Rule: Federal Implementation Plan For The Billings/Laurel, MT, Sulfur Dioxide Area

Monday,
April 21, 2008
Part II
Environmental
Protection Agency
40 CFR Part 52
Federal Implementation Plan for the
Billings/Laurel, Montana, Sulfur Dioxide
Area; Final Rule
jlentini on PROD1PC65 with RULES2
VerDate Aug<31>2005 18:55 Apr 18, 2008 Jkt 214001 PO 00000 Frm 00001 Fmt 4717 Sfmt 4717 E:\FR\FM\21APR2.SGM 21APR2
21418 Federal Register / Vol. 73, No. 77 / Monday, April 21, 2008 / Rules and Regulations
ENVIRONMENTAL PROTECTION Program, Environmental Protection (vi) The initials COPC mean or refer
AGENCY Agency (EPA), Region 8, 1595 Wynkoop to ConocoPhillips.
Street, Denver, Colorado 80202–1129, (vii) The words EPA, we, us or our
40 CFR Part 52 (303) 312–6437, ostrand.laurie@epa.gov. mean or refer to the United States
[EPA–R08–OAR–2006–0098; FRL–8551–2] SUPPLEMENTARY INFORMATION: Environmental Protection Agency.
(viii) The initials FIP mean or refer to
RIN 2008–AA01 Table of Contents Federal Implementation Plan.
Definitions (ix) The initials H2S mean or refer to
Federal Implementation Plan for the hydrogen sulfide.
I. Background of the Final Rules
Billings/Laurel, MT, Sulfur Dioxide (x) The initials MBER mean or refer to
II. Issues Raised by Commenters and EPA’s
Area Response the Montana Board of Environmental
AGENCY: Environmental Protection A. FIP Not Necessary Review.
Agency (EPA). B. EPA Exceeded Its Authority in (xi) The initials MDEQ mean or refer
Proposing a FIP to the Montana Department of
ACTION: Final rule. C. Flare Monitoring Environmental Quality.
SUMMARY: The Environmental Protection
D. Flare Limits (xii) The initials MPA mean or refer to
E. Concerns With Dispersion Modeling
Agency (EPA) is promulgating a Federal the Montana Petroleum Association.
F. Miscellaneous Comments
Implementation Plan (FIP) containing G. MSCC Specific Issues
(xiii) The initials MSCC mean or refer
emission limits and compliance H. ConocoPhillips Specific Issues to the Montana Sulphur & Chemical
determining methods for several sources I. ExxonMobil Specific Issues Company.
located in Billings and Laurel, Montana. J. CHS Inc. Specific Issues (xiv) The initials NAAQS mean or
EPA is promulgating a FIP because of III. Summary of the Final Rules and Changes refer to National Ambient Air Quality
our previous partial and limited From the July 12, 2006, Proposal Standards
disapprovals of the Billings/Laurel A. Flare Requirements Applicable to All (xv) The initials NEDA/CAP mean or
Sulfur Dioxide (SO2) State Sources refer to the National Environmental
B. CHS Inc. Development Association’s Clean Air
Implementation Plan (SIP). The C. ConocoPhillips
intended effect of this action is to assure Project.
D. ExxonMobil
attainment of the SO2 National Ambient E. Montana Sulphur & Chemical Company
(xvi) The initials NPRA mean or refer
Air Quality Standards (NAAQS) in the (MSCC) to the National Petrochemical & Refiners
Billings/Laurel, Montana area. EPA is F. Modeling to Support Emission Limits Association.
taking this action under sections 110, IV. Statutory and Executive Order Reviews (xvii) The initials SCAQMD mean or
301, and 307 of the Clean Air Act (Act). A. Executive Order 12866, Regulatory refer to the South Coast Air Quality
DATES: Effective Date: This final rule is
Planning Review Management District.
B. Paperwork Reduction Act (xviii) The initials SIP mean or refer
effective May 21, 2008. The C. Regulatory Flexibility Act
incorporation by reference of certain to State Implementation Plan.
D. Unfunded Mandates Reform Act (xix) The initials SO2 mean or refer to
publications listed in this regulation is E. Executive Order 13132, Federalism
approved by the Director of the Federal sulfur dioxide.
F. Executive Order 13175, Coordination
Register as of May 21, 2008. With Indian Tribal Governments
(xx) The words State or Montana
G. Executive Order 13045, Protection of mean the State of Montana, unless the
ADDRESSES: EPA has established a
Children From Environmental Health context indicates otherwise.
docket for this action under Docket ID (xxi) The initials SRU mean or refer
Risks and Safety Risks
No. EPA–R08–OAR–2006–0098. All to sulfur recovery unit.
H. Executive Order 13211, Actions That
documents in the docket are listed on Significantly Affect Energy Supply, (xxii) The initials SWS mean or refer
the http://www.regulations.gov Web Distribution, or Use to sour water stripper.
site. Although listed in the index, some I. National Technology Transfer and (xxiii) The initials WETA mean or
information is not publicly available, Advancement Act refer to the Western Environmental
e.g., Confidential Business Information J. Executive Order 12898: Federal Actions Trade Association.
(CBI) or other information whose To Address Environmental Justice in
(xxiv) The initials WSPA mean or
disclosure is restricted by statute. Minority Populations and Low-Income
Populations refer to the Western States Petroleum
Certain other material, such as Association.
K. Congressional Review Act
copyrighted material, is not placed on (xxv) The initials YCC mean or refer
L. Petitions for Judicial Review
the Internet and will be publicly to the Yellowstone County
available only in hard copy form. Definitions Commissioners.
Publicly available docket materials are For the purpose of this document, we (xxvi) The initials YVAS mean or
available either electronically through are giving meaning to certain words or refer to the Yellowstone Valley
http://www.regulations.gov or in hard initials as follows: Audubon Society.
copy at the Air and Radiation Program, (i) The words or initials Act or CAA
Environmental Protection Agency I. Background of the Final Rules
mean or refer to the Clean Air Act,
(EPA), Region 8, 1595 Wynkoop Street, unless the context indicates otherwise. The Clean Air Act (Act) requires EPA
Denver, Colorado 80202–1129. EPA (ii) The initials API mean or refer to to establish national ambient air quality
requests that if at all possible, you the American Petroleum Institute. standards (NAAQS) that protect public
contact the individual listed in the FOR (iii) The initials BAAQMD mean or health and welfare. NAAQS have been
FURTHER INFORMATION CONTACT section to refer to the Bay Area Air Quality established for SO2 as follows: 0.030
view the hard copy of the docket. You Management District. parts per million (ppm) annual
may view the hard copy of the docket (iv) The initials CEMS mean or refer standard, not to be exceeded in a
Monday through Friday, 8 a.m. to 4 to continuous emission monitoring calendar year; 0.14 ppm 24-hour
p.m., excluding Federal holidays. system. standard, not to be exceeded more than
FOR FURTHER INFORMATION CONTACT: (v) The initials CO mean or refer to once per calendar year; and 0.5 ppm 3-
Laurie Ostrand, Air and Radiation carbon monoxide. hour standard, not to be exceeded more
Federal Register / Vol. 73, No. 77 / Monday, April 21, 2008 / Rules and Regulations 21419
than once per calendar year. See 40 CFR final action on this FIP. The case is The FIP provides an additional incentive
50.4 and 50.5. The Act also requires captioned Montana Sulphur & Chemical for state compliance because it rescinds state
states to prepare and gain EPA approval Company v. United States authority to make the many sensitive
of a plan, termed a State Environmental Protection Agency, No. technical and political choices that a
pollution control regime demands. The FIP
Implementation Plan (SIP), to assure 02–71657. No petitions for judicial provision also ensures that progress toward
that the NAAQS are attained and review were filed regarding EPA’s May NAAQS attainment will proceed
maintained. 22, 2003, SIP action. notwithstanding inadequate action at the
Dispersion modeling completed in On July 12, 2006 (71 FR 39259), EPA state level.
1991 and 1993 for the Billings/Laurel proposed Federal Implementation Plan
area of Montana predicted that the SO2 Natural Resources Defense Council,
(FIP) provisions for the Billings/Laurel,
NAAQS were not being attained. As a Inc. v. Browner, 57 F.3d 1122, 1124
Montana area because of our
result, in March 1993 EPA (pursuant to (D.C. Cir. 1995).
disapproval of portions of Montana’s
sections 110(a)(2)(H) and 110(k)(5) of When EPA promulgates a FIP, courts
Billings/Laurel SO2 SIP. In our proposal,
the Act, 42 U.S.C. 7410(a)(2)(H) and have not required EPA to demonstrate
we indicated that our FIP would not
7410(k)(5)) requested the State of explicit authority for specific measures:
replace the SIP entirely, but instead
Montana to revise its previously ‘‘We are inclined to construe Congress’
would only replace elements of, or fill
approved SO2 SIP for the Billings/Laurel broad grant of power to the EPA as
gaps in, the SIP.
area. See 58 FR 41450, August 4, 1993. including all enforcement devices
In promulgating today’s rules, EPA is reasonably necessary to the achievement
In response, the State submitted fulfilling its mandatory duty under
revisions to the SO2 SIP on September and maintenance of the goals
section 110(c) of the Act. Under section established by the legislation.’’ South
6, 1995, August 27, 1996, April 2, 1997, 110(c), whenever we disapprove a SIP,
July 29, 1998, and May 4, 2000. Terminal Corp. v. EPA, 504 F.2d 646,
in whole or in part, we are required to 669 (1st Cir. 1974). As the Ninth Circuit
On May 2, 2002 (67 FR 22168) and promulgate a FIP. Specifically, section
May 22, 2003 (68 FR 27908), we stated in a case involving a FIP with far-
110(c) provides: reaching consequences in Los Angeles:
partially approved, partially
‘‘(1) The Administrator shall promulgate a ‘‘The authority to regulate pollution
disapproved, limitedly approved, and Federal implementation plan at any time
limitedly disapproved the Billings/ carries with it the power to do so in a
within 2 years after the Administrator— manner reasonably calculated to reach
Laurel SO2 SIP. In those actions we (A) Finds that a State has failed to make
disapproved the following: that end.’’ City of Santa Rosa v. EPA,
a required submission or finds that the plan
• The attainment demonstration due or plan revision submitted by the State does 534 F.2d 150, 155 (9th Cir. 1976),
to issues with various emission limits, not satisfy the minimum criteria established vacated and remanded on other grounds
inappropriate stack height credit, and under [section 110(k)(1)(A)],1 or sub nom. Pacific Legal Foundation v.
lack of emission limits on flares. (B) Disapproves a State implementation EPA, 429 U.S. 990 (1976).
• The emission limits for Montana plan submission in whole or in part, unless In addition to giving EPA remedial
Sulphur & Chemical Company’s the State corrects the deficiency, and the authority, section 110(c) enables EPA to
Administrator approves the plan or plan assume the powers that the state would
(MSCC’s) sulfur recovery unit (SRU)
revision, before the Administrator have to protect air quality, when the
100-meter stack and the stack height promulgates such Federal implementation
credit on which those limits were based. state fails to adequately discharge its
plan.’’
• The emission limits for MSCC’s Thus, because we disapproved
planning responsibility. As the Ninth
auxiliary vent stacks due to lack of an Circuit held, when EPA acts to fill in the
portions of the Billings/Laurel SO2 SIP,
adequate limit on fuel burned in the gaps in an inadequate state plan under
and the attainment demonstration, we
associated heaters and boilers and lack section 110(c), EPA ‘‘ ‘stands in the
are required to promulgate a FIP.
of a reliable compliance determining shoes of the defaulting State, and all of
Section 302(y) defines the term the rights and duties that would
method. ‘‘Federal implementation plan’’ in
• The emission limits for MSCC’s 30- otherwise fall to the State accrue instead
pertinent part, as: to EPA.’ ’’ Central Arizona Water
meter stack due to lack of an adequate
limit on fuel burned in the associated ‘‘[A] plan (or portion thereof) promulgated Conservation District v. EPA, 990 F.2d
heaters and boilers, and lack of a by the Administrator to fill all or a portion 1531, 1541 (9th Cir. 1993). As the First
reliable compliance determining of a gap or otherwise correct all or a portion Circuit held in an early case:
of an inadequacy in a State implementation
method. plan, and which includes enforceable ‘‘[T]he Administrator must promulgate
• Provisions that allowed sour water emission limitations or other control promptly regulations setting forth ‘an
stripper overheads to be burned in the measures, means or techniques (including implementation plan for a State’ should the
flares at CHS Inc. and ExxonMobil. economic incentives, such as marketable state itself fail to propose a satisfactory one
• ExxonMobil’s refinery fuel gas permits or auctions or emissions allowances) * * * The statutory scheme would be
combustion device emission limits and * * *.’’ unworkable were it read as giving to EPA,
associated compliance determining when promulgating an implementation plan
More simply, a FIP is ‘‘a set of for a state, less than those necessary
methods.
• ExxonMobil’s Coker CO Boiler stack enforceable federal regulations that measures allowed by Congress to a state to
emission limits and associated stand in the place of deficient portions accomplish federal clean air goals. We do not
of a SIP.’’ McCarthy v. Thomas, 27 F.3d adopt any such crippling interpretation.’’
compliance determining methods.
• CHS Inc.’s combustion source 1363, 1365 (9th Cir. 1994). As the Court South Terminal Corp. v. EPA, supra,
emission limits and certain associated of Appeals for the D.C. Circuit noted in at 668 (citing previous version of section
compliance determining methods. a 1995 case, FIPs are powerful tools to 110(c)).
On June 10, 2002, MSCC petitioned remedy deficient state action: The Billings/Laurel SO2 FIP
the United States Court of Appeals for establishes emission limits and
1 Section 110(k)(1) requires the Administrator to
the Ninth Circuit for review of EPA’s compliance determining methods for
promulgate minimum criteria that any plan
May 2, 2002, final SIP action. submission must meet before EPA is required to act
four sources located in Billings/Laurel,
Subsequently, MSCC and EPA agreed to on the submission. These completeness criteria are Montana, to replace/fill gaps in portions
a stay of the litigation pending EPA’s set forth at 40 CFR 51, Appendix V. of the SIP we disapproved, and to
support our attainment demonstration. stacks and SRU 30-meter stack. In usually are not representative of the air
Three of the sources are petroleum addition to mass limits, the FIP quality for an area. (See reference
refineries: CHS Inc., ConocoPhillips establishes concentration limits on fuel document GGGGG, April 21, 1983,
(including the Jupiter Sulfur facility), burned in the units that vent to the memorandum from Sheldon Meyers,
and ExxonMobil. The fourth source is auxiliary vent stacks and SRU 30-meter Director, Office of Air Quality Planning
Montana Sulphur & Chemical Company, stack. These concentration limits are and Standards (OAQPS), to Regional Air
which provides sulfur recovery for the 160 ppm H2S per 3-hour period and 100 and Waste Division Directors, titled
ExxonMobil refinery. ppm H2S per calendar day. When trigger ‘‘Section 107 Designation Policy
The following is a summary of the events specified in the rule occur, Summary,’’ and reference document
major components of our FIP rule: MSCC must measure the H2S HHHHH, September 4, 1992,
(1) The FIP establishes flare emission concentration in the fuel using length- memorandum from John Calcagni,
limits at all four sources (150 lbs SO2/ of-stain detector tubes on a once-per-3- Director, Air Quality Management
3-hour period at all but the Jupiter hour period. Division, OAQPS, to Regional Air
Sulfur flare, 75 lbs SO2/3-hour period (7) The FIP establishes various Division Directors, titled ‘‘Procedures
shared limit for the Jupiter Sulfur flare recordkeeping and reporting for Processing Requests to Redesignate
and the Jupiter Sulfur SRU/ATS stack) requirements. Areas to Attainment.’’)
and monitoring methods to determine It is important to note that, in cases Typically, modeling estimates of
compliance with those limits. The FIP where the provisions of the FIP address maximum ambient concentrations are
includes an affirmative defense to emissions activities differently or based on a fairly infrequent combination
penalties for violations of the flare establish different requirements than of meteorological and source operating
limits that occur during malfunction, provisions of the SIP, the provisions of conditions. To capture such results on
startup, and shutdown periods. To the FIP take precedence. We also an ambient monitor would normally
determine flare emissions, the FIP caution that if any of the four sources require a prohibitively large and
requires concentration monitoring are subject to requirements under other expensive network. Therefore,
(which can consist of continuous provisions of the Act (e.g., section 111 dispersion modeling is generally
monitoring, grab sampling, or integrated or 112, part C of title I, or SIP-approved necessary to comprehensively evaluate
sampling) and continuous flow permit programs under part A of title I), sources’ impacts and to determine the
monitoring. our promulgation of the FIP does not areas of expected high concentrations.
(2) The FIP prohibits the burning of excuse any of the sources from meeting (Id.) Air quality modeling results would
sour water stripper overheads in CHS such requirements. Finally, our be especially important if sources were
Inc.’s main crude heater and requires promulgation of the FIP does not imply not emitting at their maximum level
CHS Inc. to keep the valve between the any sort of applicability determination during the monitoring period or if the
old sour water stripper and the main under other provisions of the Act (e.g., monitoring period did not coincide with
crude heater closed, chained, and section 111 or 112, part C of title I, or potentially worst-case meteorological
locked. SIP-approved permit programs under conditions. Further, ambient monitoring
(3) The FIP provides that emission part A of title I). data are not adequate if sources are
limits for identified ExxonMobil using stacks with actual heights greater
refinery fuel gas combustion units are II. Issues Raised by Commenters and
than good engineering practice stack
contained in the SIP, and establishes EPA’s Response
height (which indeed is the case with
compliance determining methods for A. FIP Not Necessary MSCC and ConocoPhillips) or other
instances in which the H2S dispersion techniques for which SIP/FIP
concentration in the refinery fuel gas 1. Ambient Data and Historical
modeling credit is not allowed. (See also
stream exceeds 1200 ppmv. These Modeling Show Attainment
our discussion of related issues in our
methods involve the use of length-of- (a) Comment (CHS Inc., COPC, final action on the Billings/Laurel SO2
stain detector tubes on a once-per-hour ExxonMobil, NPRA, MPA, MDEQ, SIP (67 FR 22168, 22185–22187, May 2,
frequency. MSCC, WETA): The FIP is not necessary 2002.))
(4) The FIP provides that emission for attainment of the NAAQS because Ambient monitoring data and air
limits for ExxonMobil’s Coker CO Boiler ambient data show that the Billings/ quality modeling data for a particular
stack, when ExxonMobil’s Coker unit is Laurel area has been for many years and area can sometimes appear to conflict.
operating and Coker unit flue gases are continues to be in attainment with both This is primarily due to the fact that
burned in the Coker CO Boiler, are the Federal and State SO2 ambient air modeling results may predict maximum
contained in the SIP. The FIP quality standards for all averaging SO2 concentration at receptors where no
establishes compliance determining periods. monitors are located.
methods for these emission limits that Response: EPA does not agree that a Moreover, our SIP Call for the
require measurement of the SO2 FIP is not necessary because ambient Billings/Laurel area was based on
concentration and flow rate in the Coker data show attainment of the SO2 modeled violations of the SO2 NAAQS,
CO Boiler stack using CEMS. NAAQS. Ambient monitoring is limited not monitored violations. (See reference
(5) The FIP establishes emission in time and in space. Ambient documents Y and Z.) We took final
limits on MSCC’s SRU 100-meter stack, monitoring can measure pollutant action on the SIP Call in our May 2,
based on good engineering practice concentrations only as they occur; it 2002, action on the Billings/Laurel SIP
(GEP) stack height credit of 65 meters, cannot predict future concentrations (67 FR 22168, 22173), and we are not
with compliance with these limits to be when emission levels and revisiting it in this FIP action. It would
determined using methods already meteorological conditions may differ be inconsistent and inappropriate to
approved in the SIP. The FIP does not from present conditions. now rely solely on monitoring to
provide variable emission limits for this EPA has long held that ambient determine necessary measures and
stack. monitoring data alone generally are not demonstrate attainment.
(6) The FIP establishes emission adequate for SO2 attainment It is especially important to recognize
limits and compliance determining demonstrations. Additionally, a small that, as a result of our partial and
methods for MSCC’s auxiliary vent number of ambient SO2 monitors limited disapproval of the Billings/
Laurel SO2 SIP, we are legally obligated submissions. As explained in those We note that we have not reopened
to promulgate a FIP for the area. See actions, EPA does not agree that the our SIP actions as part of this action.
section 110(c)(1) of the CAA, 42 U.S.C. State’s SIP modeling, along with Thus, to the extent the commenters are
7410(c)(1). However, the SIP appropriate emission limits, show expressing their disagreement with
deficiencies that triggered our partial attainment of the NAAQS. EPA’s formal EPA’s actions on the SIP, their
and limited disapproval were varied determinations regarding the attainment comments are not relevant to this
and were not necessarily associated demonstration and emission limits were action, and EPA is not re-considering
with problems that could be measured made in final actions on May 2, 2002 them here.
at an ambient monitor. For example, one (67 FR 22168) and May 22, 2003 (68 FR
(b) Comment (CHS Inc., WETA,
basis for disapproval of the SIP was the 27908). The FIP fills the gaps for the
COPC, MDEQ, ExxonMobil, NPRA): In
State’s use of improper (too tall) stack provisions we disapproved.
We note that we have not reopened addition to the SIP, SO2 emissions in
height credit for MSCC in modeling
our SIP actions as part of this action. the Billings/Laurel area have decreased
attainment of the NAAQS. In the real
world, emissions at the actual (100 Thus, to the extent the commenters are as a result of Consent Decrees and
meter) height of the stack create less expressing their disagreement with Montana Air Quality Permit changes.
impact on monitored ambient EPA’s actions on the SIP, their These limits are all federally enforceable
concentrations in the Billings/Laurel comments are not relevant to this because there are Title V operating
area than if the emissions were emitted action, and EPA is not re-considering permit conditions (CHS Inc.). EPA did
from a lower stack. Nonetheless, we had them here. not consider these emission reductions
to partially disapprove the SIP due to (c) Comment (ExxonMobil): EPA’s in making its determination that the FIP
the State’s inappropriate grant of stack proposed FIP ignores the substantial was necessary. The FIP proposal does
height credit, and section 110(c) of the improvement in air quality in the not otherwise acknowledge the practical
CAA requires that we correct the Billings/Laurel area and instead predicts effects of the recent consent decrees
deficiency. Since the State did not exceedances of NAAQS based upon between the primary refinery parties
model attainment at the proper stack modeling performed as long as 15 years subject to regulation as well as other
height credit for MSCC’s stack, it was ago. EPA’s FIP proposal must be further permitting actions that have occurred
necessary that we do so and set examined in light of subsequent over the past eight years (MSCC, COPC).
emission limits for the stack consistent developments, including correct Response: EPA did not consider the
with our attainment demonstration. We modeling and consideration of currently emission reductions that resulted, or
believe MSCC has consistently been available information indicating will result, from the consent decrees
meeting the emission limits we are compliance. and/or State permit revisions to
adopting, so there may be no reduction Response: See response to comment determine that the FIP was necessary or
in actual emissions from the stack, but II.A.1.(a), above, regarding ambient data include the emission reductions in our
that does not mean the CAA allows us and response to comments in section modeling for several reasons.
to forego this aspect of the FIP. II.E., below, regarding modeling.
Likewise, CAA sections 110(a)(2)(A) First, the FIP is required because we
2. Existing Controls Sufficient disapproved the SIP, and the State has
and (C) require that SIP control
measures be enforceable. We (a) Comment (MDEQ, MSCC, COPC, not made revisions to the SIP to address
disapproved several source monitoring ExxonMobil, MPA, NPRA, WETA): The the SIP’s flaws. As noted in other
methods because they were not FIP offers questionable improvements responses, because we disapproved the
adequate to determine compliance because the existing control plan SIP, we have a legal obligation to
under all operating conditions. It may provisions submitted by the state are promulgate a FIP. See CAA section
be impossible to measure the impact adequate and contain sufficient SO2 110(c), 42 U.S.C. 7410(c).
these SIP deficiencies may have on emission controls and strategies and Second, even though permits and
ambient SO2 concentrations in the area, provide for the implementation, consent decrees are federally
but the CAA still requires that we maintenance, and enforcement of the enforceable, some permits can be
correct the deficiencies. Regarding the SO2 NAAQS. revised without EPA approval and
emission limits and compliance Response: EPA addressed the consent decrees have a limited
determining methods for the flares, the adequacy of Montana’s SIP submissions lifespan.2 To protect the integrity of the
State-only flare limits, which the State in its final actions on the SIP. As attainment demonstration, and our
relied on to demonstrate attainment, explained in those actions, EPA does statutory role in assessing SIP/FIP
may have positively impacted flare not agree that the State’s SIP control adequacy, we believe that stationary
emissions in the past few years. plan provisions are adequate and source emission limits necessary to
However, the State did not include the contain sufficient SO2 emission controls demonstrate attainment must be
State-only flare limits or adequate to show attainment of the NAAQS. included in the FIP (or approved SIP).
compliance determining methods in the EPA’s formal determinations regarding See, e.g., CAA sections 110(a)(2)(A),
SIP. Thus, the SIP remains deficient. We the attainment demonstration and 110(i), 110(k)(3)–(6), and 110(l), 42
now have the responsibility to ensure emission control plan were made in U.S.C. 7410(a)(2)(A), (i), (k)(3)–(6), and
that emission limits relied on to final actions on May 2, 2002 (67 FR (l). This ensures that changes to those
demonstrate attainment are included in 22168) and May 22, 2003 (68 FR 27908). limits will only be made with EPA’s
the SIP and are practically enforceable, In our May 2002 and May 2003 actions approval as a SIP or FIP revision,
consistent with the requirements of we disapproved various control plan
section 110 of the Act. provisions. The FIP fills the gaps for the 2 The State can revise construction permits
(b) Comment (MSCC, MDEQ): The provisions we disapproved. The FIP without EPA approval, and, while EPA has
State’s SIP modeling, along with offers necessary improvements to the authority to object to Title V permits, that authority
appropriate emission limits, show SIP by imposing new emission limits is only available to ensure that underlying
applicable requirements are included in the Title V
attainment of the NAAQS. and reliable compliance determining permits. Thus, if those underlying requirements
Response: EPA addressed this issue in methods to ensure attainment of the SO2 change, EPA may have no recourse at the Title V
its actions on Montana’s SIP NAAQS. stage.
following notice and comment Montana’s SO2 SIP for Billings/Laurel necessary to meet or assure SO2 NAAQS
rulemaking. did not fully satisfy CAA requirements. compliance.
Third, the consent decrees and See 67 FR 22168, May 2, 2002 and 68 Response: See our responses to
permitting actions, for some emission FR 27908, May 22, 2003. Thus, pursuant comments II.A.1.(a) and II.B.1.(a), above.
points, do not contain SO2 emission to section 110(c) of the CAA, 42 U.S.C. Much of this comment pertains to our
limits that are consistent with the 7410(c), we are required to promulgate actions on Montana’s SIP. We are not
averaging times of the SO2 NAAQS, a FIP. In doing so, we stand in the state’s revisiting or reopening comment on
specifically, the 3-hour and calendar shoes and have authority to determine those actions here. Our basis for finding
day averaging periods. For example, the emissions limitations and other that the SIP was not adequate to ensure
SIP establishes 3-hour, calendar day, measures for specific sources to fill gaps attainment and meet other CAA
and calendar year emission limits for in the SIP. Central Arizona Water requirements is described in our actions
CHS Inc.’s FCC regenerator/CO boiler Conservation District v. EPA, 990 F.2d on the SIP. Once we disapprove part or
stack. The January 17, 2007, final State 1531, 1541 (9th Cir. 1993); South all of a required SIP, section 110(c) of
construction permit (reference Terminal Corp. v. EPA, 504 F.2d 646, the Act requires that we issue a FIP. Our
document IIIII) and the consent decree 668 (1st Cir. 1974) (citing previous obligation in this action is to correct the
(reference document JJJJJ) indicate that version of CAA section 110(c)). SIP deficiencies we previously
the FCC regenerator stack SO2 emissions We note that we have not reopened identified. Thus, the findings that
shall not exceed 50 ppm by volume our SIP actions as part of this action. triggered our responsibility to
(corrected to 0% O2) for a 7-day rolling Thus, to the extent the commenters are promulgate a FIP were established in
average [or a fresh feed of 0.3 percent by expressing their disagreement with the prior rulemaking actions reviewing
weight] and 25 ppm by volume EPA’s actions on the SIP, their Montana’s SIP. EPA is not required to
(corrected to 0% O2) for a 365-day comments are not relevant to this repeat those findings in the FIP
rolling average. None of the commenters action, and EPA is not re-considering rulemaking itself.
has suggested these limits be converted them here. (e) Comment (ExxonMobil): EPA
to FIP mass limits that would apply over (b) Comment (WETA): Since the State cannot propose a FIP to replace a SIP,
a 3-hour averaging period, and the State of Montana has already taken unless the SIP is substantially
has not submitted a SIP revision with appropriate actions to reduce sulfur inadequate to ensure compliance with
such limits. dioxide emissions, EPA does not have the CAA.
