MACT limit value is checked against a value
equivalent to three times the representative method detection level (RDL), which is the average
of best performing EGUs' MDLs, and increased to the 3xRDL value as necessary, is not
appropriate because it relies on what Petitioner believes are perceived or estimated laboratory
and testing company MDL values.
Response to Issue 78: The EPA proposed to establish standards equivalent to the product of 3
times the MDL (3xMDL) if the calculated MACT floor was lower than the level 3xMDL. The
EPA received comments on this issue and responded to those comments. See, e.g., RTC Vol. 1,
pp. 541 - 548). In the final rule, the EPA used the representative method detection level or
"RDL" instead of the MDL. As explained further below, the RDL approach is a refinement of
the proposed MDL approach and for this reason the approach used in the final rule is a logical
outgrowth of the proposed rule. Because Petitioner has not demonstrated that it was impractical
to comment on the non-detect approach, we are denying the Petition for Reconsideration on this
issue.
As explained in the RTC, no emissions standards in the final rule rely solely on measurements
below a MDL. Rather, measurements below the relevant MDL are initially adjusted to the MDL
for the testing contractor that conducted the test, and then those data are averaged with other data
from the best performing sources and adjusted for variability using the upper predictive level
(UPL) technique. In the proposed rule, the EPA compared that MACT floor value to the value
equivalent to three times the MDL (3xMDL) (the value 3xMDL has an uncertainty comparable
to values obtained by EPA test methods, which is why the Agency makes these comparisons for
MACT floors established in part on non-detect data). If the calculated MACT floor was less than
3xMDL, then the EPA established the standard at the level 3xMDL. Thus, the EPA proposed
MACT emission limits with a value that was at least equivalent to 3xMDL.
In the final rule, the EPA determined that an average of all the MDLs (which can vary by
laboratory and by test run) for the sources in the MACT floor pool provided a more appropriate
mechanism for identifying an alternative to the calculated MACT floor, and the Agency
determined it was reasonable to use the RDL and the 3xRDL value instead of the MDL and the
3xMDL value. The RDL approach is a logical outgrowth of the MDL approach, and, because we
proposed that approach and responded to public comments submitted on the MDL approach in
the proposed MATS rule, the public has had ample opportunity to comment on the issue. (See,
e.g., RTC, Vol.1, pp. 522, 527, and 695).
In addition, the 3xRDL approach was challenged in National Association of Clean Water
Agencies (1VACWA) v. EPA, 734 F.3d 1115 (D.C. Cir. 2013). The Court upheld the Agency's
3xRDL approach for addressing non-detect data as reasonable. Id. at 1155. The EPA continues to
believe that the 3xRDL approach is the correct approach for addressing non-detect data, and the
Court has confirmed that the approach is reasonable.
101
-------
For all these reasons, the EPA is denying the petition for reconsideration of this issue.
Issue 79: Petitioner 20174 claims that: 1) use of the 3xRDL approach for 10 of 16 new source
emissions limits is not appropriate 2) the 3xRDL approach does not consider actual accuracy of
the methods, 3) the EPA has not demonstrated that the required methods can produce quality
data to provide reliable demonstrations of compliance, 4) the EPA has provided no guidance to
States for enforcing limits so close to the MDL, and (5) the new source Hg limit is close to the
lower detection limit for monitoring and testing.
Response to Issue 79: As a preliminary matter, the EPA provided opportunity for comment on
its approach for using non-detect data when setting standards. As explained above, while the
Agency switched to a 3xRDL approach instead of a 3xMDL approach as a logical outgrowth, no
change was made regarding use of non-detect data.
As also explained above, the 3xRDL approach was challenged in National Association of Clean
Water Agencies (1VACWA) v. EPA, 734 F.3d 1115 (D.C. Cir. 2013). The Court upheld the
Agency's 3xRDL approach for addressing non-detect data as reasonable. Id. at 1155.
In addition, the EPA received comments on the first four issues raised in the petition in response
to the proposed MATS rule. The EPA's responses to the comments are in the RTC (see, e.g.,
Vol. 1, pp. 522 - 527). Because the RDL approach is a logical outgrowth of the MDL approach,
the comments the EPA received on the MDL issues are generally applicable to the RDL
approach such that the public had ample opportunity to comment on the Agency's approach to
addressing non-detect data. For these reasons, the EPA is denying the petition for reconsideration
of the first four issues discussed above.
Regarding the last claim, the EPA reconsidered the new source Hg limit and provided an
additional opportunity for comment on that issue in the MATS reconsideration of certain new
source limits. The petition is therefore moot as to this issue (see 78 FR 24073; April 24, 2013).
102
-------
14.0 Emission Standards - Existing IGCC Pb limit
Issue 80: Petitioner 20174 states that the lead emission standards for existing IGCC EGUs in
Section 3 of Table 2, Emission Limits for Existing EGUs, are not correct. According to the
Petitioner, the input-based and output-based existing IGCC lead standards in the rule are
respectively 1,000,000 times and 1,000 times greater than the MACT analysis results. Because
this was a change in the final rule from proposal, Petitioner contends that it was unable to
comment on the error.
Response to Issue 80: The issue noted by Petitioner was a typographical error in one table of the
final rule. This error was corrected in the technical corrections notice published on April 19,
2012 (77 FR 23399). Therefore, this issue in the petition for reconsideration is now moot.
103
-------
15.0 General Provisions - Applicability
Issue 81: Petitioner 20180 states that the EPA proposed to require submission of a Notification
of Intent to conduct a performance test "at least 30 days before" the performance test is
scheduled to begin. The EPA also proposed to require compliance with the General Provisions,
which also address notification of performance testing. Those General Provisions require
submission of such notices "at least 60 calendar days before" the test is scheduled to begin.
Petitioner states that in its comments, it identified the conflict, asserted that requiring compliance
with both deadlines is not reasonable, and asked the EPA to identify the inconsistent General
Provisions as inapplicable. However, according to Petitioner, the EPA finalized the rule without
addressing the conflict. Petitioner states that although it appears the EPA intended to revise the
rule to clarify that only the 30-day notice provision applies, the EPA did not implement that
intent. Petitioner objects to this provision in the final rule and asserts that the EPA should either
revise Table 10 to state that those sections of the General Provisions are inapplicable or revise 40
CFR Part 63, subpart UUUUU, to explicitly state that 40 CFR Part 63, subpart UUUUU,
supersedes the General Provisions in the case of a conflict.
Response to Issue 81: Comments on this issue were submitted (see, e.g., RTC, Vol. 2, p. 283).
Although the EPA stated its intent to include a consistent period of 30 days in response to the
comments (see, e.g., RTC, Vol. 2, p. 283), that clarification was not made in the final MATS
rule. However, Table 9 to Subpart UUUUU of Part 63 was revised in the final new source
reconsideration (see 78 FR 24073; April 24, 2013) to indicate that the MATS provisions
supersede those of the General Provisions. (Note that Petitioner is in error in referring to Table
10 in that there are only 9 tables to Subpart UUUUU of Part 63.) Petitioners had an opportunity
to comment and did comment on this issue during the period for public comment. This issue was
further addressed in the Technical Corrections notice proposed on February 17, 2015 (see 80 FR
8448) by proposed revisions to equations Table 9 which would allow a 30 day notice. However,
because this issue has been clarified in the MATS new source reconsideration rule, the petition
for reconsideration of this issue is now moot.
Issue 82: Petitioner 20180 claims that no changes to the General Provisions table were made in
response to Petitioner's comments and suggests that the EPA should respond specifically to each
of Petitioner's comments and revise the rule to implement any changes that the EPA intended to
make. Petitioner notes that the EPA said that the rule and that Table 9 had been revised where
appropriate to remove inconsistencies.
Response to Issue 82: The EPA disagrees with Petitioner. As noted by Petitioner, this issue was
raised in comment on the proposed rule and the EPA responded that the Petitioner's concerns
had been reviewed and that the rule and the table were revised as necessary to remove
inconsistencies (see, e.g., RTC, Vol. 2, p. 731). The EPA did not state that it was making all of
the suggested changes when responding to the comments on this issue. This issue was further
addressed in the Technical Corrections notice proposed on February 17, 2015 (see 80 FR 8448)
by proposed revisions to Table 9 to correct an inadvertent omission of 30-day notification of
§63.9. Because the Petitioner has not demonstrated that it was unable to comment or that the
Agency erred in responding to the comments, we are denying reconsideration of this issue.
104
-------
In addition, most, if not all, of the Petitioner's perceived concerns involved items such as opacity
or visible emission, which the Petitioner acknowledges are not applicable to the rule. Table 9's
title clearly excludes those portions of items cited in sections of the General Provisions that are
inapplicable to the rule, as well as not including other complete cites of specific sections of the
General Provisions that are inapplicable to the rule (such as 63.6(e)(iii)). For these reasons, the
issues raised are not of central relevance. Moreover, EPA has subsequently made in technical
corrections or on reconsideration, all the appropriate changes to the provisions identified by the
Petitioner. For this reason, the petition for reconsideration on these issues is now moot.
105
-------
16.0 Heath-based Emission Limit for Acid Gases
Issue 83: Petitioner 20180 (att. 6) claims that it was inappropriate for the EPA to consider non-
HAP health co-benefits in its decision not to exercise its discretion under CAA section 112(d)(4)
to establish health-based emission limits for acid gases from U.S. EGUs.
Response to Issue 83: Issues regarding health-based emission limits were raised in public
comments submitted in response to the proposed MATS rule. Petitioner raises the same issue
that was raised by other commenters regarding the proposed rule, and the issue was addressed by
the response the EPA provided in the RTC (Vol. 1, pp. 10 - 12). The comments received on this
issue demonstrate that the public had ample opportunity to comment on this issue, and the EPA
responded to those comments.
Because Petitioner has not demonstrated that it was impracticable to comment on this issue
during the comment period on the proposed rule, the EPA is denying reconsideration of this
issue.
The EPA's decision to establish MACT standards for acid-gas HAP instead of health based
emission limits was challenged in the White Stallion case. When affirming the EPA's decision,
the Court noted that "[petitioners dispute EPA's weighing of the evidence, but petitioners offer
no compelling basis for second-guessing EPA's analysis." 748 F.3d at 1248. Included among the
Petitioners' arguments was that the EPA incorrectly considered non-HAP co-benefits. As the
Court may only consider issues raised during the period for public comment, issues raised in the
litigation and addressed by the Court clearly do not meet the criteria for reconsideration in CAA
section 307(d)(7)(B). Moreover, Petitioner does not provide any new information that became
available after the close of the comment period but within the time for judicial review, and the
Court has affirmed the EPA's approach to establish the acid-gas HAP standards (including its
method of weighing the evidence). Petitioner's objection thus does not provide support for the
argument that the rule should be revised, and this issue is not of central relevance.
106
-------
17.0 Low Emitting EGU - Bypass Stack
Issue 84: Petitioners 20183 and 20180 object to changes from the proposal that prohibit
participation in the Low Emitting EGU (LEE) program for EGUs equipped with acid gas
scrubbers that have main and bypass stacks.
Response to Issue 84: The EPA proposed to allow units that qualify as LEE to conduct reduced
testing and included the approach in the final MATS rule. 76 FR 25029-30. The EPA received
comments on bypass stacks and the requirements applicable to LEE. (See, e.g., RTC, Vol. 2, pp.
16, 182, 211-212, and 216). In response to comments and consideration of the appropriate
compliance assurance measures for LEE designation, the EPA modified the compliance
assurance provisions applicable to LEE. Because the Petitioners have not demonstrated that it
was impractical to comment on LEE monitoring issues generally, the EPA is denying
reconsideration on this issue.
The EPA provides reduced testing and reporting requirements for units that qualify as LEE in
part because sources must control their HAP emissions to a level more stringent than the
promulgated standard to qualify as LEE and that over-control provides benefits to the
environment. LEE status also provides a benefit to owners or operators because the testing costs
are significantly reduced for LEE. However, the EPA must have confidence that sources are
meeting the LEE requirements or the purpose of providing the option - reduced HAP emissions
- may be subverted. The changes made to the final LEE provisions are necessary to assure that
LEE status is properly assigned and EPA provided ample opportunity for comment on the LEE
program including how LEE status is assigned. The LEE provisions in the final rule are a logical
outgrowth of the proposed rule and Petitioners have not demonstrated that they did not have an
adequate opportunity to comment.
In addition, the statute does not require the EPA to provide EGUs with the opportunity to
become LEE, and determining whether or not to participate in the optional LEE program is a
decision made by an EGU owner or operator (see 63.10000(c)(l)(i)(A) and (B). An EGU owner
or operator has the option of complying with the non-LEE testing requirements if he does not
believe his unit can comply with the LEE provisions because of the bypass stack provisions that
were included in the final rule. The EPA believes the bypass stack provisions are necessary
additions to the rule. Further, an EGU owner or operator can always request specific alternative
emissions limits or use of a specific alternative monitoring method, provided the owner or
operator follows the procedures established in the General Provisions at 40 CFR 63.8(f) and
63.6(g). This issue was further addressed in the Technical Corrections notice proposed on
February 17, 2015 (see 80 FR 8444) by proposed revisions to 10000(c)(l)(i)(A) and 10005(h) to
clarify the provisions of units designated as being LEE when an acid gas scrubber and a bypass
stack are present. For all of these reasons, the issues raised by the Petitioner are not of central
relevance to the final rule, and we are denying reconsideration on that basis as well.
For the reasons above, the EPA is denying reconsideration of this Petition.
107
-------
18.0 Monitoring and Testing
18.1 Definition of "deviation"
Issue 85: Petitioner 20187 states that, according to the final rule, if an EGU "ha[s] a deviation
from any emission limit, work practice standard, or operating limit," the owner or operator need
only "submit a brief description of the deviation." Petitioner notes that the Rule expressly states
that such deviations are "not always violations" and that the determination of whether a deviation
"constitutes a violation of the standard is up to the discretion of the entity responsible for the
enforcement of the standards."
Petitioner maintains that, in combination with the newly devised parametric monitoring options
and the affirmative defense for malfunctions, those definitions undermine the enforceability of
the standards. Petitioner asserts that the rule monitors parameters that are associated only with
the emission levels exhibited during an annual or triennial stack test, not the limits themselves.
Hence, according to Petitioner, a facility that relies on parametric monitoring and violates its
specified parameters reports no information that would reliably allow EPA, a state, or the public
to readily discern whether HAP emissions during that "deviation" violate the emission standard,
and the final rule's definitions provide no basis by which a violation could be enforced.
As a result, Petitioner alleges that the EPA's final standard fails to establish monitoring sufficient
to ensure continuous compliance with its standards and states that the EPA should revise its
monitoring to clearly tie exceedances of all of the rule's requirements to violations of the
underlying standards.
Response to Issue 85: The issue of deviations was raised in public comments submitted in
response to the proposed MATS rule. The EPA's responses to the comments are in the RTC (see,
e.g., Vol. 2, pp. 714 - 715). Language stating that "[a] deviation is not always a violation" was
included in the proposed rule (76 FR 25122). The proposed rule contained a number of
parameter monitoring requirements such as control device operation and fuel analysis in fuel in
conjunction with continuous emissions and Hg sorbent trap monitoring requirements. Many of
these parameters were indicators, rather than direct compliance measurements. As such, failure
to meet the appropriate indicator level was not necessarily a violation of an associated emissions
limit, but it was a deviation from the appropriate indicator level. Apart from continuing to
require liquid oil-fired EGUs to identify and use parameter monitoring for acid gas emissions,
the final rule removed most other parameter monitoring requirements in response to public
comments concerning duplicate monitoring (see 77 FR 9386, February 16, 2012).
The definition of violation was also revised in the final rule to clarify that the authority delegated
with responsibility for the rule is the body with the ability to determine the condition where one
or more deviations constitute one or more violations. For instance, a missed daily calibration
check for a PM CEMS is a deviation. By itself, that deviation may not invalidate a day's worth
of PM emissions recorded by the CEMS, particularly if a subsequent successful calibration check
occurs before the end of the day. Moreover, given that the rule's emissions limit is based on a
30-boiler operating day rolling average, even if the PM emissions for a particular day turned out
to be high based on a missed daily calibration check, it is possible that when that value is
108
-------
averaged with others over the appropriate period, compliance may be assured - meaning that the
missed daily calibration check, while important, did not constitute a violation of the emissions
limit.
The EPA finds the Petitioner's specific stated concern - that parameter monitors are associated
only with the emission levels exhibited during an annual or triennial stack test, not the limits
themselves - is misapplied. In designing this and other continuous parameter monitoring
situations, the EPA requires parameters to be identified and their operating limits determined via
concurrent emissions testing. Further, such parameters and operating limits are to be revised
when subsequent operation or testing shows revisions to meet the emissions limits are necessary.
The EPA disagrees with Petitioner's supposition that a facility that relies on parametric
monitoring and violates its specified parameters reports no information that would reliably allow
the EPA, a state, or the public to readily discern whether HAP emissions during that "deviation"
violate the emission standard, and the final rule's definitions provide no basis by which a
violation could be enforced. As specified in the final rule, each deviation is to be identified and
included in semiannual reports (see 40 CFR 63.10031(e)). Moreover, records of each deviation,
as well as the conditions coincident with each deviation, are to be kept (see 40 CFR
63.10032(b)(4)).
The comments the EPA received on this issue demonstrate that the public had ample opportunity
to comment on the issue, and the EPA's response provided clarity in a logical fashion. Because
Petitioner has not demonstrated that it was impracticable to comment on this issue during the
comment period on the proposed rule, the EPA is denying reconsideration of the issue.
18.2 Monitoring system requirements
Issue 86: Petitioner 20180 asserts that because a requirement for monitoring system installation
and operation, as well as testing, for EGUs that cease to be subject to the rule was added after
proposal, it was unable to provide comments. Moreover, Petitioner states that the requirement
makes no sense and that the EPA must issue and solicit comment on a proposal.
Response to Issue 86: The EPA disagrees with Petitioner. The proposal at 76 FR 25026
specifically solicited comment on how the EPA should address reclassification of sources
between MATS and section 129 standards, particularly with respect to initial and ongoing
compliance, and directed the public to provisions in the CISWI rule (76 FR 15753) that are
almost identical to the provision included in the final MATS rule. 76 FR 25026. For this reason,
the public had ample opportunity to comment on the provisions included in the final rule,
including the monitoring provisions identified by the Petitioner. Indeed, one commenter noted
that EGUs might not have the additional monitoring required by CAA section 129 standards if
they were to begin combusting solid waste. In response to this and other comments, the EPA
revised the final rule to include provisions based on the CISWI provisions identified in the
proposed MATS rule and contained at 40 CFR 60.2145(a)(2) through (6) (see 40 CFR
63.10000(f) through (k), and, e.g., RTC, Vol. 1, pp. 274 - 283). Because the Petitioner has not
demonstrated that it was impracticable to comment on these issues during the comment period,
the EPA is denying the petition for reconsideration of these issues.
109
-------
18.3 Citati on correcti on
Issue 87: Petitioner 20180 asserts that the EPA should revise paragraph 4.1 of Appendix A to
correct a citation (from 40 CFR 63.10005(g) to 63.9984(f)) or to eliminate the deadline for Hg
CEMS or sorbent trap monitoring systems.
Response to Issue 87: The EPA has addressed this issue in the MATS new source
reconsideration rulemaking (see 78 FR 24073; April 24, 2013).
18.4 Testing
Issue 88: Petitioner 20180 states that EPA proposed that for CEMS (and sorbent trap monitoring
systems) the "performance test" consist of the "first 30 operating days of data collected with the
certified monitoring system." In comments, Petitioner asked EPA to allow EGUs to specify
when a performance test based on CEMS starts, explaining that there are good reasons why an
EGU may need to delay the performance test to some point after monitoring system certification
and that some monitoring systems may have been certified years ago (noting that the EPA did
not define in the proposed rule the point at which new systems are deemed "certified"). In
response to Petitioner's comments, Petitioner notes that 40 CFR 63.10005(a)(2) was changed to
state that the performance test consists of "30 boiler operating days of data collected by the
initial compliance demonstration date... with the certified monitoring system." However,
Petitioner indicates that 40 CFR 63.10011(c)(1) and (2) still describe the performance test for Hg
CEMS (or a sorbent trap monitoring system) and for SO2 and PM CEMS as "the first 30-boiler
operating day rolling average emission rate obtained with certified CEMS after the applicable
date in 40 CFR 64.9984 (or, if applicable, prior to that date, as described in 40 CFR
63.10005(b)(2))." Petitioner asserts that to implement its stated intent, the EPA must remove the
reference to the "first" average in 40 CFR 63.10011(c)(1) and (2) as well.
Response to Issue 88: Petitioner acknowledges that it raised this issue during the comment
period for the proposal and that the EPA responded to the comments on this issue (see RTC, Vol.
2, pp. 450 - 451). The EPA noted that the final rule had been revised so that the 30-day test is no
longer tied to the certification of the CEMS or sorbent trap monitoring system and that the
requirements for the mandatory use of PS-11 had been removed from the final rule. The
comments the EPA received on this issue demonstrate that Petitioner had ample opportunity to
comment on these issues, and the EPA responded to those comments. Because Petitioner did not
demonstrate that it was impracticable to comment on this issue during the comment period on the
proposed rule, the EPA is denying reconsideration of this issue.
Issue 89: Petitioner 20180 claims that the EPA should issue a proposal that explains why a grace
period would compromise data integrity and why the single-level system integrity check for Hg
CEMS should be treated differently than other Hg CEMS quality assurance tests. Petitioner also
claims that the EPA should revise other portions of the rule in order to clarify that the single-
level system integrity check for Hg CEMS does not have to be performed on the same day each
week and could be performed early.
110
-------
Petitioner 20180 suggests that the rule contain a grace period for quality assurance testing
intervals for PM CEMS, as the rule does for HC1 and HF CEMS.
Response to Issue 89: Petitioner acknowledges that it raised this issue during the comment
period for the proposal, and the EPA responded to the comments on this issue in the RTC (see,
e.g., Vol. 2, pp. 25 - 26, 106 - 109, 116 - 117, 127 - 128, and 227). As mentioned in the
responses, the EPA agrees that a grace period similar to the 40 CFR Part 75 RATA grace period
provisions should be provided for RATA testing and these provisions were added to Appendices
A and B; however, the EPA disagrees that grace period provisions should be provided for the
performance testing requirements because the tests need to occur at the scheduled intervals in
order to maintain the rule's compliance timeframe. The comments the EPA received on these
issues demonstrate that Petitioner had ample opportunity to comment on these issues. Because
Petitioner did not demonstrate that it was impracticable to comment on these issues during the
comment period on the proposed rule, the EPA is denying reconsideration of these issues.
Issue 90: Petitioner 20180 claims that the EPA said in the response to comment document that it
would amend the rule to allow use of Method 3 0B sorbent traps for up to 15 days for LEE testing
and Appendix A sorbent trap monitoring systems, but that the EPA did not change the rule.
Petitioner requests that the EPA revise the rule consistent with its stated intent.
Response to Issue 90: The comments the EPA received on this issue, including those from
Petitioner, demonstrate that Petitioner had ample opportunity to comment on these issues, and
the EPA responded to those comments (see, e.g., RTC, Vol. 2, pp. 238 and 247). This issue was
further addressed in the Technical Corrections notice proposed on February 17, 2015 (see 80 FR
8448) by proposed revisions to paragraph 5.2.1 of Appendix A which proposes to correct the
number of days for sorbent trap use from 14 to 15 days. Because Petitioner did not demonstrate
that it was impracticable to comment on this issue during the comment period on the proposed
rule, the EPA is denying reconsideration of this issue.
Issue 91: Petitioner 20180 claims that in response to its comment, the EPA agreed to revise the
final rule to provide procedures for calculating total concentrations under Method 29 but
finalized the rule without providing the procedures. Petitioner requests that the EPA revise Table
5 to reflect the appropriate procedure.
Response to Issue 91: The Petitioner sought instruction on how to combine the front and back
half fraction of metals determined using Method 29, in order to determine total metals emissions.
The EPA disagrees with Petitioner's claim that additional instruction are needed. Method 29,
available at Appendix A-8 to 40 CFR 60, already contains requirements for reporting metals, so
no additional direction should be needed. However, the EPA responded to the Petitioner's
comments by changing Table 5 to require reporting of both the front half and back half value
(see, e.g., RTC, Vol. 2, p. 20 and section 2 of Table 5 to the MATS rule). The fact that Petitioner
did comment on this issue during the comment period demonstrates that it was not impracticable
for them to raise the issue. The EPA responded to the comments. For these reasons we are
denying reconsideration of this issue.
Ill
-------
Issue 92: Petitioner 20180 claims that the rule's requirement to use Methods 2, 2F, or 2G to
determine velocity and the flow rate of stack gas may not be "necessary, or reasonable" because
all of the "methods specified in Table 5 to determine concentrations of those pollutants also
provide measurements of stack gas velocity and flow;" that the EPA, in response to Petitioner's
comment, said the rule had been amended to clarify that use of Method 2 or its equivalent was
only necessitated by the calculation of the emissions limit; and that it is unclear where the EPA
addressed this change. Petitioner requests that the EPA revise Table 5 with the clarification.
Response to Issue 92: As mentioned above, the Petitioner suggests that the EPA clarify in Table
5 which methods are acceptable for determining velocity and flow rate of stack gas, suggesting
that the EPA has not amended the rule to provide such specificity.
The EPA disagrees with Petitioner. The EPA responded to those comments (see the list of
acceptable methods for determining velocity and flow rate of stack gas (2, 2A, 2C, 2F, 2G, and
2H) in sections 1, 2, 3, and 4 of Table 5 to the MATS rule, e.g., RTC, Vol. 2, pp. 5 and 6).
Because the EPA responded to the comment, the EPA is denying reconsideration of this issue.
18.5 Bypass Stack
Issue 93: Petitioners 20183 and 20180 object to changes from the proposal that allow EGUs with
main and bypass stacks to count bypass hours as deviations from monitoring requirements in lieu
of installing CEMS on the bypass stacks (if it is infeasible to certify and quality-assure data from
such CEMS), because Petitioners claim they were unable to comment on the changes. Petitioners
recommend use of other alternatives, including use of default values, clean fuel use (for EGUs
that can use clean fuels), or other, unspecified work practices during events in which bypasses
occur (especially State-declared emergencies) be allowed.
Response to Issue 93: The EPA disagrees with the Petitioners regarding their claimed inability
to comment on the provisions included in the final rule on bypass stack monitoring. The
proposed rule would have required EGUs with a main and a bypass stack to install CEMS and
other monitoring systems on both stacks. See 76 FR 25111 (40 CFR 63.10010(a)(4)). The
proposed rule did not provide an alternative to this requirement so noncompliance with the
CEMS and other monitoring requirements would have been a deviation and likely a violation of
the standard. The EPA received comments on bypass stacks (see, e.g., RTC, Vol. 2, pp. 16, 182,
211-212, and 216) suggesting alternative approaches to addressing bypass stacks and expressing
concern with the feasibility of installing CEMS on all bypass stacks. In the final rule, the EPA
maintained the CEMS requirement and added an option for EGUs that could not feasibly certify
and quality assure a monitoring system on a bypass stack. The final rule further stated that
sources that used this option would be required to treat any hours of bypass stack emissions as
periods of deviation from the monitoring requirements. Thus, the proposed and final rule contain
almost identical provisions in regard to the main requirement (i.e., the requirement to monitor
emissions from bypass stacks) The additional flexibility in the final rule, added in response to
comments, simply provides sources that can demonstrate an inability to monitor their bypass
stack an additional compliance mechanism. To the extent the petitioners are concerned about the
labeling of bypass hours as deviations, that issue does not warrant reconsideration because, as
112
-------
stated above, noncompliance with the proposed rule provision would also have been considered a
deviation.
Regarding Petitioners' suggestions for use of other alternatives to monitoring during bypass,
those suggestions were not accompanied by any data or information that was unavailable at the
time the EPA proposed the MATS rule and the Petitioners have not explained why they were
unable to raise these issues during the period for public comment. In addition, EGU owners or
operators that believe a specific form of alternative monitoring is suitable for use during periods
of bypass operation can submit an alternative monitoring request to the Administrator in
accordance with the procedures given in the MACT rule general provisions at 40 CFR 63.8(f).
This issue was further addressed in the Technical Corrections notice proposed on February 17,
2015 (see 80 FR 8444 - 8447) in which we proposed how LEEs could bypass their stacks under
limited circumstances by proposed revisions to 10000(c)(l)(i)(C), 10005(h), and 10031(c)(6). In
addition, proposed 10010(a)(4) which would require exhaust gas that bypass control devices be
routed to stacks with measurement instruments and to clarify that hours that a bypass stack is in
operation are to be counted as hours of deviations from monitoring requirements.
Because Petitioners have not demonstrated that it was impracticable to comment on the bypass
stack monitoring requirements, we are denying the petition for reconsideration on this issue. We
are also denying this petition because Petitioners have not provided new information that was
unavailable during the comment period. Finally, we are denying this petition because it is not of
central relevance because the EPA's regulations provide a mechanism for sources to obtain
alternative monitoring requirements if appropriate given source-specific issues.
113
-------
19.0 Non-continental liquid oil-fired EGUs
19.1 Subcategory size and applicability
Issue 94: Petitioner 20188 asserts that the final rule has three new subcategories, including
"liquid oil-fired, non-continental" (LONC) EGUs and "limited-use oil-fired" EGUs. According
to Petitioner, because the scope of these two new subcategories and, in the case of the LONC
subcategory, final emission limits, were not available during the public comment period, issues
related to the emission limits and other requirements for these two new subcategories are ripe for
reconsideration.
Petitioner 20188 states that 31 non-continental EGUs burn liquid oil. Petitioner notes that at the
time that the MATS rule was being finalized, the EPA had received ICR data from 15 of these 31
EGUs. Petitioner believes that at least two (Costa Sur 3 and 4) and possibly more of its 14 EGUs
actually are or will be in the newly created limited-use liquid oil-fired subcategory. Petitioner
also believes that one or more of the EGUs owned by other non-continental utility companies
may be in the limited use liquid oil-fired subcategory. Petitioner states that removal of only 2
limited-use EGUs from the data set reduces the actual size of the LONC subcategory to 29
EGUs.
Petitioner states that given the MACT-setting requirements of CAA section 112(d)(3), whether
the subcategory's size is more or less than 30 sources is of critical importance in setting MACT
emission limits for a subcategory like the newly created LONC subcategory. Petitioner states that
because the EPA had data from 15 of the 31 LONC EGUs, it applied CAA section 112(d)(3)'s
12 percent requirement and used the data from two EGUs to set MACT limits. However, asserts
Petitioner, with the LONC subcategory actually containing less than 30 EGUs because of two or
more EGUs actually being in the limited-use subcategory, CAA section 112(d)(3)(B) dictates
that MACT be set at the average emission limits achieved by the best performing five EGUs of
the 15 EGUs for which EPA had data.
Petitioner 20188 also requests reconsideration of EPA's decision to establish the limited-use
liquid oil-fired subcategory capacity factor threshold at less than 8 percent on two grounds.
Petitioner asserts that it did not have the opportunity to comment on the methodology utilized by
the EPA to cull its data set and select less than 8 percent rather than another number as the
capacity factor threshold for the limited-use liquid oil-fired subcategory. Petitioner believes that
the EPA should grant reconsideration on the less than 8 percent capacity factor threshold because
that threshold fails to reflect the differences between operation of continental electric systems
and non-continental electric systems. According to Petitioner, on isolated non-continental grids,
"peaking units likely will run more frequently for reliability purposes than comparable units on
the continental grid." Regardless of the EPA's methodology for selecting the less than 8 percent
capacity factor threshold, states Petitioner, a data set dominated by continental peaking EGUs,
which, because of the nature of the continental grid, operate differently than similar EGUs that
must support an isolated non-continental grid, influenced that threshold. Petitioner suggests an
amendment to the definition of "limited-use liquid oil-fired subcategory" that would include non-
continental oil-fired EGUs with a capacity factor of less than 25 percent in the subcategory.
114
-------
Response to Issue 94: The EPA specifically identified limited-use or peaking EGUs as a
potential subcategory in the proposal notice (76 FR 25037). Further, the EPA solicited comment
on other potential subcategory approaches (see, e.g., 76 FR 25037, 25047). In the proposal, the
EPA included requests for comments on the bases upon which the subcategories could or should
be established and provided guidance on the type of information that would be useful in an
analysis of potential subcategories. Id. Based on comments received on the proposed rule that the
EPA should establish subcategories for non-continental liquid oil-fired EGUs (including
comments from Petitioner) and limited-use liquid oil-fired EGUs, the EPA established
subcategories for both non-continental liquid oil-fired EGUs (see, e.g., RTC, Vol. 1, pp. 349 -
353, Vol. 2, pp. 334 - 335 and 382 - 383) and limited-use liquid oil-fired EGUs (see, e.g., RTC,
Vol. 1, pp. 366 - 389) in the final rule. Because the EPA solicited comments on establishing
additional subcategories, including specifically identifying limited use EGUs as a potential
subcategory, the EPA maintains that the final rule is a logical outgrowth of the proposed rule. In
addition, the comments received on potential subcategories demonstrate that it was not
impracticable to comment on these issues during the comment period. For these reasons, the
EPA is denying reconsideration on this issue.
The EPA disagrees with Petitioner's contention that limited-use liquid oil-fired EGUs
inappropriately are included in the population of non-continental liquid oil-fired EGUs and that
the analysis should be changed. Based on Petitioner's responses to the EPA's 2010 ICR, the two
EGUs noted by Petitioner as potentially being in the limited-use liquid oil-fired EGU
subcategory had average annual capacity factors of 46 and 57 percent for the 3-year period 2007
- 2009, above the value for inclusion in the limited-use subcategory (and well above even
Petitioner's suggested cut-off value for inclusion in a limited-use subcategory). The EPA
appropriately populated the non-continental liquid oil-fired EGU population based on the data in
the record, and the Petitioner has not provide new information or data for these or other units in
the category that would support the argument that the units met the limited use liquid oil-fired
EGU definition at the time the EPA promulgated the final rule (e.g., by providing capacity factor
data from years 2010 - 2012 that shows a capacity factor less than 8 percent). Further, Petitioner
did not present any new information justifying why a higher capacity factor threshold for the
limited-use liquid oil-fired EGU subcategory for non-continental areas is warranted.
Commenters on the proposed MATS rule provided a range of potential capacity factors for the
limited use subcategory, and the EPA considered the comments when establishing the capacity
factor threshold for limited use EGUs (see, e.g., RTC, Vol. 1, pp. 366 - 389). The fact that the
EPA received comments on these issues during the rulemaking process demonstrates that it was
not impracticable to comment during the rulemaking process. Petitioners also have not provided
new information or data, much less new information or data of central relevance to the rule.
For all of these reasons, the EPA is denying reconsideration on this issue.
19.2 Subcategory data
Issue 95: Petitioner 20188 asserts that when the EPA solicited test data from liquid oil-fired
EGUs, the EPA assumed that the subcategory comprised 180 EGUs. From that population, the
EPA requested test data from approximately 100 EGUs, or about 56 percent. Petitioner maintains
that with a sample size of 180, collecting data from 100 members made sense from a statistical
115
-------
perspective. However, states Petitioner, when the sample size was reduced to 31, basing MACT
floor calculations on data from only 50 percent of the population (15 EGUs) does not make good
statistical sense. Petitioner indicates that the EPA should use the reconsideration process to
gather data from the 16 EGUs in the LONC subcategory that previously had not been surveyed.
Petitioner states that this collection effort could include data from compliance tests already
conducted on these EGUs, as the EPA did with continental oil-fired EGUs. According to
Petitioner, if EPA had test data from those EGUs, the MACT floors and emission limitations for
the subcategory would be based on the best performing four or five LONC EGUs, rather than
only one or two, and Petitioner asserts that this is a more statistically sound outcome.
Response to Issue 95: As noted above, the EPA solicited comment on potential additional
subcategories and the EPA included a new subcategory for non-continental liquid oil-fired EGUs
in response to comments requesting such an approach, including comments from this Petitioner
(see, e.g., RTC, Vol. 1, pp. 349 - 353). The comments requesting this subcategory were primarily
from commenters that own and operate such units and if they believed the EPA should have
additional data from untested units they could have provided those data to EPA. (In fact, the EPA
received comments on the proposed rule that it already had data sufficient to establish a non-
continental liquid oil-fired subcategory; see, e.g., RTC, Vol. 1, p. 351.) The EPA did not receive
new data from non-continental liquid oil-fired EGUs during the comment period, after the
comment period prior to the final rule, or during the period for judicial review of the final MATS
rule. In any case, the statute does not set a minimum data requirement for subcategories with 30
or more sources as the Petitioner suggests. Specifically, CAA section 112(d)(3)(A) provides, in
part, that the EPA shall establish standards that are not less stringent than "the average emission
limitation achieved by the best performing 12 percent of the existing sources (for which the
Administrator has emissions information)." (Emphasis added). For these reasons, the issue is not
of central relevance to the final rule and we are denying reconsideration for that reason. In
addition, the Petitioner has not demonstrated that it was impracticable to comment on this issue
during the comment period, and the Petitioner did not provide new data or information that
became available after the close of the comment period but during the period to request judicial
review. The EPA is denying reconsideration for these reasons as well.
19.3 Exclusion of data
Issue 96: Petitioner 20188 contends that the EPA's calculation of the LONC MACT floor for
"Total Metals" is flawed, because the EPA excluded all of Petitioner's EGUs from that data set.
According to Petitioner, this exclusion reduced the number of EGUs in the data set to seven
EGUs, and thus the EPA calculated the floor based on the best performing single boiler.
Petitioner surmises that the absence of cobalt data from Petitioner's EGUs in the EPA data base
was the reason for the data omission. Petitioner asserts that this absence resulted from the
omission of cobalt from the original version of the EPA's Emissions Reporting Tool (ERT),
which caused the need for separate entry of cobalt into files prior to submittal to the EPA and
which data Petitioner states was in the eight written test reports Petitioner submitted to EPA.
Petitioner surmises that the EPA did not notice the omission and states that the EPA did not ask
Petitioner about the missing data.
116
-------
Petitioner contends that inclusion of its EGUs in the Total Metals data set would increase the
total number of tested EGUs to 15 and the number of EGUs used to calculate the floor to 2.
According to Petitioner, the MACT floor recalculated with these data would result in a limit of
0.003 lb/MMBtu (based on a calculated floor of 0.0025 lb/MMBtu) in contrast to the emission
limit of 0.00006 lb/MMBtu adopted in the final MATS Rule. Petitioner maintains that on
reconsideration, the EPA should include all available data, including Petitioner's, and reassess
the MACT emission limits for the LONC subcategory.
Response to Issue 96: The EPA has completed an exhaustive search of its hard copies and
archived electronic copies with regard to Petitioner's submittal of responses to the 2010 ICR.
Petitioner's final certified ERT files contained run-level data in the ERT format necessary for
bulk uploading, but their final submittals did not contain cobalt data in any form, for any of
Petitioner's eight EGUs (i.e., there were no cobalt data in the ERT tables, and there were no
attachments included with any of the final ERT files). The EPA is denying the Petition for
Reconsideration on this issue because the Petition does not provide sufficient support for its
argument to reconsider the rule on this basis, thus the issue is not of central relevance. We are
also denying the petition because the issue is also not of central relevance because the EPA is
unable to confirm the alleged data exist and, therefore, the EPA could not consider the data even
if the issue were reconsidered. Moreover Petitioner had ample opportunity to submit relevant
data during the period for public comment and has not demonstrated that it was impracticable to
do so.