It should be noted that EPA did solicit the authority under the CAA to adopt Response: The commenter misstates
comment on whether we should limit the proposed FIP. the standard for promulgation of a FIP.
the main flares to 500 pounds of SO2 per Response: See response to comment Section 110(c) of the CAA is
calendar day. This value is consistent II.B.1.(a), above. The adequacy of the straightforward—a FIP is required if (1)
with the trigger point for certain State of Montana’s actions has already EPA finds that a state has failed to make
analyses contained in settlements (i.e., been considered by EPA in other a required submission; (2) EPA finds
consent decrees) between the United rulemaking actions that addressed the that a plan submission does not satisfy
State’s SIP submission. Those actions the completeness criteria established
States and CHS Inc., ConocoPhillips,
are not the subject of EPA’s present under section 110(k)(1)(A) of the CAA;
and ExxonMobil. We received limited
rulemaking, which promulgates the or (3) EPA disapproves a SIP in whole
comments on this proposal and have
necessary measures to remedy the or in part. EPA partially disapproved
decided to keep the limit at 150 pounds
deficiencies EPA identified in its prior the Billings/Laurel SO2 SIP; thus, a FIP
of SO2 per 3-hour period to maintain
SIP reviews. is required. Contrary to the commenter’s
consistency with the State’s State-only
(c) Comment (MSCC): States have assertion, the obligation to promulgate a
limit.
primacy, and because EPA did not FIP is not contingent on an EPA finding
B. EPA Exceeded Its Authority in choose to exercise its rights in the of substantial inadequacy. As explained
Proposing a FIP comprehensive and competent state above, the findings triggering our
decision process, EPA may not default responsibility to promulgate a FIP were
1. State’s Responsibility
and then act. made in the prior actions reviewing
(a) Comment (WETA, MPA, Response: Under section 110(c) of the Montana’s SIP.
ExxonMobil): EPA’s role is limited to Act, EPA is not required to participate (f) Comment (MSCC): The commenter
determining whether or not a SIP is in a state’s administrative process before claims EPA’s action violates the Tenth
attaining and maintaining the NAAQS. promulgating a FIP. Amendment to the Constitution. The
Selecting the source mix and various (d) Comment (MSCC, MDEQ, commenter also claims EPA’s FIP is
control measures to achieve these ends ExxonMobil): EPA has no authority to dictating the required controls in
has been determined by courts to be the question the wisdom of a state’s choices contravention of the holdings in
sole responsibility of the state. EPA’s of emission limitations if they are part Commonwealth of Virginia v. EPA, 108
proposed action intrudes on the primary of a plan that satisfies the standards of F.3d 1397 (D.C. Cir. 1997) and
responsibility of the state and local § 110(a)(2) of the Act. As long as the Bethlehem Steel v. Gorsuch, 742 F.2d
governments to implement the Clean ultimate effect of a state’s choice of 1028 (7th Cir. 1984).
Air Act (MSCC). emission limitations is compliance with Response: Our FIP compels no action
Response: The commenters’ the NAAQS, the state is at liberty to on the part of the State and is not
characterization of EPA’s role regarding adopt whatever mix of emission coercive vis-à-vis the State. Our FIP
SIPs is not accurate. We lack authority limitations it deems best suited to its contains requirements applicable to four
to question a state’s choices of particular situation. There is no private companies. The Tenth
emissions limitations if they are part of evidence provided by EPA that Montana Amendment is not implicated. Nor do
a plan that satisfies the standards of the reached its material conclusions or our actions contravene Commonwealth
Clean Air Act. Train v. Natural choices in the SIP unreasonably. of Virginia or Bethlehem Steel. The
Resources Defense Council, 95 S.Ct. Additionally, EPA has not shown that former case held that EPA cannot, in a
1470, 1481–1482 (1975). In our 2002 additional controls beyond the SIP SIP Call, dictate that a state adopt a
and 2003 actions, we found that measures adopted by Montana are particular control measure to
demonstrate attainment of the NAAQS. separate showing that the State plan is the commenter is suggesting that the
EPA had issued a SIP Call finding that not working or does not meet CAA entire SIP should be re-done based on
the SIPs of 12 states were inadequate to requirements is needed as part of this more current modeling and more up-to-
meet the ozone NAAQS and in its SIP action. Commenters’ comments date information. On the contrary, the
Call rule, specified that the states regarding EPA’s SIP actions are not commenter seems satisfied with the
needed to submit SIPs that included the relevant for this rulemaking. EPA-approved emission limitations in
California Low Emission Vehicle (b) Comment (ExxonMobil): Even the SIP,4 which were based on the very
Program. In this matter, we are when the EPA has statutory authority modeling that the commenter now
promulgating a FIP, not issuing a SIP for a particular rule, its technical claims is unreliable.
Call. We are not directing any action by decisions about the level of pollutant (c) Comment (ExxonMobil): Citing
the State. Thus, the Commonwealth of reduction needed to comply with the Hall v. United States Environmental
Virginia case is not relevant to our FIP. CAA and the control strategies Protection Agency, 273 F.3d 1146, 1159
Bethlehem Steel is also not relevant to necessary to meet the level of pollutant (9th Cir. 2001), the commenter states
our FIP action. In that case, the 7th reduction must be rational. Courts that in acting on a SIP, the test EPA
Circuit held that it was improper for ‘‘confronted with important and applies is to ‘‘measure the existing level
EPA to partially approve an Indiana SIP seemingly plausible objections going to of pollution, compare it with the
revision so as to render it more stringent the heart of a key technical national standards, and determine the
than the State intended. We are determination * * * ’’ will not presume effect on this comparison of specified
promulgating a FIP in this action, not that EPA would never behave emission modifications.’’ The
acting on a SIP; thus, Bethlehem Steel irrationally. South Terminal commenter argues that in the FIP
does not apply. As we note elsewhere, Corporation v. Environmental proposal, EPA did not correctly identify
once we disapprove a SIP, we are Protection Agency, 504 F.2d 646, 665 the existing level of pollution and
required to promulgate a FIP, and in (1st Cir. 1974). In South Terminal ignored the substantial evidence of
promulgating the FIP, we stand in the Corporation, various interested parties permanently reduced SO2 emissions
state’s shoes. See section 110(c) of the challenged EPA’s FIP on technical and levels in the Billings/Laurel area.
CAA, 42 U.S.C. 7410(c); Central Arizona grounds. Id. at 662–66. The court held The commenter also argues that EPA’s
Water Conservation District v. EPA, 990 that EPA failed to adequately support its authority is limited by its mandate
F.2d 1531, 1541 (9th Cir. 1993). decision to promulgate the rules under the CAA to ensure attainment and
(g) Comment (MSCC): The commenter contained in the FIP and remanded the maintenance of the NAAQS as well as
argues that the cases EPA cited in the case to EPA to develop the record. Id. the CAA’s other general requirements.
preamble to the proposed Billings/ at 666. The court questioned EPA’s Response: See responses to comments
Laurel FIP, regarding its FIP authority, position in light of contradictory II.A.1.(a), II.A.2(b), and II.E.1.(e) and (g).
do not speak to the central question— modeling and data, concluding that ‘‘it Also, the Hall case involved a challenge
‘‘When and on what authority may the is not clear whether or not the ambient to EPA’s approval of a SIP revision for
EPA undertake the draconian act of air at Logan meets, or will without Clark County, Nevada, and EPA’s
displacing a state’s implementation controls by mid-1975 will meet, the interpretation of section 110(l) of the
plan?’’ The commenter argues that the national primary standard.’’ Id. 664. CAA, which provides that EPA may not
question is particularly sensitive in this Similarly, in the present FIP proposal, approve a SIP revision if it would
case because the State and the sources EPA has neither determined appropriate interfere with attainment or other
spent years negotiating the SIP. current modeling nor used currently applicable requirements of the CAA.
Response: As noted in response to available information. EPA asserted that its approval of the
comment II.B.1.(e), the CAA requires Response: The standards for judicial
Clark County SIP revision was
that we promulgate a FIP whenever we review of this rulemaking action are
consistent with section 110(l) because
disapprove a SIP, in whole or in part. contained in section 307(d)(9) of the
the revision did not relax the existing
While we are sensitive to the fact that CAA, 42 U.S.C. 7607(d)(9). We believe
SIP. The Court disagreed, holding that
the State and sources spent years the emission limitations and other
110(l) requires more—a determination
negotiating the SIP, that does not change requirements in this FIP are reasonable
that the specific revision, when
our obligation under the CAA. and that the situation in the cited case
considered in the context of the SIP
is not analogous.3 The commenter has
2. No Adequate Basis for FIP elements already in place, can meet the
not identified any modeling that
(a) Comment (MSCC, ExxonMobil): Act’s attainment requirements. Hall at
contradicts our attainment
Because EPA must find substantive 1152, 1159. It was in these
demonstration, which forms the basis
noncompliance with some provision of for the FIP’s emission limitations; nor circumstances that the Court expected
the Clean Air Act, specifically, failure to has the commenter shown that a EPA to determine the extent of pollution
attain NAAQS, and because that finding different model would result in reductions required and evaluate
of substantial inadequacy must be substantially different emission whether the reductions resulting from
clearly stated, the present FIP decision limitations. Our responses pertaining to the revision would be sufficient to attain
must fall. It is inadequate on both model selection and input data are the NAAQS.
counts. EPA has not provided any contained in section II.E., below. In its reference to Hall, the commenter
evidence that the State plan is not Further, we note that it does not appear appears to be conflating two disparate
working. concepts. The Hall Court was
Response: See our response to 3 In South Terminal Corporation, EPA had addressing EPA’s action on a SIP
comment II.B.1.(e), above. The evidence determined emissions reductions needed to achieve revision and indicating that EPA was
supporting EPA’s determinations
the ozone and carbon dioxide NAAQS based on not adequately evaluating whether Clark
monitored values that the Court found highly County’s rule change would interfere
regarding the adequacy of Montana’s questionable (petitioners claimed the ozone monitor
SIP is contained in the record for those was defective). South Terminal Corporation, 504
4 Among other things, the commenter asserts that
rulemaking actions, and need not be F.2d 646, 662 (1974). The commenter seems to
suggest that the Court rejected EPA’s modeling the state SIP requirements are adequate to protect
repeated here. EPA’s disapproval of the approach, but in fact, the Court was satisfied with the NAAQS. See reference document YYYY, page
SIP triggered the obligation for a FIP. No the rollback modeling that EPA used. Id. 27.
with attainment and other CAA State could adopt SIP limits that which include attainment and
requirements. The Court was not correlate to refinery consent decree maintenance of the SO2 NAAQS. Using
establishing a standard for a FIP or limits.6 If the State were to submit such ISC, the same model the State used to
indicating that EPA was requiring more a revision, we would evaluate the set the commenter’s emission limits in
than necessary for the area, which revision according to the Act, our the SIP, we have determined emission
seems to be what the commenter is regulations, and the relevant cases. levels consistent with attainment and
suggesting in the case of the Billings/ (d) Comment (ExxonMobil): EPA’s established corresponding emission
Laurel FIP. As we explain in greater proposal imposes costly technology limits on the flares, MSCC’s main stack,
depth elsewhere in this notice, we are requirements not rationally designed to and other emission units, whose
not starting from scratch with our FIP. achieving their stated objectives. While emission limits we disapproved in our
Instead, we are working within the EPA has authority to impose an SIP action. While the authority to
framework of the existing Billings/ emission limitation, the emission require monitoring, recordkeeping, and
Laurel SIP to fill the gaps resulting from limitation must be necessary to attain reporting requirements can be inferred
our partial and limited disapproval of NAAQS. City of Santa Rosa v. EPA, 534 from CAA sections 110(a)(2)(A) and (C),
discrete SIP elements. In this unique F.2d 150, 155 (9th Cir. 1976), vacated on section 110(a)(2)(F) of the Act
circumstance, where only discrete other grounds, 429 U.S. 990 (1976). The specifically indicates that the EPA
elements of the SIP were deficient, the EPA derived its authority in City of Administrator may prescribe the
CAA does not require us to reevaluate Santa Rosa from its statutory mandate installation, maintenance, and
or replace the entire SIP or the basic to ensure compliance with NAAQS and replacement of monitoring equipment
modeling approach upon which it was the fact that no alternative to its by stationary sources, as well as
based. Nothing in the CAA requires EPA proposal was adequate to ensure reporting requirements. Our
to reject an entire SIP when only certain compliance with NAAQS. It is clear that requirement for the refineries and MSCC
elements within it are not approvable, Montana’s existing SIP, supplemented to install monitoring equipment to
and doing so, where that is not as it is by further state and federally measure flare gas flow and
necessary to address a discrete enforceable consent decrees are a more concentrations is consistent with this
deficiency, would be inconsistent with than adequate alternative. authority and is rationally related to the
the basic scheme of cooperative Response: The cited case actually goals of the FIP, i.e., to ensure
federalism embodied in the CAA. stands for the proposition that EPA’s attainment and maintenance of the SO2
Nor are we required as part of this FIP authority to adopt measures to meet the NAAQS. We do not believe estimating
to revisit our SIP Call or the bases for NAAQS is expansive. EPA adopted a flare emissions or emissions from other
our SIP disapproval. Our task is to fix FIP provision that would have required units is a sufficient substitute for real-
the portions of the SIP that were a substantial reduction (up to 100%) in time monitoring for purposes of this
deficient. It is reasonable to continue to the supply of gasoline to major
FIP; estimation is not an equally
treat as valid the factors we found metropolitan areas in California,
effective technique.
adequate to support the portions of the including Los Angeles. Even the EPA
acknowledged that the rule would cause The commenter argues that the
SIP we approved, and augment and/or
severe social and economic disruption, existing SIP and the State and federally
replace those factors that we found
and the EPA Administrator at the time enforceable consent decrees are a more
inadequate. In fact, based on the holding
publicly advocated amendments to the than adequate alternative to our FIP
in Train v. NRDC, 421 U.S. 57 (1975),
CAA to provide relief from EPA’s own requirements. This comment ignores the
recited by this commenter and others, it
FIP rule. Nonetheless, the Court held fact that we disapproved portions of the
would be inappropriate for EPA to now
that economic and social disruption are SIP as not meeting the CAA’s
reject or replace the portions of the SIP
that we approved as meeting the CAA’s not cognizable if (1) a measure is requirements. Elsewhere we explain
requirements, because to do so would be necessary to attain the NAAQS; (2) there that the consent decree provisions are
to intrude on the State’s authority under is no statutory limitation on EPA’s not sufficient to meet the CAA’s
the CAA to establish the mix of controls authority to adopt the measure; and (3) requirements under section 110 related
for the area.5 The State, of course, there are no equally effective, less to attainment and maintenance of the
remains free to submit a SIP revision burdensome alternatives. City of Santa NAAQS. See, e.g., sections II.A.2.(b),
that reflects a different mix of controls Rosa at 151–154. II.D.4., and II.E.1.(e).
across all the sources. This would be the The measures EPA is promulgating in (e) Comment (MSCC): EPA’s failure to
mechanism, for example, whereby the this FIP are in no way comparable to the issue the FIP within the CAA’s two-year
reduction in gasoline supply at issue in deadline is important in this case. As a
5 To the extent the commenter is arguing that we the City of Santa Rosa case. Our FIP is result of EPA’s delay, EPA should have
may do no more in this FIP than appears minimally narrowly tailored to fill the gaps in the to consider the cleanup of emissions
necessary to attain the NAAQS, we reject that Billings/Laurel SIP. Section 110(c) that has occurred and significant
notion as well. See, e.g., Central Arizona Water
Conservation District v. EPA, 990 F.2d 1531, 1541 requires us to promulgate the FIP. There changes in modeling technology.
(9th Cir. 1993) (EPA ‘‘stands in the shoes of the is no statutory limitation on our Response: We regret that it has taken
defaulting State, and all of the rights and duties that authority to adopt the measures we are this long to issue the FIP. We disagree
would otherwise fall to the State accrue instead to adopting. On the contrary, section that missing the two-year deadline
EPA.’’) Under the CAA, states are not restricted to
barely meeting the NAAQS. In fact, the opposite is 110(a)(2)(A) of the Act requires obviates our duty or the need for the
true—states may exceed minimum requirements. enforceable emission limitations as FIP. The State has not submitted a SIP
See CAA section 116, 42 U.S.C. 7416. In any event, necessary or appropriate to meet the revision correcting the portions of the
our modeled attainment demonstration resulted in SIP that we disapproved, despite the
projected values just at the 24-hour SO2 NAAQS
applicable requirements of the Act,
(365 µg/m3) and just below the 3-hour SO2 NAAQS passage of time. Regarding the argument
(1291.5 µg/m3). However, we think we had 6 As we allude to in sections II.A.2.(b), II.D.4., and that we should have considered the
discretion to adopt limits (to replace those we II.E.1.(e), the consent decree limits would need to reduction in emissions since we
disapproved) consistent with modeled ambient be translated into limits that support an attainment
concentrations further below the NAAQS, if we had demonstration for the SO2 NAAQS. In sections
disapproved the SIP, see our responses
felt a larger margin of safety was justified to ensure II.A.2.(b) and II.D.4., we identify some of our to comments in section II.A. In section
attainment and maintenance. concerns with the consent decree limits. II.E, we respond to comments arguing
that we should have used newer Ultrasonic Flow Meter,’’ J.W. Smalling, of the instrument or 5% of the measured
modeling technology. L.D. Brawsell, L.C. Lynnwoth and D. flow. In the rule, we have clarified that
Russel Wallace, Proceedings Thirty- ‘‘accuracy’’ of the instrument is the
C. Flare Monitoring
Ninth Annual Symposium on accuracy of the measured flow and not
1. Flare Flow Monitoring Instrumentation for the Process the ‘‘full-scale range’’ of the instrument.
(a) Comment (MSCC): The core Industries, 1984, the authors reported The commenter also suggests some
‘‘initially, a modest objective was changes to the rule. Apart from adding
flowmeter technology application for
established to develop an ultrasonic a separate accuracy range for the
flare systems seems to be an established
flow switch capable of detecting leaks in velocity range of 0.1 to 1 fps and
technology, with thousands of
flare lines corresponding to flow clarifying that accuracy is based on the
installations completed around the
velocity on the orders of 0.3 ms/ (1 ft/ measured flow, we are not making any
world on other types of gas and liquid
s). As testing continued, however, it additional changes to this aspect of the
streams. However, none was identified
became apparent that the equipment rule. We explain our reasoning in the
that is following the precise
could measure flows below 0.03 m/s response to this comment II.C.1.(a) and
specifications of the FIP proposal.
(0.1ft/s) and up to at least 6 m/s (20 in the responses to comments II.C.1.(b)–
Installation and operation of a flow
ft/s) in flare stacks * * *’’ (see reference (d), below.
meter in flare gas service at MSCC are (b) Comment (ExxonMobil, WSPA):
document KKKKK). See also reference
probably achievable today, but not at Manufacturers of flow monitoring
document OO, ‘‘the DigitalFlowGF868
the flow range below 1 fps, and not with instrumentation publish impressive
meter achieves rangeability of 2750 to 1.
conventional QA/QC procedures. Flow performance specifications regarding
It measures velocities from 0.1 to 275
monitors have a difficult time velocity measurement range and
ft/s (0.03 to 85 m/s) in both directions,
measuring or reliably detecting low flow accuracy, but often manufacturers’
in steady or rapidly changing flow, in
velocities (under approximately 1.0 fps) claims are not actually achieved in
pipes from 3 in. to 120 in. (76 mm to
without false positives or false 3 m) in diameter.’’ practice over the long term. To achieve
negatives. EPA should revise the Additionally, the BAAQMD (see a high level of measurement
proposed rule that currently indicates: reference document LL) and SCAQMD performance in the field requires
‘‘[t]he minimum detectable velocity of the (see reference document CCC) require adequate lengths of straight flare header
flow monitoring device(s) shall be 0.1 feet flow meters on flares. BAAQMD pipe upstream and downstream of the
per second (fps). The flow monitoring requires that the minimum detectable monitor, the absence of flow
device(s) shall continuously measure the velocity shall be 0.1 fps and the disturbances, etc. Where these criteria
range of flow rates corresponding to cannot be met, the advertised or
SCAQMD requires monitors with a
velocities from 0.5 to 275 fps and have a
velocity range of 0.1 to 250 fps. Based predicted performance of the flow
manufacturer’s specified accuracy of ±5%
over the range of 1 to 275 fps. on conversations with the BAAQMD, it monitoring system may not be fully
The revised rule should read ‘‘[t]he appears that the refineries in the Bay realized in practice. MSCC claimed that
minimum resolution of the flow monitoring Area have installed flow meters meeting significant piping modifications and
device(s) shall be 0.1 feet per second (fps) the requirements of the rule (see possible flare relocation would be
when measuring flow rates above 1.0 fps. The reference document OOOOO). required to provide such runs at
device(s) shall continuously measure the Based on the above, we conclude that accessible locations. CHS Inc. asserted
range of flow rates corresponding to flow meters are available that can that it is likely that the CHS refinery
velocities from 1.0 to 275 fps and have a measure in the velocity range below 1.0 flare header will not have adequate
manufacturer’s specified accuracy of ±5% fps, and other regulatory authorities are distances of undisturbed piping for
over the range of that range.’’
requiring such flow meters with ideal installation. In this case, either
The rule should also clarify if success. major, costly piping modification will
‘‘accuracy’’ is intended to be 5% of the The commenter also claims that be required or the accuracy criteria will
full-scale range of the instrument (13.7 installation and operation of a flow not be achievable.
fps is 5% of 275 fps), or if this is meter are probably not achievable with Response: The commenters are correct
intended to be 5% of the measured flow, conventional QA/QC procedures. The that piping modifications may be
which would be 0.05 fps at a flow of 1 QA/QC procedures are discussed below appropriate to optimize the
fps, and would clearly be non- in response to comment II.C.1.(d). measurements. Each flare system will
achievable with a resolution of 0.1 fps. The commenter argues that flow have unique flow measurement location
Response: EPA proposed the monitors have a difficult time issues and will have to be addressed on
volumetric flow monitoring measuring or reliably detecting low flow a case-by-case basis. Sources may need
specifications based on what we saw velocities (under approximately 1.0 fps) to work with the flow monitor
was achievable in vendor literature (see without false positives or false manufacturer and flow testers to assure
reference documents NN and OO) and negatives. As indicated in the response that the monitors meet the FIP’s
what was being required by regulation to comment II.C.1.(b) below, there are specifications for accuracy and
in the Bay Area Air Quality approaches available for improving representativeness and manufacturer’s
Management District (BAAQMD) (see measurement accuracy in the 0.1 to 1.0 requirements for assuring ongoing
reference document LL) and South fps range. In addition, as the response equipment performance.
Coast Air Quality Management District to comment II.C.1.(b) indicates, in the In addition to making piping
(SCAQMD) (see reference document final FIP we are specifying a separate modifications (e.g. flow straighteners),
CCC). accuracy range for the velocity range of other approaches are available to
The commenter asserts that 0.1 to 1 fps. Finally, we describe how improve the measurement accuracy in
installation and operation of a flow we are addressing the false positive and the 0.1 to 1.0 fps range. Among the
meter at the flow range below 1 fps are false negative flows in response to approaches are the use of additional
not achievable. However, various comment II.C.1.(c). monitoring paths, monitoring paths of
sources indicate that ultrasonic flow The commenter asked that the rule longer length, and unconventional
meters can measure in the range of 0.1 clarify if ‘‘accuracy’’ of the instrument is monitor configurations and path
to 1 fps. For example, in ‘‘Flare Gas intended to be 5% of the full-scale range locations. Another approach involves
the use of Computer Fluid Dynamics the ability to use other secondary means 2. Mechanical Inspection of Flowmeter
(CFD) for the existing piping. CFD to determine whether flow is reaching Transducers—the purpose is to visually
analysis has been used to provide the flare when the flow monitor verify the integrity of the flare gas flowmeter
correction factors for a series of indicates low flow. If the secondary transducers and to clean any accumulated
debris from the transducer faces;
velocities across the range of flow device indicates that no flow is going to 3. Zero Flow Verification—the purpose is
velocities. For example, these factors the flare, yet the continuous flow to evaluate the operation of the transducer
have been used to correct flow monitor is indicating flow, the pair in the flare gas process (the integrity of
measurement data for disturbances presumption will be that no flow is the original process transducers is tested in
caused by upstream pipe irregularities. going to the flare. We have revised the a controlled environment);
These approaches are discussed in ‘‘A final rule to allow the use of flare water 4. Input/Output Verification—the purpose
Total Approach to Flare Gas Flow seal monitoring devices to determine is to verify the calibration of the analog I/O
Measurement for Environmental whether there is flow going to the flare, of the flare gas flowmeter;
Compliance,’’ Gordon Mackie, Jed in addition to the continuous flow 5. Electronic Flow Simulation—the
Matson and Mike Scelzo, Institute of monitoring device. See response to purpose is to demonstrate the operation of
the flare gas flowmeter over the full
Measurement and Control— comment II.F.1.(a) regarding the measurement range of the instrument; and
Environmental Conference 2006. (See comment seeking a stakeholder process. 6. Flowmeter System Reinstallation and
reference document LLLLL.) (See also (d) Comment (ExxonMobil, WSPA): A Test—the purpose is to verify that all
Note to Billings/Laurel SO2 FIP File limitation of flare gas monitoring mechanical systems were properly aligned.
regarding conversations with GE systems is the inability to provide for an
It should also be noted that since
Sensing (reference document independent ‘‘in situ’’ verification of
ultrasonic flow monitors do not contain
MMMMM)). accuracy. For example, there is no
Finally, to address concerns regarding any moving parts, their performance is
practical way to vary the flare gas flow
the measurement accuracy in the 0.1 to not expected to deteriorate over time.
that the monitor sees, and no practical
1.0 fps range, we are revising the rule to One ultrasonic flow monitoring vendor
way to utilize a reference method.
indicate that the flow monitor must provided information on the reliability
Consequently, the calibration of a
have a manufacturer’s specified and availability of the transducers
monitor is performed electronically, and
accuracy of ± 20% over the range 0.1 to (sensors in the flare that transmit and
the demonstration of accuracy is based
1 fps. Based on conversations with a receive the ultrasound) they have
on that calibration method. MSCC
vendor, we believe this is achievable. installed. The information indicates that
asserted that the proposed FIP does not
The vendor indicated that they have the 3,998 transducers installed between
provide adequate guidance to allow
provided methodologies for sources to first quarter 2005 and first quarter 2007
development of an acceptable QA/QC
meet the SCAQMD rule, which also had a reliability percentage of 94.32%
system for routine calibration or daily
requires 20% accuracy in the 0.1 to 1.0 and an availability percentage of
checks of the system. Without clear
fps range. Methodologies include a 99.96%. (See reference documents
guidance, it is not possible to specify a
second interrogation path or MMMMM and XXXXXX.) (See also
system for a systems integrator (DAS/
straightening of pipe. (See reference reference document LLLLL, ‘‘A Total
reporting) or an end-user to design or
document MMMMM.) Approach to Flare Gas Flow
build a system to accomplish these
(c) Comment (ExxonMobil, WSPA, Measurement for Environmental
checks.
NPRA, MSCC): Consistently achieving Response: Since refinery flares Compliance,’’ Gordon Mackie, Jed
low flow detection limits can be very contain highly variable flows and highly Matson and Mike Scelzo, GE Sensing,
difficult. Spurious signal, resulting in combustible material, in situ Institute of Measurement and Control,
‘‘eddy’’ currents and back-and-forth verification of flow measurement Environmental Conference 2006, and
flows in the flare header, can easily accuracy is difficult. For that reason, the reference document NNNNN, April 5,
limit the detection and accuracy of low performance specifications in the FIP 2007, email from Jed Matson, GE
flow readings. Furthermore, sometimes rely in large part on procedures Sensing, to Laurie Ostrand, EPA,
a flow monitor will show an indication developed by the ultrasonic flow containing flare gas flow meter
of flow even though water seals ahead monitor manufacturers 7 for procedures.
of the flare stack remain intact (i.e., (e) Comment (COPC): ConocoPhillips
commissioning monitors to assure the
there is not flow to the flares). Other asserts it would need to replace a GE
monitors will meet performance
regulations in other jurisdictions allow Panametrics flare flow monitor that is
specifications on an ongoing basis.
the sources other means to positively well-suited to the variable flow
Manufacturers have established
determine when the flare is not conditions it experiences, but does not
procedures for conducting annual or
operating (e.g., flare on/off monitoring conform precisely to the proposed
more frequent verifications of the
device, pressure of water seal). specifications. It is difficult to quantify
performance of installed flow monitors
ExxonMobil recommends that similar what additional benefit this change
as well as for the initial installation and
language be considered by the would provide although the cost is
performance verification (see reference
stakeholder process for inclusion in the significant and quantifiable. The benefit
document NNNNN). Based on
EPA’s proposed FIP, and thereby evaluation is further clouded because of
manufacturer established procedures
remove the uncertainty of low flow the relatively recent installation of the
(Id.), we expect that the annual
reading. MSCC claimed that the EPA Flare Gas Recovery Unit (FGRU). There
verification procedures will address
proposed FIP language should be is no flow to measure in the flare header
elements such as:
revised to allow flare operations to be when the FGRU is operating. The FGRU
monitored by other means, and to 1. Verification of the Flowmeter with operates on a full-time basis, with the
disregard low flow readings when the Reference Transducers—the purpose is to exception of nominal periods of
evaluate all flowmeter subsystems with
flare is not operating to eliminate falsely factory-certified ultrasonic transducers; malfunction or maintenance.
reported SO2 emissions, when in fact Response: As indicated above, each
there are none. 7 Ultrasonic flow monitors will most likely be the source will have unique issues that will
Response: We agree that it is monitors installed to meet the FIP’s flow have to be addressed on a case-by-case
appropriate to include in the regulation monitoring performance specifications. basis.