19.4 Use of outlier data
Issue 97: Petitioner 20191 maintains that the filterable PM data from one of Petitioner's stack
tests conducted during the 2010 ICR was abnormally low compared to the test results at some of
Petitioner's other EGUs that were combusting the same oil. Petitioner states that it conducted
new filterable PM stack testing on the outlier EGU and confirmed that the filterable PM
emissions data from the test with abnormally low data values were inaccurate and clearly not
representative. Petitioner contends that the EPA should have recognized the abnormally low
filterable PM emissions from the subject EGU, classified the stack test as an outlier, and
discarded the data. Petitioner acknowledges it cannot explain with certainty why the data
appeared to be low. Petitioner also states that it did not identify the subject test results as
anomalous when it submitted the data to the EPA through the 2010 ICR because the total HAP
metals results for the EGU were consistent with the total HAP metals results for its other EGUs,
and Petitioner was unaware that the EPA intended to establish a filterable PM standard. Thus,
Petitioner claims it could not have provided comments earlier in the rulemaking process.
Response to Issue 97: The EPA disagrees with Petitioner's contention that because more recent,
additional data show different results than the data submitted in response to the 2010 ICR, the
initial data are somehow incorrect and notes Petitioner's acknowledgement that it cannot
"explain with certainty" the differences. The discrepancy instead shows that there may be
variability in the data, which the EPA accounts for in establishing MACT floor standards.
Petitioner was obligated to certify through the ERT and the electronic submittal process that the
data were correct to the best of its knowledge, and the Petitioner has not claimed that the
certification was in error. The EPA conducted a standard outlier test on the data as part of the
117
-------
UPL analysis and the data passed the test. Petitioner notes that it did not identify the subject test
results as being anomalous because "the total HAP metals results for the unit were consistent
with the total HAP metals results for its other units." However, the limit in question is the PM
limit, not the HAP metals limit, and Petitioner has not identified any similar analyses of its PM
data. In addition, Petitioner's assertions that the standard would have been less stringent if the
supposed "outlier" was excluded is unsubstantiated because it is possible that its EGU would not
have remained in the MACT floor after exclusion of the data. For this reason, the issue is not of
central relevance.
Further, the data submitted by Petitioner and used by the EPA for the final rule were included in
the MACT floor spreadsheets at proposal and Petitioner could have commented on the data
during the public comment period. That is, the filterable PM data included in the total PM
analysis (EPA-HQ-OAR-2009-0234-3040; column CI in "Metallic oil" tab) at proposal are the
same data as were used for the filterable PM analysis (EPA-HQ-OAR-2009-0234-20132,
attachment a5; column CU in "Metallic oil Metals and Hg OTUS" tab) at final. Thus, because
the EPA established a non-continental liquid oil-fired EGU subcategory and moved to filterable
PM based on comments received (including supporting comments from Petitioner on both issues;
see EPA-HQ-OAR-2009-0234-18477, pp. 8-10, 20), Petitioner, as a potential member of this
subcategory, should and could have commented on its alleged data issues at proposal as it was
aware that the EPA could accept either request made in comments. Petitioner had ample
opportunity to raise its data issues in public comments on the proposed rule. Because Petitioner
has not demonstrated that it was impracticable to comment on this issue during the comment
period on the proposed rule and because Petitioner does not provide sufficient support for its
argument that the rule needs to be reconsidered on this basis, the EPA is denying the Petition for
Reconsideration of the issue.
The EPA is also denying the petition because it is not of central relevance. The EPA maintains
that it is reasonable to rely on the certified data submitted by industry, and in this case the
Agency has no tangible reason to question the validity of the data and Petitioner has not provided
new data to support its contention that reconsideration is required. See White Stallion, 1222 F.3d
at 1249 (finding that the EPA acted reasonably in relying on certified data).
118
-------
20.0 PM Limits
20.1 Filterable PM as a surrogate
Issue 98: Petitioner 20187 alleges that filterable PM is not a valid surrogate for As, Se, and other
non-Hg metals and states that the EPA should revise the filterable PM standard used as a
surrogate for non-Hg metals. Petitioner states that during the comment period, the EPA received
comments pointing out several problems with its use of total PM as a surrogate for non-Hg
metals and that rather than correct these problems, the EPA exacerbated them by announcing in
the final rule that it would instead use filterable PM as a surrogate for all non-Hg metals.
Petitioner notes that the EPA's rationale for switching to filterable PM as a surrogate for non-Hg
metals is that "no specific particulate form provided a consistently superior indicator of better
metals control" (77 FR 9402) and characterizes this rationale as "cursory." Petitioner maintains
that the relevant test is not whether filterable PM is a better or worse surrogate than total PM but
whether (1) all the metals for which EPA seeks to use filterable PM as a surrogate (here, all non-
Hg metals) are "invariably" present in the surrogate pollutant; (2) methods to control or capture
filterable PM "indiscriminately" control or capture all non-Hg metals as well; and (3) controls of
filterable PM are the "only means" by which facilities "achieve" reductions of non-Hg metals.
Petitioner maintains that a surrogate must meet all three prongs of this test to be valid and that
filterable PM meets none of them. According to Petitioner, first "vapor-phase" metals are not
"invariably" present in filterable PM because they are not filterable particles; Petitioner states
that the EPA "admits" this. Second, states the Petitioner, the methods to control or capture
filterable PM do not "indiscriminately" control or capture non-Hg metals as well as filterable
PM. Petitioner states that filterable PM also fails the third part of the National Lime Ass 'n
surrogacy test; the control of filterable PM, as with total PM, is not the only means by which
facilities achieve reductions in non-Hg metals.
Petitioner alleges that the EPA claims that some metals might be controlled by acid gas control
measures in attempt to dismiss the flaws in its use of filterable PM as a surrogate for non-Hg
metals. Petitioner maintains that the EPA does not claim that any acid gas is a surrogate for any
of the non-Hg metals and that accordingly the EPA's discussion of the alleged effects of acid gas
control on non-Hg metals is irrelevant. Petitioner states that if the EPA wished to use some acid
gas as a surrogate for specific non-Hg metals, it could have proposed to do so and provided
support for the reasonableness of such a surrogacy relationship in the record. Petitioner states
that the only purported surrogate for non-Hg metals in the final rule, however, is filterable PM,
characterizes that surrogacy relationship as "unlawful" and "unreasonable," and asserts that the
possibility that emissions of some non-Hg metals might also be affected by acid gas controls
does not render that relationship any less "unlawful" or "unreasonable."
Response to Issue 98: The issue of the use of PM as a surrogate for the non-Hg metal HAP was
raised in public comments submitted in response to the proposed MATS rule. The EPA's
responses to the comments are in the RTC (see, e.g., Vol. 1, section 4F01a, pp. 728 - 735). The
comments the EPA received on this issue demonstrate that the public had ample opportunity to
comment on the issue, and the EPA responded to those comments. Petitioners thus have not
119
-------
demonstrated that it was impracticable to comment on whether total PM or filterable PM should
be used as the surrogate. Indeed, a number of commenters also specifically raised the issue of
whether total PM or filterable PM should be used as a surrogate. Commenters argued that the
EPA should use filterable PM as a surrogate for non-Hg metal HAP instead of total PM, in part
because of testing concerns associated with total PM testing. (RTC, Vol. 1, pp. 736 - 751.) In
addition, the EPA has historically used filterable PM as a surrogate for non-Hg metal HAP. (See,
e.g., Industrial Boilers NESHAP (80 FR 3114 - 3119; January 21, 2015); Portland Cement (75
FR 54971-2, 54974, 54987, 55020; September 9, 2010)). Moreover, Petitioner's fundamental
objection goes to the use of any form of PM as a surrogate for the metallic HAP, not specifically
to whether total PM or filterable PM should be used as the surrogate. For all of these reasons,
Petitioner has not demonstrated that it was impracticable to comment on this issue during the
comment period for the proposed rule, and the EPA is therefore denying the Petition for
Reconsideration of the issue.
Further, the EPA specifically addressed the relationship between HAP metals and both filterable
and total PM in the RTC (Vol. 1, pp. 733 - 734). As noted in the RTC, the EPA re-assessed the
relationships between individual metal emissions, filterable PM emissions, total PM emissions
and total PM2.5 emissions, compared the measured emissions of metals and particulate with the
uncontrolled emissions estimates, and found that the control of metals emissions (i.e., removal
from the flue gas stream) was generally consistent with that of filterable PM. In addition, the
EPA compared the correlations associated with non-Hg HAP metal emissions and the three
forms of PM and found that no specific particulate form provided a consistently superior
indicator of better metals control. In addition, the EPA noted that the vapor fraction of non-Hg
metals form acid gases and that the controls for HC1 and SO2 would indiscriminately control the
non-Hg metal acid gases. For these reasons, the EPA maintains that the switch from total to
filterable PM is not of central relevance to the rulemaking, and the EPA is denying the petition
for reconsideration of the issue.
20.2 Requirement for fuel sampling
Issue 99: Petitioner 20187 states that the EPA's final rule does not require sampling of the HAP
metal concentrations in coal burned in boilers for EGUs that adopt the PM surrogate and that
emissions of PM cannot be used to predict the release of As, Cr, and other metals in the absence
of reasonably accurate information about the concentration of specific metals in the coal that is
being burned. Petitioner maintains that concentration of metals in coal can vary widely, and
testing for the metal content of coal as it is fired is critical to understanding the amount of metals
emitted during coal firing and that the range in metals concentrations can have a significant
impact on the amount of metals emissions released at power plants. Petitioner indicates that it
understands that the EPA's final rule is meant to take variability into account, and that metal
concentrations measured at the mine may be higher than they are in coal that is actually burned
due to coal washing and other practices, but Petitioner maintains that the variability in HAP-
content of coal is significant enough to justify sampling of specific metals during the annual PM
stack tests the rule requires.
Response to Issue 99: The issue of fuel sampling was raised in public comments submitted in
response to the proposed MATS rule. The EPA's responses to the comments are in the RTC (see,
120
-------
e.g., Vol. 2, section 5A03, pp. 45 - 60). The comments EPA received on fuel sampling
demonstrate that it was not impracticable to comment on the issue during the comment period.
Many of the comments pointed to the technical issues associated with analysis of the fuel
samples at the low HAP constituent concentrations expected to be found for certain pollutants at
various times. The requirement for fuel sampling was changed based on the legitimate challenges
to fuel sampling that were identified in the public comments. In addition, the Agency did solicit
comment on whether the Agency should require stack testing for total metals or PM from liquid
oil-fired EGUs instead of requiring fuel sampling only, so EPA was considering whether fuel
sampling alone was appropriate. For these reasons, the petition on this issue is not of central
relevance.
In addition, for coal-fired EGUs, the EPA proposed to require stack testing for all the standards
and also proposed to require fuel sampling. The compliance mechanism for coal-fired EGUs was
and is stack testing, and the fuel sampling was additional testing that was not associated with
demonstrating compliance with the final standards. While the Agency maintains that it has the
authority to require this testing, because it was not actually associated with demonstrating
compliance, it is not of central relevance.
For the reasons above, the EPA is denying the petition for reconsideration of this issue.
20.3 Methodology for calculating PM floor
Issue 100: Petitioner 20187 maintains that the EPA used inconsistent measures of EGUs' actual
performance, which resulted in a standard that "exceeds the statutory maximum." Petitioner
states that when choosing its "best performing" EGUs, the EPA defined the EGUs' performance
as each EGU's single lowest test (the lowest result from all available three-run stack tests in the
last 5-plus years for that EGU). Petitioner asserts that when calculating the "average emissions"
of those best-performing EGUs, however, the EPA measured the EGUs' performance differently
by first including additional stack tests in the Agency's record for each EGU, rather than the
single lowest test, and secondly, by adjusting the EGUs' performance (via an Upper Prediction
Limit (UPL)) to account for purported variability within and between those EGUs' emissions.
Response to Issue 100: The issue of the approach used to establish the MACT floors was raised
in public comments submitted in response to the May 3, 2011, proposed MATS rule (76 FR
24976). The EPA's responses to the comments are in the RTC (see, e.g., Vol. 1, pp. 435 - 436,
519 - 520, and 577 - 578). The comments the EPA received on this issue demonstrate that the
public had ample opportunity to comment on the issue, and the EPA responded to those
comments. Because Petitioner did not demonstrate that it was impracticable to comment on this
issue during the comment period on the proposed rule, the EPA is denying the Petition for
Reconsideration of the issue.
In addition, the EPA ranked the EGUs and established all the MATS MACT standards in the
same manner and the issue could have been raised in the White Stallion litigation. As the Court
may only consider issues raised during the period for public comment, issues that could have
been raised in the litigation and addressed by the Court clearly do not meet the criteria for
121
-------
reconsideration in CAA section 307(d)(7)(B). Moreover, Petitioner has not presented
information that provides substantial support for an argument that the rule should be revised or
that any error warranting reconsideration exists. For this reason, we are also denying the petition
as not of central relevance.
20.4 The Agency improperly assumed that its data captured all best performing EGUs
Issue 101: Petitioner 20187 states that the EPA based its PM floor for existing coal-fired plants
on data from 130 EGUs, which represent 12 percent of the total fleet (1,091 EGUs) rather than
12 percent of the EGUs for which the Agency had "emissions information" (approximately 720
EGUs). Petitioner notes that in the RTC the EPA states that it based that decision on its belief
that it "identified and required testing from the best performing 15 percent of EGUs for non-Hg
metal HAP" (RTC, Vol.1 at pp. 473 - 474) by collecting data from "the 175 units with the
newest PM controls installed."
Petitioner maintains that the results of EPA's analysis, however, indicate that the best performing
EGUs are not those with "newest PM controls installed." Petitioner states that of the 130 EGUs
constituting EPA's floor pool for coal-fired EGUs, 60 have PM controls (including devices such
as SO2 scrubbers which incidentally control PM) installed before 2000. Petitioner alleges that
demonstrates that a significant fraction of low PM-emission coal-fired EGUs use older controls.
Petitioner states that the EPA cannot therefore assume that its data includes all the best-
performing EGUs and that the Agency's record suggests that a significant number of best-
performing EGUs lie outside its data-set. Petitioner states that at a minimum, the Agency should
use the data in its possession to craft a better estimate of the number of sources that should be
included in its floor pool, in light of its knowledge that a significant fraction of sources outside
its data-set are performing at levels lower than those for which it has data.
Response to Issue 101: As Petitioner notes, the issue of the pool of sources used to establish the
MACT floor was raised in public comments submitted in response to the proposed MATS rule.
The EPA's responses to the comments are in the RTC (see, e.g., Vol. 1, pp. 473 - 474). The
comments the EPA received on this issue demonstrate that the public had ample opportunity to
comment on the issue, and the EPA responded to those comments. Because Petitioner did not
demonstrate that it was impracticable to comment on this issue during the comment period on the
proposed rule, the EPA is denying the Petition for Reconsideration of the issue.
In addition, the EPA used this approach in both the proposed and final rules and the issue could
have been raised in the White Stallion litigation. No parties challenged this aspect of the standard
setting process and we do not believe any errors warranting reconsideration exist.
20.5 Beyond-the-floor (BTF) analysis for PM
Issue 102: Petitioner 20187 notes that the EPA does not revisit its refusal to undertake a BTF
analysis for total filterable PM. Petitioner notes that at proposal, the EPA based its refusal to set
a BTF standard for existing coal-fired EGUs on two rationales. Petitioner states that first, the
EPA stated that the proposed standard required use of a fabric filter (FF), and that the only
available method of further reducing PM emissions would be combining multiple FFs, which the
122
-------
EPA deemed to be not cost effective, and second, the EPA asserted that "there is simply not
enough long term performance data under actual power plant conditions to justify establishing a
beyond-the-floor standard based on filter efficiency." Petitioner asserts that regardless of their
merit as to the EPA's total PM surrogate initially proposed, those rationales do not justify the
EPA's refusal to set a BTF standard for the filterable PM standard adopted by the final rule.
Petitioner notes that the final standard does not require use of a FF and that the EPA
"maintain[s]" instead "that the standards are achievable at all time[s], even with an [electro-static
precipitator (ESP)]." Petitioner states that most - 76 out of 130 - of the coal-fired EGUs
identified as best performers by EPA lack a FF (they instead utilize an ESP). Petitioner maintains
that FFs provide superior reductions to ESPs, and can achieve PM emissions below the final
limits. Petitioner asserts that review of recent permit limits set pursuant to BACT showed that
multiple EGUs had filterable PM BACT determinations well below the final limit using
baghouses. Petitioner alleges that the EPA should establish a BTF standard for existing, coal-
fired EGUs at a rate lower than the final limit and/or an averaging time below 30 boiler-
operating days.
Response to Issue 102: The EPA received comments on the BTF analysis in the proposed rule,
including comments on the PM BTF analysis, and the EPA responded to the comments (see, e.g.,
RTC, Vol. 1, section 4C05). We do not agree with the Petitioner's contention that the change
from total to filterable PM undermines the legitimacy of the PM BTF analysis, and the petition
does not provide new data and information that cause us to question our conclusion. As stated in
the RTC (Vol. 1, page 668), the EPA believes that its decision to not set beyond-the-floor
emission limits for PM is reasonable based on the available information. Although the EPA does
encourage use of the best performing filters, there is simply not enough long term performance
data under actual power plant conditions to justify establishing a beyond-the-floor standard based
on filter efficiency. Because the public had an opportunity to comment on the BTF analysis for
PM, we are denying reconsideration of this issue.
In addition, the EPA has utilized permit data in MATS and in other MACT standards. However,
the EPA believes that where sufficient emission test data are available, as is the case with most
of the MATS categories, those data should be used over permit data. Therefore, the EPA used
the data available to the Agency in establishing the coal-fired EGU PM emission limit. A review
of the MACT floor pool of sources used to set the final limit (130 EGUs) indicates that both FFs
and ESPs are employed by the best performing sources. Our review revealed 67 units using FF
only, 54 units using ESP only, and another 9 units that employed some combination of ESP +
another downstream polishing PM control device. Units using ESPs only were distributed
throughout the top 130 best performing units. At proposal the EPA indicated that"... we could
not identify better HAP emissions reduction approaches that could achieve greater emissions
reductions of HAP than the control technology combination(s)... that we expect will be used to
meet the MACT floor levels of control (and that are already in use on EGUs comprising the top
performing 12 percent of sources), though we did consider duplicate controls (e.g., multiple
scrubbers) in series and found the cost of that option unreasonable." In modeling for the
proposed rule, the EPA predicted installation of 166 GW of FF (beyond that predicted for the
base case). However, some of the FF installations were tied directly to use of dry sorbent
injection (DSI) (predicted 56 GW over base case) and activated carbon injection (predicted 93
GW over base case). In the final rule, the EPA recognized in the modeling that well performing
123
-------
ESPs could meet the filterable PM emission limit. In documentation supporting the IPM
modeling for the final MATS rule, the filterable PM compliance technologies for existing units
are described as follows:
In the MATS policy case all coal units with a capacity greater than 25 MW must
meet the filterable PM compliance requirement. Units that have an existing fabric
filter are assumed to meet the requirement. Depending on the incremental
filterable PM reduction needed to bring a unit into compliance, uncontrolled units
and units with electrostatic precipitators (ESPs) for PM control that do not
currently meet their compliance requirement are assigned either a fabric filter or
one of three tiered ESP upgrades to bring them into compliance.
Thus, the EPA assumed that an EGU that is not in compliance would choose to install a fabric
filter or, depending upon the incremental PM reduction needed, would choose to upgrade their
ESP. In modeling for the final rule, the EPA predicted installation of 102 GW of FF (beyond that
predicted for the base case). However, as with the modeling for the proposal, some of the FF
installations were tied directly to use of DSI (predicted 43 GW over base case) and activated
carbon injection (predicted 141 GW over base case). For the final rule, the EPA predicted that 34
GW of ESP capacity would be upgraded in order to meet the final PM standard. Consistent with
our position at proposal, the EPA continued to believe - and the modeling supported it - that the
PM emission standard would be met primarily by use of existing and newly installed FFs or by
upgraded ESPs that achieve a performance very similar to existing FFs. For these reasons, the
issue is not of central relevance to the final MATS rule and the EPA is denying the Petition for
Reconsideration of this issue.
20.6 Final PM "MACT" floor for existing coal-fired EGUs is weaker than existing PM limits
at more than 12 percent of EGUs subject to the rule
Issue 103: Petitioner 20187 notes that, as the EPA has acknowledged, "[ujnits complying with
stringent PM limits represent the top performers with regard to non-Hg metallic HAP
emissions..Yet in establishing final PM standards, Petitioner alleges that the EPA apparently
did not evaluate the PM limits already in effect.
Petitioner states that, through a partial review of existing PM standards at existing plants, it has
identified 161 units that already have to meet PM emission limits at 0.03 lb/MMBtu or below on
an hourly basis, which is well above 12 percent of the 1,100 sources subject to the rule.
Petitioner notes that compliance determinations for such units are usually based on a 3-hour
rolling average, which is far more stringent than the 30-day rolling average adopted in the final
MATS. Petitioner maintains that these limits are found in Title V permits or consent decrees,
sources that are readily available to EPA and that some have been in place for over two decades.
Petitioner states that many of these are based on the hourly limits established under subpart Da of
the new source performance standards (NSPS). Those NSPS apply to any unit for which
construction commenced after September 18, 1978. The Subpart Da limit adopted at that time
applies on a unit specific basis, i.e., facilities cannot average compliance across multiple units.
Compliance is generally determined by stack testing, but all Title V permits are required to
124
-------
include additional methods for determining compliance with such limits on an hourly basis (or 3-
hour average).
Petitioner alleges that, in other words, the proposed MATS limits are less stringent than the
limits that the EPA has imposed on any new or modified coal unit 34 years ago, and less
stringent than limits with which at least 161 units in operation today claim to continuously
comply. Petitioner maintains that by allowing compliance with an 0.03 lb/MMBtu filterable PM
limit to be averaged across 30 days and multiple units, the MATS limit adopted by EPA is much
less stringent than NSPS standards in effect since 1978, and already met or exceeded at more
than 12 percent of existing sources.
Response to Issue 103: CAA section 112(d)(3) requires the EPA to establish MACT floors
based on the performance of the best performing 12 percent of existing sources in a subcategory
for which the Agency has data for subcategories with 30 or more sources (and the best 5 sources
for subcategories with less than 30 sources). In establishing the standard, the EPA utilized the
extensive emission test data available to the Agency to determine the MACT floors. The EPA
also determined that a beyond-the-floor level of control was not appropriate. The fact, if correct,
that more than 12 percent of coal-fired EGUs meet the final limit does not make the standard
invalid or insufficiently stringent. Coal-fired EGUs are almost universally subject to a PM
standard and the fact that a large number of such units meet the calculated MACT standard does
not make the standard invalid, and it is to be expected.
In addition, the EPA requested data through the 2010 ICR and used the data available to the
Agency in establishing the coal-fired EGU PM emission limits. That is, the EPA evaluated PM
emissions through the available emissions data from more than the 161 EGUs noted by Petitioner
rather than evaluating what these EGUs were supposed to be achieving as evidenced by their
permit or consent decree requirements. The EPA believes that Petitioner is misinterpreting the
information - a stack test conducted over a 3-hour period does not equate to an hourly emission
limit, and most of the limits Petitioner provides in its Attachment are not indicated as being
hourly limits (nor are the provided data sufficient to reassess the final MATS emission limits).
The EPA believes that the final rule, which requires continuous compliance on a 30-day rolling
average basis, is more protective of the environment than one based on an annual manual stack
test.
For these reasons, the issue is not of central relevance to the final MATS rule, and EPA is
denying the Petition for Reconsideration of this issue. In addition to not being of central
relevance, because the Petitioner has not explained why it could not have commented on this
issue during the comment period, we are also denying reconsideration for that reason as well.
20.7 Solid oil-derived fuel-fired EGUs
Issue 104: Petitioner 20173 alleges that, based on technical documents that revealed a significant
number of datasets, 32 out of the 47, were excluded from the EPA's emissions floor analysis.
The omitted data included reported filterable PM test results from both Phase II and Phase III of
the EPA's 2010 ICR databases that were used in the final rule. Petitioner maintains that the EPA
technical documents made available along with the final rule reveal that the EPA did not
125
-------
consider all of the available PM emission data collected from existing petroleum coke EGUs in
setting the filterable PM standard. Petitioner states that it and other commenters were unable to
evaluate EPA's calculation of the filterable PM standard until these technical documents were
released along with the final rule, well after the public comment period closed. Petitioner thus
alleges that Commenters were denied a meaningful opportunity to submit comments regarding
the final filterable PM standards and the floor analysis, as the EPA's proposal was an entirely
different approach imposing emission standards for total PM. Petitioner also alleges that the
EPA's original floor analysis contained a number of significant flaws that resulted in standards
that were biased low in most cases, including inappropriate handling of non-detect values,
determining the units in the emissions floor pool (e.g., "pool determinate method") based on
minimum test averages, and using inadequate criteria to determine the appropriate UPL
methodology to address variability. For the PM standards, Petitioner states that there were
additional problems such as inconsistencies in calculating the UPL compared with other
subcategories and the exclusion of significant amounts of PM emissions data obtained under
Phase II of the ICR.
Response to Issue 104: Petitioner cites to the EPA's 2010 ICR to support this assertion, but fails
to identify which datasets the EPA "ignored." It appears from Petitioner's attachment that the 47
datasets quoted are run-by-run data from the top 5 units; however, the EPA was not able to
match the number of runs Petitioner says are available for each EGU (Petitioner's Table 2 in the
attachment) as the text is not detailed enough. Petitioner also has claimed to screen the ICR data
to exclude runs where (1) the petcoke firing rate was less than 90 percent of total rated capacity,
(2) there was missing fuel data during the test, and (3) there were potential test issues identified
in the database or by the source test contractor. The EPA tried to replicate Petitioner's item (1)
with the boiler rated capacity data available in the database to see if this brought our data set
closer to Petitioner's count, but we could not see how Petitioner arrived at 47 run-level data
points. When the EPA looked to screen the available data for the 90 percent of total heat input,
we did not see any data drop out. Absent Petitioner's spreadsheet, showing the data used and the
data rejected, we cannot replicate their MACT floor result.
The data used between proposal and final (as presented in the MACT floor spreadsheets) did not
change. The data used by the EPA at final in establishing the filterable PM limit were included in
the MACT floor spreadsheets at proposal and stakeholders could have commented on the data
and analyses during the public comment period. That is, the filterable PM data included in the
total PM analysis (EPA-HQ-OAR-2009-0234-3040; column CM in "Filterable PM_petcoke"
tab) at proposal are the same data as were used for the filterable PM analysis (EPA-HQ-OAR-
2009-0234-20132, attachment a7; column CI in "Filterable PM_petcoke" tab) at final. Further,
the EPA moved to filterable PM based on comments received (including supporting comments
from Petitioner), and Petitioner should and could have commented on the alleged data issues at
proposal as it was aware that the EPA could accept the request to switch formats made in
comments. Petitioner had ample opportunity to raise its data issues in public comments on the
proposed rule. Because Petitioner has not demonstrated that it was impracticable to comment on
this issue during the comment period on the proposed rule and because the issue is not of central
relevance, the EPA is denying the Petition for Reconsideration of the issue.
126
-------
The EPA did not ignore relevant data sets related to pet-coke units, as Petitioner argues.
Additionally, the EPA did not rely on data where the pet-coke firing rate was less than 90 percent
of the total heat input during a particular performance test in order to ensure that the standards
were not based on EGUs that fired a significant amount of a different fuel, which potentially
excluded some of Petitioner's data sets. As for all the subcategories in the MATS Rule, the EPA
also excluded 2010 ICR data where there was missing fuel data, or where the source test
contractor identified potential test issues. See Docket ID EPA-HQ-OAR-2009-0234-20220,
Reconsideration MACT Floor Memo. The EPA did not subject pet-coke units to disparate
statistical treatment, and its approach was reasonable because the EPA used this same approach
for other small data sets (e.g., for the IGCC subcategory).
Responses to Petitioner's other issues (e.g., use of non-detects, selection of the MACT floor pool
of sources, and use of the UPL) are addressed elsewhere in this document.
Because the Petitioner has failed to demonstrate it was impractical to comment on this issue in
the proposed rule, and because the Petitioner has not provided new relevant information or data
that became available after the close of the comment period but during the time specified for
seeking judicial review, we are denying this petition. In addition, we are denying the petition
because it is not of central relevance for the reasons discussed above.
127
-------
21.0 Regulatory Impact Analysis and Statutory or Executive Order Analyses
21.1 Regulatory Impact Analysis (RIA) for non-continental liquid oil-fired EGUs
Issue 105: Petitioner 20188 states that the EPA should complete a separate RIA to address the
specific impacts for the non-continental liquid oil-fired EGU sub-category. Petitioner cites CAA
section 112(d)(2) to support its assertion that the EPA was required to consider costs, benefits,
and other impacts when setting the MATS emission limits. Petitioner believes that this mandate
extends to subcategories that may have different economic considerations than other
subcategories; in this situation this includes the facts that each non-continental area is an isolated
grid without interconnections that allow least-cost dispatch, current non-continental electric rates
are substantially higher than continental rates, and access to alternative fuels is more limited and
costly.
Response to Issue 105: EPA fully complied with the executive order requirements to prepare a
regulatory impact analysis (RIA) for this final rule. The purpose of the RIA is in part to evaluate
the cost and benefits of the promulgated standards, but the RIA is not directly relied upon during
the rulemaking process. Moreover, even if EPA had prepared a separate analysis of impacts on
the non-continental liquid oil-fired EGU sub-category, it would not have altered the MACT
standards for sources in that category, which were set at the minimum level of stringency
permitted by the statute. CAA section 112(d)(2) states that emission standards set under that
section shall require the maximum degree of reduction in emissions that the Administrator
determines is achievable. CAA section 112(d)(2) does allow for consideration of costs and other
impacts. CAA section 112(d)(3), however, specifies that the level of emission reduction deemed
achievable for new sources shall not be less stringent than the emission control achieved by the
best controlled similar source (the minimum stringency level of control is the MACT floor).
Similarly, the CAA requires that the existing source MACT floor shall not be less stringent than
the average level of HAP emission control achieved by the top 12 percent of sources for each
category or subcategory with 30 or more sources and the best performing 5 sources for each
category or subcategory with less than 30 sources. All of the MACT standards for non-
continental EGUs are set at this minimum stringency level, and thus would not have been
impacted by additional analyses of other considerations. For these reasons, Petitioner's objection
is not of central relevance, and we are denying the petition for reconsideration on this issue.
Additionally, the EPA disagrees that it is obligated under any law or executive order to produce a
separate RIA for any individual subcategory of regulated sources and notes that it has not
produced such an analysis for any other subcategory of sources subject to MATS. The EPA
agrees with Petitioner's view that the unique considerations faced by non-continental EGUs,
including their limited ability to obtain alternative fuels that produce different emissions
characteristics, warrant a separate subcategory for these EGUs and, in response to data received
after publication of proposed rule, the EPA included such a subcategory in the final rule. The
final MATS RIA (EPA-HQ-OAR-2009-0234-20131) includes costs for oil-fired EGUs that were
not included in the proposal analysis. Although the EPA does not estimate specific costs or
benefits for non-continental liquid oil-fired EGUs, in Chapter 3 of the RIA the EPA notes that
the data suggest the final MATS emission limitations for non-continental EGUs will likely allow
these EGUs to continue firing residual fuel oil without additional air pollution controls (p. 3-29).
128
-------
Petitioner does not introduce any additional data that would cause the EPA to reassess this
conclusion. Although Petitioner notes that it would incur significant costs if forced to achieve
compliance with MATS by burning distillate rather than residual oil, MATS does not force the
use of any specific compliance method. Further, Petitioner admits that it is not planning to
comply by that method, making the estimated cost data irrelevant.
21.2 Regulatory Flexibility Act (RFA) and Small Business Regulatory Enforcement Fairness
Act (SBREFA) issues
Issue 106: Petitioner 20189 states that the EPA failed to meet its obligations under the
Regulatory Flexibility Act (RFA) and Small Business Regulatory Enforcement Fairness Act
(SBREFA). Specifically, Petitioner states that the SBREFA panel EPA conducted was
inadequate, the Initial Regulatory Flexibility Analysis (IRFA) and Final Regulatory Flexibility
Analysis (FRFA) failed to meet the requirements set out in RFA sections 603 and 604, and the
FRFA does not recognize the unique problems faced by public power utilities.
Response to Issue 106: These issues were raised in public comments submitted in response to
the May 3, 2011, proposed MATS rule (76 FR 24976). The EPA's responses to the comments
are in the RTC (see, e.g., Vol. 2, pp. 669 - 675). The comments that the EPA received on these
issues demonstrate that the public had ample opportunity to comment on these issues, and the
EPA responded to those comments. Because Petitioner did not demonstrate that it was
impracticable to comment on these issues during the comment period on the proposed rule, the
EPA is denying the Petition for Reconsideration of these issues.
Petitioner had ample opportunity to comment on the Initial Regulatory Flexibility Analysis, and
those comments were taken into account in developing the FRFA, which accompanies the final
rule. The commenter does not raise any new issues beyond those already addressed in the RTC
document regarding the IRFA and does not introduce new information related to the FRFA. The
EPA is not required to provide additional opportunity to comment on the agency's responses to
comments received on these issues.
In addition, this issue is not of central relevance under the CAA because the statute does not
require consideration of the FRFA when acting under section 112. Alleged noncompliance by the
EPA with these requirements could also have been raised in the White Stallion litigation. No
parties challenged the Agency's actions in implementing the FRFA requirements and we do not
believe any errors exist in the manner we implemented these separate statutory mandates.
21.3 Unfunded Mandate Reform Act (UMRA) and Executive Order 13132 (Federalism) issues
Issue 107: Petitioner 20189 states that the EPA failed to meet its obligations under the Unfunded
Mandates Reform Act of 1995 (UMRA) and Executive Order 13132 (Federalism). Specifically,
Petitioner states that the EPA did not conduct required consultation with municipal governments
and violated the requirements of Title II of the UMRA.
129
-------
Response to Issue 107: These issues were raised in public comments submitted in response to
the proposed MATS rule published May 3, 2011 (76 FY 24976). The EPA's responses to the
comments are in the RTC (see, e.g., Vol. 2, pp. 676 - 680). The comments that the EPA received
on these issues demonstrate that the public had ample opportunity to comment on these issues,
and the EPA responded to those comments. Alleged noncompliance by the EPA with these
requirements could also have been raised in the White Stallion litigation. As the Court may only
consider issues raised during the period for public comment, issues that could have been raised in
the litigation and addressed by the Court clearly do not meet the criteria for reconsideration in
CAA section 307(d)(7)(B). Because Petitioner did not demonstrate that it was impracticable to
comment on these issues during the comment period on the proposed rule, the EPA is denying
the Petition for Reconsideration of these issues.
In addition, these issues are not of central relevance under the CAA because the statute does not
require consideration of any UMRA or Executive Order 13132 analyses when acting under CAA
section 112. Moreover, the EPA does not believe Petitioner has identified any errors in the
manner we implemented this separate statutory mandate.
21.4 Methodological issues in estimates of PM2.5 co-benefits
Issue 108: Petitioner 20180, att. 6 claims that the EPA did not respond to methodological issues
identified in the Greven, et al., (2011) study92 regarding the epidemiology studies used to
estimate PM2.5 co-benefits in the RIA. Petitioner also expresses concerns with the EPA's
treatment of uncertainty in benefits estimates, differential toxicity of particle species, and
mortality valuation that differs from other federal agencies.
Response to Issue 108: The purpose of a RIA is to inform the public about the potential costs,
potential benefits, and other potential impacts associated with a regulatory action. As stated in
the RTC (Vol. 1, p. 62), "[t]he RIA is not a requirement of the CAA generally, or CAA section
112 specifically. The agency's legal and factual bases for this rule are set forth in the preambles
to the proposed rule and in this final action." Further, as stated in the RTC (Vol. 1, p. 64), "[t]he
EPA does not use the benefits estimates in the RIA to support the Appropriate and Necessary
Finding. As shown in the Hg and non-Hg risk assessments supporting the finding, Hg and non-
Hg HAP continue to pose hazards to public health and the environment, and U.S. EGU emissions
cause and contribute to these hazards.'" Therefore, RIA issues are not of central relevance to the
"Appropriate and Necessary" finding or to the emission standards established in the final MATS
rule, and we are denying reconsideration for that reason.
Further, issues regarding the EPA's methods for calculating PM2.5 co-benefits were raised in
public comments submitted in response to the proposed MATS rule. The EPA's responses to
these health benefits comments are in Section 6K (Vol. 2, pp. 624 - 632) of the RTC. The EPA
also addressed each of these issues extensively in Chapter 5 of the proposal RIA (EPA-HQ-
OAR-2009-0234-3051) and the final RIA (EPA-HQ-OAR-2009-0234-20131). The comments
92 Greven S, Dominici F, Zeger S. 2011. "An Approach to the Estimation of Chronic Air
Pollution Effects Using Spatio-Temporal Information." J. Amer. Stat. Assoc., 106, 494, 396-406.
130
-------
received on these issues demonstrate that the public had ample opportunity to comment on this
issue, and the EPA responded to those comments.
Because the issue is not of central relevance to the final MATS rule and because Petitioners did
not demonstrate that it was impracticable to comment on this issue during the comment period
on the proposed rule, the EPA is denying reconsideration of this issue.
131
-------
22.0 Reliability and Compliance Time
22.1 Electricity reliability
Issue 109: Petitioner 20180 stated that the EPA did not include in the docket for public comment
certain studies that Petitioner alleges show that the EPA's regulations would have serious effects
on reliability. As an example, Petitioner presents a 2010 study by the North American Reliability
Corporation (NERC). Petitioner further states that the EPA did not properly consider the impact
of MATS on electricity reliability in finding that it was appropriate and necessary to regulate
coal and oil-fired power plants under CAA section 112 and cites a FERC staff analysis to support
its assertion.
Petitioner argues that the EPA dismissed information that MATS would seriously compromise
electricity reliability and asserts that the EPA should grant reconsideration to address the impacts
on system reliability and take public comment on all "relevant information." Petitioner states that
the EPA should reconsider the rule because the EPA underestimated the number of EGUs that
would retire as a consequence of MATS. Petitioner further asserts that the EPA did not do a
cumulative analysis of retirements due to MATS and other regulations that were recently or soon
to be promulgated.
Petitioner states that there would be short-term reliability impacts at the end of the compliance
period, as many plants complete their retrofits during the spring of 2015.
Petitioner also states that the EPA did not account for the effect of impacts to local reliability
beyond acknowledging that there could be an impact. Petitioner adds that the EPA is not the
proper agency to analyze reliability but regardless, the EPA should reconsider its decision not to
analyze local reliability.
Response to Issue 109: The argument put forth by Petitioner that the EPA did not consider the
reliability impacts of MATS in developing the rule is not a new comment, and the EPA
addressed this at length in the preamble to the proposed rule (76 FR 25054, section V.M.), in the
RTC (Vol. 2, p. 287, section 5C01), and in the preamble to the final rule (77 FR 9406, section
VII.F.). For this reason, the EPA is denying reconsideration of this petition.
The preambles of both the proposed and final rules as well as of the actions the agency was
taking in concert with its federal partners include lengthy discussions of the potential impact of
MATS and thus demonstrate that the EPA did consider potential impacts of the rule on
electricity reliability. The EPA undertook analysis of the amount of generating capacity that
would be retired rather than retrofitted with pollution controls to meet the requirements of the
rule. This analysis showed that although local reliability issues could not be ruled out, the bulk
power system has adequate resource capacity, even with projected retirements. The EPA also
analyzed recent investments by the industry and compared that to what MATS would require,
and the EPA concluded that the industry was capable of meeting MATS without compromising
reliability. The preamble to the proposed rule discussed these analyses, and they were in the
docket, allowing the opportunity for public comment. The DOE also performed a study of
resource adequacy under MATS and reached the same conclusion as the EPA.