We understand that ConocoPhillips concentration measurements of sulfur present in the gas stream to the flare. In
has a FGRU and ExxonMobil will be are also made on a ‘‘wet’’ basis. cases when the total sulfur analyzer is
installing one. We do not agree that a Response: The commenter is correct. not working or where the concentration
source with a FGRU should be We are revising the regulatory text to of the total sulfur exceeds the range of
exempted from monitoring flow to the read: ‘‘The flare gas stream volumetric the monitor, methods established in the
flare. We still believe it is reasonable to flow rate shall be measured on an actual flare monitoring plan required by the
include this requirement to gain an wet basis, converted to Standard FIP shall be used to determine total
accurate picture of occasions when flow Conditions, and reported in SCFH.’’ sulfur concentrations, which shall then
is going to the flare. We note that other (g) Comment (several commenters): be used to calculate SO2 emissions. In
areas that have required refinery flare Several commenters express a general quarterly reports, sources shall indicate
monitoring (SCAQMD and the concern that the technology will not be when these other methods are used.
BAAQMD) have not eliminated the flare able to meet the performance (b) Comment (ExxonMobil, WSPA):
monitoring requirements at sources with specifications. SCAQMD Rule 1118 had an important
FGRUs. (See Note to Billings/Laurel SO2 Response: See responses to comments provision requiring an analyzer pilot
FIP File regarding conversations with II.C.1.(a)–(c), above. project, and one Los Angeles area
BAAQMD, reference document (h) Comment (YVAS): YVAS concurs refiner is currently engaged with a
OOOOO.) However, as indicated below, with the proposed volumetric flow sulfur analyzer demonstration project. It
we are providing sources other means to monitoring requirements. is conceivable that the pilot project
determine total sulfur concentrations in Response: We acknowledge receipt of could result in the conclusion that the
the gas stream to the flare. the supportive comment. analyzer being evaluated could not
Additionally, we note that the 2. Flare Total Sulfur Analyzers provide sufficient accuracy, that the
ConocoPhillips refinery in Rodeo, system was not maintainable, or that
(a) Comment (ExxonMobil, WSPA, there were other problems.
California has installed flare flow meters COPC): SCAQMD staff was not able to
and that the refinery also has a flare gas Response: On June 1, 2007, the
identify a single commercial sulfur SCAQMD presented to its Governing
recovery system. The ConocoPhillips analyzer in service on a refinery flare
San Francisco Refinery’s July 2007 Flare Board an ‘‘Implementation Status
system. It is unreasonable for EPA to Report for 2006 for Rule 1118—Control
Minimization Plan (FMP), pages 3–7, conclude that sulfur analyzer
indicates that flow meters have been of Emissions from Refinery Flares.’’
technology is either ‘‘available’’ or Agenda No. 27 discusses the total sulfur
installed on the Main and MP30 flares ‘‘reliable.’’ MSCC was not able to
per the BAAQMD Regulation 12–11– (TS) analyzer pilot project at the BP
identify any installations where flare gas refinery in Carson and indicates:
501. EPA’s Billings/Laurel FIP contains monitoring was, in fact, covering a range
flare flow monitoring specifications very from 0–100% sulfur. The TS pilot project is in the final step
similar to the specifications in prior to certification of the analyzer.
Response: EPA has identified two Although several adjustments and redesign of
BAAQMD Regulation 12–11–501. The sources where analyzers are on lines sampling equipment were required; [sic]
July 2007 FMP indicates ‘‘The leading to the refinery flare. preliminary results have demonstrated the
installation of the flow meters provides Specifically, the Tesoro refinery in the feasibility of measuring total sulfur emissions
for enhanced recognition of flaring Bay Area, California, has two Thermo from vent gases directed to flares. Based on
events. The flow meters help reduce Electron Tracker XP continuous H2S these results, two refineries have already
flaring by providing an accurate means analyzers. The Tesoro analyzers are dual placed purchase orders for their TS
to measure and provide indication as to range instruments, 0–1% and 0–5% (see analyzers.
when flaring is occurring. The flow reference document OOOOO). In the May 15, 2007, ‘‘Implementation
meters are especially useful for small Additionally, the Shell refinery in Puget Status Report for 2006 for Rule 1118—
flaring events which may not be Sound, Washington, uses an analyzer Control of Emissions From Refinery
detectable from visual flare stack that thermally oxidizes total sulfur to Flares,’’ attached to Agenda No. 27, the
monitoring only. The meters help to SO2 and then measures the SO2. The SCAQMD concludes:
track and record all instances of flaring analyzer can measure up to 40,000 ppm Although they are behind schedule to
as well as giving Unit Operators of SO2 (see reference document comply with the July 1, 2007 monitoring
immediate indication that flaring is QQQQQ). Finally, as indicated in the requirements, the pilot projects are moving
occurring so that they can take action to response to comment II.C.2.(b) below, ahead convincingly towards completion by
reduce flaring.’’ (See reference the SCAQMD recently reported on a the end of 2007. As the rule is forcing new
document PPPPP.) pilot project study, testing a total sulfur technologies for flare emission reporting,
analyzer vendors have responded to the
(f) Comment (MSCC): The proposed analyzer at the BP Carson facility in challenge and several options are now
40 CFR 52.1392(h)(2)(iii) appears to be southern California, and indicated that available, such as calorimeters, gas
in error. The rule indicates that ‘‘The the ‘‘preliminary results have chromatographs, mass spectrometers and
flare gas stream volumetric flow rate demonstrated the feasibility of Pulsed UV Fluorescence analyzers, for
shall be measured on an actual wet basis measuring total sulfur emissions from continuously measuring HHV [higher heating
in SCFH.’’ Actual wet basis would be vent gases directed to flares.’’ value] and TS. Therefore, staff expects full
abbreviated as ACFH. SCFH means The proposed FIP did not specifically implementation of the continuous
standard cubic feet per hour, meaning require that an analyzer be capable of monitoring provisions of the rule once the
pilot projects are complete. Since the
that the data has been corrected to measuring in the range from 1–100% refineries could not meet the monitoring
standard temperature and pressure. The sulfur, although the preamble implied requirements by July 1, 2007, the refineries
SCFH could be replaced with ACFH. and the record reported conversations petitioned and were granted variances in late
Alternately, the term ‘‘actual’’ could be with vendors indicating that analyzers April 2007 by the AQMD Hearing Board to
removed from the section, leaving ‘‘wet could measure in the range from 1– install and operate their flare monitoring
basis in SCFH.’’ SCFH (corrected for 100% sulfur. We are clarifying the final systems over the next two years.
temperature and pressure) can also be FIP to indicate that the total sulfur See reference document RRRRR.
used to compute a mass emission rate of analyzers should measure in the range Based on the above information, the
sulfur dioxide, provided that any of concentrations that are normally total sulfur pilot project did not
conclude that the analyzer being SCAQMD in fact resulted in reduced Response: See response to comment
evaluated could not provide sufficient flaring (72 FR 27178, at 27195) (see II.C.2.(a), above.
accuracy, that the system was not reference document SSSSS). (g) Comment (MSCC): Since H2S is
maintainable, or that there were other (e) Comment (ExxonMobil, WSPA): believed to be the principal
insurmountable problems. Cost of installing total sulfur analyzers (overwhelming) sulfur component of
(c) Comment (ExxonMobil): EPA and should be further evaluated given that candidate flares, further consideration is
industry need more time to review the the analyzers themselves do not provide warranted as to whether the ‘‘total’’
SCAQMD pilot project test results and an air quality benefit. Costs of total sulfur component is the appropriate
conclusions as they become available sulfur analyzer pilot project in the methodology, given the clear lack of
over the next few months and to South Coast area expected to be in the existing equipment for the full potential
determine if the technology that was range of 3 to 5 million dollars. range of concentrations of flare gases,
tested is technically viable and whether Response: See response to comment and the complexity involved in
or not a more cost effective alternative II.C.2.(d), above. Additionally, the cost continuously converting a variable
technology may be available. MSCC of the South Coast pilot project was mixture into a single component such as
recommends that the implementation of higher than expected because it was a SO2 or H2S. EPA should evaluate
total sulfur monitoring on the flares be pilot study and because some whether there is a real, necessary, and
delayed at least until the full results difficulties were encountered during the significant need to require total sulfur
from the long-term program in study. (See also note to Billings/Laurel analysis instead of allowing a somewhat
California are available, and the SO2 FIP File regarding conversations simpler H2S analysis of flare gases.
capability of the market to supply and with SCAQMD, reference document Response: The commenter has not
support such systems in severe weather TTTTT.) provided any technical analyses
locations such as Montana is Also, in its ‘‘Implementation Status supporting the notion that H2S is the
demonstrated. At that point EPA should Report for 2006 for Rule 1118—Control overwhelming component of the total
revise and then issue the final rule, after of Emissions From Refinery Flares,’’ sulfur in the gas stream to its flares or
full stakeholder involvement in the May 15, 2007, the SCAQMD reported other flares in the area. EPA reported in
process and full consideration of that refineries involved in the pilot the May 14, 2007, proposed new source
realistically available options. projects reported that monitoring costs performance standards (NSPS) for
Response: See responses to comments were estimated to be about 2 to 4.7 Subpart Ja (72 FR 27178, at 27194) (see
II.C.2.(a) and (b), above. Also, as noted million dollars per flare. After looking at reference document SSSSS) that ‘‘based
in response to comment II.C.3.(a), the breakdown of the costs, SCAQMD on available data, we understand that a
below, EPA is revising the proposed FIP staff concluded that the total sulfur and significant portion of the sulfur in fuel
to allow other methods to determine higher heating value analyzer costs were gas from coking units is in the form of
total sulfur concentration in the gas comparable to staff’s original estimates. methyl mercaptan and other reduced
stream to the flare. See response to However, the costs to design and build sulfur compounds. These compounds
comment II.F.1.(a) regarding the request the monitoring system were will also be converted to SO2 in the fuel
for a stakeholder process. significantly different. Research and gas combustion unit, which means the
(d) Comment (ExxonMobil): development (R&D), engineering, labor/ SO2 emissions will be higher than the
Recognizing that these total sulfur oversight, piping/electrical, analyzer amount predicted when H2S is the only
analyzer systems do not, by themselves, shelters, and contingencies stated by the sulfur-containing compound in the fuel
provide any air quality benefit, and refineries represented approximately 75 gas.’’ See also the response to comment
considering that there are alternatives to to 85 percent of the flare monitoring II.C.2.(a), above. Therefore, in the FIP
continuous analyzers (e.g., individual system cost. (See reference document we are still requiring that the gas stream
grab samples, etc.), ExxonMobil submits RRRRR.) to the flare be analyzed for total sulfur.
that the proposed requirement to install SCAQMD also indicated that in a (h) Comment (ConocoPhillips, MSCC):
continuous analyzers requires further related development, ExxonMobil In a typical CEMS installation, the
evaluation in the stakeholder process. informed staff in January 2007 that analyzers are subjected to frequent
Response: As discussed under ExxonMobil was taking a different testing with gases intended to represent
response to comment II.C.1.(a), below, approach and was going to use a a ‘‘zero’’ condition and a ‘‘span’’
our final FIP allows other methods to different technology, namely, gas condition which is specified as a
determine total sulfur concentration in chromatography (GC) for both the TS significant percent of full scale of the
the gas stream going to the flare, and the HHV analyzer; the estimated analyzer. ‘‘Total Sulfur’’ analyzers,
including grab or integrated sampling cost given to SCAQMD staff was 1 to 2 operating over a wide range of
methods. This should address the million dollars. ExxonMobil advised concentrations, present some special
commenter’s concerns. However, we SCAQMD staff that similar instruments concerns for span gases. If the proposed
note that whether or not total sulfur had been used at ExxonMobil’s flares in FIP requires high concentration
analyzer systems provide any air quality Baytown, TX, and Chalmette, LA, for analyzers, it also needs to incorporate
benefit by themselves is immaterial; the monitoring H2S and the BTU content of protocols to establish calibration
FIP establishes emission limits to assure vent gases for compliance with EPA and standards for these analyzers.
that the SO2 NAAQS are attained and Texas Commission on Environmental ConocoPhillips indicates that flare gas
maintained and it is essential that the Quality (TCEQ) regulations. (Id.) sulfur concentrations can be highly
FIP include reliable mechanisms to (f) Comment (CHS Inc.): Analysis of variable, which makes the comparison
determine compliance with the limits. total sulfur in a flare system is required by the Relative Accuracy Test
See, e.g., CAA section 110(a)(2)(F), 42 challenging because of the wide range of Audit (RATA) difficult. The sulfur
U.S.C. 7410(a)(2)(F). Finally, as we sulfur concentrations possible as well as analyzer captures samples in a series of
noted in our May 14, 2007, proposal to the number of individual sulfur periodic discrete ‘‘grab’’ samples, to be
revise subpart J of the new source compounds potentially present. It is the averaged over the period of total sample
performance standards (NSPS), and to understanding of CHS that there is not time. Comparison sample techniques
adopt new subpart Ja, the requirement one commercial total sulfur analyzer in vary, but in general involve getting a
to monitor flare emissions in the service on a refinery flare. continuous sample over a period of
time, with the concentration averaged 77. All continuous or intermittent, quality control requirements makes it
over that time period. Depending on the routinely-generated refinery fuel gas streams reasonable for EPA to allow for AMPs
variability of the concentration over this that are routed to the flare header at Cenex similar to other EPA regulations. MSCC
time period, the average of the discrete shall be equipped with a CEMS as required indicated that it calculates and reports
by 40 CFR § 60.105(a)(4) or with a parametric
‘‘grab’’ samples has the potential to be monitoring system approved by EPA as an
the amount of SO2 emitted during each
different than the average of the alternative monitoring plan (‘‘AMP’’) under flaring event based on the recent
continuous RATA sample. When the 40 CFR § 60.13(i), at the combined juncture content, and estimated flow gas(es)
concentrations are numerically low, this prior to the flare. Cenex shall comply with flared, based on reasonable technical
difference is compounded and skews the reporting requirements of 40 CFR Part 60, judgment and indirect metering
the accuracy calculations. This poses a Subpart J, for the Refinery Flare. calculations. MSCC asserted that EPA
significant risk of failing the RATA We also note that the proposed NSPS has failed to show any significant errors
specifications, thereby voiding the Subpart Ja includes a total sulfur or omissions with these methods.
monitor data and imposing a Response: EPA is revising the
standard and CEMS requirements for
compliance issue (even if the difference proposed FIP to allow other methods to
fuel gas combustion devices, which are
is a few parts per million). determine total sulfur concentration in
defined to include flares. (See 72 FR
ConocoPhillips believes that this the gas stream going to the flare. The
27178 (May 14, 2007), reference
requirement is not technically valid for other methods allow sources to use grab
document SSSSS.)
the operations for which it is being or integrated sampling, followed by
(j) Comment (MSCC): MSCC is aware
proposed. sample analysis, to determine total
that it may be possible to use gas sulfur concentration of the gas stream
Response: As indicated in response to chromatography systems to attempt to
II.C.2.(b), above, the BP Pilot Project is going to the flare. These grab and
meet the proposed FIP requirements. integrated sampling methods are
nearing completion and expected to be Due to time constraints, they were not
a success. Also, see note to Billings/ currently allowed in the BAAQMD rule
able to investigate this subject (see reference document LL), and
Laurel SO2 FIP File regarding thoroughly.
conversations with SCAQMD (reference similar methods have been allowed by
Response: As indicated in response to the SCAQMD. Two of the refinery
document TTTTT). With respect to the II.C.2.(e), ExxonMobil reported to the
calibration of the analyzer, SCAQMD companies (ConocoPhillips and
SCAQMD that it is using gas ExxonMobil) in the Billings area also
indicated that there are several issues chromatography for its total sulfur and
that need to be addressed. Specifically, have refineries in the Bay Area and/or
higher heating value analyzers. the South Coast Area and should be
one needs to assure that (1) the correct ExxonMobil has advised SCAQMD staff
calibration gas is in the bottle, (2) the familiar with these manual methods.
that similar instruments have been used Specifically, we are revising the rule
sample lines do not absorb or desorb on its flares in Baytown, TX, and to indicate that the total sulfur
sulfur, (3) the probe is positioned Chalmette, LA, for monitoring H2S and concentration of the gas stream going to
appropriately, and (4) all flow testing or the BTU content of vent gases for the flare can be determined by: (1) A
other sample collection is correlated compliance with EPA and Texas total sulfur concentration monitoring
temporally with the analyzer Commission on Environmental Quality system as we proposed on July 12, 2006,
measurements to ensure representative (TCEQ) regulations. (See reference and including the changes we have
comparisons. document RRRRR.) Also, see note to identified here; or (2) grab sampling or
(i) Comment (ExxonMobil): EPA Billings/Laurel SO2 FIP File regarding integrated sampling.
recognized the impracticality of conversations with SCAQMD (reference If a source chooses to use the grab or
concentration monitoring for flares document TTTTT). integrated sampling methods, the
during the recent Consent Decree (k) Comment (several commenters): A requirement to obtain a grab or
negotiations. CEMS were deemed general concern is expressed that the integrated sample will be triggered if the
unnecessary and impractical for flares, technology is not there to meet velocity of the gas stream to the flare in
unless the flare was in continuous use. performance specifications. any consecutive 15-minute period
Response: The basis for the FIP is Response: See responses to above continuously exceeds 0.5 feet per
different than the consent decrees. The comments II.C.2.(a) and (b). second (fps) and shall continue until the
FIP assures attainment of the SO2 (l) Comment (YVAS): YVAS concurs flow rate of the gas stream to the flare
NAAQS, a health-based standard, and that total sulphur concentrations and in any consecutive 15-minute period is
the consent decrees assure that the new not just H2S be monitored. continuously 0.5 fps or less.
source performance standards (NSPS), Response: We acknowledge receipt of Additionally, the rule indicates that a
technology-based standards, are met. the comment and the support for our grab or integrated sample will not be
Because of these differences, we believe proposal. required if any water seal monitoring
it is appropriate to take a different device indicates that flow is not going
approach. 3. Miscellaneous Flare Monitoring
to the flare. See discussion in response
We disagree with the commenter’s Concerns
to comment II.C.1.(c). Under these
statement that ‘‘EPA recognized the (a) Comment (COPC, CHS Inc., conditions, if the water seal monitoring
impracticality of concentration MSCC): The proposed FIP should allow device indicates that there is no flow
monitoring for flares during the recent for Alternative Monitoring Plans (AMPs) going to the flare, yet the continuous
Consent Decree negotiations. CEMS to determine compliance. flow monitor indicates flow, the
were deemed unnecessary and ConocoPhillips argued that AMPs are presumption will be that no flow is
impractical for flares.’’ The CDs technically sound data gathering plans going to the flare.
required that compliance with 40 CFR that are developed based on site-specific For grab sampling, a sample shall be
60.104(a) be determined by several factors. These AMPs allow a facility to collected within 15 minutes after the
options, one of which was to install and comply based on equivalent but triggering conditions occur (see above),
operate a CEMS per 40 CFR supbart J customized criteria. CHS Inc. claimed and the sampling frequency, thereafter,
(e.g. see paragraph 77 of CHS Inc.’s CD, that uncertainty of the monitoring shall be one sample every 3 hours. For
reference document JJJJJ): capabilities and the quality assurance/ integrated sampling, a sample shall be
collected within 15 minutes after the sulfur analyzer systems within 180 days refinery fuel gas is used as a purge gas.
triggering conditions occur (see above), after receiving EPA approval of a Refinery fuel gas can have high sulfur
and the sampling frequency, thereafter, monitoring plan is a requirement that content. Because of the potential for SO2
shall consist of a minimum of 1 aliquot would simply be impossible to meet. emissions from the burning of pilot and
for each 15-minute period until the Response: Based on the comments purge gas, we believe it is necessary to
sample container is full, or until the end received, we have revised the FIP to account for these emissions and include
of a 3-hour period is reached, whichever allow 365 days, rather than 180 days, them when determining the total
comes sooner. Within 30 minutes after EPA approval of the flare emissions from the flare.
thereafter, a new sample container shall monitoring plan to install continuous
flow monitors and to begin determining We agree that the proposed FIP
be placed in service. For grab and
total sulfur concentrations on the gas implied that the pilot and purge gas
integrated sampling, sampling shall
stream to the flare. Based on should be monitored by the analyzers
continue until sampling is no longer
conversations with an ultrasonic flow on the flare line used to measure flow
required (see above).
Samples obtained by either grab or monitor manufacturer, BAAQMD, and and concentration of the gas stream to
integrated sampling shall be analyzed SCAQMD (see reference documents the flare. We are revising the FIP to
for total sulfur concentration using MMMMM, OOOOO, and TTTTT, require flow and H2S concentration
ASTM Method D4468–85 (Reapproved respectively), we believe this additional monitoring of the pilot and purge gas as
2000) ‘‘Standard Test Method for Total time is reasonable to install continuous one possible method to determine sulfur
Sulfur in Gaseous Fuels by flow monitors and total sulfur analyzers dioxide emissions from the burning of
Hydrogenolysis and Rateometric or to initiate grab or integrated such gas in the flare. However, the FIP
Colorimetry’’ (see reference document sampling. allows sources to forego monitoring if
MMMMMM); ASTM Method D5504–01 (c) Comment (MSCC, ExxonMobil): certain requirements are met. First, if
(Reapproved 2006) ‘‘Standard Test The FIP implies that pilot and purge gas facilities certify that only natural gas or
Method for Determination of Sulfur must be monitored. Pilot and purge gas an inert gas is used for the pilot and/or
Compounds in Natural Gas and Gaseous lines are separate from the main header purge gas, then the gas does not need to
Fuels by Gas Chromatography and vent gas lines. Monitoring these other be monitored. Second, if facilities can
Chemiluminescence’’ (reference relatively small gas flows to the flare is measure other parameters so that
document NNNNNN); or 40 CFR part a waste of effort and resources. The pilot volumetric flows, expressed in SCFH, of
60, Appendix A–5, Method 15A gas is usually a small natural gas stream pilot and purge gas can be calculated
‘‘Determination of Total Reduced Sulfur of low flow and essentially zero sulfur (based on the design and the
Emissions From the Sulfur Recovery content. The small purge gas line parameters), then the flows do not need
Plants in Petroleum Refineries.’’ Total usually is natural gas, refinery fuel gas, to be monitored. Third, if the H2S
sulfur concentration shall be reported as or inert gas such as carbon dioxide or concentration of the pilot or purge gas
H2S or SO2 in ppm. Proper QA/QC nitrogen, or mixtures of such gases with can be determined through other
procedures shall be used to assure that air or steam. In either case, the flow is methods, then the H2S concentration
the samples are obtained and analyzed not high and usually ExxonMobil does does not need to be monitored. Once
appropriately. not expect high sulfur content. These flow and H2S concentration of the pilot
We chose the trigger level for two two stream types (pilot gas, purge gas) and purge gas are determined, sources
reasons. First, the rule indicates that the cannot physically be mixed with the must then calculate the SO2 emissions
minimum detectable velocity of the flow main vent gas stream for measurement from the pilot and purge gas. The
monitoring device(s) shall be 0.1 fps and of flow and sulfur content by one set of calculated SO2 emissions will then be
the flow monitoring devices shall monitors, without defeating their added to the other SO2 emissions from
continuously measure the range of flows essential purposes of safety. Given the the flare to determine compliance with
corresponding to 0.5 to 275 fps. Since nature of the pilot gas and purge gas the flare SO2 emission limits. Also, we
0.5 fps is the minimum flow measure streams, it is not reasonable to require are revising the reporting requirements
required, it is a reasonable trigger level flow and sulfur monitors which meet to require sources to: (1) Certify in the
to ensure protectiveness. Second, flow the proposed FIP specs on these quarterly reports if pilot or purge gas is
monitoring software averages all the streams. Regulations from other areas not monitored because only natural gas
readings in a 15-minute timeframe and allow the flow and sulfur content of or an inert gas is used as the pilot and/
records/reports the average flow. Using pilot and purge gas to be estimated/ or purge gas; or (2) report flow and H2S
the minimum recorded/reported monitored by other devices or sampling concentration of the pilot and/or purge
timeframe is reasonable to ensure means. It is recommended that the
gas and the resultant SO2 emissions.
protectiveness. proposed FIP language be re-written to
With respect to using estimations, clearly exempt pilot gases and purge (d) Comment (MSCC): Flow and
technical judgment, and indirect line gases from the proposed FIP concentration monitoring would be
metering to calculate emissions from the monitoring requirements. Neither can costly and there is no justification for
flare, because this FIP is designed to reasonably be considered as a such costs and complexity given that
protect the NAAQS, we are choosing to significant source of sulfur dioxide. the area is in attainment for the NAAQS.
require real-time direct monitoring ExxonMobil asserted that EPA’s Response: See response to comments
methods to determine emissions. We do proposed FIP requirement for the II.C.2.(d) and II.C.3.(c), above.
not believe estimations, technical Billings/Laurel area is neither
(e) Comment (YVAS): YVAS concurs
judgments, and indirect metering are reasonable nor legally supportable.
adequate substitutes for real-time Response: In conversations with the that each source submit for EPA review
monitoring for purposes of the FIP. SCAQMD, we learned that in some a quality assurance and quality control
plan for each of the continuous
(b) Comment (ExxonMobil, WSPA, instances they had seen copious

COPC, CHS Inc., MSCC): The proposed emissions due to flare pilot and purge monitors.
requirement for a facility to install, gas (see reference document TTTTT). Response: We acknowledge receipt of
commission, and calibrate flow SCAQMD indicated, as do the the comment and the support for our
monitoring systems and continuous commenters above, that in some cases proposal.
D. Flare Limits that allows for an automatic exemption (l), 42 U.S.C. 7410(a) and (l); reference
for excess emission is prohibited.8 document RRR, September 20, 1999,
1. Concerns With Flare Emission Limit
(b) Comment (NEDA/CAP, MSCC, memorandum titled ‘‘State
(a) Comment (CHS Inc, MSCC): The ExxonMobil): The capriciousness of Implementation Plans: Policy Regarding
proposed flaring limit of 150 lbs SO2/3 EPA’s proposed FIP provision affecting Excess Emissions During Malfunctions,
hour period was used in the model to flaring is that EPA recognizes in the Startup, and Shutdown,’’ from Steven
represent routine flaring and proposed notice that sources likely will A. Herman and Robert Perciasepe, to
background SO2 concentrations. This be unable to comply with the Regional Administrators (hereafter
threshold was never intended to and did continuous flaring emission limitations. ‘‘1999 excess emissions
not account for malfunctions, startups, Yet the proposed FIP would allow memorandum’’); City of Santa Rosa v.
or shutdowns. citizens to bring actions for violations of EPA, 534 F.2d 150, 155 (9th Cir. 1976),
Response: The FIP fills the gap for the unattainable limits when EPA or the vacated on other grounds, 429 U.S. 990
provisions of the SIP that were State likely would choose to exercise its (1976). Thus, we have long held that
disapproved. In its attainment prosecutorial discretion. Such a outright or ‘‘automatic’’ exemptions
demonstration modeling, the State regulatory ‘‘Catch-22’’ is both from emission limits needed to
modeled emissions from flares at 150 unreasonable and unlawful. demonstrate attainment of the NAAQS
lbs of SO2/3-hour period, yet the SIP did Response: We respectfully disagree are not appropriate, something we
not contain corresponding emission with the commenter. First, in our indicated in our proposed FIP. See our
limits for the flares. This was the basis proposal we did not say that sources 1999 excess emissions memorandum,
for our disapproval of part of the SIP. will be unable to comply with the reference document RRR, and our
We believe we have appropriately continuous flaring emission limitations. proposed FIP, 71 FR 39264, col. 1, July
addressed malfunction, startup, and We note that, after receiving the 12, 2006. Our interpretation on this
shutdown in this final rule. See section refineries’ estimates of routine flare issue has been upheld by the U.S. Court
II.D.3., below. emissions, the State established as a of Appeals for the 6th Circuit: in a 2000
Certain assumptions were made in the decision, the Court rejected a challenge
State-only limit the same numerical
State’s attainment demonstration for the to EPA’s disapproval of a Michigan SIP
flare limit we are adopting, and the
Billings/Laurel SO2 SIP. Included in the revision that provided an automatic
refineries and MSCC agreed to the
assumptions was that flares had routine exemption from SIP limits during
stipulations containing those limits. See
emissions of 150 lbs of SO2/3-hour malfunction, startup, and shutdown
67 FR 22180, col. 2, May 2, 2002, and
period. To assure attainment and periods. Michigan Department of
reference documents UUUUU,
maintenance of the NAAQS, the SIP or Environmental Quality v. EPA, 230 F.3d
OOOOOO, PPPPPP, QQQQQQ, and
a FIP must contain enforceable emission 181 (6th Cir. 2000).
SSSSSS. Also, at the time of our SIP
limits on the flares. This is fully As we explained as long ago as 1977,
explained in our proposed action on the action, Conoco indicated to us that
routine emissions from its flare were the appropriate approach in SIPs/FIPs is
Billings/Laurel SO2 SIP (64 FR 40791, to require continuous compliance in
40801, July 28, 1999) and in the expected to be less than 150 lbs SO2/3-
hour period. See 67 FR 22180, col. 2, order to create an incentive for sources
response to comments contained in our to properly operate and maintain their
final action on the Billings/Laurel SO2 May 2, 2002, and reference document
RRRRRR. Based on this information, we facilities and to improve their operation
SIP (67 FR 22168, 22179, May 2, 2002). and maintenance practices over time.
The State of Montana has flare have concluded that the refineries and
MSCC will be able to comply with the See, e.g., 42 FR 21472, April 27, 1977
provisions that apply to CHS Inc., (reference document VVVVV), and 42
ConocoPhillips, ExxonMobil, and 150 lbs SO2/3-hour flare limit under
normal operating conditions. FR 58171, November 8, 1977 (reference
MSCC. See CHS Inc.’s, ConocoPhillips’, document WWWWW). We explained
ExxonMobil’s, and MSCC’s exhibit A–1, We did say in our proposal that we
recognize flares are sometimes used as that an automatic exemption would
adopted by the Montana Board of encourage the source to claim after
Environmental Review on June 12, 1998 emergency devices and that it may be
difficult to comply with the flare limits every period of excess emissions that
(reference documents QQQQQQ, the exemption applied, and that instead
PPPPPP, UUUUU, and OOOOOO). during malfunctions. See 71 FR 39264,
col. 1, July 12, 2006. However, contrary the proper means to provide relief to
Exhibit A–1 contains additional State sources was through the exercise of
requirements that were not submitted to the commenters’ assertions, our
decision to require an emission limit enforcement discretion in appropriate
for inclusion in the SO2 SIP. Among circumstances. Id.
these is an emission limit on flares of that may be difficult to meet under
Later, in 1999, we indicated that
150 lbs of SO2/3-hour period, the value certain conditions is not capricious,
states could include in their SIPs, as an
the State relied on to model attainment. unreasonable, or unlawful. alternative to the enforcement discretion
These flare provisions do not and would There is often a conflict, which is not
approach, narrowly tailored affirmative
not satisfy the SIP/FIP requirements of limited to refinery flare emissions,
defense provisions to address source
the CAA for two reasons. First, they between a source’s ability to control difficulties meeting emission limits
were never submitted to EPA to be emissions during certain operating during malfunction, startup, and
included as part of the SIP. Second, the conditions and the CAA’s requirement shutdown periods. See reference
flare provisions contain automatic to attain and protect the NAAQS. Our document RRR, our 1999 excess
exemptions for malfunction, startup, fundamental responsibility under the emissions memorandum. In this 1999
and shutdown. This is inconsistent with CAA with respect to SIPs/FIPs, memorandum we reiterated our long-
EPA’s longstanding interpretation of the however, is to ensure the NAAQS are held view that, ‘‘because excess
CAA, which is that, since SIPs must attained and other CAA requirements emissions might aggravate air quality so
are met. See CAA sections 110(a) and
provide for attainment and maintenance as to prevent attainment or interfere

of the NAAQS and the achievement of 8 See reference document RRR, September 20,
with maintenance of the ambient air
the PSD increments, all periods of 1999, memorandum entitled ‘‘State Implementation
quality standards, EPA views all excess
excess emission must be considered Plans: Policy Regarding Excess Emissions During emissions as violations of applicable
violations. Accordingly, any provision Malfunctions, Startup, and Shutdown.’’ emission limitation[s].’’ We also
repeated our recognition that some limits for flares are not made less (e) Comment (MSCC): There is no
malfunctions may be unavoidable. stringent, the FIP must recognize in its reasonable basis to believe that flaring,
Thus, while flares may have unique final action that flares must be available as practiced in this air-shed, prevents
characteristics, the underlying conflict for use during malfunctions and attainment and maintenance of NAAQS,
between the ability to comply and need emergencies to protect the safety of or that it is inadequately regulated, or
to meet the NAAQS is the same. We do employees and the public, as well as that it has an impact on health, welfare,
not believe the nature of the emission equipment integrity, regardless of the or commerce among states, as years of
point should dictate a different mass emission rate of the time. experience confirm. The State of
approach to protection of the NAAQS. Response: The FIP is not intended to Montana flare provisions are adequate.
Whether considering stack emissions at jeopardize the safety of refineries, their No federal action is needed.
a power plant or other source, or flare workers, or neighbors. Our SIP policy 9 Response: This comment goes to the
emissions at a refinery, the SIP/FIP has long recognized that imposing validity of our SIP action and is not
should be structured to provide the penalties for violations of emission relevant here. See our response to
source with the incentive to properly limitations for sudden and unavoidable comment II.B.2.(a), above.
design, operate, and maintain its (f) Comment (MDEQ): Imposing a
malfunctions caused by circumstances
facility. An outright exemption from the mass-based emission limit (and the
entirely beyond the control of the owner
emission limits would not do this. necessary and ancillary requirements for
or operator may not be appropriate.
To provide relief to the sources for measuring flows and concentration) on
truly unavoidable violations, while still States, EPA, and citizens have the a flare increases the regulatory workload
maintaining appropriate incentives for ability to exercise enforcement while providing a marginal benefit.
compliance, we are providing an discretion to refrain from taking Currently, Montana’s Malfunction rule
affirmative defense to penalties for enforcement action in these (ARM 17.8.110) provides Montana with
violations of flare limits during circumstances. In addition, EPA has enforcement discretion during
malfunctions, startups, and shutdowns. revised the FIP to provide sources with malfunction events.
The elements of the defense, which a the ability to assert an affirmative Response: We note that the State has
source would have to prove in court or defense to penalties for violations of mass-based emission limits on the flares
before an administrative judge, are flare limits during malfunction, startup, in the Billings/Laurel SO2 area. See CHS
enumerated in our final rule and are and shutdown. However, while we Inc.’s, ConocoPhillips’, ExxonMobil’s,
consistent with the elements described recognize some violations may be and MSCC’s exhibit A–1, adopted by the
in our 1999 excess emissions unavoidable, we also believe that Montana Board of Environmental
memorandum. The gist of these sources have a responsibility to do their Review on June 12, 1998 (reference
elements is that a source must take all best to achieve continuous compliance documents QQQQQQ, PPPPPP,
possible steps to prevent exceedances of and to minimize the number, duration, UUUUU, and OOOOOO). Exhibit A–1
the limits and to minimize the amount, and severity of malfunctions and other contains State requirements that were
duration, and impact of those events leading to excess emissions. not submitted for inclusion in the SO2
exceedances. These same or similar (d) Comment (MSCC): Various SIP. The provisions of exhibit A–1 also
criteria have been adopted by other jurisdictions have attempted to address appear in the sources’ Title V permits
regulatory agencies, including the State flare emissions. There is no uniform and are labeled as State-only provisions.
of Colorado and Maricopa County, federal requirement or regulation See, for example, ConocoPhillips’ Title
Arizona, in excess emissions rules. See, requiring such limits or monitoring, V permit (see reference document
e.g., Colorado Air Quality Control particularly for short term limits, or for XXXXX).
Commission Common Provisions malfunction, startup, and shutdown The exhibit A–1 requirements
Regulation, 5 CCR 1001–2, Sections II.E. controls. It is difficult to understand any indicate that the facilities shall not
and J. (reference document TTTTTT); reason that the Montana SIP for allow SO2 emissions from any flare,
Maricopa County Air Pollution Control Billings/Laurel is ‘‘substantially unless the emissions are a minor flaring
Rules, Rule 140, ‘‘Excess Emissions’’, inadequate’’ regarding flaring or for event (defined as less than or equal to
Section 400 (reference document proposing restrictions going far beyond 150 pounds per 3-hour period), or the
ZZZZZ). those in effect in any jurisdiction or result of start-up, shutdown, or a
Finally, we reject commenters’ federal rule. malfunction. Exhibit A–1 does not
assertion that citizens will necessarily Response: Regardless of what other indicate how compliance with the
pursue enforcement where the State and areas are doing with respect to flare emission limit is determined and only
EPA do not, but in any event, this emissions, we must fulfill our requires reporting of flare emissions that
possibility is inherent in the structure of responsibility to fill the gaps of the are not minor flaring events.
the CAA; Congress provided citizens provisions of the SIP that we Presumably, the additional workload
with the ability to enforce SIPs and disapproved. Each area must be provided by the FIP, that the State is
FIPs. This inherent structure is not a addressed on a case-by-case basis. The referring to, is in evaluating the
reason for us in this rulemaking action response to comment II.D.1.(a) and our continuous analyzers and receiving
to change our longstanding notice of proposed rulemaking express quarterly reports. We believe the
interpretations regarding the proper why we believe the FIP should contain additional workload is warranted and
treatment of excess emissions. emission limits for flares in the Billings/ necessary to determine compliance with
(c) Comment (NEDA/CAP): Industry the flare emission limits and assure that
Laurel area. Regarding the comment
contends that it is virtually impossible the SO2 NAAQS will be attained and
about substantial inadequacy, please see
to meet the proposed limits during maintained. See, e.g., CAA sections
our response to comment II.B.2.(a),
flaring, since flares themselves are not 110(a)(2)(A), (C), and (F), 42 U.S.C.
above.
process units when they are treating 7410(a)(2)(A), (C), and (F).
excess gases during malfunction events. 9 See reference document RRR, September 20,
We do not understand the intent of
EPA has presented no information in 1999, memorandum entitled ‘‘State Implementation
the comment that indicates MDEQ has
this notice or elsewhere to the contrary. Plans: Policy Regarding Excess Emissions During enforcement discretion under its
On this basis alone, if the mass emission Malfunctions, Startup, and Shutdown.’’ malfunction rule in ARM 17.8.110
(reference document YYYYY). Before justification criteria are being used to Response: See responses to comments
MDEQ could decide whether or not to establish the 500 lb. minimum per day II.D.1.(b) and (c), above. As we indicate
pursue an enforcement action for base in the Proposed Rule. And, as in our response to comment II.D.1.(c),
violations of the State-only flare limit, noted on page 39264 of the Federal the FIP is not intended to jeopardize the
MDEQ would need to evaluate Register dated July 12 announcing the safety of refineries, their workers, or the
information submitted by sources. FIP, EPA says ‘‘if we adopted the 500 community. However, we believe it
Additionally, we note that in response pound value in this FIP, we would would be inconsistent with CAA
to our proposed action on the Billings/ impose it as an enforceable emission sections 110(a) and (l) to provide an
Laurel SIP, the State said the following: limit.’’ If there are still questions outright exemption from the flare limits
‘‘The State agrees with EPA that the SIP concerning the 500 lb per day emission during malfunction, startup, and
is incomplete without enforceable limit, why is it being proposed? Is there shutdown periods. Instead, to provide
emission limitations applicable to flares, a lower and perhaps ‘‘better’’ emission some measure of relief to the sources,
and that such limitations should limit per day that should be considered? we have included an affirmative defense
correspond to the emission rates used in Response: The current SO2 NAAQS to penalties in our final FIP rule. If a
the attainment demonstrations. were set to protect public health and source takes steps consistent with the
However, after significant effort to welfare after consideration of various elements of the affirmative defense,
address the issue, the State was unable scientific data. It is not our role here to excess flaring emissions during
to find a workable solution that would re-evaluate the NAAQS, but to ensure malfunction, startup, and shutdown
meet EPA’s concerns.’’ See document they are met. Through modeling we periods would not be penalized. We
#IV.A–23, comment #3, from docket determined that both limits would have considered several additional
#R8–99–01; 67 FR 22183, col. 1, May 2, protect the SO2 NAAQS. While a lower
factors: First, historically, the sources
2002; and reference document ZZZZZZ. limit might be attractive, we are setting
(g) Comment (YVAS): YVAS concurs have used the flares as part of their
the limits at 150 lbs of SO2/3-hour
with EPA’s further assumption (page routine operations, i.e., in non-
period, a level sufficient to meet the SO2
39264), that ‘‘the 3-hour SO2 NAAQS emergency conditions. See September
NAAQS; we think this is reasonable.
would be attained’’ if ‘‘the limit for the 28, 1995, letter from Bob Raisch to
See response to comment II.A.2.(b). See
main flares was established at 500 Douglas Skie (reference document
also our response to comments
pounds of SO2 per calendar day.’’ Since SSSSSS); 67 FR 22180, col. 2, May 2,
pertaining to SO2 NAAQS and SO2
there is apparently precedent (as noted 2002. Also, in its comments on the FIP
Health Effects (II.F.9. and 10.,
on page 39263 FR) ‘‘contained in respectively) below. (reference document QQQQ), CHS Inc.
settlements between the United States (i) Comment (MDEQ): MDEQ believes indicated that the 150 lbs/3-hour value
and CHS Inc, ConocoPhillips and that hard cap emission limits on flares was used in the original model to
ExxonMobil,’’ YVAS further agrees to are good but believes that the flare represent routine flaring and
and accepts EPA’s reasoning that ‘‘the emission limits will be more accepted if background SO2 concentrations. MSCC
500 pound value for this FIP (should) be malfunction, startup, and shutdown indicated in its comments on the FIP
imposed as an enforceable limit and not exemptions are introduced. (reference document WWWW) that
just a trigger point for further analysis’’ Response: We acknowledge MDEQ’s flares can be used for handling streams
as a starting point. However, the ‘‘500 support for hard cap emission limits on other than those arising from
lbs per day limit,’’ if extended for any flares. Regarding exemptions for malfunction, startup, and shutdown.
length of time, is not acceptable. Based malfunction, startup, and shutdown, see Second, flaring events have not
on acquired information, YVAS does our responses to comments II.D.1.(b) necessarily been as infrequent as the
not think this limit would be punitive, and (c), above. commenter implies. From the first
nor would it be impossible for industry As indicated above, to address quarter of 2005 through the second
sources to attain. It is accepted that zero industry concerns regarding quarter of 2007, source reports indicate
emissions may not be possible or malfunctions, startup, and shutdown, that MSCC and the 3 refineries
attainable, but any lower emissions rate we are revising the FIP to provide experienced over 150 flaring events
would be a public benefit. And, sources the ability to assert an with SO2 emissions greater than 150
although a compliance drop could affirmative defense to penalties for pounds over 3 hours. See reference
create greater industry noncompliance violations of flare limits during document HHHHHH. Third, the
and require more enforcement action, malfunction, startup, and shutdown. emissions during these events can be
YVAS does not believe the more very high—the State estimated that
2. Safety Device emissions during malfunctions could be
stringent standards would create more
noncompliance problems for the (a) Comment (CHS Inc., WETA, MPA, as high as 6,000 pounds/3-hour period,
sources. NPRA): From a safety standpoint, there and the sources’ own reports for first
Response: We have decided to retain are concerns with flare limits applying quarter 2005 through second quarter
the proposed limit of 150 lbs of SO2/3- at all times, including malfunction, 2007 reflect emissions as high as 12,400
hour period. A more stringent limit than startup, and shutdown. Flares are pounds over a 2-hour period. See
either proposed is unnecessary to primarily safety devices, designed as a reference documents SSSSSS and
ensure attainment of the NAAQS. Thus, means to ensure the safety of employees HHHHHH. The maximum value
we believe it is reasonable not to impose and the community and to maintain the reported for a flaring event during the
a more stringent limit as the commenter integrity of refinery equipment during period was 40,800 pounds of SO2 over
suggests. situations that are not representative of an unknown duration, and there were
(h) Comment (Citizen): The proposed normal operations. It will be precedent numerous events in the thousands of
rule should not be adopted unless setting if the EPA views these infrequent pounds. See reference document
recognized medical opinion concerning events as enforcement situations. It HHHHHH. Fourth, we want to ensure
the cumulative health risks of the would, in essence, require facilities to that the owners/operators design,
release of 500 lbs per day of sulphur choose between maintaining a safe, operate, and maintain their facilities to
dioxide into the area’s airshed is controlled refinery and violating the minimize flare emissions by minimizing
analyzed. Specifically, what FIP. the conditions that lead to malfunctions,
startups, and shutdowns. In the FIP historically, and they can be the source the release of vent gases during
context, the appropriate way to do this of very large quantities of emissions in malfunctions, or, to treat the emergency
is by establishing a flare emission limit a short period of time. We believe it is vent gases to remove sulfur compounds
that is not subject to outright necessary and appropriate to impose prior to combustion in the flare.
exemptions. Fifth, the State and EPA limits on the flare emissions to fill one Response: See responses to comments
have already viewed these events as of the gaps in the SIP, to support our II.D.1.(b) and (c), II.D.2.(a), and
enforcement situations in the context of attainment demonstration, and to create II.D.3.(a), above. Also, we understand
the refinery initiative and, through the appropriate incentives for the sources in that while a malfunction is underway, it
consent decrees, have created the the design, operation, and maintenance may be impossible to treat the gases
expectation that the refineries will of their facilities. prior to combustion in the flare.
minimize flare emissions. We explain in However, we do not agree that all
3. Malfunction, Startup, and Shutdown
this preamble why the conditions of the malfunctions are categorically
consent decrees, while beneficial, are (a) Comment (WSPA, MSCC, unavoidable. We are concerned with the
not sufficient for purposes of the FIP. ExxonMobil): In working with the South causes leading to the malfunctions and
See, e.g., responses to comments Coast Air Quality Management District, the steps taken after the malfunction
II.A.2.(b), II.D.4., and II.E.1.(e). We also they were careful not to compromise begins to mitigate its effects. We are
note that MSCC is not subject to a safety by restricting, either explicitly or promulgating an affirmative defense
consent decree. Finally, the air does not implicitly, the use of flares during provision along with the flare emission
care whether emissions come out of a emergencies through the imposition of limits that should ensure sources take
flare that is used as a safety device at a mass emission limits or otherwise. all steps within their control to avoid
refinery or a stack at a power plant or Response: See responses to comments
malfunctions and minimize their
other facility.10 In both cases, the II.D.1.(b) and (c), and II.D.2.(a), above.
impacts on air quality once they occur.
emissions of SO2 impact air quality, and Our FIP does not require or direct the
We believe this is reasonable and
EPA’s charge is to address those impacts sources to not use their flares during
appropriate to ensure protection of the
so as to protect the NAAQS. emergencies. Unlike the South Coast or
NAAQS.
(b) Comment (WSPA, MSCC, Bay Area,11 however, we are required to
promulgate a FIP that demonstrates (c) Comment (WETA): Pursuing the
ExxonMobil): EPA proposes that flare adoption of this FIP could potentially
limits apply at all times without attainment of the SO2 NAAQS.
Consequently, it is necessary and result in the setting of an inconsistent
exception. It would be virtually national policy for malfunction, startup,
impossible to comply with SOx mass appropriate that we impose emission
limits on the flares that are consistent and shutdown.
emission limits at all times and for all Response: We do not agree with the
malfunctions for the simple reason that with our modeled attainment
demonstration. To address industry comment. The FIP would not be setting
the primary function of a refinery flare inconsistent national policy for
is to serve as a safety device. Flares concerns, we are providing an
affirmative defense to penalties for malfunction, startup, and shutdown
must be available for use during occurrences. To the contrary, we are
malfunctions and emergencies to protect excess flare emissions during
malfunction, startup, and shutdown following our national policy with
equipment and the safety of employees respect to malfunctions, startup, and
and the public. periods.
We note that SCAQMD’s rule 1118(d) shutdown as expressed in the 1999
Response: See responses to comments
imposes annual SO2 performance targets excess emissions memorandum (see
II.D.1.(b) and (c), and II.D.2.(a), above.
(c) Comment (NPRA): The U.S. for flare emissions (caps on the amount reference document RRR).
Chemical Safety Board (CSB) urges the of SO2 emitted from flares in one year). (d) Comment (MSCC): MSCC believes
installation of flares. The CSB sites The performance targets are based on that the approach taken by the State of
flares as a ‘‘safer alternative’’ when the crude processing capacity and Montana in providing for minimization
compared to other techniques. Clearly become more stringent over time. of flaring, above a reasonably
the CSB recommendation is at odds Malfunction, startup, and shutdown determined de minimis threshold, and
with Agency’s proposal. emissions count towards the annual clear exceptions for malfunctions,
Response: See responses to comments performance targets unless they meet startup, shutdowns and other
II.D.1.(b) and (c), and II.D.2.(a), above. certain narrowly defined exemptions in operational needs is the sound
Also, we do not believe our action is at rule 1118(k). Sources that exceed their approach, to address the reality that
odds with the CSB’s recommendations. annual performance targets must submit there are, and will be situations such as
In this action, we are not opining on the a flare minimization plan and are malfunctions, startups, and shutdowns
use of flares versus other techniques. subject to mitigation fees of up to four and emergencies that are beyond the
We are not telling the refineries or million dollars a year (see reference reasonable control of a source, in the
MSCC to stop using their flares. document CCC). operation of flares.
However, flares are an emission point at (b) Comment (WSPA, MSCC, Response: We recognize there may be
the refineries and MSCC, they have been ExxonMobil): It is essential for EPA to violations of flare emission limits
the source of routine emissions recognize the true nature of during malfunctions, startups,
malfunctions at refineries, and the fact shutdowns, and emergencies that are
10 In theory, a smokestack could also be that there is no practical way to regulate beyond the control of a source;
characterized as a safety device; among other accordingly, we are providing sources
things, a stack is used to prevent harmful ground 11 The Bay Area prohibits all refinery flaring
with the ability to assert an affirmative
level concentrations of pollutants. In addition, gases unless the flaring is consistent with a flare
are sometimes bypassed around control devices minimization plan or is caused by an emergency.
defense to penalties for violations of
directly to the stack to avoid damage to control See BAAQMD rule 12–12–301 (reference document flare limits that occur during
devices and/or other dangerous conditions. In the AAAAAAA). The South Coast rule requires malfunction, startup, and shutdown
SIP/FIP context, we do not believe it is appropriate minimization of flaring and prohibits combustion of periods. We believe this is a reasonable
to automatically exempt these stack emissions, even vent gas in the flare except during emergencies,
though the stack may serve a safety purpose. See shutdowns, startups, turnarounds or essential
approach, consistent with our views
our 1999 excess emissions memorandum, reference operational needs. See SCAQMD rule 1118(c)(4) that automatic exemptions are not
document RRR. (reference document CCC). appropriate for emission limits relied on
to demonstrate attainment of the limit, even absent the additional authority to seek injunctive relief for all
NAAQS. concerns we discuss below, would exceedances of emission limits so that
(e) Comment (COPC): The rule as assure attainment of the SO2 NAAQS. we remain able to protect the NAAQS,
written will ultimately put Second, during certain situations, as regardless of source ‘‘culpability’’ for
ConocoPhillips in the position of having indicated in 40 CFR 60.8(c) and any specific exceedance.
to choose between compliance with an 60.104(a)(1), the H2S limit does not We note that in our proposed FIP
environmental regulation and apply. Specifically, the consent decree preamble, we invited comment
maintaining safe operating conditions. indicates that the CHS Inc. refinery flare regarding whether it would be
This is an untenable position which can is an affected facility under 40 CFR part appropriate to extend an affirmative
be avoided by acknowledging in rule 60, subparts A and J for fuel gas defense to the FIP sources for
language that flare SO2 emissions can combustion devices and that fuel gases exceedances of their flare limits during
occur during periods of malfunction, combusted in the refinery flare shall malfunctions, startup, and shutdown.
startup, and shutdown, provided that comply with the emission limit of 40 See 71 FR 39264, July 12, 2006. There
accepted management systems are CFR 60.104(a)(1). However, 40 CFR we said the following:
followed. 60.104(a)(1) exempts process upset ‘‘We do interpret the CAA to allow owners
Response: See responses to comments gases and certain types of fuel gas from and operators of sources to assert an
II.D.1.(b) and (c), and II.D.2.(a), above. the emission limit. Additionally, the affirmative defense to penalties in
We believe the provision of the provisions in 40 CFR 60.8(c) indicate appropriate circumstances, but normally we
affirmative defense to penalties for that emissions in excess of the level of would not view such an affirmative defense
excess emissions during malfunction, the applicable emission limit during as appropriate in areas where a single source
startup, and shutdown periods periods of malfunction, startup, and or small group of sources has the potential
appropriately and reasonably addresses shutdown shall not be considered a to cause an exceedance of the NAAQS. See
violation of the applicable emission 1999 policy statement. We solicit comment
the commenter’s concerns.
limit unless otherwise specified in the on whether it would be appropriate to
(f) Comment (COPC): A FIP program
include in our final FIP the ability to assert
that adopts the same evaluation applicable standard. Emission limits for an affirmative defense to penalties only (not
procedures for malfunctions, startups, demonstrating attainment and injunctive relief) for violations of the flare
and shutdowns for flares is preferred to maintenance of the NAAQS must apply limits.’’
a fiction that a facility can maintain a at all times. (See responses to comments
II.D.1.(b) and II.D.2.(a), above, and We have decided to provide an
flare emission limit in all malfunction,
reference document RRR.) affirmative defense for violations of the
startup, or shutdown events regardless
Third, the alternative monitoring plan flare limits during malfunction, startup,
of size or magnitude.
Response: See response to comments (AMP), that was approved pursuant to and shutdown. We believe this
II.D.3.(a), (b), (c), (d), and (e), above. the consent decree and NSPS represents a deviation from our 1999
(g) Comment (YVAS): Specific to requirements (see reference document excess emissions memorandum because
flaring emergencies by the sources, any LLLLLL) for the refinery flare fuel gas in the Billings/Laurel area, one or more
added controls on flaring to protect the combustion device, primarily relies on of the FIP sources may have the
public (from SO2 exceedences) is quarterly measurement of the H2S potential to cause an exceedance of the
essential and is common sense. content of some of the refinery fuel gas SO2 NAAQS. In the unique
Response: We acknowledge the streams that go to the flare using stain circumstances of this FIP, with the rule
comment and support for our proposal. tubes; more frequent measurement may language we are adopting, we believe a
be required for a limited time depending deviation from the 1999 excess
4. Subject to NSPS emissions memorandum is warranted.
on the concentration measured.
Comment (CHS Inc.) It should be Although this may be acceptable under For example, we have included rule
noted that the CHS refinery flare is the terms of the consent decree and the language that indicates the affirmative
subject to NSPS Subpart J as a result of NSPS, we believe more frequent testing defense is not available if, during the
the consent decree. This limits the H2S is necessary for determining compliance period of the excess emissions, there
content of the routine refinery fuel gas with an emission limit set to assure was an exceedance of the SO2 NAAQS
streams routed to the flare and requires attainment and maintenance of the that could be attributed to the emitting
monitoring to demonstrate compliance NAAQS. source. At least one other EPA Region
with the limit. has approved an affirmative defense
Response: As indicated by the 5. Affirmative Defense/1999 Excess provision with this language. See
commenter, the consent decree limits Emissions Memorandum Maricopa County Rule 140 (reference
the H2S content of the routine refinery (a) Comment (WSPA): The availability document ZZZZZ), which Region 9
fuel gas streams routed to the flare. of an affirmative defense is desirable. approved on August 27, 2002 (67 FR
However, there are several reasons why Even though EPA may allow for the 54957) (reference document AAAAAA).
the H2S ppm limit alone is not sufficient assertion of affirmative defenses, the Although not identical to the 1999
to support the FIP’s attainment affirmative defense would only be excess emissions memorandum, this
demonstration. allowed for the mitigation of penalties. rule language should provide a
First, flow information is needed to This is an unreasonable position in significant incentive to the facilities to
translate H2S ppm values into pounds of which to place refiners subject to the take steps to avoid and reduce flaring
SO2 for a given period of time. Flow proposed requirements. whenever possible.
rates to the flares can vary widely. Response: We are providing an Also, based on our experience since
Without knowing potential worst-case affirmative defense to penalties in the the 1999 excess emissions
flows to the flare, we cannot determine final rule, but not to injunctive relief. memorandum was issued, we believe
whether the consent decree H2S ppm This is consistent with the Clean Air that the elements of the affirmative
limit would assure compliance with the Act interpretations expressed in our defense delineated in the 1999 excess
FIP 150 pounds of SO2/3-hour limit at 1999 excess emissions memorandum. emissions memorandum, which
the 3 refineries. Therefore, we cannot See reference document RRR. We elements we have adopted in this FIP,
conclude that the consent decree H2S believe it is reasonable to retain the provide a very significant incentive for
facilities to do all they can to comply to the treatment of excess emissions case, EPA has made no demonstration to
with their emission limits. It is not clear during malfunction, startup, and justify an exception to the general
that the incentive is significantly shutdown. See 71 FR 39264, col. 1, July allowance for affirmative defenses for
different than would be present under a 12, 2006. We have considered the malfunction events. Consequently, API
traditional enforcement discretion commenter’s comments along with all urges EPA to allow the assertion of
approach, particularly when sources other comments. affirmative defenses in the final FIP.
assume that enforcement action will (e) Comment (NEDA/CAP): NEDA/ Additionally, ConocoPhillips indicated
rarely be taken for infrequent or small CAP is also concerned that EPA has that because of the harsh consequences,
violations. Finally, we have considered made no demonstration that ‘‘a single EPA should only apply this exception to
industry comments regarding safety source or small group of sources has the its policy where it is clearly
concerns, and while we do not agree potential to cause an exceedence of the demonstrated that there is very real,
that emissions from flares should be NAAQS,’’ or that the NAAQS in this air extended potential for a single or small
treated entirely differently from basin is in fact, any more vulnerable to group of sources to cause an exceedence
emissions from stacks and other points, a NAAQS exceedence from these of the NAAQS. This is not present in
we think our resolution of this issue sources than any other nonattainment this case. In fact, actual monitoring has
appropriately and reasonably addresses areas is from a small group of sources. shown that even during malfunction,
industry concerns. If finalized, the failure to provide an ambient NAAQS violations do not
(b) Comment (WETA): Any flare affirmative defense for malfunctions occur. ConocoPhillips urges EPA to
emission limitations should include, at would be entirely arbitrary and allow the assertion of affirmative
the least, an allowance for an affirmative unreasonable. Moreover, as a national defenses for both penalties and
defense for malfunction, startup, and precedent with severe legal injunctive relief in the final FIP.
shutdown circumstances. consequences for sources in other Response: See our prior responses to
Response: See response to comment nonattainment areas, adoption of this comments II.D.5.(a), (d), and (e). Also,
II.D.5.(a), above. proposed FIP provision would be highly we note that on two occasions, one in
(c) Comment (NEDA/CAP): EPA vulnerable to legal challenge for failure 1985 and one in 1995, flaring resulting
should adopt a broad affirmative to meet the Clean Air Act’s notice and from malfunctions at ConocoPhillips
defense for penalties and injunctive comment procedures under a federal caused ambient exceedances of the SO2
relief for malfunctions as part of the court’s recent decision in NAAQS (see reference documents
mass emission limit for flares. MPA Environmental Integrity Project v. EPA, DDDDDDD and EEEEEEE).
indicated that the FIP should not be 425 F.3d 992 (D.C. Cir. 2005). (g) Comment (NEDA/CAP, MSCC):
adopted in the proposed form because Response: In our final action, we are The proposed FIP appears to
the failure to include an affirmative providing an affirmative defense to misinterpret the 1999 Malfunction
defense for flaring resulting from penalties for the flare limits. We Policy. The July 12 preamble for
malfunctions poses a significant safety disagree with the commenter’s assertion adoption of the FIP appears to suggest
risk to employees and the public with regarding notice and comment that prosecutorial discretion would
no corresponding benefit. procedures; we believe we have met all never be allowed in a nonattainment
Response: See response to comment applicable requirements and provided area where the agency decides that ‘‘one
II.D.5.(a), above. fair notice regarding our intentions in or a group of sources are directly
(d) Comment (NEDA/CAP): NEDA/ our notice of proposed rulemaking. We implicated in nonattainment of a
CAP is concerned about the potential for proposed that the flare limits would NAAQS.’’ In fact, the 1999 Policy
EPA’s establishment of any precedent apply at all times and also invited recommends that such situations have
with regard to limiting the availability comment on whether it would be to be addressed in the underlying
of affirmative malfunction defenses in appropriate to extend an affirmative standards themselves through narrowly-
nonattainment areas generally. NEDA/ defense for the flare limits to the four tailored SIP revisions. Moreover, in no
CAP is also concerned with the sources subject to the FIP. Our final event does the 1999 Malfunction Policy
application of the 1999 Malfunction action is a logical outgrowth of our ever prohibit the use of prosecutorial
Policy in the Billings/Laurel proposed proposal; we have decided to provide an discretion.