132
-------
The EPA explained in the preamble to the final MATS (77 FR 9406, section VII.F.) and in the
RTC (Vol. 2, p. 287, section 5C01) that several studies conducted by groups outside of the EPA
were seriously flawed. The NERC study and the FERC Preliminary Staff Analysis that Petitioner
asserts should have been considered by EPA as part of the rulemaking were both based on
assumptions about requirements for the EPA's rules that differed substantially from what the
EPA proposed. As such, these studies are not relevant to this rulemaking, and the EPA has no
obligation to solicit comment on them and reconsider the rule on this basis. Regarding the FERC
staff analysis that Petitioner cited, which remains a staff draft, the FERC chairman stated that the
analysis was "inadequate to be used as a basis of decision-making."93 The EPA agrees with the
Chairman of FERC and further notes that this staff draft is not relevant to the appropriate and
necessary finding.
The possibility that plants may not be able to retrofit by the end of the compliance period and
thus cause a reliability problem was raised during the comment period and is discussed in the
preamble to the rule (77 FR 9406, section VII.F.) and in the RTC (Vol. 2, p. 287, section 5C01).
The EPA notes that although the EPA expects that most plants would be able to meet the MATS
requirements within the compliance period, when the EPA issued the final rule the EPA offered
flexibilities allowable under the CAA in order to avoid the unlikely situation that a plant needed
for reliability could not timely comply and hence compromise local reliability. The EPA also
encouraged early planning and action to ensure timely compliance with MATS.94
The EPA disagrees with Petitioner's assertion that the EPA did not account for local reliability
issues. The EPA has worked closely with interested parties since promulgation of the MATS rule
to monitor compliance, identify potential reliability concerns, and ensure that any potential
problems are addressed, and reliability is not compromised. Parties have included DOE, FERC,
states (both their public utility commissions and administrative environmental departments),
system operators and other planning authorities, and power generators. Consistent with the
analyses of EPA and DOE performed during the rulemaking process, it appears that the tools that
these parties have at their disposal are adequate to ensure electric reliability related to MATS
compliance and that all parties are moving forward to address reliability concerns related to EGU
retirements and outages to install controls. Further, though more cumulative retirements of coal
capacity are occurring due to market conditions, existing mechanisms appear to be adequately
addressing these retirements.
93 Testimony of Chairman Jon Wellinghoff, FERC before the House Subcommittee on Energy
and Power of the Committee of Energy and Commerce, US House of Representatives,
September 14, 2011.
94 US Environmental Protection Agency, An Assessment of the Feasibility of Retrofits for the
Mercury and Air Toxics Standards Rule, EPA-HQ-OAR-2009-0234-20001;
US Environmental Protection Agency, National Emission Standards for Hazardous Air
Pollutants from Coal- and Oil-fired Electric Utility Steam Generating Units and Standards of
Performance for Fossil-Fuel-Fired Electric Utility, Industrial-Commercial-Institutional, and
Small Industrial- Commercial-Institutional Steam Generating Units, February 16, 2012, 77 FR
9406-9411.
133
-------
The EPA is denying reconsideration because the Petitioner has not demonstrated that it was
impracticable to raise these issues during the comment period, because the Petitioner has not
provided new data or information that became available before the end of the period for judicial
review, and because the issue is not of central relevance to the final MATS rule.
22.2 Policies and procedure for extending the compliance time beyond 3 years
Issue 110: Petitioner 20189 maintains that the EPA brushed aside its concerns that public power
entities need additional time for compliance with the rule due to their additional procedural
requirements for procuring the equipment, engineering and labor needed for compliance.
Petitioner cites the need for public and competitive bidding, procedures for obtaining public
bonds, and the use of eminent domain, which might be necessary in some cases. Petitioner also
states that because some of its members are small entities, they will not be as able to compete for
services, equipment and labor as larger entities are. Petitioner also maintains that the
enforcement response policy memorandum that outlines a procedure for the use of
Administrative Orders for an additional year to come into compliance does not address
Petitioner's members' special needs and notes that it was not subject to public notice and
comment.
Petitioner 20180 maintains that the EPA has not provided sufficient time for compliance with the
rule and failed to adequately consider comments to that effect. They maintain that not even 5
years is sufficient time to achieve compliance. Petitioner maintains that remedies the EPA
presented to address potential issues related to the MATS compliance time frame were
inadequate. Petitioner asserts that the fourth year that states can grant for compliance is
inadequate, as is the enforcement policy. Petitioner also asserts that the enforcement policy was
never subject to notice and comment, and therefore the EPA should grant reconsideration.
Petitioner argues that states may take too long to issue a fourth year, and hence the EPA should
issue a "general permit" that could be in the rule itself explaining that an additional year is
necessary and the conditions under which it applies. Petitioner also states that the requirements
for a Presidential Exemption are met and that the EPA should seek a Presidential Exemption.
Response to Issue 110: The EPA provided ample opportunity to comment on the amount of
time that public power entities need to install controls. In fact, numerous commenters raised this
issue in comments submitted in response to the proposed MATS rule. The EPA's responses to
the comments are in the RTC (see, e.g., Vol. 2, pp. 342 - 343). Because Petitioners have not
demonstrated that it was impractical to raise this issue during the public comment period, we are
denying reconsideration of this issue.
Regarding the enforcement policy memorandum which is the subject of comments from both
Petitioners, this memorandum is not a part of the MATS rule. This comment is therefore
irrelevant to the final MATS rule and, thus, not of central relevance. Moreover, the enforcement
memorandum is not a rulemaking under the CAA and is not subject to notice and comment
requirements. For similar reasons, Petitioner 20180's assertion that the EPA should seek a
Presidential Exemption is irrelevant to the petition. The President possesses the authority to grant
a presidential exemption under CAA section 112(I)(4), not the EPA. Because these issues are not
134
-------
of central relevance to the MATS rule, we are denying the petitions for reconsideration on this
issue.
In addition, the Agency's decision not to issue a blanket extension to public utility companies
was challenged in the White Stallion case, and Petitioners in that case raised many of the same
arguments identified in the petition for reconsideration. The Court affirmed the EPA's decision
to reject a blanket extension based on the information in the record. White Stallion, 748 F.3d at
1251-52. As the Court may only consider issues raised during the period for public comment,
issues raised in the litigation and addressed by the Court clearly do not meet the criteria for
reconsideration in CAA section 307(d)(7)(B). In addition, petitioners have not provided new
information or data relevant to that decision and thus the petition does not provide substantial
support for an argument that the rule should be revised and the issue is thus not of central
relevance.
For these reasons, the EPA is denying the petitions for reconsideration on these issues.
22.3 EPA made erroneous assumptions with regard to ERCOT
Issue 111: Petitioner 20174 states that EPA made erroneous assumptions regarding ERCOT and
the Texas electric grid when it analyzed the effects of the rule on electric reliability and that the
EPA refused to correct those assumptions for the final rule. Specifically Petitioner maintained
that the EPA assumed unlimited transmissions capability with ERCOT and that the EPA used
different estimates of capacity, demand, and reserve margins than ERCOT was using.
Response to Issue 111: Issues related to the regional agencies, including ERCOT, were raised in
public comments submitted in response to the proposed MATS rule. The EPA's responses to the
comments are in the RTC (see, e.g., Vol. 2, pp. 289 - 296 and 324 - 331). Because Petitioner did
not demonstrate that it was impracticable to comment on these issues during the comment period
on the proposed rule, the EPA is denying the petition for reconsideration of the issues.
Furthermore the assumptions that the EPA used in its analysis reflect factors specific to each
NERC region, including ERCOT. These factors include the transmission dynamics of Texas and
the NERC region ERCOT. See section 3.3 of IPM 4.10 Documentation which details the
assumptions regarding the transfer capabilities, wheeling costs, and inter-regional transmission
used in EPA's IPM modeling (available at: http://www.epa.gov/airmarkets/progsregs/epa-
ipm/BaseCasev410.html#documentation). The EPA's assumptions regarding load modeling,
capacity and reserve margins are documented in sections 3.2, 3.5 and 3.6 respectively.
In addition, as explained in the preamble to the final rule, the EPA has made every effort to
provide the fullest compliance period allowed by the CAA to assure that plants will have
adequate time to come into compliance. The EPA notes that since the rule was promulgated
agency staff along with staff from the DOE and the FERC have met regularly with ERCOT to
monitor MATS compliance and the efforts of all parties to assure electricity reliability in the
region. Because the EPA has provided the maximum time for compliance authorized by the
statute, the issue is not of central relevance to the final MATS rule, and, therefore, we are
denying reconsideration of this issue for that reason as well.
135
-------
23.0 Reporting
Issue 112: Petitioner 20180 objects to the requirement to report Hg, HF, and HC1 emissions data
in a greater number of significant figures than required by the proposed standards and claims it
was unable to comment on this requirement because the number of significant figures required to
be reported by the final rule changed from those proposed. Petitioner claims that the EPA has
provided no data to suggest that the additional figure is significant, or consistent with its
calculation of the final emission standards. Petitioner similarly claims it was unable to comment
on the rule's requirement to round the 30-day average Hg emissions rate to two significant
figures and that no language is found in Appendix A to warrant rounding to two significant
figures for Hg emissions.
Response to Issue 112: The EPA disagrees with Petitioner's objection that it was unable to
comment on requirements concerning the number of significant figures required to be reported
by the final rule for Hg, HF, and HC1 emissions. Comments were received on the number of
significant figures required to be reported in the proposed rule (see, e.g., RTC, Vol. 1, pp. 465,
786; Vol. 2, pp. 126); Petitioner does not explain how or why it was unable to provide
comments. Based on comments received concerning reporting of values such that a specific
number of decimal places would be provided, the EPA revised the rule's requirement to instead
report values to a specific number of significant figures. Such an approach contains three
significant figures for hourly concentrations and emission rates, which are then rounded to two
significant figures for 30-day averages (which is consistent with the number of significant
figures contained in the emissions limits). Those changes to the rule, which are based on public
comments, as well as the EPA responses, can be found in the RTC, Vol. 2, p. 126. As the
Petitioner has not demonstrated that it was impracticable to comment on the HF and HC1
significant figures reporting requirement during the comment period on the proposed MATS
rule, the EPA is denying reconsideration of these issues.
Issue 113: Petitioner 20180 states that although the EPA finalized provisions that allow use of a
diluent cap for Hg emissions during periods of startup and shutdown, the EPA did not finalize
provisions that allow use of a diluent cap for SO2, HC1, and HF or a megawatt default value for
Hg, HC1, HF, PM, and non-Hg metals under low load conditions and that Petitioner was unable
to comment on the implications of recording and reporting CEMS values without the benefit of
the proposed but not finalized default diluent and megawatt values.
Response to Issue 113: This issue is moot, as the EPA addressed this issue as part of the
reconsideration. 77 FR 68785. EPA provided an additional opportunity to comment on the issue
in that context. For this reason, the EPA is denying reconsideration of this issue.
Issue 114: Petitioner 20180 objects to the final rule's requirement to evaluate HC1 and/or HF
monitoring plans using the Emissions Collection and Monitoring Plan System (ECMPS) and
claims it was unable to comment on this requirement, as the EPA did not propose it.
Response to Issue 114: In the proposed rule preamble, the EPA mentioned that with just a few
tweaks, the ECMPS could be used to evaluate the other monitoring plans, including those for
HC1 and HF, and the EPA asked for comment on such an approach (see 76 FR 25033). A
136
-------
commenter asked the EPA to include HC1 and HF monitoring plan evaluation as part of ECMPS,
and after consideration, the EPA made that expansion in the final rule (see, e.g., RTC, Vol. 2, pp.
266 - 271). The EPA disagrees with Petitioner's objection that it was unable to comment on
requirements to evaluate HC1 and/or HF monitoring plans using the ECMPS. The Petitioner
provided comments on requirements to evaluate Hg monitoring plans using the ECMPS, but
does not explain how or why it was unable to provide the same type of comments for HC1 or HF
monitoring plans, especially since the proposed rule pointed out that HC1 and HF monitoring
plan evaluation could be accomplished via ECMPS and sought comment on the proposal's
approach. As the Petitioner has not demonstrated that it was impracticable to comment on this
issue during the comment period on the proposed MATS rule, the EPA is denying
reconsideration of this issue.
Issue 115: Petitioner 20180 claims that the EPA added requirements for EGU owners or
operators to submit reports of startup or shutdown periods without proposing to require such
reports; that some of the rule's cites appear to be misplaced, as the cited material does not require
report submission; and that it was unable to comment on the requirements because they were not
proposed. Petitioner claims that the EPA must propose and solicit comment on such
requirements.
Response to Issue 115: The EPA reconsidered the requirements applicable during periods of
startup or shutdown, and provided an additional opportunity for public comment on this issue in
that context. 77 FR 68777 Therefore, this issue is moot.
Issue 116: Petitioner 20180 claims the nature and purpose of the requirement to include in the
notice of compliance a justification for the selection of fuel(s) burned during the performance
test is unclear since the EPA did not finalize a requirement from the proposal to conduct
performance tests while combusting the worst case fuel or fuel mixture. Petitioner claims the
EPA must explain and solicit comment on these requirements.
Response to Issue 116: The EPA denies reconsideration of the rule on this issue. The Petitioner
is incorrect to believe that the requirement to have the notification of compliance justify why a
particular fuel was used during a performance test is a relic associated with now-removed
independent operating limits. Knowing what fuel was used and why that fuel remains an
appropriate choice is important for emissions limit applicability determinations, for alternative
monitoring option selections, for establishing a baseline for determining when retesting for
parametric monitoring operating limits may be required, and, for the EGUs that combust fuels
other than fossil fuels or waste, for complying with those fuel combustion requirements.
Moreover, the Petitioner has not demonstrated that it was impracticable to comment on this
issue. In fact, other commenters did comment on the issue and EPA's responses to those
comments are in the RTC (see, e.g., RTC, Vol. 2, p. 263). For these reasons, we are denying
reconsideration.
Issue 117: Petitioner 20180 claims that the EPA finalized a requirement to have EGU owners
and operators report new types of fuel being combusted, as well as dates of performance testing
using the new types of fuel, without having proposed such requirements. Petitioner claims the
EPA must explain and solicit comment on these requirements.
137
-------
Response to Issue 117: The EPA denies reconsideration of the rule on this issue because the
promulgated reporting requirements under 40 CFR 63.10031(c)(3) are consistent with the
proposed initial testing requirements under 40 CFR 63.1001 l(b)(l)(i), (2)(i), (3)(i), and (4)(i)
that require affected source operators to "determine the fuel type or fuel mixture that you could
burn in your EGU that has the highest content" of each regulated HAP prior to conducting the
initial performance test, and the promulgated reporting requirements under
40 CFR 63.10031(c)(3) are also consistent with the proposed continuous compliance requirement
under 40 CFR 63.10021(a)(2) to "keep records of the type and amount of all fuels burned in each
EGU during the reporting period to demonstrate that all fuel types and mixtures of fuels burned
would either result in lower emissions of HC1, HF, S02, non-Hg HAP metals, or Hg, than the
applicable emission limit for each pollutant (if you demonstrate compliance through fuel
analysis), or result in lower fuel input of chlorine, fluorine, sulfur, non-Hg HAP metals, or Hg
than the maximum values calculated during the last performance tests (if you demonstrate
compliance through performance stack testing)." The EPA provided an opportunity to comment
on reporting requirements for the rule. Changes to the reporting requirements were based on
public comments the EPA received regarding fuel use from EGU owners or operators whose
EGUs combust other than fossil fuel or waste and who self-determine or have the EPA determine
via petition that their fuel is a non-waste under 40 CFR 241.3 or who combust fuel processed
from discarded non-hazardous secondary materials (see, e.g., RTC, Vol. 2, p. 263). The
Petitioner did not demonstrate that it was impracticable to comment on this issue during the
comment period on the proposed rule. For this reason, the EPA is denying reconsideration of this
issue.
Issue 118: Petitioner 20180 claims that the EPA proposed many duplicative reporting
requirements but did not provide regulatory language to implement the reporting using ECMPS,
that Petitioner provided extensive comments on the reporting requirements, and that the EPA
rejected most of the Petitioner's comments but claimed to have revised the rule to address the
duplicative requirements. Petitioner claims it was unable to provide comments on the
requirements in the final rule, that the EPA should remove what Petitioner believes to be other
duplicative requirements, and that the EPA should provide mechanisms for others to use data
already reported to the EPA.
Response to Issue 118: Petitioner acknowledges that it submitted comments during the
comment period for the proposal and that the EPA provided responses to the comments on this
issue. (See, e.g., RTC, Vol. 2, pp. 128 - 129, 283 - 284, 200 - 201, and 264 - 271). In response to
the Petitioner's and other comments, the EPA revised the rule to remove the duplicative
requirement to submit hard copies of the Notification of Compliance Status and Compliance
Reports to EPA Regions. The EPA also provided the "delegated authority" discretion to require
submission of performance test results and RATA reports, believing these changes will address
the Petitioner's concerns. As the Petitioner has not demonstrated that it was impracticable to
comment on these issues during the comment period, the EPA denies reconsideration of these
issues.
Issue 119: Petitioner 20180 claims that in some cases the proposal was not clear with respect to
reporting requirements for deviations or excess emissions, that the final rule is still unclear, and
138
-------
that it was unable to comment on the reporting requirements for deviations or excess emissions
contained in the final rule. Petitioner objects to these reporting requirements and suggests that
the EPA issue and solicit comment on a proposal that clearly defines these reporting
requirements using consistent terminology.
Response to Issue 119: Petitioner appears to believe the terms 'deviation' and 'excess emission'
are exclusive; this misunderstanding appears to be the basis of its issue. As defined in the
proposed and final rules, a deviation means any occurrence outside what is normal or expected
and a deviation may or may not be a violation of the rule. In response to comments received on
the proposed rule - including those from the Petitioner - the final rule now contains a definition
of excess emissions (see, e.g., RTC, Vol. 2, p. 285). Although a deviation may or may not cause
excess emissions or a violation, an excess emission event is both a deviation and a violation.
Moreover, in response to comments submitted by this Petitioner, the EPA revised the final rule
to clarify what constitutes an excess emissions report, as referenced in the General provisions of
part 63 (see, e.g., RTC, Vol. 2, pg. 284). Because the Petitioner did not demonstrate that it was
impracticable to comment on these issues during the comment period on the proposed rule, the
EPA is denying reconsideration of these issues.
Issue 120: Petitioner 20180 claims that the EPA finalized changes to the proposal concerning
monitoring systems deviations that it was unable to comment on because the EPA did not
propose this language or propose its interpretation with respect to scheduled maintenance.
Petitioner seeks revision of the rule language to exclude monitoring system maintenance from
being a deviation unless the activities are included in a site-specific monitoring plan.
Response to Issue 120: As mentioned in the preamble to the final rule, commenters suggested
that the EPA needed to clarify the excused reasons for monitor downtime (including an exclusion
for scheduled and unscheduled maintenance, as suggested by the Petitioner) and the final rule
was revised to require that an EGU report as deviations to the rule a failure to collect data during
required periods if the deviations were not covered by exceptions allowed in the final rule (see
77 FR 9421). More specifically, the final rule was revised to clarify that data from PM CEMS,
PM CPMS, or HAP metals CEMS need not be collected during periods of scheduled
maintenance (see 63.10010(h)(6)(ii), (i)(5)(i)(B), and (j)(4)(i)(B)). Should an EGU owner or
operator want additional periods of scheduled maintenance for data exclusion, he should specify
such periods in the site-specific monitoring plan. Because the Petitioner did not demonstrate that
it was impracticable to comment on these issues during the comment period on the proposed
rule, the EPA is denying reconsideration of these issues.