FIP because the Policy has never been affirmative defense to penalties for Response: Enforcement discretion or
subject to notice and comment violations of the flare limits during prosecutorial discretion is always
rulemaking, but the application of the malfunction, startup, and shutdown. available. The question in this case was
policy results in clear legal While our action on this FIP may have whether it was appropriate to codify an
consequences for regulated entities in some impact on other SIPs and FIPs affirmative defense, which we have
contravention of Appalachian Power v. based on the logic we have applied, our done in our final rule. We have not
EPA, 208 F.3d 1015 (D.C. Cir. 2000). rule is only directly applicable to the misinterpreted our 1999 policy.
Response: See response to comment four sources subject to the FIP. It is (h) Comment (NEDA/CAP, API):
II.D.5.(a), above. Also, we respectfully possible EPA may reach a different There is no rational basis in the
disagree with the commenter that we are decision in future rulemaking. proposed FIP or the 1999 Malfunction
contravening the Appalachian Power (f) Comment (API, COPC, MSCC, Policy to limit the affirmative defense to
case holding. In our proposal, we ExxonMobil): While EPA’s 1999 penalties. NEDA/CAP asserts that such
proposed that the flare limits would Malfunction policy does state EPA’s a limitation is not reasonable since the
apply at all times but took comment on position that affirmative defenses are malfunction condition during which the
the application of an affirmative defense not appropriate ‘‘where a single source exceedence of the applicable limitation
to penalties for those limits. In this final or small group of sources has the occurs would be unavoidable.
rulemaking, we have decided to provide potential to cause an exceedence of the Response: We respectfully disagree.
the affirmative defense to penalties. The NAAQS,’’ API and others are unaware There could be instances in which
commenter had a full opportunity to of any instance where EPA has utilized malfunctions are unavoidable based on
comment on our proposal, which this exception from its general policy current plant layout and operating
included a discussion of our allowing for the assertion of affirmative parameters but in which some form of
interpretations of the CAA with respect defenses during malfunctions. In this corrective action would still be
appropriate. We cannot predict the affirmative defense to penalties for E. Concerns With Dispersion Modeling
exact nature of those circumstances, but exceedances of the flare limits during 1. Policy Issues
protection of the NAAQS and public malfunction, startup, and shutdown.
health is not an intermittent obligation; Under this approach all excess (a) Comment (MSCC, ExxonMobil):
we are required to assure attainment emissions are considered violations. Out-of-Date and Invalid Model Choice.
and maintenance of the NAAQS at all However, if we or anyone else brings an (i) The proposed FIP uses the same
times, not just when sources are in model as that used in the SIP. EPA’s
enforcement action, the facility may
normal operation mode or when models have changed since the time the
then assert the defense to penalties. To
attainment is convenient. See, e.g., City SIP was developed. It is inappropriate to
establish the defense, the facility must propose and justify more restrictive
of Santa Rosa v. EPA, 534 F.2d 150 (9th
demonstrate to the judge that it took requirements on sources without
Cir. 1976), vacated and remanded on
appropriate steps to avoid the excess considering more current modeling
other grounds sub nom. Pacific Legal
Foundation v. EPA, 429 U.S. 990 (1976) emissions and met other requirements, techniques and requirements. The older
(‘‘ ‘Neither EPA nor this court has any the details of which are contained in our model may be more appropriate to
right to decide that it is better to final rule. If the facility cannot establish confirm an existing situation or permit
maintain pollutants at a level hazardous the defense, it may be subject to CAA minor changes. However, the FIP goes
to health than to require the degree of penalties up to $32,500 per day. We do beyond minor changes.
public sacrifice needed to reduce them not typically advise the public when a Response: The commenter is correct
to tolerable limits’ ’’, citing South limit is exceeded or which facility has that a newer model is now available. For
Terminal Corp. v. EPA, 504 F.2d 646, at exceeded a limit, although we often new SIPs, we would require states to
656 (1st Cir. 1974); South Terminal alert the public through the press when use EPA’s most recent model. However,
Corp. v. EPA, 504 F.2d 646, 675 (1st Cir. we bring an enforcement action. Under this is a unique situation. The State
1974) (‘‘[I]t seems plain that Congress the FIP, the subject sources must submit developed the Billings/Laurel SO2 SIP
intended the Administrator to enforce reports to EPA identifying their using the ISC model, which was current
compliance with air quality standards emissions. Those reports are available to at that time, and we approved various
even if the costs were great.’’) Preserving source-specific emission limits in the
the public through the Freedom of
injunctive remedies ensures that we SIP based on the State’s modeling effort.
Information Act (FOIA). The
remain able to protect air quality The purpose of this FIP is to fill gaps in
establishment of flare requirements the approved SIP. We are not intending
standards and PSD increments in should help reduce flaring incidents.
accordance with our fundamental or required to re-do the entire SIP. See,
responsibilities under the CAA. See 6. Installation of Additional SO2 e.g., section 302(y) of the CAA, 42
CAA sections 110(a) and (l), 42 U.S.C. Reduction Equipment U.S.C. 7602(y) (‘‘Federal
7410(a) and (l). See, also, the discussion implementation plan’’ means a plan (or
of this issue in our 1999 excess Comment (ExxonMobil): EPA’s portion thereof) promulgated by the
emissions memorandum, reference proposed FIP does not allow for time for Administrator to fill all or a portion of
document RRR. the design and installation of facilities a gap or otherwise correct all or a
(i) Comment (MSCC, ExxonMobil): An necessary to comply with the proposed portion of an inadequacy in a State
exception and affirmative defense flare emissions limitations. The implementation plan * * *’’); McCarthy
should be available under the FIP that facilities required for compliance with v. Thomas, 27 F.3d 1363, 1365 (9th Cir.
is at least consistent with the consent the proposed FIP go above and beyond 1994) (A FIP is ‘‘a set of enforceable
decrees executed by EPA and the State what was built for the SIP or what will federal regulations that stand in the
of Montana with most of the affected be built for the Consent Decree. For place of deficient portions of a SIP.’’)
sources. EPA’s proposed FIP, the required Accordingly, we think it is reasonable to
Response: As we have noted rely on the same model the State used
controls have not yet been identified.
previously, the consent decrees and the to develop the SIP. That way, all
FIP serve different purposes. We have Response: It is not clear what facilities emission limits in the SIP and FIP will
adopted an affirmative defense the commenter is envisioning. Without have been established on the same basis.
provision that is consistent with the greater detail, it is difficult to respond We note that MDEQ tested the
protection of the NAAQS. to the comment. However, the FIP performance of the ISC model when the
(j) Comment (Citizen): On page 39264 imposes no specific requirement for the Billings/Laurel SO2 SIP was being
is the statement ‘‘We are proposing that sources to install control equipment to developed, and the results showed that
the flare limits will apply at all times limit flare emissions, and the limit we the model performance exceeded the
without exception.’’ Laudable as that are imposing is the same one the State performance criteria for models of this
seems, EPA then subsequently states, imposed on the sources, and which type. The FIP modeling represents a
‘‘We solicit comment on whether it continues to be included in their minor change to MDEQ’s basic
would be appropriate to include in our permits. Our expectation is that sources approach. The sources in the SIP
final FIP the ability to assert an will take all steps within their control modeling are characterized in the
affirmative defense to penalties only to avoid flaring events and minimize modeling inputs as 25 point and volume
(not injunctive relief) for violations of sources and, except for minor
their impacts on air quality if they do
flare limits.’’ If the former statement is corrections provided by the sources, the
occur.
accepted, what are the penalties for major FIP-related change in modeling
exceeding flare limits and how will they To the extent that the commenter is involves only one source: The MSCC
be imposed and will the public be referring to the time needed to design 100-meter stack. We had to change the
advised which refinery exceeds a flare and install flare monitoring systems inputs for MSCC’s 100-meter stack
limit and how often could that happen required by the FIP, we have extended because the State gave too much stack
to the detriment of air quality in this the deadline for installation from 180 height credit to MSCC’s stack in the SIP
area? days to 365 days after EPA approval of modeling, and we, consequently,
Response: In this final rulemaking the flare monitoring plan. disapproved MSCC’s SIP emission
action, we have promulgated an limits and the SIP attainment
demonstration. Otherwise, the FIP the proposed FIP does not meet EPA’s demonstration (ExxonMobil). Without
modeling uses meteorology data, own guidelines and requirements including these existing emission
receptors, and stack parameters for because of the model used, lack of reductions from the SIP and near term
sources other than MSCC that are nearly current building profile, and numerous future reductions from consent decrees,
identical to those used in the SIP other problems found elsewhere. EPA’s proposed FIP ignores state and
modeling. Response: See our response to federally enforceable SO2 emission
We also note that ISC remained an comment II.E.1.(a)(i), above. The reductions already in place.
accepted EPA model at the time we modeling approach was extensively Response: See our responses to
proposed our FIP, and it is reasonable discussed with regulatory agencies and comments II.A.2.(b), II.B.2.(d) and II.D.4.
to finalize the FIP based on the same the public when the SIP was developed, The FIP modeling accounts for the
model. Switching models after our and the ISC-based modeling approach limits that we approved in the Billings/
proposal would have required us to re- met the requirements of EPA’s Laurel SO2 SIP and those we are
propose the FIP and would have Guideline on Air Quality Models. promulgating in the FIP. We cannot
delayed the FIP further. (d) Comment (MSCC): Modeling File include State requirements that were
(ii) A newer model, ‘‘AERMOD,’’ has Naming Convention. EPA’s modeling not submitted with the SIP.
been adopted as the EPA regulatory files and Technical Support Document, Additionally, the ExxonMobil consent
default model. It is clear that AERMOD both contained in the docket, do not decree limits have not been translated
is now preferred for regulatory use over provide a reference to the naming into short term emission limits by
the model used in the SIP development. conventions used in the modeling effort. MDEQ and made a part of the SIP. Short
Consideration needs to be afforded to While it is possible to dissect some of term emission limits are required to
models available today, and particularly the naming conventions, it was not ensure compliance with the 3-hour and
to the model reasonably believed to give possible to discern each and every file 24-hour average SO2 NAAQS. Also, the
the most accurate results. The and its purpose. Therefore, the consent decrees do not address all of the
stakeholder process should be used to reviewers are not certain that all the stacks/sources involved in the SIP/FIP.
determine which dispersion model modeling attempts, purposes and (f) Comment (MSCC): MSCC has
should be used for the FIP nuances have been accounted for in the concerns with using the SIP modeling.
(ExxonMobil). analysis. The commenter recommends a The predecessor model routines had
Response: See our response to more complete description of the been discredited (‘‘invalidated’’) in this
comment II.E.1.(a)(i), above. We also naming convention and the purpose valley following a study done years
note that AERMOD has more complex behind each modeling effort needs to be earlier by the State. The model, even in
software than ISC and, as a result, it explained. the 1990’s, did not represent state of the
would be extremely difficult to perform Response: At the recommendation of art in modeling science and was
the 1320 model simulations necessary to industry, MDEQ allowed the use of admittedly prone to serious over-
establish emission limitations that buoyancy flux in establishing emission predictions, particularly in so-called
would address buoyancy flux variations limits, which made the modeling far intermediate and complex terrain.
that were included in the State’s SIP. A more complex. As a result, many more Response: As noted above, the
stakeholder process is not required by modeling files are included than is modeling was EPA’s preferred model at
the CAA and would merely serve to typically the case in SIP modeling the time of the SIP, has been validated
delay issuance of the FIP. applications. To improve for use in the Billings/Laurel area, and
(b) Comment (MSCC): Out-of-Date documentation, some extraneous has been used extensively throughout
Model Input. Any dispersion modeling modeling files have been removed and the United States in setting emission
used for the proposed FIP must include a text file added to explain naming limits for nearly two decades. The
improved techniques regarding building conventions. The naming convention
model has not been ‘‘invalidated’’ for
downwash. A new method for used for the Billings/Laurel SO2 FIP
use in the Billings/Laurel area. See also
calculating the downwash effects modeling files is typical of that used by
our discussion of related issues in our
buildings have on predicted ambient the modeling community. To a modeler,
May 2, 2002, final action on the
concentrations has been developed. The the naming convention helps define the
Billings/Laurel SO2 SIP, 67 FR 22168,
new technique is known as ‘‘Plume Rise purpose behind the modeling effort. On
22183.
Model Enhancement’’ (PRIME) July 13, 2007, the revised modeling files (g) Comment (ExxonMobil): Only the
algorithm. This technique is now were indexed in the electronic docket
current actually existing emission
commonly in practice in both ISC– contained on http://
sources with proper geographical
PRIME and AERMOD. EPA’s FIP www.regulations.gov, and a compact
coordinates should be used as inputs to
modeling does not use this technique. disk containing the modeling files was
the dispersion model.
Response: The PRIME downwash placed in the docket for this action. See Response: We do not understand what
technique was never formally adopted reference document FFFFF. the commenter is referring to when they
by EPA for use in ISC. In order for states (e) Comment (MSCC, ExxonMobil):
indicate ‘‘only the current actually
to employ this technique, EPA regional Out-of-date and Invalid Emissions
existing emission sources * * * should
offices needed to authorize its use on a Rates. Federally enforceable emission
be used as inputs to the dispersion
case-by-case basis until ISC was rates from refinery consent decrees have
model.’’ With respect to geographical
replaced as the reference model on not been included in the FIP modeling.
coordinates used in the modeling, they
December 9, 2006. The plume rise EPA has used 10-year-old emission
were provided by the sources in
technique used in ISC was the inventory data that compromise the
response to EPA’s CAA Section 114
recommended approach at the time the accuracy of the results. Reductions that
information request. The incorrect
State developed the SIP, and the have occurred in the past ten years have
source coordinate for MSCC in the
technique served the modeling been ignored. The settlement documents

modeling files has been corrected.12 On
community well for many years. related to the 1998 SIP contain
(c) Comment (MSCC, ExxonMobil): requirements that substantially change 12 In reference document WW, Technical Support
Modeling Violates EPA’s Own the SO2 emission limits, and, therefore, Document, Dispersion Modeling to Support Sulfur
Requirements. The modeling used for the results of any modeling Dioxide (SO2) Emission Limits in Federal
July 13, 2007, the revised modeling files height value we used in our modeling background concentrations of 11 µg/m3
were indexed in the electronic docket was incorrect. are considered.
contained on http:// (i) Comment (ExxonMobil): The (k) Comment (MDEQ): Montana
www.regulations.gov and a compact meteorological data to be used as input continues to affirm the use of the ICS3
disk containing the modeling files was to the dispersion model should reflect model.
placed in the docket for this action. See the most representative information. Response: We acknowledge receipt of
reference document FFFFF. To the The meteorological set to be used the comment and the support for the
extent the commenter is asserting that should be chosen based on availability model used.
actual emission rates should be used as and based on those monitored (l) Comment (ExxonMobil): EPA has
inputs to the dispersion model, we parameters that are best able to take full not used current accurate process and
respectfully disagree. As described more advantage of the latest dispersion meteorological inputs in its modeling.
fully in our response to comment modeling techniques. This is contrary to EPA’s assurance in
II.E.1.(e), above, potential emissions Response: EPA believes that the its May 2002 final rule that: ‘‘Any future
rather than actual emissions are used in meteorological data from the Billings modeling in the Billings/Laurel area
SO2 attainment demonstrations, per airport that was used in the SIP/FIP should incorporate all corrections. The
longstanding EPA policy and 40 CFR SIP limitations are based on the best
modeling is representative of conditions
part 51, Appendix W requirements. information available at the time the
within the modeling domain. The
Accordingly, in our attainment attainment demonstration was modeled,
Billings airport is located in an open
demonstration, we modeled the and the same will be true for any FIP
area with good exposure to prevailing
emission limits we approved in the SIP limitations that are developed.’’ 67 FR
wind flow and has a long period of
and any new emission limits we are 22189. Also, in its May 2002 final rule,
record. Five years of historical weather
promulgating in the FIP. Thus, with the EPA stated that: ‘‘We agree that future
data (1984, 1986, 1987, 1988, and 1989)
exception of certain units at MSCC, we modeling should include all corrected
were used in the modeling to ensure
modeled the same emission rates that data.’’ 67 FR 22189. However, EPA has
that the full range of possible
the State used in its SIP modeling. ignored critical factual data for purposes
meteorological conditions were
of developing the proposed FIP.
(h) Comment (ExxonMobil): Only the evaluated in the modeling. To our Response: The commenter ignores the
verified actual stack heights should be knowledge the Billings airport data have context and meaning of EPA’s
used as inputs to the dispersion model. the longest period of record of any site statements in its 2002 SIP action. The
Response: Stack height regulations in the Billings area. When the State cited quotes were part of our response
determine the stack height values that developed the SIP modeling approach to specific comments from one source
are used as inputs to dispersion models that EPA has now used for the FIP, the that there were errors in the State
in SIP attainment demonstrations. In State tested ISC model performance modeling numbers used for that source’s
some cases this value may not be the using the Billings airport data. That stack parameters. The comment was:
same as the actual stack height. See 40 evaluation showed acceptable model ‘‘CEMS data now indicate an error in
CFR 51.118. For example, under our performance. the assumed buoyancy flux for MSCC’s
stack height regulations, 65 meters is the (j) Comment (ExxonMobil): EPA main stack; the current modeling
appropriate stack height value for should be modeling emission rates to protocol contains an assumption which
MSCC’s SRU stack, even though the levels that predict values slightly less significantly underestimates the average
stack is 100 meters tall. We believe we than the NAAQS. This modeling rise in emissions. Any revised modeling
have used the correct stack height concept is referred to as ‘‘pushing the should correct this assumption.’’ 67 FR
values in all cases, and the commenter model to failure.’’ This approach is 22189. We were merely agreeing that
did not indicate that any specific stack designed to determine the maximum future modeling should include
emission limits allowed by regulation corrected stack parameters based on
Implementation Plan (FIP) for Billings/Laurel, under acceptable modeling protocol. By CEMS measurements: ‘‘CEMS
Montana, June 2006, we indicated that one proposing mass emission limits on
suggested change that was not incorporated into the measurements of flow and temperature
EPA FIP modeling involved the coordinate system
flares of 150 pounds of SO2 per 3-hour data provide the best estimates of stack
used in the model to identify source location. period or 500 pounds of SO2 per parameters, and values based on CEMS
MDEQ developed the original source locations calendar day, EPA has chosen to use, data should be used in any future SIP
based on the UTM NAD27 (North America Datum without further consideration, mass
of 1927) coordinate system, and EPA has retained modeling for Billings provided the
that coordinate system in our modeling. It appeared
emission limits that do not ‘‘push the CEMS data are accurate.’’ Id. We were
that several of the suggested changes to source model to failure’’ but instead arbitrarily not indicating we would use a new
locations were based on NAD83 values. The newer limit the sources to mass emission model, different meteorological data, or
coordinate system can affect source locations by up limits that go far beyond protecting the
to 200 meters. In dispersion modeling on the scale consider entirely new structures. In fact,
of the current modeling domain, consistency
NAAQS. on the same page of our 2002 notice, we
between the source and receptor locations is the Response: Emission inputs to the said the following:
most important consideration. For this reason, model were established using criteria
suggested changes that appeared to be based on the ‘‘In addition, dispersion models and data
contained in 40 CFR part 51, Appendix bases are continually being improved. The
NAD83 were not included in the modeling.
However, changes that address local inconsistencies
W, Section 8. The emission limits set by task of demonstrating attainment could never
in measured distances between fixed stacks (such the modeling analysis are based on be completed if we or the State were
as at MSCC) on a specific property were emission rates that would just meet the compelled to update the analysis with each
incorporated in EPA’s modeling using UTM NAAQS. They are not based on new refinement. For the FIP, we intend to
NAD27. Sensitivity testing of the model showed continue to use ISC2 as the applicable model
that even the NAD27/NAD83 differences did not
‘‘arbitrary limits’’ that go ‘‘far beyond
protecting the NAAQS’’. For example, to fill in the gaps in the State’s attainment
significantly affect total predicted concentrations;
demonstration created by our disapproval of
the principal effect was, in some instances, to shift with the limits we are establishing and
the location of the maximum impact to a different the emission limitations for MSCC’s 100-
the SIP limits we approved, our meter stack. Some source parameters have
receptor. An electronic record (compact disk) of
EPA’s sensitivity testing of the model is contained
modeling resulted in a high value of 354 been corrected since the 1994 modeling
in the docket. See reference document EEE in µg/m3 which would exactly meet the 24- analysis (see Response V.D.4.(d), above), but
Docket Number EPA–R08–OAR–2006–0098. hour SO2 NAAQS of 365 µg/m3 when we intend to use the same meteorological
data and modeling protocols the State used, (c) Comment (MSCC, ExxonMobil): by EPA in the model runs were ignored
so that the results will be comparable.’’ Missing Modeling Files. Three source by IGM (in favor of other information)
For a more complete discussion of our input files (SRI files) were not included and therefore of no value. Instead,
basis for selecting the model and data in Reference Document EEE, the basis specific building data (discussed above)
inputs we have used, please refer to the for the modeling conclusion and the should have been entered into the
other responses to comments in this proposed emission limit for MSCC’s 100 program. There is at least one
section II.E, our proposed FIP, and our meter stack. It appears that these files substantial building, the YELP coke
TSD for the proposed FIP. were actually used in model runs. barn, that should have been included in
Response: We have added the the 2006 model runs.
2. Technical Issues referenced modeling files. On July 13, Response: See responses to comments
(a) Comment (MSCC, ExxonMobil): 2007, the revised modeling files were II.E.1.(a) and II.E.2.(e), above. As noted
Incorrect Source Location. The location indexed in the electronic docket above, to the extent possible, EPA is
of the small boiler stacks at MSCC that contained on http:// using the model inputs and model
are modeled as a volume source is www.regulations.gov and a compact settings selected by the State at the time
incorrect. The error occurs by the nature disk containing the modeling files was of SIP preparation and used in the IGM
in which the X and Y coordinates are placed in the docket for this action. See code. The model input selections reflect
entered into the SRI file. The entry is off reference document FFFFF. modeling practice and conditions at the
by one column. (d) Comment (MSCC, ExxonMobil): time of the SIP. The coke barn did not
Response: This has been corrected. Hanging Modeling Files. A source input exist at the time the SIP was prepared.
On July 13, 2007, the revised modeling file (ref_5t.sri) is included in Reference HB and PW values reflect the
files were indexed in the electronic Document EEE. However, this input file dimensions of the facilities that had
docket contained on http:// does not appear to be used in any input large structures nearby and that MDEQ
www.regulations.gov and a compact (RUN) and output files (OPF) files. It is included for downwash processing in
disk containing the modeling files was not possible to comment effectively on their SIP modeling. While the
placed in the docket for this action. See the adequacy of the model without commenter is correct that, in the IGM
reference document FFFFF. knowing the file’s purpose. model, these values were superseded by
(b) Comment (MSCC): Incorrect Response: This was a test file other data, obtaining these values was
Emission Rate. Table 2 of EPA’s inadvertently included in the electronic useful to us as a screening tool, and
Dispersion Modeling Technical Support record. It has now been deleted. On July inputting these values into the model
Document shows the modeling value of 13, 2007, the revised modeling files did not affect the validity of the results.
136.21 g/sec for MSCC’s SRU-100-meter were indexed in the electronic docket (g) Comment (MSCC, ExxonMobil):
stack. An emission rate of 150.0 g/sec contained on http:// Compliance Analysis Not Valid. The FIP
was modeled in the majority of the EPA www.regulations.gov and a compact proposal notes that there is a ‘‘trigger
modeling. If the proposed emission disk containing the modeling files was point’’ of 500 lb/calendar day in various
limit of 3003.1 lb/3-hours (126.13 g/sec) placed in the docket for this action. See ‘‘settlements’’ between EPA and
is correct, then the number that should reference document FFFFF. refineries. The proposal goes on to
appear in both the table and the input (e) Comment (MSCC, ExxonMobil): assert that a modeling analysis was
files is 126.13 (g/sec) to be consistent Outdated Building Profile Data. The conducted assuming the flares emitted
with the emission limit. dispersion modeling runs do not SO2 at a rate of 500 lb/3-hours and that
Response: In the State’s original SIP contain up-to-date information the model demonstrated compliance to
modeling submittal there were 1,320 regarding building profile data. EPA’s this alternative. A review of the
modeling scenarios with various use of 10-year old historical data is not modeling files, however, indicates that
buoyancy flux combinations that were logical considering the agency requested the ‘‘controlling’’ model run that
tested, and it was determined that only and received certain building data in its defined MSCC’s emission limit for the
a few of these resulted in concentrations December 2003 request. 100-meter stack (modeled at 65 meters)
that threatened the NAAQS. EPA Response: Building profile data were did not include this 500 lb/3-hour flare
conducted screening to eliminate the current at the time the MDEQ prepared emission rate option.
need for refined modeling of those the SIP. EPA is not updating the inputs Response: We solicited comment on
scenarios where the NAAQS were not to reflect recent changes in building whether we should limit the flares to
threatened. The 150 g/sec emission rate dimensions or changes in dispersion 500 lbs of SO2 per calendar day. We
was used provisionally to determine models. We are simply correcting have not adopted that option. But, for
which modeling scenarios would result deficiencies in the MDEQ’s SIP purposes of the attainment
in the maximum ground level modeling. If we were to follow the demonstration, we modeled the 500 lbs
concentrations, and was not used to set commenters’ suggestion, we would have as if it were emitted over a 3-hour
MSCC’s proposed emission limit. Once to revisit the entire SIP, including SIP period rather than a calendar day. We
the appropriate modeling scenarios limits we approved. The CAA does not wanted to assure that if all the calendar-
were determined by EPA, only those require us to re-open the entire SIP. See day allowed emissions were emitted in
scenarios were used to conduct the response to comment II.E.1.(a), above. a 3-hour period, the 3-hour NAAQS
refined modeling to establish an (f) Comment (MSCC): Variable ‘‘HB’’ would still be protected. Those
emission limit of 126.13 g/sec. The and ‘‘PW’’ Not Used. In order to execute modeling files are contained in the
commenter is correct that there is a the FIP model, EPA requested source docket.
discrepancy between Table 2 in EPA’s specific information including the However, the controlling model run
Dispersion Modeling Technical Support modeling terms HB and PW. These that defined MSCC’s emission limit for
Document (reference document WW) values may be input into the IGM the SRU 100-meter stack was for the 24-
and the modeling input files. The input model, however, this information is hour NAAQS. There was no need to
files for the limited modeling scenarios superseded by direction-specific model the 500 lbs of SO2/calendar day
reflected the correct value, 126.13 g/sec. building parameters by the model while to show compliance with the 24-hour
Table 2 of the TSD contains the wrong executing in all cases (stacks) of interest. NAAQS since we had already modeled
value. In other words, the data that was coded the flares at 1200 lbs of SO2/calendar
day. Since attainment of the 24-hour developing the SIP should be used in public hearing was held in Billings,
NAAQS was shown at 1200 lbs of SO2/ the FIP. A stakeholder process will Montana, on August 10, 2006. EPA
calendar day, the area would still show allow all parties an opportunity to believes it provided sufficient time and
attainment at 500 lbs of SO2/calendar ensure that the best available opportunity for all commenters to
day. information is considered in provide comments on the proposed FIP.
formulating any proposed requirements.