139
-------
24.0 Statistical Analyses
Issue 121: Petitioner 20188 asserts that in other MACT rules, the EPA evaluates whether the
data were more consistent with a normal or log normal distribution and assumes log normality,
unless the skewness and kurtosis tests indicate that normality better describes the data
distribution. Petitioner states that for the MATS Rule, however, the EPA has argued that the
analysis of data should be based on test averages, and that, per the Central Limit Theorem, the
average of individual measurements tends toward normality as the number of samples increases.
Petitioner believes that the EPA had insufficient data to support an assumption that the test data
used to set the MACT floor levels for LONC EGUs is normally distributed and quotes the EPA's
consultant as saying, "[w]hen the sample size is less than 15 the Central Limit Theorem cannot
be used to support the normality assumption." Petitioner states that given the amount of existing
data that shows that environmental samples tend to be log normally distributed (reflecting the
fact that many reasons can result in occasionally higher than "normal" readings and that lower
than "normal" readings are constrained by the physical reality that low readings cannot be less
than zero and by the detection limit), using the Central Limit Theorem to argue that a small data
set consisting of data from one or two EGUs is normally distributed is not statistically defensible.
Because of the small sample size used for the LONC subcategory, Petitioner contends that the
EPA should evaluate the data assuming log normal distribution.
Petitioners 20188 and 20191 contend they have found examples of where the EPA did not
correctly use its chosen analytical methods. Petitioners note that the EPA chose to base its
MACT floor analysis on the use of average emission rates and the assumption that such averages
are normally distributed. However, Petitioners state, when determining the MACT floor for
filterable PM for LONC EGUs, the Excel spreadsheet documenting the EPA's analysis shows
that the EPA based the calculations on six individual test values from two boilers, not on the
average emission rates from these two boilers, to develop a MACT emission limit of 0.03
lb/MMBtu. According to Petitioners, if the average emission rates for these two boilers were
used in the analysis, the calculated UPL for filterable PM would be 0.134 lb/MMBtu, which
would round upward to 0.14 lb/MMBtu. Petitioners state that although the EPA may have some
discretion in selecting the methodology to be used for analyzing data, once the EPA selects
methodology, the EPA must consistently apply that methodology and that the EPA did not do so
with the LONC subcategory.
Response to Issue 121: The issue of the statistical analysis was raised in public comments
submitted in response to the proposed MATS rule. The EPA's responses to the comments are in
the RTC (see, e.g., Vol. 1, pp. 490 - 521). The comments the EPA received on this issue
demonstrate that the public had ample opportunity to comment on the issue, and the EPA
responded to those comments. The EPA followed the same procedures on this issue at proposal
and final. Because Petitioner did not demonstrate that it was impracticable to comment on this
issue during the comment period on the proposed rule, the EPA is denying the Petition for
Reconsideration of the issue.
140
-------
Issue 122: Petitioner 20191 maintains that the PM standard does not properly reflect EGU
variability over a 30-day rolling average. Petitioner states that the EPA determined the 30-day
rolling average PM standard using the UPL formula and six data points from two stack tests but
that the UPL has been demonstrated to produce seriously flawed results when based on a
significantly larger data set and similar assumptions. According to Petitioner, the EPA relied on
an extremely limited data set to set the PM standard for the existing LONC subcategory, and this
reliance exacerbates these flaws. For these reasons, Petitioner contends that the PM standard for
existing LONC EGUs in the final rule is so stringent that it is not achievable by any of
Petitioner's EGUs.
Additionally, Petitioner contends that the EPA used a different methodology to set the PM
standard for the LONC subcategory than it used to set standards for other HAP for the
continental liquid oil-fired EGUs. According to Petitioner, for total metals, individual HAP
metals, HC1, and HF, the EPA calculated the continental liquid oil-fired standards based on the
average emissions reported by the best performing EGUs; for filterable PM, however, the EPA
used data from each of the six individual test runs at two of Petitioner's EGUs, rather than the
average. Petitioner alleges that if the EPA had applied the same methodology to filterable PM as
it did for the other HAP at continental liquid oil-fired EGUs, the PM standard for LONC EGUs
would be significantly higher. Petitioner states that the application of different approaches to
setting limits is arbitrary and that the EPA should recalculate the PM standard for the LONC
subcategory by using the average emissions from each test rather than using each test run.
Response to Issue 122: The issue of the methodology used to account for variability (i.e., the
UPL) was raised in public comments submitted in response to the proposed MATS rule. See 76
FR 24976, 25041-46 (May 3, 2011) (describing EPA's approach to establishing MACT floors
for public comment). The EPA's responses to the comments are in the RTC (see, e.g., Vol. 1, pp.
459 - 460 and 491 - 496). The comments the EPA received on this issue demonstrate that the
public had ample opportunity to comment on the issue, and the EPA responded to those
comments. Because Petitioners did not demonstrate that it was impracticable to comment on this
issue during the comment period on the proposed rule, the EPA is denying the Petition for
Reconsideration of this issue.
The EPA believes that its analysis procedures were consistent among both the coal and oil data
sets. Further, the EPA disagrees with Petitioner with regard to the "achievability" of the filterable
PM standard for existing non-continental liquid oil-fired EGUs as the final limit (0.030
lb/MMBtu) is 89 percent higher than the maximum run for the EGUs in the non-continental
subcategory floor data set.
Issue 123: Petitioner 20187 states that the EPA reached its PM floor by taking a 99 percent UPL,
based upon the emissions of its best-performing EGUs. Petitioner maintains that the resulting
floors are not "the average emission limitation achieved by the best performing 12 percent" of
coal-fired EGUs. Petitioner notes that the UPL, according to the EPA, is the emission level at or
below which it would "expect [a three-run stack test] to fall" (in this case, with 99 percent
confidence); in other words, it is a number representing the highest extreme in EPA's data.
Petitioner states that Merriam Webster defines "average" as "a single value (as a mean, mode, or
141
-------
median) that summarizes or represents the general significance of a set of unequal values,"
"equaling an arithmetic mean," or "being about midway between extremes."
Petitioner alleges that a UPL and an average thus pose different questions: one seeks the highest
extreme within in a data-set, the other seeks the middle. And they result in correspondingly
different answers. Petitioner maintains that by using a UPL, the EPA has set its floor at the high
extreme of its projected emissions from the best-performing plants, but that the statute requires
the EPA to set the floor at a level mid-way between extremes.
Response to Issue 123: The issue of the use of the UPL was raised in public comments
submitted in response to the proposed MATS rule. See 76 FR 24976, 25041-46 (May 3, 2011)
(describing EPA's approach to establishing MACT floors for public comment). The EPA's
responses to the comments are in the RTC (see, e.g., Vol. 1, pp. 459 - 460 and 491 - 496). The
comments the EPA received on this issue demonstrate that the public had ample opportunity to
comment on the issue, and the EPA responded to those comments. Because Petitioners did not
demonstrate that it was impracticable to comment on this issue during the comment period on the
proposed rule, the EPA is denying the Petition for Reconsideration of this issue.
Issue 124: Petitioner 20187 states that the EPA should establish an averaging time consistent
with its floor analysis. Petitioner asserts that the EPA calculated its floors based on emissions
limitations achieved over three operating-hour stack tests whereas the rule's final standards
require facilities to comply on a 30 boiler-operating-day rolling average. Petitioner maintains
that emissions over 30 days vary less than emissions over 3 hours and that a UPL calculated
using 30 operating-day averages will, as a result, be much lower than one calculated using 3-
hour averages. Petitioner states that the formula the EPA uses to calculate its UPL depends upon
a specific number of runs (n), as well as variability over those runs (s2). Petitioner indicates that
in using that formula, the EPA assumes that "compliance with [the] limitations will be based on
the same" three-run test average as used to derive the UPL. Petitioner alleges that that formula
and assumption do not support an emission limit reflecting 717 additional hours of emissions.
Petitioner alleges that as a practical matter - because it allows short-term emissions well above
the specified limit - an emission limit using a 30 operating-day average is much less stringent
than an otherwise equivalent limitation using a shorter averaging period. According to Petitioner,
the Proposed Rule justified that lower stringency by noting that the standard would apply during
startup and shutdown, when hourly emission rates may spike above normal, but in the final rule,
the EPA exempted startups and shutdowns, leaving no justification for use of a 30-day average.
Petitioner additionally notes that if the EPA had wanted to consider the most relevant data for
determining an appropriate PM limit based on a 30-day rolling average, the EPA could have
evaluated emissions data from PM CEMS that Petitioner alleges are currently installed at more
than 70 EGUs. Petitioner indicates that data from these CEMS should have been used to test the
EPA's assumption that emission limits based on hourly stack tests are a reasonable reflection of
emissions measured over much longer periods of time.
In short, Petitioner maintains that the Agency has no reasonable basis to conclude that the final
standards over 30 boiler-operating days reflect the "average emission limitation achieved by the
142
-------
best performing 12 percent of existing" coal-fired EGUs. Petitioner maintains that the EPA
should revise its PM standards to reflect the emissions average upon which it based its
calculation of the MACT floor.
Response to Issue 124: The issue of the use of short-term emission test data to establish 30-day
rolling average emission limits was raised in public comments submitted in response to the
proposed MATS rule. See 76 FR 24976, 25041-46 (May 3, 2011) (describing EPA's approach to
establishing MACT floors for public comment). The EPA's responses to the comments are in the
RTC (see, e.g., Vol. 1, pp. 462 - 463, 494 - 496, 500, and 549 - 551). The comments the EPA
received on this issue demonstrate that the public had ample opportunity to comment on the
issue, and the EPA responded to those comments. Because Petitioner did not demonstrate that it
was impracticable to comment on this issue during the comment period on the proposed rule, the
EPA is denying the petition for reconsideration of this issue.
The EPA also disagrees with Petitioner's assertion that CEMS data could have provided
additional insight. In any case, no PM CEMS data were provided to the EPA through the 2010
ICR. Further, Petitioner identified, but did not provide data from, a number of EGUs with PM
CEMS installed. Absent data from Petitioner, the EPA reviewed the provisions of three publicly
available consent decrees that required the installation of PM CEMS on coal-fired EGUs.95'96'97
Contrary to Petitioner's assertion that data from these CEMS would be useful, the EPA generally
found that the method for compliance with the PM emission limits was an annual Reference
Method 5 (a short-term manual stack test method) test.
The EPA also notes that though the consent decrees require the companies to "install, correlate,
maintain, and operate PM CEMS," the consent decrees also state that each company "maintains
that its agreement in this Consent Decree to install, correlate, maintain, and operate PM CEMS
shall not prevent...in any future proceedings from challenging the relationship between the data
generated from such PM CEMS, including the averaging period for which such data is
reported...and the results of performance tests for PM..." or "may attempt to demonstrate that it is
infeasible to continue operating PM CEMS..." Therefore, the EPA does not consider the PM
CEMS data unchallengeable. Further, the EPA has no information to confirm that these PM
CEMS were calibrated according to specified procedures, and thus the data obtained would have
questionable value for establishing emission limits. Hence, in addition to the reason stated above,
95 United States District Court, Eastern District of Kentucky, Central Division, Lexington. United
States of America, Plaintiff, v. East Kentucky Power Cooperative, Inc., Defendant. Civil Action
No. 04-34-KSF. Consent Decree. EPA-HQ-OAR-2009-0234-20347.
96 United States District Court for the District of Arizona. United States of America, Plaintiff, v.
Salt River Project Agricultural Improvement and Power District, Defendant. Civil Action No.
2:08-cv-1479-JAT. Consent Decree. EPA-HQ-OAR-2009-0234-20348.
97 United States Environmental Protection Agency. In the matter of: Tennessee Valley Authority,
400 West Summit Hill Drive, Knoxville, Tennessee 37902. Allen, Bull Run, Colbert,
Cumberland, Gallatin, Johnsonville, John Sevier, Kingston, Paradise, Shawnee, and Widows
Creek Fossil Plants. Federal Facilities Compliance Agreement Between the United States
Environmental Protection Agency and the Tennessee Valley Authority. Docket No. CAA-04-
2010-1760. EPA-HQ-OAR-2009-0234-20349.
143
-------
the EPA is denying the petition for reconsideration of this issue because it is not of central
relevance.
144
-------
25.0 Subcategorization - Coal Refuse-fired Circulating Fluidized Bed (CFB) EGUs
Issue 125: Petitioner 20175 maintains that existing coal refuse-fired CFB boilers cannot satisfy
the final emissions standard for HC1 for coal-fired EGUs through the current CFB design and
control configuration. Petitioner alleges that the EPA has not identified, nor is Petitioner aware
of,98 any situation where an existing coal refuse-fired CFB has installed additional back-end
scrubbing technology in a retrofit setting to further reduce SO2 or HC1 emissions or otherwise
achieved through any control measure the HC1 (or corresponding SO2 surrogate) emission
standard included in the final rule (i.e., 0.002 lb/MMBtu). Petitioner maintains that its member
plants would generally be required to reduce HC1 emissions between 73 and 99 percent beyond
the level of control already achieved through the effective use of limestone injection, to meet the
HC1 emission rate of the final rule.
Petitioner also states that increased limestone injection, consistent with current design and
operational constraints, cannot further reduce HC1 emissions from these existing EGUs to levels
that meet the final limit. Petitioner's members also extensively evaluated alternative means to
satisfy the stringent emission standard for HC1 imposed through the MATS rule and Petitioner
indicates that such means necessarily would require either the installation of additional emission
control equipment, or the use of alternative sorbent additives to achieve the substantial degree of
further emission reduction mandated by the regulation. Petitioner evaluated each of these
alternatives and concluded that neither option is both technically and economically available for
application to Petitioner's member units.
Petitioner states that to the extent that installation and operation of add-on emission control
equipment could be technically feasible for retrofit application to coal refuse-fired CFB EGUs,
the capital costs associated with such equipment would render the facilities economically non-
viable in a competitive electricity generation market, largely due to the relatively small size of
these facilities (generally due to the "size restrictions pursuant to the Public Utility Regulatory
Policy Act (PURPA)"), the inconsistency of the add-on equipment platform with the combustion
zone-based emission control inherent in the distinct design characteristics of the EGUs, and the
long-term fixed-price power purchase agreements that are common to the sources. Petitioner
cites an analysis it undertook indicating that capital costs for installation of an add-on flyash
conditioning and reinjection system (similar to that used at the Gen-On Seward facility) would
exceed $54 million. Petitioner states that this analysis indicated the capital cost of necessary
retrofits, which would include the replacement of existing bag houses, would be cost prohibitive
for application to Petitioner's members' existing plants and that a vendor that supplies both
scrubber and fly ash reinjection technologies estimated that the costs required for retrofit
installation at a typical member plant of a Seward-type fly ash reinjection system would be even
higher than for dry scrubbing technology. Petitioner further alleged that it is uncertain whether
the retrofit installation of such systems at the Petitioner's member plants, which are older and
otherwise incorporate distinct design characteristics than the Gen-On Seward facility, would be
effective in reducing HCI emissions to the levels required under the final MATS rule.
98 Petitioner notes that because it is not aware of a single retrofit installation of back-end
scrubbing technology for a coal refuse-fired CFB EGU, the technology cannot be said to be
"available" for application to this source category.
145
-------
Petitioner maintains that introduction of sodium-based sorbents, at a level effective at
sufficiently reducing HC1 emission rates, would render the ash generated by coal refuse-fired
CFB EGUs unsuitable for beneficial reuse in the context of mine reclamation. According to
Petitioner, the resulting disposal costs, especially considering the significant additional quantity
of ash that these EGUs would generate, would make continued operation of these sources
economically non-viable and eliminate the significant environmental benefits provided by coal
refuse-fired CFBs. Petitioner alleges that in order for its member plants to achieve the HC1
emission reductions required under MATS through the use of DSI, the plants would be required
to add significant quantities of either trona or sodium bicarbonate, both of which are sodium-
based sorbents. According to Petitioner, the addition of such sorbents would alter the leaching
characteristics of the ash, such that the Petitioner's member plants would no longer be able to
beneficially reuse the ash to reclaim abandoned mines in Pennsylvania (because of degradation
of the chemical characteristics of the CFB-derived ash and increased leaching of sulfate and
chloride from the ash). Instead, Petitioner maintains that the plants would be forced to dispose of
the ash in lined landfills. Petitioner states that the costs associated with disposing of ash in lined
landfills range between 10 and 50 times higher than those associated with beneficially reusing
ash in the reclamation of mine sites. Accordingly, Petitioner states that the increased ash disposal
costs resulting from the use of DSI to control HC1 emissions would render economically
infeasible the use of such control technology by coal refuse-fired CFB units.
Petitioner states that, based on these considerations, the Agency should revise the final rule to
eliminate or revise application of the current HC1 emissions limitation for this source type.
According to Petitioner, the EPA could accomplish this objective by eliminating any specific
numeric HC1 emissions limit for coal refuse-fired CFB EGUs from the final regulation or by
adjusting the limit to a level appropriate (including under relevant provisions of the CAA) for
these sources. Specifically, Petitioner requested that the EPA apply the methodology, consistent
with the statutory directive, for calculating an HC1 emission rate for coal refuse-fired CFB EGUs
for which the ash is beneficially used (i.e., establish a subcategory for such sources).
In addition, Petitioner notes that the definition of "coal refuse" included within the final rule
includes criteria for ash content and heating value. Petitioner alleges that these specifications are
inconsistent with the actual characteristics of coal refuse, contrary to the standards identified by
other federal agencies, and conflict with the EPA's own regulatory definitions. Petitioner notes
that the EPA should revise the definition of "coal refuse" within the final rule to be consistent
with pre-existing definitions of such terms previously promulgated by the EPA through NSPS.
Petitioner maintains that revising the definition of coal refuse in this manner would also serve to
resolve the ambiguity currently existing within the MATS related to the applicability of the
subcategories for coal-fired EGUs to units that combust coal refuse.
Response to Issue 125: EPA provided ample opportunity for public comment on issues related
to the subcategorization of EGUs covered by the MATS rule. In fact, the issue of establishing a
separate subcategory for coal refuse-fired CFB EGUs was raised in public comments submitted
in response to the proposed MATS rule. The EPA's responses to the comments are in the RTC
(see, e.g., Vol. 1, pp. 358 - 365). The issue of a separate HC1 limit for coal refuse-fired EGUs
was also raised in public comments submitted in response to the proposed rule. The EPA's
146
-------
responses to the comments are in the same RTC (see, e.g., Vol. 1, pp. 584 - 587). The issue
relating to the inability of coal refuse-fired EGUs to add on additional controls was also raised in
public comments submitted in response to the proposed MATS rule. The EPA's responses to the
comments are in the same RTC (see, e.g., Vol. 1, p. 761). The issue of the impact of sodium-
based sorbents on the ability of coal refuse-fired EGUs to reuse their ash was also raised in
public comments submitted in response to the proposed MATS rule. The EPA's responses to the
comments are in the RTC (see, e.g., Vol. 2, pp. 509 - 511). Because the Petitioner has not
demonstrated it was unable to comment on these issues, we are denying reconsideration. In
addition, the decision not to subcategorize CFBs and coal refuse-fired CFBs was challenged in
the White Stallion case, and the court rejected those challenges. 748 F.3d at 149-50. As the
Court may only consider issues raised during the period for public comment, issues raised in the
litigation and addressed by the Court clearly do not meet the criteria for reconsideration in CAA
section 307(d)(7)(B). The court has affirmed EPA's decision not to create the requested
subcategory and the Petitioner has not presented new, information or data relevant to that
determination. For this reason, the petition does not provide substantial support for an argument
that this aspect of the rule should be revised and it is therefore also being denied because it is not
of central relevance.