F. Miscellaneous Comments Response: See response to comment 4. EPA’s Strategic Plan
1. Stakeholder Process II.F.1.(a), above. Comment (COPC): The proposed FIP,
which contains inflexible flare emission
(a) Comment (CHS Inc.): If EPA 2. Ripple Effect limits and strictly-specified monitor
intends to regulate malfunctions, (a) Comment (WETA): The commenter installations requirements, is
startups, and shutdowns, a stakeholder’s is concerned not only with the impact inconsistent with EPA’s Strategic Plan,
process should be used to accurately of the FIP on the refineries in the area which commits EPA to ‘‘finding
develop a reasonable flare limit. but the potential ripple effect on the innovative solutions and collaborating
Response: EPA announced its businesses, workers, and other with others.’’
proposed FIP in the Federal Register on consumers who daily use and depend Response: We acknowledge the
July 12, 2006, invited public comment, on the variety of products produced by commenter’s concerns. However, we are
and identified the time and place for a the petroleum refineries in the Billings/ charged with meeting the CAA’s
public hearing. A public hearing was Laurel area. requirement to assure that the SO2
held in Billings, Montana, on August 10, Response: We acknowledge the NAAQS are met and maintained.
2006. Only one person from industry commenter’s concerns. We recognize Accordingly, the FIP adopts flare
spoke at the hearing. Prior to the hearing that our FIP will result in costs to MSCC emission limits and compliance
and at the hearing itself, no one and the refineries, which they may or determining methods.
mentioned the concept of a stakeholder may not pass on to consumers. We have It should be noted that the discussion
process. In addition, we provided nearly tried to be sensitive to the costs MSCC on Innovation and Collaboration in the
four months for the affected facilities and the refineries may incur to meet the ‘‘2006–2011 EPA Strategic Plan,
and other members of the public to FIP’s requirements, which potentially Charting Our Course,’’ September 2006
submit written comments and would affect the costs of products to (reference document BBBBBB), pertains
suggestions regarding our proposed FIP, consumers. For example, where we to complex environmental challenges
including a substantial extension to our determined less costly methods to where broad-based problems cannot be
original 60-day comment period in an monitor SO2 concentrations could solved with conventional regulatory
attempt to reasonably accommodate achieve similar results, we are allowing controls. We do not think this is
State and industry requests. We have these other methods to be used. relevant here. We are merely
made a number of changes in response However, our ultimate charge under the establishing limits on flares and
to comments received. If the affected CAA is to protect the SO2 NAAQS, methods to determine compliance with
facilities had other ideas about how we recognizing that cost impacts to sources those limits.
could better structure the FIP, they had and consumers may occur. See, e.g., City
ample opportunity to express those 5. FIP Provisions in Title V Permits
of Santa Rosa v. EPA, 534 F.2d 150 (9th
concepts. Cir.1976), vacated and remanded on Comment (MDEQ): Montana
We have complied with the other grounds sub nom. Pacific Legal acknowledges that the FIP provisions, if
requirements of the CAA as set forth in Foundation v. EPA, 429 U.S. 990 (1976). promulgated, will be incorporated in
section 307(d) regarding public (b) Comment (citizen): The Title V permits. However, Montana
participation for the FIP. We are not commenter is a dryland farmer and uses expects EPA will take the lead on
required to hold a stakeholder process. an ammonium sulfate (thiasol) fertilizer, implementing and enforcing the FIP
Issues regarding malfunctions, startups, which is a by-product of the refinery provisions.
and shutdowns are addressed above. process. He says he is doing as much as Response: EPA intends to assume
(b) Comment (CHS Inc., ExxonMobil, he can to be environmentally primary responsibility to implement
MPA): It would be in the best interest of conscientious and not introduce metals and enforce the FIP. However, the FIP
all involved that a stakeholder process into the soils found in other fertilizers. requirements will be ‘‘applicable
be used to determine what, if any, This requires him to use the thiasol that requirements’’ under Title V, which,
enhancements to the Montana SIP are is refinery-produced. He requests that therefore, must be included in Title V
appropriate. EPA not exacerbate a bad situation for permits for the affected sources and be
Response: See response to comment agriculture, which increases costs to a enforceable by the State.
II.F.1.(a), above. major industry which is marginal in
(c) Comment (WETA, COPC): If the 6. Length of Time it Took EPA To
profitability and major in importance to Propose FIP
EPA feels strongly that consideration the State of Montana.
should be given to different controls for Response: See response to comment Comment (YVAS): Since the 1990
SO2, then a stakeholder process should II.F.2.(a), above. Clean Air Act requires NAAQS for SO2
be utilized to consider issues and to protect public health, YVAS deplores
relevant information in deciding if a 3. Extend Comment Period this ‘‘inadequacy [sic] and ‘‘non-
further SIP or FIP is necessary. Comment (COPC, ExxonMobil, MSCC, attainment’’ and deplores further that
Response: See response to comment WETA, YCC): Commenters asked for the EPA did not adequately and in
II.F.1.(a), above. additional time to comment on the timely fashion, take necessary steps to
(d) Comment (MSCC, ExxonMobil): proposed FIP, until at least December enforce the CAA’s provisions to protect
EPA has developed the proposed FIP in 11, 2006. the air quality in the Billings/Laurel
a vacuum as to the affected parties. It is Response: The public comment area in a reasonably suitable time period
inappropriate for EPA to not consult the period on the FIP proposal ran from July regardless of any mitigating
affected facilities in any meaningful 12, 2006, through November 3, 2006— circumstances. A specific justification
way. The process used by Montana in almost four months. Additionally, a explaining this lapse in EPA’s
responsibilities for not acting in the Nowhere in this FIP is there an attempt not seeking comment on any changes to
public interest is essential to the to address the issue of a further the NAAQS in this FIP action.
residents of the Billings/Laurel area reduction in the total emissions of the Two sections of the CAA govern the
given that at the time, the Billings/ industrial sources in the Billings/Laurel establishment and revision of NAAQS.
Laurel Sulphur Dioxide Area was area. Accordingly, YVAS believes that Section 108 (42 U.S.C. 7408) directs the
subject to excessive amounts—estimated all anti-lower SO2 emission arguments Administrator to identify pollutants
to be over 35,000 tons (1993)—of SO2 are irrelevant against the demand for which ‘‘may reasonably be anticipated
atmospheric pollution. protecting public health standards and to endanger public health or welfare’’
Response: We believe EPA’s SIP Call additional reduction of SO2 emissions is and to issue air quality criteria for them.
and subsequent State and EPA actions mandatory under the CAA. Failing to These air quality criteria are to ‘‘reflect
to address the SIP Call have helped address a further SO2 emissions the latest scientific knowledge useful in
reduce SO2 emissions in the Billings/ reduction should be considered another indicating the kind and extent of all
Laurel area. There is no question that serious breach of your responsibility to identifiable effects on public health or
this process has taken longer than it the Billings/Laurel public. Why did EPA welfare which may be expected from the
should have. not include a discussion towards presence of [a] pollutant in the ambient
reducing the total SO2 emissions in the air.’’
7. EPA Enforcement Section 109 (42 U.S.C. 7409) directs
Billings/Laurel Sulphur Dioxide area in
Comment (YVAS): YVAS insists that this FIP and since EPA did not include the Administrator to propose and
the EPA consistently monitor industry that discussion here, does EPA plan to promulgate ‘‘primary’’ and ‘‘secondary’’
emissions in order that industry sources do that and if so, when? NAAQS for pollutants identified under
continue to comply with the SIP and/or Response: The 1970 CAA established section 108. Section 109(b)(1) defines a
the ‘‘more stringent requirements under the air quality management process as a primary standard as one ‘‘the attainment
other provisions of the CAA’’ or ‘‘SIP- basic philosophy for air pollution and maintenance of which, in the
approved permit programs.’’ control in this country. Under this judgement of the Administrator, based
Response: EPA intends to take the system, we establish air quality goals on the criteria and allowing an adequate
lead in enforcing the emission limits (NAAQS) for criteria pollutants. States margin of safety, [is] requisite to protect
and monitoring requirements contained develop control programs (termed SIPs) the public health.’’ A secondary
in the FIP. Congress intended that states to attain and maintain these NAAQS. standard, as defined in section
have primary responsibility for Our fundamental obligation in the SIP/ 109(b)(2), must ‘‘specify a level of air
implementing and enforcing their SIPs. FIP context is to ensure that the NAAQS quality the attainment and maintenance
Additionally, states may take the lead in are met, not reduce emissions to zero. of which, in the judgement of the
implementing and enforcing other CAA Thus a reduction of SO2 emissions is Administrator, based on [the] criteria, is
programs (e.g., News Source mandatory only to the extent needed to requisite to protect the public welfare
Performance Standards (NSPS), attain the NAAQS. However, under from any known or anticipated adverse
Maximum Achievable Control section 116 of the CAA, states may effects associated with the presence of
Technology (MACT) standards, Title V adopt and enforce any air pollutant [the] pollutant in the ambient air.’’
permitting), either through EPA standard, limitation, or control Welfare effects are defined in section
delegation or program approvals. In the requirement so long as it is no less 302(h), 42 U.S.C. 7602(h), to include
latter cases, we have an oversight role stringent than that required by the CAA. ‘‘effects on soils, water, crops,
and may take enforcement action under Put another way, states can require that vegetation, manmade materials,
section 113 of the CAA for violations of the air be cleaner than the NAAQS. Our animals, wildlife, weather, visibility and
a SIP or other CAA requirements when goal in the FIP is to ensure attainment climate, damage to and deterioration of
a state does not take action or when its of the SO2 NAAQS. property, and hazards to transportation,
action is considered ineffective. as well as effects on economic values
EPA Region 8 communicates regularly 9. SO2 NAAQS and on personal comfort and well-
with the MDEQ regarding sources. We (a) Comment (YVAS): Nowhere in this being.’’
have regular meetings with MDEQ FIP is any reference made to what clean On April 30, 1971 (reference
regarding sources that are violating air standards should be under the CAA document CCCCCC), the Environmental
emission limit requirements and discuss or NAAQS. Commenters should have Protection Agency (EPA) promulgated
the MDEQ’s proposed or ongoing been informed as to those standards in primary and secondary NAAQS for
actions to address these violations. We this FIP in order to fairly judge as sulfur oxides (SOx) (measured as SO2)
intend to continue to carry out our acceptable or non-acceptable the release (then codified as 40 CFR 410.4 and
oversight responsibility for the SIP and standards proposed for the sources in 410.5). The primary standards were set
other CAA requirements for the this FIP. How can the public adequately at 365 micrograms per cubic meter (µg/
Billings/Laurel sources. If we determine comment on clean air issues when those m3) (0.14 parts per million (ppm)),
that the MDEQ is not taking appropriate standards are unknown to the public? averaged over a 24-hour period and not
action for violations of the SIP, or other Further, referring the general public to to be exceeded more than once per year,
CAA requirements, we will take sources where those standards would be and 80 µg/m3 (0.03 ppm) annual
appropriate action. found is a disservice to the public since arithmetic mean. The secondary
many of those sources of such standard was set at 1,300 µg/m3 (0.5
8. Further Emission Reductions information may be unattainable or ppm) averaged over a period of 3 hours
Comment (YVAS): Although the unavailable. and not to be exceeded more than once
industry is attaining lower yearly Response: The July 12, 2006, per year. In accordance with sections
decreases of SO2 since 1994, with proposed FIP did identify the 24-hour 108 and 109 of the CAA, in the 1990’s,
presumably a better and ‘‘healthier’’ air and 3-hour SO2 NAAQS under the EPA reviewed and revised the health
quality in the area thereby, the modeling discussion (71 FR 39259, and welfare criteria upon which these
assumption logically follows that starting at 71 FR 39270, col. 1). The SO2 primary and secondary SO2 standards
industry should be required to comply NAAQS were previously established were based. On April 21, 1993 (58 FR
with further reduced SO2 release levels. (see discussion below), and EPA was 21351) (reference document DDDDDD),
EPA announced its final decision under contains the Montana ambient air public disclosure and the administrative
section 109(d)(1) of the CAA that the quality standards (MAAQS). The requirements for proposing the FIP. We
revisions of the secondary SO2 NAAQS MAAQS are not contained in the are announcing this final FIP in the
were not appropriate at that time. On federally-approved SIP; the CAA does Federal Register as well. A discussion
May 22, 1996 (61 FR 25566) (reference not require that the standards be in the of the SO2 NAAQS is provided above.
document EEEEEE), EPA announced its federally-approved SIP. The SO2 (b) Comment (Citizen): Plans for
final decision under section 109(d)(1) of MAAQS are contained in ARM 17.8.210 controlling emissions ‘‘at the source’’
the CAA that the revision of the primary (see reference document FFFFFF) and must be provided by the EPA at any
SO2 NAAQS was not appropriate at that are as follows: (1)(a) Hourly average— public meeting announced by the EPA
time. EPA is currently reviewing the 0.50 ppm, not to be exceeded more than and those plans should be announced
primary and secondary standards again 18 times in any 12 consecutive months; publicly in advance of the meeting in
to determine whether they should be (1)(b) 24-hour average—0.10 ppm, not to order for the public to understand what
revised. be exceeded more than once per year; the effects and results of such plans will
The Code of Federal Regulations and (1)(c) annual average—0.02 ppm, be on the air shed quality of the
(CFR) is available at most public not to be exceeded. The 24-hour and Billings/Laurel metropolitan area.
libraries and on the internet at: http:// annual SO2 MAAQS are more stringent Response: See response to comment
ecfr.gpoaccess.gov/. Likewise, the CAA than EPA’s 24-hour and annual SO2 II.F.11.(a), above.
is also available at most public libraries NAAQS. The State has a 1-hour average 12. Stack Height
and on the internet at EPA’s Web site: SO2 MAAQS and EPA has a 3-hour
http://www.epa.gov/air/caa/. average SO2 NAAQS. The State does not (a) Comment (Citizen): Included in
(b) Comment (citizen): The rejection require that plans be developed to EPA’s emission control plans must be a
of Montana’s Plan to control air quality assure attainment and maintenance of stringent requirement that none of the
in the Billings/Laurel air shed 4 years the MAAQS, whereas, EPA does require three area refineries or the Montana
previously has left a serious gap in the plans to assure that the NAAQS are Sulphur and Chemical company may
air quality in this air shed. attained and maintained. construct any emissions stack or flaring
Response: We acknowledge this (c) Comment (Citizen): Commenter system of 100 meters or higher.
comment. See response to comment works the evening shift near the Information concerning the probable
II.F.6., above. industrial sector and the refineries and effects, distance, wind patterns, content
the coke plant. He notices that at night etc. of the dispersal plumes of stacks of
10. SO2 Health Effects this height should be provided to the
the air becomes more sour. Depending
(a) Comment (Citizen): The air is so upon which way the wind is blowing or public at any hearing in order that
bad near the commenter’s house that whatever is occurring in the area, it will public comment on this crucial aspect
she needs to close the windows. She has burn his eyes and nose. It will start to of the emission control plan may be
headaches and burning eyes and burn his lungs and inflame his chest properly analyzed. Under no
sinuses. How safe is it for the families? and it will make it harder for him to circumstances should the 100-meter
Commenter is concerned that air breathe. The air is like a smoke-filled height be considered as a minimum
emissions affect landscape and river barroom. He used to live in this area as permissible height by the EPA or by the
areas. Commenter would like EPA to well. Commenter feels it degrades the companies involved for any stack or
assure that refineries do not off-gas quality of his life. He’s standing up for flaring system.
unmeasureable blasts of pollution as she his lungs. Response: EPA does not restrict the
has seen them do over her water, Response: See response to comment physical height of a smoke stack. See 40
county, and home. II.F.10.(a)., above. CFR 51.118(a). However, we do restrict
Response: We acknowledge this the credit a company receives for its
comment. The FIP, along with other 11. Public Process stack height in the modeling used to
requirements contained in the SIP, will (a) Comment (Citizen): Since there has determine whether a SIP will meet
provide an enforceable mechanism to been no public disclosure of the EPA’s national standards for specific air
assure that the SO2 NAAQS in the plans for complying with the standards pollutants. Id. The stack height credit is
future will be protected in the Billings/ (considered as minimal by local public based on the greater of the following: (1)
Laurel area. Since EPA initially health advocates) as set forth in the A height of 65 meters, (2) a height based
requested the State to revise the National standards (which also have not on a formula that considers the
Billings/Laurel SO2 SIP, actual SO2 been provided publicity to create public surrounding buildings, or (3) a height
emissions from companies have been awareness of those standards), the EPA based on technical modeling studies
cut by more than half and there have should not proceed with any rule which show a certain height is
been measured improvements in air making unless the public receives an necessary to avoid high levels of
quality. The SIP and FIP contain an opportunity to comment. pollutants in the nearby area. See 40
enforceable control strategy to help Response: EPA announced its CFR 51.100(ii).
ensure that the SO2 NAAQS are attained proposed FIP in the Federal Register on EPA has rules that apply to tall stacks;
and maintained. July 12, 2006. In the July 12, 2006 otherwise, companies could avoid
(b) Comment (Citizen): Since national Federal Register notice, EPA provided installing needed pollution control
air quality standards are more stringent for the opportunity of a public hearing. equipment. Industry could simply build
than Montana requires, serious health A public hearing was held in Billings, higher stacks and emit into the air
risks to area residents is probable and Montana on August 10, 2006. At the additional pollutant levels that would
cannot be ignored. hearing, EPA discussed its proposed not violate local air quality standards,
Response: See response to comment FIP. Additionally, EPA’s proposed but could eventually affect the air
II.F.10.(a), above. Note that the State’s notice indicated that detailed quality of communities farther
ambient standards, in some cases, are information regarding the proposed FIP downwind. This is because the higher
more stringent than the national was available on the Internet. We have the stack height, the greater the
standards. Subchapter 2 of the complied with the requirements of dispersion of pollutants and the less
Administrative Rules of Montana (ARM) section 307(d) of the CAA regarding likely they will reach the ground in the
vicinity of the stack. EPA does allow Response: We acknowledge receipt of have been adopted, approved, and
increases to stack height credits when the comment and the support for our implemented. The thousands of other
the stacks meet the conditions noted proposal. emission limitations nationwide are
above. based on a single fixed buoyancy flux
14. SIP Escape Clause
EPA disapproved part of the Billings/ value similar to what we proposed for
Laurel SO2 SIP because MSCC’s stack Comment (MSCC): The SIP contains MSCC.
height credit did not meet the an important ‘‘escape clause’’ by which (b) Comment (MSCC): Complicated to
conditions noted above. EPA believes there was a general agreement that if the Model. (i) MSCC agrees that it is more
that the appropriate stack height credit State provided more favorable treatment complicated to model a variable
for the MSCC SRU 100-meter stack is 65 to one facility, the same accommodation emission rate than a fixed emission rate.
meters. The 65-meter stack height credit would be offered to the other facilities. That alone is not sufficient reason to
was used in the modeling for the FIP. The present proposed FIP which deny MSCC the variable emission rate.
We did not identify any other concerns proposes to reduce MSCC’s stack height Also, much has changed since the
with the stack height credit used for credit and drastically reduce MSCC’s original modeling effort. Computer
other sources in the SIP. emission limits will violate that clause. speed, memory, data handling, and
This unwarranted intrusion into a storage are all improved.
(b) Comment (Citizen): Studies,
carefully-bargained agreement among Response: Modeling was one of the
including wind roses of the dispersal
multiple parties, violates both the letter reasons we offered for not providing a
pattern of all stacks of 65 meters and
and the spirit of the CAA. variable emission limit; however, it was
higher should be provided to the public
Response: We are not bound by the not the only reason. Although computer
at a hearing of the final FIP, in order that
escape clause that the State approved; in speed, data handling, and storage are
the public comment on this crucial
fact, we disapproved this aspect of the improved since the MDEQ developed
aspect of the emission control plan may
SIP. See 67 FR 22168, May 2, 2002. the Billings/Laurel SO2 SIP, there would
be properly analyzed.
Instead, we are obligated to correct the still be a considerable effort on EPA’s
Response: The CAA directs EPA to portions of the SIP we disapproved. We
take public comment on proposed FIPs, part to model a variable emission limit
disapproved MSCC’s main stack for the SRU 100-meter stack. Therefore,
not final FIPs. See CAA section 307(d). emission limits because they were based
EPA’s modeling studies for the we used EPA’s historical practice of
on inappropriate stack height credit. selecting mean values of historical data.
proposed FIP were contained in the The FIP establishes new limits for
docket for the proposed FIP and Individual stationary sources in SIP
MSCC’s main stack that are consistent attainment demonstrations are typically
available for review during the comment with our modeled attainment
period on the proposed FIP. modeled assuming a single
demonstration, based on a Good representative value for the model input
Additionally, on July 13, 2007, the Engineering Practice (GEP) stack height
revised modeling files were indexed in parameters that affect plume rise. Model
credit of 65 meters. While it is not clear input parameters that affect plume rise
the electronic docket contained on to us how this violates the State-
http://www.regulations.gov and a include stack gas temperature and
approved escape clause, setting volume flow, or buoyancy flux. If
compact disk containing the modeling emission limits for MSCC’s main stack
files was placed in the docket for this emissions are held constant, ground
consistent with our stack height level concentrations would tend to
action. See reference document FFFFF. regulations and necessary to decrease during periods with higher
13. General Support demonstrate attainment of the NAAQS plume rise associated with elevated
does not violate the CAA. On the stack gas temperature and increased
(a) Comment (Citizen): The contrary, setting such limits is required
commenter wants to lend support to stack flow velocities. Conversely,
by the CAA, regardless of the State- ground level concentrations would tend
what EPA is trying to do here and the approved escape clause.
proposals that EPA is making, and he to increase during periods with reduced
thinks it is very much on target and for G. MSCC Specific Issues stack gas temperatures and stack flow
his benefit, and he would hope the velocities. The State opted to set
1. Variable Emission Limit emission limitations based on variable
industries who are being regulated in
this sense will find a way to make it (a) Comment (MSCC): EPA offers buoyancy flux values for three of the
worth their while to do it also. surprisingly little discussion as to why sources. MDEQ identified a total of 11
a variable limit was not proposed for buoyancy flux modeling scenarios for
Response: We acknowledge receipt of
Montana Sulphur. EPA’s reasoning MSCC, 12 for ExxonMobil, and 10 for
the comment and the support for our
seems to ignore that MSCC has been the Corette Power Plant. Modeling all
proposal.
operating under a variable emissions possible combinations of scenarios
(b) Comment (Citizen): Commenter limit that has been modeled, monitored, required the State to model a total of
encourages EPA to carry on the work we and enforced for close to a decade. 1,320 combinations for each year of
have been doing, to encourage Response: EPA’s reasoning for not meteorological data processed. EPA
movement in the positive direction of offering a variable limit is discussed in used a fixed buoyancy flux value for
reducing emissions. the July 12, 2006, proposal notice (see modeling MSCC and that reduced the
Response: We acknowledge receipt of 71 FR 39259, starting at 39268, col. 2) number of potential modeling scenarios
the comment and the support for our and reference document WW to 120. EPA reviewed the modeling
proposal. ‘‘Technical Support Document’’ results in the State’s attainment
(c) Comment (Citizen): Commenter contained in EPA Docket No. EPA–R08– modeling to identify which scenarios (of
appreciates the changes that EPA is OAR–2006–0098. Additionally, to our the 120 possible scenarios) would
making and thinks the people in knowledge, the SIP limits for two produce the highest concentrations.
Billings deserve them. Commenter feels sources in Billings (ExxonMobil and Based on this selection process, EPA
the industries need to step up to the Montana Power) are the only instances modeled approximately 50 scenarios in
plate and be responsible for their in the United States where variable the FIP modeling, and we believe that
emissions. emission limits based on buoyancy flux these scenarios represent the limiting
(i.e. maximum predicted concentration) to place an increased burden on determined through modeling as the
case. ourselves to enforce a variable limit. See limit needed to assure attainment of the
(ii) It is completely arbitrary to create, also the response to comment II.G.1.(e), SO2 NAAQS. Since the NAAQS are
model, approve, monitor, and enforce below. health-based standards, as a general
variable limits at other Billings facilities (e) Comment (MSCC): Variable Limit matter, SIPs/FIPs must assure
but to deny the same courtesy for MSCC is Better Science. Though it involves attainment of the NAAQS on a
claiming that it is, in this case alone, too incremental initial work, from a continuous basis.
complicated a modeling effort. modeling perspective, the use of We note that apparently MSCC was
Response: Again, modeling was not variable limits is better science. It able to conduct maintenance on the
the sole reason for not providing a replaces a false assumption in modeling SuperClaus unit in 2003, 2004, and
variable emission limit for MSCC’s SRU (constant, average stack conditions 2005 without exceeding the proposed 3-
100-meter stack. Although EPA under all operating scenarios) with hour and 24-hour FIP SRU 100-meter
approved the variable emission limits at factual information so that plume limits. MSCC may be able to perform its
other Billings facilities, we did so with height, which is variable, can be more maintenance on the SuperClaus unit
reservations. (See our July 28, 1999, accurately represented. Plume height, when other process equipment at
proposed rulemaking action on the just like mass emissions, is normally ExxonMobil is down for maintenance.
Billings/Laurel SO2 SIP, 64 FR 40791, variable and is critical to calculation of Additionally, we understand that MSCC
starting at 40794, col. 3, and our May 2, downwind concentrations. intends to install a second SuperClaus
2002, final rulemaking action, 67 FR Response: In addition to looking at air unit to provide redundancy to the
22168, starting at 22206, col. 2, for a full quality impacts of the FIP, we also need existing SuperClaus equipment.
discussion of our concerns with the to assure that the FIP is enforceable. Installation is expected to begin in the
variable emission limit concept.) Since Although we may agree with the fourth quarter 2007, at the earliest
EPA is taking the lead in establishing commenter that the variable emission (reference documents GGGGGG and
emission limits for MSCC’s SRU 100- limitation will result in fewer emissions BBBBBBB). Concerns about additional
meter stack and will take the lead in when the buoyancy of the plume is emissions during maintenance should
enforcing the FIP, EPA has chosen not lower, it will also result in higher be eliminated with the addition of a
to model and provide a variable emissions when the buoyancy of the second SuperClaus unit.
emission limit. We believe our exercise plume is higher. Additionally, a variable
of discretion so as to simplify FIP emission limit is more difficult to 2. 100-Meter Stack Height Credit and
development and enforcement is enforce. Granted the same instruments Emission Limit
reasonable, particularly where the data would be used to determine compliance (a) Comment: MSCC submitted
indicate MSCC will be able to comply whether the emission limit is fixed or summary comments regarding its
with a fixed emission limit without variable. However, in addition to position concerning good engineering
additional controls and where fixed confirming that the source is in practice stack height credit for the 100-
limits are the norm in SIPs throughout compliance with a variable emission meter SRU stack. MSCC noted that these
the country. limit, agencies will also need to confirm comments had generally been submitted
(c) Comment (MSCC): Complicated to that the variable emission limitation previously to both EPA and Montana.
Monitor. Buoyancy flux has been was determined correctly. Therefore, we MSCC claimed that it has not received
measured and reported to DEQ for a believe that variable emission limits the proper stack height credit for the
period of about eight years, with very increase the workload and add a layer 100-meter SRU stack in the proposed
high reliability. It is simply illogical to of complexity that is not found with FIP.
argue or imply that monitoring fixed emission limitations. Because of Response: EPA disapproved the
buoyancy flux is a task not worthy or this enforcement complexity, we do not State’s determination of stack height
too complicated in nature. One cannot agree with the commenter that variable credit for MSCC’s 100-meter SRU stack
deny the historical evidence that it has emission limitations are a superior on May 2, 2002 (67 FR 22168). In the
been measured successfully for many approach to setting emission May 2, 2002, notice, starting on page
years and that it does not require any limitations. 22209, we responded to all the stack
monitor instrumentation not already (f) Comment (MSCC): Fixed Limit height comments MSCC previously
required to measure sulfur dioxide. Compliance. Although MSCC has been submitted. We hereby incorporate by
Response: See response to comment able to meet the proposed FIP limit for reference our responses from that
II.G.1.(b)(ii), above. several years, it must be noted that notice. We indicated in the May 2, 2002,
(d) Comment (MSCC): Complicated to MSCC has not always been able to notice that ‘‘[w]e considered the
Enforce. EPA’s reason for not proposing operate within such limits, and that comments received and still believe we
a variable limit for MSCC due to MSCC was not operating its sulfur plant should finalize our proposed
enforcement is puzzling. If EPA at maximum capacity during the time disapproval of the MSCC’s stack height
approved variable emission limits for periods cited by EPA. The primary credit and SRU 100-meter stack
other sources, even though the same reason MSCC can operate under EPA’s emission limitations. None of the
enforcement concern exists, it should proposed limit arises from MSCC’s adverse comments has convinced us
also be approved for MSCC. voluntary installation of SuperClaus TM. that our interpretation of the CAA and
Response: The State developed the The SuperClaus unit must be shut down our regulations is unreasonable or that
original SIP that allows variable periodically for repair. MSCC needs the we should change our proposed course
emissions for several sources. The State variable limit to be in compliance when of action.’’ See our May 2002 final
takes the lead in enforcing the SIP, and SuperClaus unit is shut down. MSCC action (67 FR 22168). EPA has
EPA takes an oversight role. EPA should not be punished for its good determined that the GEP stack height
approved portions of the SIP, including behavior by requiring control credit for the 100-meter SRU stack is 65
variable emission limits at two sources, technology and lower emissions than is meters and has used that height in
and we did so with reservations. Since necessary to maintain NAAQS. establishing the 100-meter SRU stack
we would be taking the lead in Response: EPA’s proposed FIP limit emission limit. Our stack height
enforcing the FIP, we have chosen not for MSCC’s SRU 100-meter stack was regulations, codified at 40 CFR 51.100
and 51.118, provide that the degree of emission limit we proposed for the SRU we have revised the method by which
emission limitation required for 100-meter stack, and that we are now MSCC shall determine the H2S content
pollutant control under an applicable promulgating in the FIP, assures that the of the fuel burned. Our final FIP
SIP shall not be affected by stack height annual SO2 NAAQS will be attained and indicates that on a once-per-3-hour
in excess of GEP stack height. The maintained. Additionally, the 3-hour period frequency until no heater or
central component of the regulations and 24-hour SO2 NAAQS are more boiler is exhausting to the 30-meter
consists of definitions of the term ‘‘good controlling than the annual SO2 stack or an auxiliary vent stack, MSCC
engineering practice stack height.’’ GEP NAAQS. This means that more stringent shall determine the H2S content of the
stack height is the greater of (1) 65 emission limits must be placed on fuel burned using length-of-stain
meters (known as ‘‘de minimis’’ stack stacks to assure that the 3-hour and 24- detector tubes with the appropriate
height), (2) the height calculated using hour SO2 NAAQS are attained and sample tube range pursuant to ASTM
a formula specified by regulations maintained than would be required to Method D4810–06, ‘‘Standard Test
(‘‘formula height’’), or (3) the height assure that the annual SO2 NAAQS are Method for Hydrogen Sulfide in Natural
demonstrated using fluid modeling or a met. Gas Using Length-of-Stain Detector
field study (‘‘non-formula height’’ or (c) Comment (Citizen): Commenter Tubes’’ (see reference document
‘‘above-formula height’’). See 40 CFR appreciates the logic of not allowing UUUUUU). The final FIP indicates that
51.100(ii)(1)–(3). Prior to our SIP action, increases in stack height credit. if the results exceed the tube’s range,
the State calculated the formula height Response: We acknowledge the another tube of a higher range must be
for the SRU 100-meter stack to be 47.8 support for our proposal. Also, please used until results are in the tube’s range.
meters (see reference documents see our response to comment II.G.2.(a), (b) Comment (MSCC): Emission
VVVVVV and WWWWWW). Per our above. Limit—100 ppm H2S—A Redundant
regulations, since this is lower than 65 3. 30-Meter Stack and Auxiliary Vent Limit. Having both a 12 lb/3-hour limit
meters, GEP stack height is 65 meters. Stack and 100 ppm H2S limit creates double-
We have not received any new jeopardy. Both limits are for solely and
information to indicate formula height (a) Comment (MSCC): Emissions exactly the same thing. If a particular 3-
should be higher than 47.8 meters, nor monitoring for 30-meter Stack and hour period were to indicate 120 ppm,
have we received a valid demonstration Auxiliary Vent Stacks. EPA has it would be in violation of both limits.
for above-formula stack height credit. proposed unnecessarily complex, This could (and is very likely to) occur
See our proposed and final actions on redundant, and unneeded monitoring even if the units were not, in fact,
the Billings/Laurel SO2 SIP, 64 FR and reporting requirements for both the operating anywhere near an actual
40791 (July 28, 1999) and 67 FR 22168 30-meter stack and the auxiliary vent emission rate of 12 lbs/3-hours. This
(May 2, 2002), respectively. In light of stacks. The emissions from these units result is overkill and is not appropriate
our prior decision on the fluid modeling have minimal impact on model results. or necessary for protection of the
in the SIP action, and in the absence of These predicted concentrations are less NAAQS.
a new, valid, GEP stack height than 1% of the NAAQS. The emission Response: In our FIP proposal, we
demonstration, it would be limit applicable is miniscule in were attempting to simplify the method
inappropriate in this FIP for us to use comparison with other uncertainties in to determine compliance with the mass
a stack height value for MSCC that is the implementation plan. Emissions emission limits. The assumption in the
inconsistent with our prior action. from these units, although authorized, proposal was that if the H2S
(b) Comment (YVAS): YVAS believes are infrequent. Venting to the boiler concentration was below 100 ppm H2S,
the annual emission limit of 9,088,000 stack is generally associated with events then the source would be in compliance
lbs of sulphur is too excessive because such as maintenance. For operational with the mass emission limits. We were
YVAS believes this ‘‘proposed’’ reliability and flexibility, MSCC needs not trying to create ‘‘double jeopardy’’
emission to be a major contribution to to be able to vent these boilers locally. for MSCC. It appears that the
the total emissions of sulphur dioxide in Monitoring these units is an expense commenter believes the 100 ppm H2S
the Billings/Laurel area and is, and requirement that serves no real or limit is too restrictive because the
therefore, not acceptable. In addition, useful purpose. Essentially the same source could be in compliance with the
EPA states that: ‘‘We (EPA) are information is already gathered under mass emission limit but out of
proposing fixed emission limits rather the State plan. compliance with the ppm limit.
than variable emission limits on MSCC’s Response: As we indicated in our July In our final FIP we are keeping the
SRU 100 meter stack because they are 12, 2006, proposed FIP (71 FR 39259, simplified method to determine
less complicated to model monitor and 39268), it is necessary for EPA to require compliance with the mass emission
enforce.’’ This proposal is inadequate methods to assure that the emission limits. We believe determining direct
and does not address the continuing limits for the 30-meter stack and compliance with the mass emission
high total SO2 emission limits you auxiliary vent stacks are met. However, limits would either require additional
intend permitting MSCC to continue to since MSCC has already established a monitoring equipment or methods and/
release. method to monitor these emissions or would be unreliable due to potential
Response: Stack emission limits are using length-of-stain detector tubes (e.g., variation in boiler use and venting
set to assure that the SO2 NAAQS are Dräger Tubes),13 and since length-of- practices. However, to address the
met. As seen in the SIP and FIP, there stain detector tubes are widely-used and commenter’s concern, we are increasing
are 3-hour, 24-hour, and annual SO2 reliable, we have revised the FIP to the H2S concentration limit to 160 ppm
emission limits on most stacks. These make its requirements similar to those per 3-hour period. We are adding a
emission limits assure that the 3-hour, MSCC must already meet under the calendar day H2S concentration limit of
24-hour, and annual SO2 NAAQS are State’s operating permit. Specifically, 100 ppm.
attained and maintained. As indicated We selected the 160 ppm H2S per 3-
13 See MSCC’s ‘‘Hydrogen Sulfide Fuel Gas
in the response to comment II.F.8., hour period limit for the following
Monitoring Plan,’’ dated September 2000, that
above, we cannot require states to adopt fulfilled requirements of Montana Air Quality
reasons. First, as explained in greater
provisions that go beyond attaining and Operating Permit 2611–00, Appendix H. (See detail below, this value will protect the
maintaining the NAAQS. The annual reference document IIIIII.) 3-hour SO2 NAAQS. Second, 160 ppm
of H2S per 3-hour period is the current action. See reference document or additional monitoring would need to
NSPS limit for fuel gas combustion KKKKKK. be performed, at greater expense to
devices. EPA reported the following in (c) Comment (MSCC): Emission MSCC, to achieve adequate methods to
its May 14, 2007, proposal to revise Limit—100 ppm H2S—Overly Stringent. determine direct compliance with the
subpart J of the new source performance The 100 ppm H2S limit, which is a mass emission limits. The concentration
standards (NSPS), and to adopt new surrogate for the pound/hour SO2 limit, limits we are imposing are reasonable,
subpart Ja: is far too restrictive. EPA developed the can be monitored at reasonable cost, and
100 ppm H2S limit based on conditions will ensure protection of the NAAQS.
after consideration of current operating
practices, we concluded that amine that have a miniscule probability of
(e) Comment (MSCC): Monitoring
scrubbing units are still the predominant occurring. It has the effect of
introducing a new, strict ‘‘performance Cost. EPA proposes imposing significant
technology for reduction of H2S in fuel gas overly burdensome on-going costs to
(and SO2 emissions from subsequent fuel gas standard’’ into the mix of limits, where
such standard is not applicable. track a minuscule amount of potential
combustion). Considering the variability of
the fuel gas streams from various refinery Response: See response to II.G.3.(b), or actual SO2 emissions.
processing units, 160 ppmv also is still a above. Also, in order to protect the Response: As we indicate in response
realistic short term H2S concentration limit. NAAQS, it is reasonable to consider to comment II.G.3.(a), we have revised
However, one California Air Quality potential worst-case conditions in the FIP to allow MSCC to use the same
Management District rule sets a 40 ppmv H2S setting emission limits and compliance devices to determine H2S concentrations
limit in fuel gas (averaged over 4 hours), and determining methods. in the gas going to the 30-meter stack
several refiners have reported that the typical (d) Comment (MSCC): Monitoring and auxiliary vent stacks as MSCC is
fuel gas H2S concentrations (after scrubbing) Requirements. The requirement to
are in the same range. using to meet State requirements
monitor the auxiliary vent stacks has (length-of-stain detector tubes). While
(See 72 FR 27178, 27193.) Third, the already been addressed through the the frequency of monitoring may be
State’s SIP indicates that MSCC shall State plan; there is no inadequacy or somewhat different than the frequency
burn only low sulfur fuel gas or natural other basis to FIP this. The current under the State’s permit, the final FIP
gas in any unit being exhausted through system already periodically measures should not result in any substantial
the 30-meter stack (see MSCC’s exhibit the H2S content in the fuel gas header additional monitoring costs for the 30-
A, reference document II). Low sulfur for gas that is not natural gas, using a meter stack and the auxiliary vent
fuel gas is not defined in exhibit A. simple portable detector (non- stacks, particularly since MSCC
However, an MDEQ staff member electronic) such as a Dräger tube or Gas- indicates emissions from these stacks
indicated that the term ‘‘low sulfur fuel Tec tube. The frequency of testing are infrequent.
gas’’ in the SIP would be gas with an necessity was determined through the
H2S concentration much lower than the State’s plan and the frequency of such H. ConocoPhillips Specific Issues
NSPS subpart J limit of 160 ppm (see testing steps up in response to high
measurements until the measurements SRU/ATS Stack and Jupiter Flare
reference document GGGGGG). This
suggests that MSCC should already be have returned to low levels. The present Comment (COPC): ConocoPhillips
achieving a daily limit of 100 ppm. plan also reasonably estimates the urges EPA to delete the proposed
volume of gas used in each boiler to prohibition of simultaneous emissions
To test the use of a 160 ppm limit, we permit calculation of the SO2 emitted by
remodeled the area assuming the from the SRU/ATS stack and the Jupiter
each auxiliary vent when in use, and flare even if the combined SO2
emissions were 1.01 g/s from the 30- logs the venting location, as the State
meter stack and auxiliary vent stacks. emissions are less than 25 lb/hr. This
plan provides.
We derived the higher emission value merely imposes a compliance risk and
Response: In large part, this comment
from the same assumptions and appears to pertain to our disapproval of produces no environmental benefit.
calculations expressed in our proposal, the relevant portion of the SIP. We note Logic does not dictate that because both
except we assumed a maximum H2S that we have not reopened our SIP sources were modeled as one point, that
concentration of 160 ppm (see 71 FR action as part of this action and are not combined, simultaneous emissions from
39259, 39268, July 12, 2006). At the considering comments on that action both are prohibited. Quite the contrary,
higher three hour emissions, the area here. To the extent the comment is having modeled both sources as one
would still show attainment of the 3- relevant to our FIP action, see response point supports and endorses the option
hour SO2 NAAQS. However, the area to comment II.G.3.(b), above. As we of both sources being able to emit a
would not show attainment of the 24- explain there, the FIP retains the combined total of the amount of SO2
hour SO2 NAAQS if all 3-hour periods requirement that MSCC measure the which was modeled.
in a calendar day were at the 160 ppm H2S content of the fuel burned but Response: EPA agrees that it is not
level. Therefore, we are revising the FIP increases the 3-hour concentration limit necessary to prohibit simultaneous
to indicate that the H2S concentration in to 160 ppm. The FIP also allows MSCC emissions from both emission points.
the fuel burned in the heaters and to use length-of-stain detector tubes in Attainment of the SO2 NAAQS would
boilers, while any of the heaters and lieu of portable analyzers. However, be assured so long as the combined
boilers are exhausting to the SRU 30- based on comments received, we are not emissions from both emission points do
meter stack or auxiliary vents stacks, convinced that MSCC’s current methods not exceed 75.0 pounds per 3-hour
shall not exceed 160 ppm per 3-hour for determining direct compliance with period. Since both emission points have
period and 100 ppm per calendar day. the mass emission limits are sufficiently methods for determining emissions,
The mass emission limits remain the reliable or accurate for purposes of the compliance with the emission limit
same as proposed. The revised modeling FIP due to potential variation in boiler would be assured. We are revising the
files are indexed in the electronic use and venting practices and lack of regulatory text to eliminate the
docket contained on http:// equipment to directly measure relevant restriction on simultaneous emissions
www.regulations.gov, and a compact parameters at or emissions from each and any corresponding language.
disk containing the modeling files has boiler. We believe additional monitoring Additionally, in the final regulatory text
been placed in the docket for this equipment would need to be installed, we are clarifying the reporting
requirements to correspond to this and MDEQ and notify EPA and MDEQ its two closest competitors, that a
change. of RATAs. significant lowering in total SO2
emissions in the Yellowstone Valley
I. ExxonMobil Specific Issues 2. Tutwiler Analysis
could be attained if ExxonMobil would
1. Coker CO Boiler Comment (ExxonMobil): The be required to use that equipment under
proposed FIP would require that either Federal EPA standards or under
Comment (ExxonMobil): The ExxonMobil measure the H2S
proposed FIP would require that the the State of Montana emissions
concentration of the fuel gas once every requirements as well. That there is no
Coker CO Boiler stack CEMS operate at three hours using the Tutwiler method
all times. This is unnecessary because requirement to insist that ExxonMobil
contained in 40 CFR 60.648 any time use equipment/refining processes that
the Coker Process gas is exhausted the refinery fuel gas H2S CEMS
through the nearby Yellowstone Energy would lower its future SO2 emissions is
measures a concentration of greater than a deplorable lack of public concern to
Limited Partnership Co-Generation 1200 ppmv. The proposed once per 3-
facility. During those hours, Coker CO YVAS’ best interests and should be
hour Tutwiler analysis is less protective publicly examined by the EPA.
Boiler stack SO2 emissions are than the existing requirement identified
monitored by the existing fuel gas CEM Response: EPA acknowledges this
in the alternative monitoring plan
for fuel gas combustion devices. The comment. See response to comment
(AMP) submitted to DEQ. The AMP
existing SO2 SIP requires that a SO2 II.F.8., above.
requires measurement of the fuel gas
CEMS be operated on the Coker CO H2S concentration with Dräger tubes on J. CHS Inc. Specific Issues
Boiler stack during those few hours that an hourly basis anytime the fuel gas H2S
the Coker Process Gas is exhausted Particulate Issues
CEMS data are expected to be
through the Coker CO Boiler and stack. unavailable for any reason for more than Comment (YVAS): YVAS is
Given that a CEMS is already required one 3-hour block. concerned that the Coker production
for this source, nothing is served by Response: In our proposed FIP, EPA unit at CHS Inc. will not have to provide
requiring ExxonMobil to report the proposed a method for determining H2S a containment system shielding the
emissions and compliance assurance concentrations when the range of the nearby area from the effects of
data for this source to both EPA and H2S CEMS is exceeded. ExxonMobil particulate pollution. This is a
MDEQ. Nothing is served by requiring commented that they currently use deplorable lack of proper protection of
ExxonMobil to notify both EPA and another method for determining H2S the public and, although addressing this
MDEQ of required Relative Accuracy concentrations when the H2S CEMS is particular issue was apparently not
Test Audits (RATA). not available. This other method has important to this FIP, since it was
Response: It was not EPA’s intent to been identified in an AMP submitted to completely omitted from this FIP, either
require that the Coker CO Boiler stack DEQ (reference document JJJJJJ). Since through oversight or deliberate
CEMS be operated at all times. Our ExxonMobil already has procedures omission, YVAS seeks a ruling from the
intent was to clarify that the Coker CO established for determining H2S EPA that could require CHS, Inc. to
Boiler CEMS already installed, in concentrations when the H2S CEMS is address this issue and provide relief to
conjunction with the appropriate not available, namely, the use of Dräger the public from this oversight.
equations, must be used to determine Tubes, a type of length-of-stain detector
compliance with the emission limits Response: EPA acknowledges the
tube, and since length-of-stain detector comment. However, the FIP addresses
established in section 3(B)(1) of tubes are widely-used and reliable, EPA
ExxonMobil’s 2000 exhibit. only the provisions of the SO2 SIP that
is revising its FIP to incorporate the we disapproved. Under CAA section
We are clarifying the FIP to indicate other method identified by ExxonMobil.
that the Coker CO Boiler CEMS only 110(c), EPA’s authority is to remedy the
Specifically, we are revising the FIP to
needs to be operating when deficiencies we identified in the SO2
indicate that when the H2S
ExxonMobil’s Coker unit is operating SIP.
concentration in the refinery fuel gas
and Coker unit flue gases are exhausted exceeds 1200 ppmv as measured by the III. Summary of the Final Rules and
through the Coker CO Boiler stack. We H2S CEMS, ExxonMobil shall measure Changes From the July 12, 2006,
are also clarifying that whenever the H2S concentration on an hourly Proposal
ExxonMobil’s Coker unit is operating basis using length-of-stain detector
and Coker unit flue gases are exhausted The following summarizes the final
tubes pursuant to ASTM Method
through the Coker CO Boiler stack, the FIP and the major changes from our July
D4810–06, ‘‘Standard Test Method for
CEMS shall immediately be operational. 12, 2006, FIP proposal. Generally, the
Hydrogen Sulfide in Natural Gas Using
Also, with respect to the SO2 CEMS, we reasons for the changes made in the
Length-of-Stain Detector Tubes.’’ The
indicate that ExxonMobil shall perform final FIP appear in section II, above,
length-of-stain detector tubes shall have
a Cylinder Gas Audit (CGA) or Relative ‘‘Issues Raised by Commenters and
the appropriate sample tube range. If the
Accuracy Audit (RAA), which meets the EPA’s Response.’’ In some cases, the
results exceed the tube’s range, another
requirements of 40 CFR part 60, reasons appear below. We also describe
tube of a higher range must be used
Appendix F, within 8 hours of when the some minor changes to the FIP in this
until results are in the tube’s range. The
Coker unit flue gases begin exhausting section.
hourly length-of-stain detector tubes
through the Coker CO Boiler stack. data will then be used to calculate SO2 A. Flare Requirements Applicable to All
Finally, for both the SO2 and flow emissions from refinery fuel gas Sources
CEMS, we indicate that ExxonMobil combustion and to determine
shall perform an annual RATA, on the compliance with the emission limits in Since the State’s attainment
CEMS. 40 CFR 52.1392(f)(3)(i). demonstration assumed that the main
flares at each source were limited to 150
Because we will have primary

responsibility to enforce the FIP, we 3. ExxonMobil Emissions pounds of SO2 per 3-hour period, and
have retained the requirements that Comment (YVAS): The question must that the Jupiter Sulfur SRU flare would
ExxonMobil submit emissions and be asked that since ExxonMobil’s share an emission limit of 75 pounds of
compliance assurance data to both EPA emissions are appreciably higher than SO2 per 3-hour period with the Jupiter
Sulfur SRU/ATS 14 stack, we proposed specified accuracy of ± 5% over the changes. The final FIP now indicates
flare emission limits that reflected the range of 1 to 275 fps; that in quarterly reports, sources shall
State’s assumption that emissions from (3) For correcting flow rate to indicate the date and time when a
these points would not exceed these standard conditions (defined as 68°F monitor is not working or the range is
levels. While we proposed that 150 and 760 millimeters of mercury exceeded, and the other methods used
pounds of SO2 per 3-hour period be the (mmHg)), temperature and pressure to determine flare emissions. We have
limit for the main flares, we also would be monitored continuously; made these revisions to the final FIP so
solicited input on whether we should (4) The temperature and pressure that these provisions are consistent with
instead limit the main flares to 500 would be monitored in the same what we require in the flare monitoring
pounds of SO2 per calendar day. The location as the flow monitoring plan.
final FIP requires that the main flares at device(s) and be calibrated to meet The final FIP also adds the ability for
each source be limited to 150 pounds of accuracy specifications as follows: sources to use means other than the
SO2 per 3-hour period and that the Temperature would be calibrated flow monitor to determine that the flare
Jupiter Sulfur flare share an emission annually to within ± 2.0% at absolute is not operating when the flow monitor
limit of 75 pounds of SO2 per 3-hour temperature and the pressure monitor registers low flow. Specifically, the final
period with the Jupiter Sulfur SRU/ATS would be calibrated annually to within FIP allows sources to use devices that
stack. ± 5.0 mmHg; monitor the integrity of the flare water
We also proposed that the flare limits (5) Flow monitoring device(s) would seal. If these devices indicate that no
would apply at all times without be calibrated prior to installation to flow is going to the flare, yet the flow
exception. We also solicited comment demonstrate accuracy to within 5.0% at monitor indicates there is flow, the
on whether it would be appropriate to flow rates equivalent to 30%, 60%, and presumption will be that no flow is
include in our final FIP the ability to 90% of monitor full scale; and going to the flare. We have also revised
assert an affirmative defense to penalties (6) After installation, the flow the flare monitoring plan and reporting
only (not injunctive relief) for violations monitoring devices would be calibrated requirements to recognize the use of,
of the flare limits. Under the final FIP, annually according to manufacturer’s and require reporting on, these other
flare limits apply at all times. However, specifications. flare flow devices.
The final FIP flow monitoring We proposed that sources install,
we have changed the proposed rule to
provisions are the same as proposed calibrate, maintain, and operate an on-
provide the ability for sources to assert
except that we are revising the following line analyzer system capable of
an affirmative defense to penalties only
provisions: continuously determining the total
(not injunctive relief) for violations of (1) With respect to the accuracy of the sulfur concentration of the gas stream
the flare limits. The affirmative defense flow monitor, the final FIP indicates sent to a flare. We proposed that the
provision includes notification that the device(s) shall continuously continuous monitoring occur at a
requirements that are distinct from the measure the range of flow rates location or locations that are
FIP’s quarterly reporting requirements. corresponding to velocities from 0.5 to representative of the gas combusted in
We proposed that compliance with 275 fps and have a manufacturer’s the flare and be capable of measuring
the flare emission limits would be specified accuracy of ± 5% of the the expected range of total sulfur in the
determined by continuous measurement measured flow over the range of 1 to 275 gas stream to the flare. We proposed that
of the total sulfur concentration and fps and ± 20% of the measured flow the total sulfur analyzer be installed,
volumetric flow rate of the gas stream to over the range of 0.1 to 1.0 fps. certified (on a concentration basis), and
the flare(s), followed by calculation, (2) With respect to measurement of operated in accordance with 40 CFR
using appropriate equations, of SO2 volumetric flow rate, the final FIP part 60, Appendix B, Performance
emitted per 3-hour period. indicates that volumetric flow rate shall Specification 5, and be subject to and
We proposed that sources install, be measured on an actual wet basis and meet the quality assurance and quality
calibrate, maintain, and operate a converted to standard conditions, and control requirements (on a
continuous flow monitoring system reported in SCFH. concentration basis) of 40 CFR part 60,
capable of measuring the total (3) With respect to temperature and Appendix F. Additionally, we proposed
volumetric flow of the gas stream pressure monitors, the final FIP that sources notify EPA in writing of
combusted in a flare in accordance with indicates that temperature and pressure each Relative Accuracy Test Audit
the specifications described below. We monitors should be calibrated prior to (RATA) a minimum of 25 working days
indicated that the flow monitoring installation according to manufacturer’s prior to the actual testing. In the final
system could require one or more flow specifications. We inadvertently omitted FIP, we are retaining the above
monitoring devices or flow this requirement in our proposal. provisions, but are allowing the use of
measurements at one or more header We proposed that in cases where the other methods to determine total sulfur
locations if one monitor could not flow to the flare exceeds the range of the concentration. See discussion below.
measure all of the volumetric flow to a monitor, other methods could be used to The final FIP also clarifies that the total
flare. determine the volumetric flow rate. In sulfur concentration monitor should
We proposed the following the final FIP, we have clarified this measure in the range of concentrations
volumetric flow monitoring provision to read that in cases when the that are normally present in the gas
specifications: volumetric flow monitor is not working stream to the flare.
(1) The minimum detectible velocity or where the flow exceeds the range of In the final FIP, we are adding
of the flow monitoring device(s) would the monitor, methods established in the provisions that indicate that, in cases
be 0.1 feet per second (fps); flare monitoring plan required by the when the total sulfur analyzer is not
(2) The device(s) would continuously FIP shall be used to determine the working or where the concentration of
measure the range of flow rates volumetric flow rate to the flare, which the total sulfur exceeds the range of the
corresponding to velocities from 0.5 to shall then be used to calculate SO2 monitor, methods established in the
275 fps and have a manufacturer’s emissions. Additionally, we have flare monitoring plan required by the
revised the quarterly reporting FIP shall be used to determine the total
14 ATS stands for Ammonium Thiosulfate. requirements to be consistent with these sulfur concentrations, which shall than
be used to calculate SO2 emissions. monitoring plan to estimate flow and installing a chain and lock on the valve
Additionally, the final FIP indicates that H2S concentration. Pilot and purge gas that supplies sour water stripper
in quarterly reports, sources shall SO2 emissions will then be calculated overheads from the ‘‘old’’ SWS to the
indicate the date and time when a and added to the other SO2 emissions main crude heater to insure that the
monitor is not working, or the range is from the flare to determine compliance valve could not be opened. The
exceeded, and the other methods used with the SO2 flare emission limits. We proposed FIP also required CHS Inc. to
to determine flare emissions. We have have revised the reporting requirements maintain the chain and lock in place,
made this addition to the FIP so that accordingly to require sources to either: keep the valve closed at all times, and
these provisions are consistent with (1) Certify in the quarterly reports if log and report any noncompliance with
what we require in the flare monitoring pilot and/or purge gas is not monitored this provision. The final FIP is the same
plan. because only natural gas or inert gas is as proposed.
In lieu of continuous total sulfur used as the pilot and/or purge gas; or (2)
concentration analyzers, the final FIP C. ConocoPhillips
report flow, H2S concentration of, and
allows sources to determine the total SO2 emissions from, the pilot and/or Flare Requirements
sulfur concentration through grab or purge gas.
integrated sampling. If a source chooses We proposed that ConocoPhillips’s
We also added provisions that main flare be limited to 150 pounds of
to use one of these methods, the final indicate that in cases when any pilot or
FIP provides a trigger by which sources SO2 per 3-hour period and that
purge gas flow monitor or H2S analyzer compliance with the limit be
must begin the sampling and indicates is not working, or where the flow or
the analytical methods to be used to determined as discussed above. We also
concentration of the H2S exceeds the proposed that at any one time,
determine the total sulfur concentration range of the monitor or analyzer,
in the sample. The final FIP also ConocoPhillips could only use either
methods established in the flare the north or south main flare. The final
provides that in cases where a grab or monitoring plan required by the FIP
integrated sample is not obtained or FIP is the same as proposed except for
shall be used to determine the pilot and the flare monitoring changes applicable
analyzed, methods established in the purge gas flow and/or H2S
flare monitoring plan required by the to all sources mentioned above.
concentrations, which shall then be We proposed an emission limit of 75
FIP shall be used to determine total used to calculate SO2 emissions. The
sulfur concentrations, which will then pounds of SO2 per 3-hour period for the
FIP indicates that in quarterly reports, Jupiter Sulfur SRU flare and SRU/ATS
be used to calculate SO2 emissions. The sources shall indicate the date and time
flare monitoring plan and reporting stack and that emissions could only be
when a monitor or analyzer is not vented from the SRU flare when
requirements have also been revised to working, or the range is exceeded, and
recognize the potential use of grab or emissions were not being vented from
the other methods used to determine the SRU/ATS stack. We proposed that
integrated sampling. flare emissions.