As noted in the RTC (Vol. 1, p. 761), the EPA is aware of at least two coal refuse-fired CFB
EGUs that reported add-on "polishing" technologies for SO2 (or, at least, technologies beyond
the CFB itself) in their responses to the 2010 ICR. One of these EGUs had the additional
equipment installed during initial construction; however, the other EGU retrofitted the
equipment. Petitioner has not provided information that would lead the EPA to believe that this
information is incorrect. Further, Petitioner's allegation that there is no "situation where an
existing coal refuse-fired CFB has...otherwise achieved through any control measure the HC1 (or
corresponding SO2 surrogate) emission standard included in the final rule" is rebutted by
Petitioner's own information. In meetings held subsequent to the filing of the Petition for
Reconsideration, Petitioner provided the EPA with a list of "twenty CFB (circulating fluidized
bed) combustion units that utilize coal refuse as a primary fuel source (anthracite culm and/or
bituminous "gob"), inject limestone in the combustion zone for the control of SO2 gases, and
beneficially utilize CFB ash."99'100 Of the 19 CFB EGUs, 12 are highlighted as being unable to
"currently meet the UMACT-MATS rule." However, by difference, 7 of the 19 (37 percent or
well above the CAA's stipulated 12 percent criteria) CFB EGUs must be presumed to currently
be meeting the MATS HC1 or SO2 limit(s). In addition, 14 of the 19 CFB EGUs are located in
Pennsylvania; of these, 9 are highlighted as being unable to currently meet the MATS limits -
leaving, by difference, 5 CFB EGUs that are presumed to be meeting the limits. Of these five
CFB EGUs presumed to be able to achieve the limits, one is in western Pennsylvania and four
are in eastern Pennsylvania; of the nine CFB EGUs listed as being unable to achieve the limits,
five are in western Pennsylvania and four are in eastern Pennsylvania.101 Thus, Petitioner's claim
that no CFB EGU has "achieved through any control measure the HC1 (or corresponding SO2
99 See EPA-HQ-OAR-2009-0234-20338; EPA-HQ-OAR-2009-0234-20373.
100 Note that 1 of the 20 listed CFB units is an Industrial Boiler not subject to the MATS rule.
101 Compare list in EPA-HQ-OAR-2009-0234-20338 with "Location of ARIPPA members in
Pennsylvania (from http://arippa.org/index.php?id=51)" in docket EPA-HQ-OAR-2009-0234.
147
-------
surrogate) emission standard included in the final rule" is not supported by the information
available to the EPA.
The EPA does not believe (as Petitioner implies) that the cost of such add on (or additional)
"polishing" equipment is prohibitive for coal refuse-fired EGUs or that the add-on technologies
are not "available." Petitioner has provided no "hard" cost information supporting its allegation
beyond a generic spreadsheet analysis that used estimated capital and operating costs, unit sizes,
and sorbent use (i.e., no cost analysis for a specific coal refuse-fired CFB EGU has been
provided). The EPA has found that at least one control device vendor markets such add-on
technology (including installations on a coal refuse-fired CFB EGU, on CFB EGUs using other
coal ranks, and on retrofit applications) that will lower HC1 (and HF) emissions to the 1 part per
million level. 102>103 One unique feature of the technology when employed on CFB EGUs is the
"ability to use the ash (which contains surplus lime) supplied by the CFB (via the flue gas
stream)" for enhanced SO2 control.104'105 The EPA believes that this "left over" lime in the ash
will lessen or eliminate the need to add lime, thus lessening the overall cost of the technology.
Although not specifically installed for HC1 control, the EPA believes that the humidification step
inherent in the process will further remove HC1. Therefore, the EPA believes that such add-on
technology is "available" and is a viable option for use on coal refuse-fired CFB EGUs. Limited
cost information for this technology installed on two coal-fired EGUs indicate that the total
capital cost ranges from approximately $56 (for a 630 MW EGU) to $72/kilowatt (kW) (for two
660 MW EGUs),106 considerably less than either the capital cost of wet FGD installation
($407/kW)107 or the $675/kW (for a hypothetical 80 MW EGU) presented by Petitioner.108
Therefore, the EPA does not believe, as Petitioner states, that the HC1 limit in the final rule is
unachievable for such EGUs and the "excessive" cost has not been demonstrated.109
102 Gatton, Lawrence, Principal Engineer, Alstom Power ECS. Next Generation NID for PC
Market. Presented at Coal-Gen Conference and Exposition, Columbus, Ohio. August 17 - 19,
2011. Found in docket EPA-HQ-OAR-2009-0234.
103 See EPA-HQ-OAR-2009-0234-20344.
104 See EPA-HQ-OAR-2009-0234-20346.
105 See EPA-HQ-OAR-2009-0234-20343.
106 See "Kiewit Chooses Alstom Clean Air Technology to Meet Stringent New Missions
Standards: Brayton Point 3 to see reduced SO2 and mercury emissions" and "Alstom to supply
NID™ emission control system for the Homer City Generating Station." Power Engineering.
10/01/2010. Found in docket EPA-HQ-OAR-2009-0234.
107 See "Acid Gas Emission Control Measures" in docket EPA-HQ-OAR-2009-0234.
108 The EPA would note that the basis years for the cost quotes are different which may introduce
a limited amount of error in the calculations. However, given the general lack of specificity in
any of the quotes, the EPA believes they are reasonable to use for comparison purposes.
109 Petitioner notes that it also evaluated "potential installation of...fly ash reinjection systems"
comparable to those installed on the Gen-On Seward facility (presumed by the EPA to be the
Alstom NID system noted elsewhere). However, Petitioner provided no definitive rationale
eliminating this technology other than an unsubstantiated vendor estimate that "the costs required
for retrofit installation at a typical ARJPPA plant of a Seward-type fly ash reinjection system
would be even higher than for dry scrubbing technology."
148
-------
Petitioner acknowledges that such coal refuse-fired EGUs are eligible for the alternate SO2
emission limit in lieu of the HC1 limit but maintains that the CFB EGUs are unable to achieve
this limit. The EPA has reviewed a number of Title V permits for coal refuse-fired EGUs and,
based on those permits, believes that the alternate SO2 emission limit is achievable by such
EGUs if they believe that they cannot achieve the HC1 emission limit.110
The EPA would note that the discussion cited by Petitioner relating to use of sodium-based
sorbents, etc. is merely an example provided by the EPA to acknowledge that there are alternate
means of achieving the emission reductions (the Alstom NID technology noted elsewhere does
not use sodium-based sorbents but achieves low levels of HC1 and SO2).
Additionally, EPA provided ample opportunity for comment on the definition of "coal refuse."
The definition was modified based on public comment (see, e.g., RTC, Vol. 2, pp. 709 - 710) and
the definition in the final rule is a logical outgrowth of the proposed rule. Further, the EPA
disagrees with Petitioner's allegation that the final definition is inconsistent with other, utility
NSPS-related definitions of "coal refuse," as the final definition is consistent with that found in
40 CFR Part 60, subpart Da. Petitioner has not demonstrated that it was impracticable to
comment on this issue during the period for public comment. Also, because the Petitioner does
not demonstrate that there is definitional inconsistency, the EPA maintains this issue is not of
central relevance.
For all of these reasons, we are denying reconsideration of this petition.
110 See "Pennsylvania Title V permits for coal refuse-fired EGUs" in docket EPA-HQ-OAR-
2009-0234 (permits cited in memo are also in docket EPA-HQ-OAR-2009-0234).
149
-------
26.0 Subcategorization - IGCC
26.1 IGCC-specific definitions
Issue 126: Petitioner 20183 claims that the ambiguity of the finalized startup and shutdown
definitions creates additional uncertainty with respect to IGCC EGUs. Accordingly, Petitioner
requests reconsideration of the final rule to provide definitions for startup and shutdown and
applicable work practice standards that are tailored specifically for IGCC EGUs.
Response to Issue 126: The EPA reconsidered definitions of startup and shutdown in the MATS
reconsideration rulemaking, and provided an additional opportunity for public comment through
that process. See 79 FR 68777. For that reason, this issue is moot.
26.2 Revision of IGCC to specifically exclude the gasifier
Issue 127: Petitioner 20183 specifically requests that the EPA confirm that it does not intend for
the gasification process in IGCC EGUs (i.e., emissions from the gasification process from, for
example, the gasifier startup stack or the gasifier flares, or other emission points from the
gasification system) to be subject to final rule. Petitioner further requests that the EPA revise its
definition of IGCC to specify that the definition excludes any gasification processes or systems.
Response to Issue 127: The definition of the covered source (i.e., "electric utility steam
generating unit") in CAA section 112(a)(8) is for a "fossil fuel fired combustion unit," and based
on this definition the gasifier is excluded from coverage by the rule. The EPA considers the
definition to be sufficiently clear. The EPA is also denying reconsideration on this issue because
the Petitioner has not demonstrated that it was impracticable to comment on this issue during the
comment period on the proposed rule.
26.3 IGCC-specific work practice standards
Issue 128: Petitioner 20183 suggests that the EPA revise the work practice standards in the final
rule to reflect IGCC-specific standards because the rule's work practice standards are not
consistent with operation of an IGCC. Petitioner points out as an example that synthetic gas (the
cleaned gasifier product consisting of a mixture of hydrogen and carbon monoxide) is a clean
fuel but is not identified as such in the work practice standards. Petitioner also states that if the
EPA intended the work practice standards for coal- and oil-fired EGUs to apply to IGCCs, those
work practice standards are not consistent with operation of an IGCC and suggests that the EPA
add language specific to IGCC EGUs (e.g., "Maintain burners associated with the combustion
turbine and duct burners in the heat recovery steam generator in accordance with vendor
recommendations, such as those contained in applicable long-term service agreements").
Response to Issue 128: The EPA reconsidered definitions of startup, shutdown, and clean fuels
and the work practice requirements for periods of startup and shutdown in the MATS
reconsideration rulemaking and provided an additional opportunity for public comment through
that process. 79 FR 68777. For that reason, these issues are moot.
150
-------
The EPA also understands that IGCC EGUs often use syngas as both a clean startup fuel and as
the production fuel; that syngas is generated in the gasifier of an IGCC EGU; and that after the
syngas achieves the desired site-specific composition, it is fired in the combustion turbine as a
clean startup fuel and as the primary production fuel. The EPA generally considers the syngas
fired in the combustion turbine a clean fuel as long as it is cleaned to remove trace contaminants
such as sulfur before its injection in the turbine. The EPA also understands that during startup
and shutdown periods, some or all of the syngas produced may not be combusted in the turbine.
In the November 2012 reconsideration proposal, we proposed two options for IGCC EGUs for
handling syngas not fired in the combustion turbine: that either (1) syngas must be flared, not
vented, or (2) syngas must be routed to duct burners, which may need to be installed, and the flue
gas from the duct burners must be routed to the heat recovery steam generator. These options
were finalized in the November 19, 2014, final startup/shutdown reconsideration notice. Id.
Further, the EPA revised the work practice standards in 40 CFR 63.10021(e)(6) in the final
MATS and believes that this revision adequately addresses Petitioner's concern related to non-
startup work practice standards and IGCC EGUs.
151
-------
27.0 Variability
Issue 129: Petitioner 20182 claims that the EPA has set new-source HAP standards that are
based on a flawed analysis of the variability in the emissions data. Petitioner further claims that
the EPA did not sufficiently address and support the variability approach in the final rule.
Petitioner points out that any normally operating coal-fired EGU has variations in the pollutant
levels emitted and maintains that the EPA did not correctly address the role of normal variability
in coal-fueled power plant operations when setting the standards, because it based calculations
on short-term test runs rather than longer tests showing variability over time. Petitioner asserts
that short-term stack tests do not represent emissions over time and that therefore the EPA has
erred in using only this approach to gather data to set the emission standards.
Petitioner acknowledges the EPA's use of the UPL to account for variability but claims that "the
stack test data to which EPA applied the UPL lacked the completeness required to ensure the
UPL formula would work correctly." Petitioner then suggests that the EPA "should have
adjusted emissions achieved by the best controlled similar source to reflect the variability
occurring over the full range of operating conditions, not just during the stack test." Petitioner
points out the availability of additional stack tests that they feel should have been used in setting
new source standards.
Response to Issue 129: The EPA disagrees with Petitioner regarding this issue. The EPA has
explained its use of the UPL methodology in the rulemaking preamble(s) and in associated TSDs
(see, e.g., RTC, Vol. 1, pp. 457 - 458, 460 - 463, 491 - 496, 496 - 497, 501 - 502, 504, and 554 -
557). The EPA continues to believe that the use of the UPL methodology adequately accounts
for variability. The EPA is denying the petition for reconsideration on this issue because the
Petition does demonstrate that it was impractical to comment on this issue during the public
comment period. Further, to the extent that Petitioner is focusing on new-source issues, those
issues are moot as the EPA reconsidered the new-source limits (see 78 FR 24073).
152
-------
28.0 Work practice Standards - Organic HAP
Issue 130: Petitioner 20178 claims that the work practice standards for dioxin/furans and other
organic HAP cannot be met by EGU owners or operators who employ CFB technology because
there are significant differences in design and operation between pulverized coal and CFB EGUs
(e.g., CFB EGUs do not use coal burners that can be tuned; CFB EGUs do not have flame whose
pattern can be adjusted). Petitioner seeks reconsideration of these requirements and guidance on
tune-ups that will satisfy the final rule for EGUs with CFBs.
Response to Issue 130: The EPA provided ample opportunity to comment on the work practice
standards for dioxin/furans and other organic HAP. In fact, the concern that a specific EGU
technology or process could have difficulty with the tune-up requirements was raised in public
comments submitted in response to the proposed MATS rule. The EPA's responses to the
comments are in the RTC (see, e.g., Vol. 2, pp. 464 - 470). The comments the EPA received on
these issues demonstrate that the public had ample opportunity to comment on these issues, and
the EPA responded to those comments. Moreover, as Petitioner admits, the rule qualifies the
portions of the tune-up requirements that Petitioner considers problematic by stating those
portions only apply "(a)s applicable." In other words, if the portions of the tune-up requirements
are inapplicable, then the EGU owner or operator does not need to comply with them.
Concerning tune-up guidance, Petitioner has not demonstrated that it was impracticable to raise
this issue during the comment period. Because Petitioner did not demonstrate that it was
impracticable to comment on these issues during the comment period on the proposed rule, the
EPA is denying reconsideration of this issue.
Issue 131: Petitioner 20187 claims that the EPA proposed standards in the MATS rule based on
a method detection level (MDL) approach, that it used but did not propose a representative
method detection level (RDL) approach for standards other than the dioxins and other organic
HAP standards in the final MATS rule, and that the proposal did not acknowledge what
Petitioner claims to be flaws with the MDL approach for the dioxins and other organic HAP
standards in the MATS rule. Petitioner requests reconsideration on this issue, as Petitioner claims
it was unable to comment on the use of an RDL approach for the dioxins and other organic HAP
standards. Petitioner suggests that the EPA must redo its analysis using an RDL approach or
provide a rational explanation for using the MDL approach for the dioxins and other organic
HAP standards.
Response to Issue 131: The EPA disagrees with Petitioner. Non-dioxin/furan organic and
dioxin/furan organic HAP measurements obtained as part of the ICR data collection effort differ
from other categories of HAP measurements with respect to amounts collected above and below
the respective MDLs. As mentioned in RTC, Vol. 1, p. 476 and as explained in the preamble to
the proposed rule, a significant majority of the measured emissions from EGUs of non-
dioxin/furan organic and dioxin/furan organic HAP were below detection levels of the EPA test
methods (76 FR 25040), so work practice standards were proposed and finalized. In contrast, the
measured emissions from other pollutants had a significant majority of measured emissions
above the respective MDLs, so numerical emissions limits were found to be appropriate. The
change from the use of 3xMDL to 3xRDL is thus not relevant to the regulation of organic HAP
153
-------
emissions. For this reason, the objection is not of central relevance to the rule, and we are
denying the petition for reconsideration of this issue.
Concerns about the quality of measurements at very low emission limits (especially for new
sources) were raised regarding standards other than non-dioxin/furan organic and dioxin/furan
organic HAP in public comments submitted in response to the proposed MATS rule. The
changes to the rule are based on public comments (see, e.g., RTC, Vol. 1, pp. 547 - 548) and thus
are a logical outgrowth of the proposed rule. Additionally, concerns about using RDLs as
opposed to MDLs are not relevant for non-dioxin/furan organic and dioxin/furan organic HAP
work practice standards because no emissions measurements (which could result in values at or
below detection levels) are made for work practice standards. Because the Petitioners have not
demonstrated that it was impractical to comment on these issues in the proposed rule, the EPA is
denying reconsideration. In addition, because these issues are not of central relevance to the final
MATS rule, the EPA is denying reconsideration of these issues.
154
-------
Appendix A. List of NESHAP Petitioners
Petitioners
Docket number
(EPA-HQ-OAR-2009-0234-)
American Public Power Association (APPA)
20189
ARIPPA
20175
Babcock & Wilcox (B&W)
20193
Basin Electric Power Cooperative
20192
Climate Policy Group (CPG)
20185
Coal Utilization Research Council (CURC)
20186
Earthjustice
20187
East Kentucky Power Cooperative (EKPC)
20178
Edgecombe/Spruance Genco
20194
Edison Mission Energy
20177
FirstEnergy
20173
Hawaiian Electric Company (HECO)
20191
Institute of Clean Air Companies (ICAC)
20176
International Brotherhood of Boilermakers (IBB)
20182
Power4Georgians
20181
Puerto Rico Electric Power Authority (PREPA)
20188
Southern Company
20183
State of Texas (Texas Commission on Environmental
Quality, Texas Public Utility Commission, Railroad
Commission of Texas)
20174
Utility Air Regulatory Group (UARG)
20180
Wolverine Power Supply Cooperative
20184
155
-------
NSPS
29.0 Limiting Opacity Standard Exemption for Subpart D EGUs
Issue 1: Petitioner 5777 requests that applicability of the exemption from the opacity standards
under the EGU NSPS be expanded to apply to all affected EGUs using a particulate matter (PM)
continuous emissions monitoring system (CEMS) to demonstrate compliance with a PM
emissions limit in the rule (i.e., EGUs subject to Subpart D and EGUs subject to Subpart Da).
Petitioner notes that the amendment to the EGU NSPS applies only to those affected EGUs for
which the owner or operator elects to comply with the PM emissions limit of 0.03 pounds per
million British thermal units (lb/MMBtu) under Subpart Da. Petitioner states that the EPA has
not clearly explained the rationale for its decision to restrict the applicability of the opacity
standards exemption and associated periodic visible emissions testing requirements and has not
fully addressed previous public comments submitted on the Agency's past proposal notices
regarding such an exemption limitation. Petitioner states that there is no basis to support the
EPA's rationale that by complying with the EGU NESHAP requirements, all affected EGUs
subject to Subpart D will have to comply with a PM emissions standard such that they would
qualify for opacity standards exemption under the EGU NSPS.
Response to Issue 1: The EPA did not amend the previous requirements in a fundamental way
and is, therefore, denying the Petition for Reconsideration on this issue because it is outside the
scope of the final rule and it therefore does not meet the criteria for reconsideration under CAA
section 307(d)(7)(B). Further, EPA has previously denied a petition for reconsideration raising
this issue submitted in the context of another rulemaking and commenter has presented no new
information that would cause the EPA to revisit that decision. As stated in previous rulemakings,
the contribution of filterable PM to opacity at emission levels of 0.03 lb/MMBtu or less is
generally negligible (see 76 FR 25071). We concluded, and continue to believe, that an opacity
standard is no longer necessary for sources that can demonstrate continuous compliance with a
PM standard of 0.03 lb/MMBtu by continually monitoring the mass of PM emissions because
this would ensure that the source is also in compliance with the otherwise applicable opacity
standard. The EPA further stated that this did not reflect an agency view regarding the efficacy
of opacity limits generally. On the contrary, at higher PM emissions rates, the contribution of
filterable PM to opacity is not negligible and we, therefore, concluded that facilities subject to a
PM emission standard greater than 0.03 lb/MMBtu should continue to be subject to the opacity
standard.
156
-------
30.0 Subpart Da Condensable PM Testing for Post-2011 EGUs
Issue 2: Petitioner 5777 requests that the condensable PM testing requirement applicable to
certain new EGUs constructed after May 3, 2011, to be removed from Subpart Da on the basis
that this requirement is flawed, burdensome, and unreasonable and was not included in the
proposed rule to allow for public comment. Petitioner notes that the requirement specifies that
condensable PM be measured using Method 202 and reported to EPA in conjunction with each
performance test conducted to determine compliance with the applicable PM emissions limit.
Reasons presented by Petitioner for removal of the condensable PM testing requirement from the
rule include: (1) the requirement does not apply to all new EGUs constructed after May 3, 2011,
but only to the subset of those EGUs that elect to conduct Method 5 performance tests to
determine compliance with the PM emissions limit; (2) for those affected EGUs using wet flue
gas desulfurization (FGD) and required to conduct the condensable PM testing, the testing
presents significant burdens, because these EGUs must use EPA Method 5B to demonstrate
compliance with the PM emissions limit although Method 202 prohibits its use in conjunction
with Method 5B; and (3) including testing requirements in the EGU NSPS that are unrelated to
any standard in the rule is unreasonable, and the appropriate use of Clean Air Act (CAA) section
114 authority to collect additional condensable PM data is instead using a properly structured
and supported ICR, of limited duration and appropriate scope.
Response to Issue 2: The EPA is denying the Petition for Reconsideration on this issue because
it does not meet the criteria for reconsideration under CAA section 307(d)(7)(B). The EPA did
provide notice of its intent to require such testing and the commenter commented on the issue.
The fact that the commenter disagrees with EPA's response to its comments does not present a
legal basis compelling reconsideration of the issue. The EPA proposed a total, i.e., filterable plus
condensible, PM standard for new EGU's constructed after May 3, 2011. The proposed rule
specified: "Total particulate matter concentration consists of the sum of the filterable and
condensable fractions. The condensable fraction shall be measured using Method 202 of
appendix M of part 51 ..." Proposed 40 CFR 60.50Da(b)(4). (76 FR 25099, column 1.) Thus,
Petitioner both had notice that EPA intended to require that condensable PM be measured using
Method 202 and an opportunity to comment on the appropriateness of the requirement. Petitioner
acknowledged that the EPA intended to require that the fraction of condensible PM be
determined using Method 202. Comments of Utility Air Regulatory Group, at 12. Section
60.50Da(b)(l)(ii) requires the use of either Method 5 (for affected facilities that do not use a wet
FGD) or Method 5 A (for affected facilities that do use a wet FGD) to determine compliance with
the applicable PM limit without exception. Because 60.50Da(b)(2) requires affected facilities for
which construction, reconstruction or modification is commenced after May 3, 2011, to measure
condensable PM using Method 202 in conjunction with a performance test performed according
to the requirement of 60.50Da(b)(l), all affected facilities constructed after May 3, 2011, are
required to test for condensable PM using Method 202. The assertion that the requirement
presents significant burdens for EGUs using wet FGD is also unfounded. Under such
circumstances the Method 5B performance test can be used in conjunction with Method 202 with
only a minor modification to the analysis procedure. The tester would perform the test by
weighing the filters twice; once before the filter is heated in a 320°F oven for six hours and once
after the filter is heated. We have concluded that this additional modification would increase the
cost of a performance test by about $150 to $200. The EPA also disagrees with Petitioner's
157
-------
characterization of the required testing for condensable PM as being unrelated to any standard in
subpart Da. The data collected through that testing will allow EPA to assess the appropriateness
of the stringency of the applicable filterable PM standard on an ongoing basis. Further, although
Petitioner may believe that it is most appropriate to collect additional condensable PM data using
an ICR, the EPA's use of its authority under CAA section 114 is not constrained in this manner.
In addition, the condensable PM testing requirement was specifically included in the ICR for
MATS, which was approved by OMB.
158
-------
31.0 Subpart Da Electronic Reporting Requirements for Facilities with PM CEMS
Issue 3: Petitioner 5777 requests that the requirement for electronic reporting of data by affected
EGU owners and operators that use a PM CEMS to demonstrate compliance with the PM
emissions limits be removed from 40 CFR Part 60, subpart Da. Petitioner asserts that as
promulgated, the requirement is inexplicable due to the fact that no such data exist for these
EGUs. Petitioner states that affected EGUs that use PM CEMS to demonstrate compliance with
the applicable PM emissions limit do not conduct periodic PM performance tests or perform
relative accuracy test audits (RATAs) comparing the PM CEMS performance to results from an
independent reference method. According to Petitioner, these EGUs use data measured by the
PM CEMS and conduct quality assurance tests under EPA Procedure 2 (which does not include a
RATA). Petitioner states that if the EPA intended to promulgate a general electronic reporting
requirement applicable to other EGUs (e.g., EGUs that conduct periodic performance tests or
conduct RATAs for SO2 or NOx CEMS), then the EPA must propose and solicit comment on
such a requirement and that if EPA intended to require affected EGU owners and operators to
report the data collected from the PM CEMS and/or reference method conducted during
Performance Specification (PS) 11 correlation testing, then the EPA must propose and solicit
comment on such a requirement. In addition, Petitioner also states that the EPA would need to
explain how those PS 11 correlation data are useful to the EPA's development of emission
factors.
Response to Issue 3: The Petitioner has not demonstrated that the issue presented meets the
criteria for reconsideration set forth in CAA §307(d)(7)(B). The Petitioner has neither
demonstrated that it was impracticable to raise the objection during the comment period, nor that
the grounds for the objection arose after the close of the comment period. The EPA, therefore,
denies reconsideration of the issue. Further, the Petitioner's assertion that EGUs that use PM
CEMS to demonstrate compliance with the applicable PM emissions limit do not conduct
periodic PM performance tests is incorrect. EGU owners or operators subject to 40 CFR Part 60,
subpart Da, who use PM CEMS are required to conduct a number of performance tests,
including periodic emissions tests. They must also conduct periodic absolute correlation audits,
response correlation audits, relative response audits, and sample volume audits - in accordance
with the requirements of 40 CFR 60.49Da(v). These tests, which as Petitioner acknowledges in
its Petition for Reconsideration for the NESHAP "are similar to RATAs...," along with any
other RATAs, must be reported electronically in accordance with the requirements of 40 CFR
60.49Da(v)(4). As Petitioner's claim is factually incorrect and Petitioner has failed to
demonstrate that the issue presented meets the criteria for reconsideration set forth in CAA
§307(d)(7)(B), the EPA is denying reconsideration of this issue. The EPA notes that it agrees
with Petitioner that the requirements as written only apply to affected facilities using PM CEMS
and has concluded that this issue is best addressed by applicability determinations, in which the
case-by-case specifics may be analyzed and addressed.
159
-------
32.0 Periods of Startup, Shutdown, and Malfunction for EGUs
Issue 4: Petitioners 5775 and 5777 request reconsideration of the startup, shutdown, and
malfunction (SSM) requirements added to the EGU NSPS and assert that the EPA did not
address alleged "fundamental" questions previously raised by public commenters on the
proposed rule; did not explain why compliance with the NSPS emissions limits is necessary
during SSM given the long standing exemption in the NSPS rules; and in the absence of any
legal mandate by the courts. Petitioners state that the decision of the U.S. Court of Appeals for
the D.C. Circuit in Sierra Club v. EPA, 551 F.3d 1019 (D.C. Cir. 2008) regarding compliance
during periods of SSM only pertained to NESHAP regulations and did not address NSPS rules.
Reasons presented by Petitioner 5775 for reconsideration of the SSM requirements include: (1)
the definition of "startup" does not allow time for all startup systems to be set into operation; (2)
the startup work practice standards require emission control devices to be used in an unsafe
manner; (3) the definition of "startup" does not recognize integrated coal combine cycle (IGCC)
unit-specific startup practices; (4) the definition of "startup" does not recognize supercritical
unit-specific practices; (5) the definition of "startup" does not take into account emissions
monitoring issues; (6) the EPA fails to recognize differences in EGU shutdown procedures; (7)
the EPA fails to reconcile application of startup and shutdown provisions to multiple EGUs at a
facility sharing a common stack; and (8) the EPA fails to account for startup fuel availability
issues. Petitioner 5775 further states that the EPA did not justify revoking the exemption in the
NSPS regarding compliance during SSM events and requiring new, modified, or reconstructed
units under Subpart Da to comply with emission standards during SSM.
Response to Issue 4: Petitioners have not demonstrated that the issue presented meets the
criteria for reconsideration set forth in CAA §307(d)(7)(B). Petitioners have neither
demonstrated that it was impracticable to raise the objection during the comment period, nor that
the grounds for the objection arose after the close of the comment period. The EPA, therefore,
denies reconsideration of the issue. The EPA proposed that the PM, NOx, and SO2 emissions
limits apply at all times. In response to comments regarding this proposed approach, including
comments from Petitioners 5775 and 5777, and based on its own further analysis of the issue, the
EPA adjusted the approach such that the NOx and SO2 emissions limits apply at all times and the
applicable PM limit applies at all times except during periods of startup and shutdown. The data
used in developing the NOx and S02 emissions limits included periods of startup and shutdown.
As a result, the EPA has concluded there is no justification for not requiring compliance with
those standards during periods of startup and shutdown. The EPA did not have data on PM
emissions during periods of startup and shutdown and, as a result, the final rule establishes work
practice standards for the control of PM that apply during such periods. These changes, together
with the responses set forth in the response to comments document (RTC) for the final rule
constitute a reasonable response to the comments received as they relate to the NOx and SO2
standards. The EPA previously finalized reconsideration of various issues associated with the
control of PM during periods of startup and shutdown and Petitioners have provided no basis for
revisiting the issue.
160
-------
33.0 Affirmative Defense for EGUs
Issue 5: Petitioner 5775 requests reconsideration for the addition of a requirement to the EGU
NSPS allowing the use of EPA-approved state affirmative defense provisions as an alternative to
the affirmative defense provisions included in 40 CFR Part 60, subpart Da, under 40 CFR
60.48Da. Petitioner states that such an alternative allows both affected EGU owners and
operators and State air regulatory agencies to avoid issues arising from duplicative affirmative
defense programs. Petitioner states that in the EGU NESHAP, the EPA allows states to petition
for an alternative approach with regard to the affirmative defense provisions (40 CFR 63.93).
Petitioner states that the general provisions of the NSPS rules in 40 CFR Part 60 do not have a
similar provision. According to Petitioner, not having this alternative will force states to either
modify existing EPA-approved state implementation plan (SIP) rules for affirmative defense
provisions to be consistent with the new NSPS provisions or to maintain two legally distinct
affirmative defense programs in a state. Petitioner asserts that the EPA has not provided any
justification for denying Petitioner's request or not at least modifying the NSPS rule to allow
states to petition EPA for an alternative affirmative defense provision.
Response to Issue 5: The EPA denies the request for reconsideration because the Petitioner has
failed to demonstrate that the issue presented meets the criteria for reconsideration set forth in
CAA §307(d)(7)(B). The Petitioner has neither demonstrated that it was impracticable to raise
the objection during the comment period, nor that the grounds for the objection arose after the
close of the comment period. As Petitioner notes, the EPA responded to comments on the
affirmative defense to civil penalties in the proposed MATS rule and made revisions to the
affirmative defense when issuing the final rule. Petitioner's request for reconsideration merely
maintains that the EPA's response was not adequate. The EPA provided a reasoned response to
the comments and the Clean Air Act requires nothing more. As further grounds for denying the
request for reconsideration, the EPA notes that the issue for which reconsideration is sought is
not of central relevance to the outcome of the final rule. In addition, as EPA has previously
stated (see EPA-HQ-OAR-2011-0044-5759, RTC, p.26), the NSPS is a nationally applicable rule
and as such the affirmative defense provisions should be the same regardless of where the
affected facility is located. Any other approach would create confusion for both the regulated
community and the implementing authority. Therefore, allowing the use of a SIP-approved
affirmative defense as an alternative is not appropriate. Finally, the revision to the NSPS
General Provisions (40 CFR part 60, subpart A) that Petitioner suggests is beyond the scope of
this rulemaking.
161
-------
34.0 Definition of Periods of Out-of-control for EGUs
Issue 6: Petitioner 5777 states that the EPA adopted a definition of "out-of-control" for the EGU
NSPS that is similar to the definition in 40 CFR Part 75 for the Acid Rain program, except that
the "out-of-control" period under the EGU NSPS begins and ends with the corresponding
"quadrant" instead of the corresponding "hour." Petitioner objects to the definition used for the
EGU NSPS because the definition is inconsistent with the definitions of the term used for
requirements applicable to EGUs under both Part 75 and Part 63. Petitioner asserts that
inconsistent definitions make data programming more difficult and could result in inconsistent
data sets and inappropriate invalidation of data. Petitioner also states that the EPA's use of the
term "quadrant" is unclear and that to the extent that the EPA intended to invalidate and validate
hourly data based on 15 minute periods, the practical implications of such action are not clear.
Petitioner requests that the EPA propose and solicit comment on a revised definition of an "out-
of-control" period provider a rationale for the definition, and discuss how the definition would
apply in practice.
Response to Issue 6: The EPA is denying the request for reconsideration of this issue because it
does not meet the criteria for reconsideration under CAA section 307(d)(7)(B). The EPA
proposed the definition, received comment thereon and adequately responded to those
comments. Further, the Petition does not demonstrate that the issue is of central relevance to the
outcome of the final rule. Finally, the fact that "out-of-control" period is defined somewhat
differently in the EGU NSPS as compared to how it is defined in 40 CFR Part 75 does not render
the EGU NSPS definition inappropriate. While the EPA strives to maintain consistent
definitions where appropriate, when dealing with programs that are fundamentally separate and
distinct, such as the NSPS and Acid Rain programs, the use of inconsistent definitions may be
appropriate.
162
-------
35.0 Frequency of Subpart D Visible Emissions Testing
Issue 7: Petitioner 5777 requests reconsideration of the revisions to the frequency of visible
emission testing for EGUs subject to 40 CFR Part 60, subpart D, in order to be consistent with
the opacity monitoring requirements used for EGUs subject to 40 CFR Part 60, subpart Da.
Reasons presented by Petitioner for revision of the visible emission testing requirements in the
rule include: (1) the EPA provided no rationale for its decision not to reduce the testing
frequency for EGUs subject to 40 CFR Part 60, subpart D, in the same manner it did for EGUs
subject 40 CFR Part 60, subpart Da; (2) the EPA concluded without discussion or record citation
that additional visible emissions testing is needed for all EGUs with opacity greater than 5
percent; and (3) the EPA does not explain why existing state visible emissions testing
requirements are not adequate for EGUs and why the more frequent visible emissions testing
required under the EGU NSPS is necessary and not redundant to assure compliance if such
testing already is required by a state Title V operating permit.
Response to Issue 7: The EPA is denying the request for reconsideration of this issue because it
does not meet the criteria for reconsideration under CAA section 307(d)(7)(B) as it was not
impractical to raise the objection during the period for public comment nor did the grounds for
the objection arise after the close of the comment period. Petitioner raised precisely this issue in
commenting on the proposed rule. (Comments of Utility Air Regulatory Group at 46 - 47.) In
response, the EPA noted that subpart D had been amended to allow the permitting authority the
discretion to establish site-specific monitoring plans for owners/operators of facilities burning
fuels that typically result in low opacity. The EPA also noted that the frequency of Method 9
performance testing for owners/operators of facilities with some visible emissions, but with all 6-
minute readings of 5 percent or less, had been reduced from every 6 months to every 12 months
for facilities subject to subpart Da. Finally, the EPA stated that the additional testing frequency
for facilities with opacities of 5 percent and higher is necessary to adequately assure compliance
with the applicable opacity standard. (RTC at 20.) In response to Petitioner's comment regarding
existing state visible emissions testing requirements, the EPA notes that the NSPS itself must
contain monitoring sufficient to assure compliance with the applicable emission standards.
Further, owners/operators of affected facilities can currently petition the agency for alternate
monitoring procedures to account for site specific conditions.
163
-------
36.0 Redundant Subpart Da Recordkeeping
Issue 8: Petitioner 5777 requests reconsideration of the addition of the requirement to the EGU
NSPS under 40 CFR 60.51Da(d), that affected owners and operators must maintain records of
the results of Method 9 performance tests and submit "excess emissions" reports. Petitioner
asserts that the requirement serves no purpose and duplicates existing recordkeeping
requirements in the rule. Petitioner notes that the requirement as promulgated applies only to
steam generating units subject to 40 CFR Part 60, subpart Dc, and states that because EGUs are
not subject to opacity limits in 40 CFR 60.43c(c), this requirement does not apply to EGUs.
Petitioner states that assuming that the EPA intended the requirement to apply to EGUs subject
to the opacity limit in 40 CFR 60.42Da(b), the requirement duplicates existing recordkeeping
requirements in 40 CFR 60.52Da(b)(l). Specifically, according to Petitioner, 40 CFR
60.52Da(b)(l) already requires affected EGU owners and operators subject to the opacity
standards in 40 CFR 60.42Da(b) to maintain all of the records identified in 40 CFR 60.51Da(d).
Petitioner requests that the EPA identify what units must comply with this new recordkeeping
requirement and remove the redundant recordkeeping requirements.
Response to Issue 8: The EPA is denying the request for reconsideration of this issue because it
does not meet the criteria for reconsideration under CAA section 307(d)(7)(B). Although the
EPA acknowledges that the reference to 40 CFR 60.43c(c) is incorrect, this error is not of central
relevance to the outcome of the final rule. Further, as Petitioner points out, EGUs are not subject
to 40 CFR Part 60, subpart Dc. As a result, the referenced provision does not apply to
owners/operators of EGUs and, therefore, imposes no burden on them. The EPA will correct the
reference to 40 CFR 60.43c(c) at some point in the future and may propose to remove the
requirement itself in a future notice and comment rulemaking.
164
-------
37.0 De Minimis Use of Natural Gas in Subpart Db Steam Generating Units
Issue 9: Petitioner 5773 asserts that the use of natural gas for safety purposes and stable flame
operation (i.e., less than 5 to 10 percent of the capacity factor) should not cause a steam
generating unit subject to 40 CFR Part 60, subpart Db, to comply with the rule's SO2 standard
when the affected unit would not otherwise be required to meet the standard. Petitioner states
that a de minimis use of natural gas for safety purposes and minimization of air pollutant
emissions from stable flame operation should not be considered to be a "mixture of fuels" for
purposes of applying the 0.32 lb/MMBtu SO2 emissions limit under 40 CFR 60.42b(k)(l).
Specifically, Petitioner states that it is not clear whether a steam generating unit subject to 40
CFR Part 60, subpart Db, that burns blast furnace gas (BFG) combined with a very limited
amount of natural gas is exempt from the SO2 emissions limit. Petitioner states that for these
types of units, there is a need for a flame stabilization fuel that provides for both lower and
controlled NOx and CO emissions during variations in the BFG composition burned and that
reduces the risk of a BFG deflagration or detonation in the unit following a flameout. Petitioner
states that the supplemental natural gas fuel provides an instantaneous source of ignition within
the burner as the BFG fuel properties change rapidly, preventing flameouts. Petitioner
recommends that de minimis use of natural gas for safety purposes and air pollutant emissions
minimization from stable flame operations should not be considered a "mixture of fuels" for
purposes of implementing 40 CFR 60.42b(k)(l). Petitioner states that the NSPS without this
revision could discourage the development of energy recovery technologies and processes.
Response to Issue 9: As currently written, the subpart Db SO2 standard would apply to a new
industrial boiler facility that burns any amount of coal, oil, or natural gas. The EPA did not
amend the industrial boiler SO2 standard in promulgating this final rule. As a result, the EPA is
denying the request for reconsideration as outside the scope of the final rule.
165
-------
Appendix B. List of NSPS Petitioners
Petitioners
Docket number
(EPA-HQ-OAR-2011-0044-)
Air Products
5773
State of Texas (Texas Commission on Environmental
Quality, Texas Public Utility Commission, Railroad
Commission of Texas)
5775
Utility Air Regulatory Group (UARG)
5777
166
------- |