We proposed that within 180 days compliance with the SRU flare emission
The flare monitoring plan limit, when Jupiter Sulfur vented
after receiving EPA approval of the flare requirements have been revised to be
monitoring plan, sources install and emissions to the SRU flare rather than
consistent with the pilot and purge gas the SRU/ATS stack, be determined by
calibrate, and thereafter calibrate,
provisions described above. measuring the total sulfur concentration
maintain, and operate continuous flow We have added definitions of Aliquot,
monitors and total sulfur concentration and volumetric flow rate of the gas
Integrated sampling, Pilot gas, and stream to the flare.15 Our final FIP is the
analyzers. The final FIP has been
Purge gas to clarify the FIP’s flare same as proposed except that we have
revised to allow sources 365 days after
monitoring requirements. Finally, we removed the restriction that emissions
receiving EPA approval of the flare
proposed quarterly reporting could only be vented from the SRU flare
monitoring plan to install and calibrate,
requirements similar to the reporting when emissions were not being vented
and thereafter calibrate, maintain, and
requirements contained in the Billings/ from the SRU/ATS stack. Our final FIP
operate the continuous volumetric flow
Laurel SO2 SIP and those contained in indicates that compliance with the
monitors and to start determining total
40 CFR 60.7(c). We added to the combined emission limit be determined
sulfur concentrations of the gas stream
reporting requirements as necessary to by summing the emissions from the
by either continuous total sulfur
concentration analyzers or grab or address the changes to other Jupiter Sulfur SRU flare and SRU/ATS
integrated sampling monitoring. requirements. stack.
We proposed that each facility submit B. CHS Inc. D. ExxonMobil
a flare monitoring plan including,
among other things, information 1. Flare Requirements 1. Flare Requirements
regarding pilot and purge gas at each We proposed that CHS Inc.’s flare be We proposed that ExxonMobil’s
flare and how the concentration and limited to 150 pounds of SO2 per 3-hour primary process and turnaround flares
volumetric flow monitors would period and that compliance with the be limited to 150 pounds of SO2 per 3-
analyze the pilot and purge gases. The limit be determined as discussed above. hour period and that compliance with
final FIP indicates that if the facility The final FIP is the same as proposed the limit be determined as discussed
certifies that only natural gas or an inert except for the flare monitoring changes above. Our proposal indicated that we
gas is used as pilot and/or purge gas, applicable to all sources mentioned understood that the turnaround flare is
monitoring the stream(s) consisting of above. only used about 30–40 days every 5 to
only natural gas or inert gas is not 6 years and is not normally operating.
required. However, if natural gas or 2. Combustion Sources Emission Limits
Therefore, we proposed to establish one
inert gas is not used for pilot and/or We proposed a prohibition in the FIP combined emission limit for the primary
purge gas, then the source must measure on the burning of SWS overheads in the process and turnaround flares. Our
the flow and H2S concentration of the main crude heater. We proposed that
gas streams that do not consist of only compliance with the prohibition to not 15 Note that the SRU/ATS stack has an SO CEMS
2
natural gas or inert gas or use other burn SWS overheads in the main crude and flow monitor to determine compliance when
methods approved by EPA in the flare heater be determined by CHS Inc. emissions are vented through that stack.
assumption was that the flow and appropriate sample tube range. If the perform an annual RATA on the flow
concentration monitoring devices results exceed the tube’s range, another and SO2 CEMS.
installed to measure the gas stream to tube of a higher range must be used We proposed that compliance with
the primary process flare would also be until results are in the tube’s range. ExxonMobil’s Coker CO Boiler emission
able to measure the gas stream to the ExxonMobil shall continue to use the limits 18 be determined using the data
turnaround flare. However, we length-of-stain detector tube method at from the CEMS mentioned above and in
indicated that if that was not the case, this frequency until the H2S CEMS accordance with the appropriate
ExxonMobil could propose another measures an H2S concentration in the calculations described in ExxonMobil’s
method to determine emissions from the fuel gas stream equal to or less than 1998 exhibit.19 We also proposed
turnaround flare. The final FIP is the 1200 ppmv continuously over a 3-hour reporting requirements similar to the
same as proposed except for the flare period. We also revised the equation requirements adopted in the Billings/
monitoring changes applicable to all used to calculate the SO2 emissions Laurel SO2 SIP and those contained in
sources mentioned above. because of the change in the H2S 40 CFR 60.7(c). Our final FIP is the same
2. Compliance Monitoring of Refinery analysis method. as proposed, except as noted above.
Fuel Gas Combustion Emission Limits We proposed reporting requirements E. Montana Sulphur & Chemical
similar to the requirements adopted by Company (MSCC)
We proposed a method for measuring the State for CHS Inc. and those
the H2S concentrations in the refinery contained in 40 CFR 60.7(c). We added 1. Flare Requirements
fuel gas when the H2S concentrations in a provision that requires ExxonMobil to
the refinery fuel gas exceed the range of We proposed that MSCC’s 80-foot
report information for periods when the west flare, 125-foot east flare, and 100-
the H2S CEMS. The method we
range of the refinery fuel gas CEMS is meter flare be limited to 150 pounds of
proposed is identical to the method
exceeded. SO2 per 3-hour period combined total
included in CHS Inc.’s 1998 exhibit.16
Specifically, we proposed that within 3. Compliance Monitoring of Coker CO and that compliance with the limit be
four hours of the initial determination Boiler Emission Limits determined as discussed above. Our
that the H2S concentrations in the final FIP is the same as proposed except
refinery fuel gas stream exceed the We proposed that existing SO2 and for the flare monitoring changes
upper range of the H2S CEMS, flow CEMS, in conjunction with the applicable to all sources mentioned
ExxonMobil would have to initiate appropriate calculations mentioned above.
sampling of the refinery fuel gas stream below, be used to determine compliance
with the emission limits established in 2. SRU 100-Meter Stack
at the fuel header on a once-per-3-hour-
period frequency using the Tutwiler section 3(B)(1) of ExxonMobil’s 2000 We proposed the following emission
method in 40 CFR 60.648. The Tutwiler exhibit. Specifically, we proposed that limits for the SRU 100-meter stack:
method determines the H2S at all times ExxonMobil operate and Emissions of SO2 not to exceed (a)
concentration in the refinery fuel gas. maintain CEMS to measure SO2 3,003.1 pounds per 3-hour period, (b)
We also proposed that the Tutwiler- concentrations from the Coker CO Boiler 24,025.0 pounds per calendar day, and
derived H2S refinery fuel gas stack and a continuous stack flow rate (c) 9,088,000.0 pounds per calendar
concentration be used in calculations to monitor to measure stack gas flow rates year. Our final FIP is the same as
determine the hourly, 3-hour, and 24- from the Coker CO Boiler stack. We proposed except that the 3-hour and
hour SO2 emission rates, in pounds, proposed that the SO2 and flow rate calendar day emission limits have been
from refinery fuel gas combustion. CEMS meet the CEM Performance slightly reduced due to minor
These emission rates would then be Specifications contained in sections corrections in the modeling. The final
used to determine compliance with the 6(C) and (D), respectively, of FIP emission limits for the SRU 100-
refinery fuel gas combustion emission ExxonMobil’s 1998 exhibit, except that meter stack are as follows: Emissions of
limits in ExxonMobil’s 1998 and 2000 ExxonMobil would have to notify EPA SO2 shall not exceed (a) 2981.7 pounds
exhibits when the H2S concentrations in in writing of each annual RATA a per 3-hour period, (b) 23,853.6 pounds
the refinery fuel gas stream exceeded minimum of 25 working days prior to per calendar day, and (c) 9,088,000.0
the upper range of the H2S CEMS.17 actual testing. pounds per calendar year
In our final FIP we have revised the Our final FIP is the same as proposed We proposed that compliance with
method by which ExxonMobil shall except that we have deleted the the above emission limits be determined
obtain the H2S concentration of the requirement that the flow and SO2 according to the methods established in
refinery fuel gas when the H2S CEMS be operated at all times and MSCC’s 1998 exhibit. Finally, we
concentrations in the refinery fuel gas added the requirement that whenever proposed quarterly reporting
exceed the range of the H2S CEMS. ExxonMobil’s Coker unit is operating requirements similar to the reporting
Specifically, our final FIP indicates that and Coker unit flue gases are exhausted requirements contained in the Billings/
within four hours after the H2S CEMS through the Coker CO Boiler stack, the Laurel SO2 SIP and those contained in
measures an H2S concentration in the flow and SO2 CEMS shall be 40 CFR 60.7(c). Our final FIP is the same
fuel gas stream greater than 1200 ppmv, immediately operational. We have also as proposed, except as noted above.
ExxonMobil shall initiate sampling of clarified that ExxonMobil shall meet the 3. SRU 30-Meter Stack
the fuel gas stream at the fuel header on specifications contained in section 6(C)
a once-per-hour-period frequency using of ExxonMobil’s 1998 exhibit, except We proposed the following mass
length-of-stain detector tubes with the that ExxonMobil shall perform a emission limits for the 30-meter stack:
Cylinder Gas Audit (CGA) or Relative Emissions of SO2 not to exceed: (a) 12.0
16 See section 6(B)(3) of CHS Inc.’s 1998 exhibit. Accuracy Audit (RAA) which meets the pounds per 3-hour period, (b) 96.0
(See reference document DD for a copy of the requirements of 40 CFR part 60,
exhibit.) Appendix F, within eight hours of when
18 See section 3(B)(1) of ExxonMobil’s 2000. (See
17 See sections 3(A)(1) and 3(B)(2) of reference document HH for a copy of the exhibit.)
ExxonMobil’s 1998 and 2000 exhibits. (See
the Coker unit flue gases begin 19 See sections 2(A)(1), (8), (11)(a), and (16) of
reference documents GG and HH for copies of the exhausting through the Coker CO Boiler ExxonMobil’s 1998 exhibit. (See reference
exhibits.) stack and that ExxonMobil shall document GG for a copy of the exhibit.)
pounds per calendar day, and (c) 35,040 we should promulgate a different (2) One commenter indicated that
pounds per calendar year. The mass compliance determining method. only proper geographical coordinates
emission limits remain the same as In our final FIP, we are keeping the should be used as inputs to the
proposed. simplified method to determine dispersion modeling. Commenters
We proposed that H2S concentrations compliance with mass emission limits. indicated that the location of the small
in the fuel burned in the boilers and However, we are increasing the H2S boiler stacks at MSCC that were
heaters, while any boiler or heater was concentration limit to 160 ppm/3-hour modeled as volume sources was
exhausting through the SRU 30-meter period and adding a calendar day H2S incorrect.
stack, be limited to 100 ppm of H2S or concentration limit of 100 ppm. EPA changes: We have corrected the
less, averaged over a 3-hour period. We proposed that the H2S incorrect source coordinate for MSCC’s
While we proposed the foregoing concentration in the fuel be measured boiler stacks in the modeling files.
approach for determining compliance using a portable H2S monitor. In our (3) One commenter indicated that
with the SRU 30-meter stack emission final FIP we have revised the method by three source input files were not
limits, we also solicited input on which MSCC shall determine the H2S included in reference document EEE.
whether we should promulgate a EPA change: We have added the three
content of the fuel burned. Specifically,
different compliance determining source input files to the compact disk
our final FIP indicates that MSCC shall
method. containing the modeling files.
determine the H2S content of the fuel (4) One commenter indicated that a
In our final FIP, we are keeping the burned using length-of-stain detector source input file (ref-5t.sri) was
simplified method to determine tubes with the appropriate sample tube included in reference document EEE but
compliance with mass emission limits. range. The final FIP indicates that if the did not appear to be used in any input
However, we are increasing the H2S results exceed the tube’s range, another and output files.
concentration limit to 160 ppm/3-hour tube of a higher range must be used EPA change: This was a test file that
period and adding a calendar day H2S until results are in the tube’s range. we inadvertently included and have
concentration limit of 100 ppm. Finally, we proposed quarterly now deleted.
We proposed that the H2S reporting requirements similar to On July 13, 2007, the revised
concentration in the fuel be measured reporting requirements contained in the modeling files were indexed in the
using a portable H2S monitor. In our Billings/Laurel SO2 SIP and those electronic docket contained on http://
final FIP, we have revised the method contained in 40 CFR 60.7(c). Our final www.regulations.gov and a compact
by which MSCC shall determine the H2S FIP is the same as proposed, except as disk containing the modeling files was
content of the fuel burned. Specifically, noted above. placed in the docket for this action. See
our final FIP indicates that MSCC shall reference document FFFFF.
F. Modeling To Support Emission Limits
determine the H2S content of the fuel Also, as noted above, with respect to
burned using length-of-stain detector Our proposal discussed the modeling the 30-meter stack and auxiliary vent
tubes with the appropriate sample tube conducted to support the emission stacks, we are keeping the simplified
range. The final FIP indicates that if the limits proposed for MSCC’s SRU 100- method to determine compliance with
results exceed the tube’s range, another meter stack. EPA received comments the mass emission limits. However, we
tube of a higher range must be used regarding our modeling files that are increasing the H2S concentration
until results are in the tube’s range. identified the need for minor technical limit to 160 ppm/3-hour period and
Finally, we proposed quarterly corrections to those files. In response to adding a calendar day H2S
reporting requirements. The quarterly several of these comments, EPA has concentration limit of 100 ppm. The
reporting requirements are similar to the revised its modeling files, as necessary, mass emission limits remain the same as
reporting requirements contained in the to omit extraneous information, add proposed.
Billings/Laurel SO2 SIP and those information that was inadvertently We remodeled the area assuming the
contained in 40 CFR 60.7(c). Our final omitted, make minor corrections, or emissions were 1.01 g/s from the 30-
FIP is the same as proposed, except as otherwise clarify the files. EPA does not meter stack and auxiliary vent stacks.
needed to address the changes noted consider any of the revisions to be We derived the higher emission value
above. significant. The only change with any from the same assumptions and
substantive impact—the correction to calculations expressed in our proposal,
4. Combined SO2 Emission Limit From the coordinates for MSCC described except we assumed a maximum H2S
the Auxiliary Vent Stacks below—results in a very slight decrease concentration of 160 ppm (see 71 FR
We proposed the following mass in our proposed emission limit for 39259, 39268, July 12, 2006). At the
emission limits for the auxiliary vent MSCC’s 100-meter stack from 126.13 g/ higher 3-hour emissions, the area would
stacks: emissions of SO2 not to exceed: second to125.23 g/second, less than a 1 still show attainment of the 3-hour SO2
(a) 12.0 pounds per 3-hour period, (b) percent change. The specific changes NAAQS. However, the area would not
96.0 pounds per calendar day, and (c) EPA has made are as follows: show attainment of the 24-hour SO2
35,040 pounds per calendar year. The (1) A commenter recommended that NAAQS if all 8 3-hour periods in a
mass emission limits remain the same as the modeling files contain a more calendar day were at the 160 ppm level.
proposed. In our proposal, we indicated complete description of the naming Therefore, we are revising the FIP to
that the issues associated with convention and purpose behind each indicate that the H2S concentration in
monitoring compliance with these modeling effort. the fuel burned in the heaters and
limits were essentially the same as those EPA changes: To improve boilers, while any of the heaters and
associated with monitoring compliance documentation, some extraneous boilers are exhausting to the SRU 30-
with the SRU 30-meter stack emission modeling files have been removed and meter stack or auxiliary vents stacks,
limits. Thus, we proposed the same a text file added to explain the naming shall not exceed 160 ppm per 3-hour
approach for monitoring compliance conventions. The naming conventions, period and 100 ppm per calendar day.
with these emission limits as we typically used by modelers, help define The revised modeling files are indexed
describe in section III.E.3, above. the purpose behind each modeling in the electronic docket contained on
Similarly, we solicited input on whether effort. http://www.regulations.gov, and a
compact disk containing the modeling number of small entities because this private sector, the costs are expected to
files was placed in the docket for this FIP applies to only four sources (CHS be less than $100 million in any one
action. See reference document Inc., ConocoPhillips, ExxonMobil and year. Thus, today’s rule is not subject to
KKKKKK. MSCC) in the Billings/Laurel, Montana the requirements of 202 and 205 of the
area. Therefore, I certify that this action UMRA.
IV. Statutory and Executive Order EPA has determined that this rule
will not have a significant economic
Reviews contains no regulatory requirements that
impact on a substantial number of small
A. Executive Order 12866, Regulatory entities. might significantly or uniquely affect
Planning and Review small governments, because it imposes
D. Unfunded Mandates Reform Act no requirements on small governments.
Under Executive Order 12866, 58 FR
Title II of the Unfunded Mandates Nor will the rule impact small
51735 (October 4, 1993), all ‘‘regulatory
Reform Act of 1995, Public Law 04 4, governments in any significant or
actions’’ that are ‘‘significant’’ are
establishes requirements for Federal unique way. Thus, today’s rule is not
subject to Office of Management and
agencies to assess the effects of their subject to the requirements of section
Budget (OMB) review and the
regulatory actions on State, local, and 203 of the UMRA.
requirements of the Executive Order. A
tribal governments and the private
‘‘regulatory action’’ is defined as ‘‘any E. Executive Order 13132, Federalism
sector. Under section 202 of UMRA,
substantive action by an agency Executive Order, 13132, entitled
EPA generally must prepare a written
(normally published in the Federal ‘‘Federalism’’ (64 FR 43255, August 10,
statement, including a cost benefit
Register) that promulgates or is 1999), requires EPA to develop an
analysis, for proposed rules and for final
expected to result in the promulgation accountable process to ensure
rules with ‘‘Federal mandates’’ that may
of a final rule or regulation, including ‘‘meaningful and timely input by State
result in the expenditure by State, local,
* * * notices of proposed rulemaking.’’ and local officials in the development of
and tribal governments, in the aggregate,
A ‘‘regulation or rule’’ is defined as ‘‘an regulatory policies that have federalism
or by the private sector, of $100 million
agency statement of general implications.’’ ‘‘Policies that have
or more in any one year. Before
applicability and future effect, * * * ’’ federalism implications’’ include
The FIP is not subject to OMB review promulgating an EPA rule for which a
written statement is needed, section 205 regulations that have ‘‘substantial direct
under E.O. 12866 because it applies to
of the UMRA generally requires EPA to effects on the States, on the relationship
only four specifically named facilities,
identify and consider a reasonable between the national government and
with requirements unique to each
number of regulatory alternatives and the States, or on the distribution of
facility, and is, therefore, not a rule of
adopt the least costly, most cost power and responsibilities among the
general applicability. Thus, it is not a
effective, or least burdensome various levels of government.’’
‘‘regulatory action’’ under E.O. 12866
alternative that achieves the objectives The final rule does not have
and was not submitted to OMB for
of the rule. The provisions of section federalism implications. It will not have
review.
205 do not apply when they are substantial direct effects on the States,
B. Paperwork Reduction Act inconsistent with applicable law. on the relationship between the national
This action does not impose an Moreover, section 205 allows EPA to government and the States, or on the
information collection burden under the adopt an alternative other than the least distribution of power and
provisions of the Paperwork Reduction costly, most cost-effective or least responsibilities among the various
Act, 44 U.S.C. 3501 et seq. Burden is burdensome alternative if the levels of government, as specified in
defined at 5 CFR 1320.3(b). Under the Administrator publishes with the final Executive Order 13132. This rule
Paperwork Reduction Act, 44 U.S.C. rule an explanation why that alternative establishes standards appropriate for
3501 et seq., OMB must approve all was not adopted. Before EPA establishes four companies in the Billings/Laurel,
‘‘collections of information’’ by EPA. any regulatory requirements that might Montana area, and, thus, does not
The Act defines ‘‘collection of significantly or uniquely affect small directly affect any State or local
information’’ as a requirement for governments, including tribal government. It does not alter the
‘‘answers to * * * identical reporting or governments, it must have developed relationship or the distribution of power
recordkeeping requirements imposed on under section 203 of UMRA a small and responsibilities established by the
ten or more persons * * * ’’ 4 U.S.C. government agency plan. The plan must Clean Air Act. Thus, Executive Order
3502(3)(A). Because the FIP only applies provide for notifying potentially 13132 does not apply to this rule.
to four companies, the Paperwork affected small governments, enabling
officials of affected small governments F. Executive Order 13175, Coordination
Reduction Act does not apply. With Indian Tribal Governments
to have meaningful and timely input in
C. Regulatory Flexibility Act the development of EPA regulatory Executive Order 13175, entitled
Under the Regulatory Flexibility Act proposals with significant Federal ‘‘Consultation and Coordination with
(RFA), 5 U.S.C. section 601 et seq., EPA intergovernmental mandates, and Indian Tribal Governments’’ (65 FR
generally must prepare a regulatory informing, educating, and advising 67249, November 9, 2000), requires EPA
flexibility analysis of any rule subject to small governments on compliance with to develop an accountable process to
notice and comment rulemaking the regulatory requirements. ensure ‘‘meaningful and timely input by
requirements unless EPA certifies that EPA has determined that this rule tribal officials in the development of
the rule will not have a significant does not contain a Federal mandate that regulatory policies that have tribal
economic impact on a substantial may result in the expenditure of $100 implications.’’ This final rule does not
number of small entities. Small entities million for State, local and tribal have tribal implications, as specified in
include small businesses, small not-for- governments, in the aggregate, or the Executive Order 13175. It will not have
profit enterprises, and small private sector in any one year. The FIP substantial, direct effects on tribal
governmental jurisdictions. 5 U.S.C. does not impose any enforceable duties governments, on the relationship
603, 604, and 605(b). on state, local, or tribal governments. between the Federal government and
This FIP will not have a significant Although the FIP would impose Indian tribes, or on the distribution of
economic impact on a substantial enforceable duties on entities in the power and responsibilities between the
Federal government and Indian tribes as practices) that are developed or adopted Enforcement Fairness Act of 1996,
specified in Executive Order 13175. by voluntary consensus standards generally provides that before a rule
This Action does not involve or impose bodies. The NTTAA directs EPA to may take effect, the agency
any requirements that affect Indian provide Congress, through OMB, promulgating the rule must submit a
Tribes. Thus, Executive Order 13175 explanations when the Agency decides rule report, which includes a copy of
does not apply to this rule. not to use available and applicable the rule, to each House of the Congress
voluntary standards. and to the Comptroller General of the
G. Executive Order 13045, Protection of This rulemaking involves technical
Children From Environmental Health United States. Section 804 exempts from
standards. We have identified three VCS section 801 the following types of rules:
Risks and Safety Risks that can be used in lieu of EPA methods. (1) Rules of particular applicability; (2)
Executive Order 13045: ‘‘Protection of The American Society for Testing and rules relating to agency management or
Children from Environmental Health Materials (ASTM) Methods D4468–85 personnel; and (3) rules of agency
Risks and Safety Risks’’ (62 FR 19885, (Reapproved 2000) and D5504–01 organization, procedure, or practice that
April 23, 1997), applies to any rule that: (Reapproved 2006) are acceptable do not substantially affect the rights or
(1) Is determined to be ‘‘economically methods for determining total sulfur obligations of non-agency parties. 5
significant’’ as defined under Executive concentrations in the gas streams going U.S.C. 804(3). EPA is not required to
Order 12866, and (2) concerns an to facility flares in lieu of using a submit a rule report regarding today’s
environmental health or safety risk that continuous total sulfur analyzer in action under section 801 because this is
EPA has reason to believe may have a accordance with 40 CFR part 60, a rule of particular applicability; it only
disproportionate effect on children. If Appendix B, Performance Specification applies to four specifically named
the regulatory action meets both criteria, 5. ASTM Method D4810–06 is an sources, with requirements unique to
the Agency must evaluate the acceptable method for determining the each facility.
environmental health or safety effects of hydrogen sulfide concentration in
the planned rule on children, and ExxonMobil’s refinery fuel gas in lieu of L. Petitions for Judicial Review
explain why the planned regulation is using the Tutwiler method described in Under section 307(b)(1) of the Clean
preferable to other potentially effective 40 CFR 60.648. We are incorporating Air Act, petitions for judicial review of
and reasonably feasible alternatives these methods by reference in 40 CFR this action must be filed in the United
considered by the Agency. 52.1392(j). States Court of Appeals for the
This FIP is not subject to the appropriate circuit by June 20, 2008.
Executive Order because it is not J. Executive Order 12898: Federal
Actions To Address Environmental Filing a petition for reconsideration by
economically significant as defined in the Administrator of this final rule does
Executive Order 12866. Further, EPA Justice in Minority Populations and
Low-Income Populations not affect the finality of this rule for the
interprets Executive Order 13045 as purposes of judicial review nor does it
applying only to those regulatory Executive Order (EO) 12898 (59 FR extend the time within which a petition
actions that are based on health or safety 7629 (Feb. 16, 1994)) establishes federal for judicial review may be filed, and
risks, such that the analysis required executive policy on environmental shall not postpone the effectiveness of
under section 5–501 of the Order has justice. Its main provision directs such rule or action. This action may not
the potential to influence the regulation. federal agencies, to the greatest extent be challenged later in proceedings to
This FIP is not subject to Executive practicable and permitted by law, to enforce its requirements. (See CAA
Order 13045 because it implements a make environmental justice part of their section 307(b)(2).)
previously promulgated health and mission by identifying and addressing,
safety-based Federal standard. as appropriate, disproportionately high List of Subjects in 40 CFR Part 52
and adverse human health or Environmental protection, Air
H. Executive Order 13211, Actions That environmental effects of their programs,
Significantly Affect Energy Supply, pollution control, Incorporation by
policies, and activities on minority reference, Intergovernmental relations,
Distribution, or Use populations and low-income Reporting and recordkeeping
This rule is not subject to Executive populations in the United States. requirements, Sulfur oxides.
Order 13211, ‘‘Actions Concerning EPA has determined that this final
Dated: March 28, 2008.
Regulations That Significantly Affect rule will not have disproportionately
Energy Supply, Distribution, or Use’’ (66 high and adverse human health or Stephen L. Johnson,
FR 28355, May 22, 2001) because it is environmental effects on minority or Administrator.
not a significant regulatory action under low-income populations because it ■For reasons stated in the preamble, 40
Executive Order 12866. increases the level of environmental CFR part 52 is amended as follows:
protection for all affected populations
I. National Technology Transfer and PART 52—[AMENDED]
without having any disproportionately
Advancement Act
high and adverse human health or ■ 1. The authority citation for part 52
As noted in the proposed rule, environmental effects on any continues to read as follows:
Section 12(d) of the National population, including any minority or
Technology Transfer and Advancement low-income population. This final rule Authority: 42 U.S.C. 7401 et seq.
Act (NTTAA) of 1995, Public Law No. establishes emission limits and
104–113 (15 U.S.C. 272 note), directs Subpart BB—Montana
compliance determining methods at
EPA to use voluntary consensus four sources in the Billings/Laurel, ■ 2. Subpart BB is amended by adding
standards in its regulatory activities Montana area to assure that the SO2 § 52.1392 to read as follows:
unless to do so would be inconsistent NAAQS are met.
with applicable law or otherwise § 52.1392 Federal Implementation Plan for

impractical. Voluntary consensus K. Congressional Review Act the Billings/Laurel Area.
standards (VCS) are technical standards The Congressional Review Act, 5 (a) Applicability. This section applies
(e.g., materials specifications, test U.S.C. section 801 et seq., as added by to the owner(s) or operator(s), including
methods, sampling procedures, business the Small Business Regulatory any new owner(s) or operator(s) in the
event of a change in ownership or Quality and that facility, adopted by the (22) Pilot gas means the gas used to
operation, of the following facilities in Montana Board of Environmental maintain the presence of a flame for
the Billings/Laurel, Montana area: CHS Review on either June 12, 1998, or ignition of gases routed to a flare.
Inc. Petroleum Refinery, Laurel March 17, 2000. (23) Purge gas means a continuous gas
Refinery, 803 Highway 212 South, (9) 1998 Exhibit means for a given stream introduced into a flare header,
Laurel, MT; ConocoPhillips Petroleum facility named in paragraph (a) of this flare stack, and/or flare tip for the
Refinery, Billings Refinery, 401 South section, the exhibit adopted by the purpose of maintaining a positive flow
23rd St., Billings, MT; ExxonMobil Montana Board of Environmental that prevents the formation of an
Petroleum Refinery, 700 Exxon Road, Review on June 12, 1998. explosive mixture due to ambient air
Billings, MT; and Montana Sulphur & (10) 2000 Exhibit means for a given ingress.
Chemical Company, 627 Exxon Road, facility named in paragraph (a) of this (24) The initials ppm mean parts per
Billings, MT. section, the exhibit adopted by the million.
(b) Scope. The facilities listed in Montana Board of Environmental (25) The initials SCFH mean standard
paragraph (a) of this section are also Review on March 17, 2000. cubic feet per hour.
subject to the Billings/Laurel SO2 SIP, as (11) Flare means a combustion device (26) The initials SCFM mean standard
approved at 40 CFR 52.1370(c)(46) and that uses an open flame to burn cubic feet per minute.
(52). In cases where the provisions of combustible gases with combustion air (27) Standard Conditions means (a) 20
this FIP address emissions activities provided by uncontrolled ambient air °C (293.2 °K, 527.7 °R, or 68.0 °F) and
differently or establish a different around the flame. This term includes one (1) atmosphere pressure (29.92
requirement than the provisions of the both ground and elevated flares. inches Hg or 760 mm Hg) for stack and
approved SIP, the provisions of this FIP (12) The initials Hg mean mercury. flare gas emission calculations, and (b)
take precedence. (13) Hourly means or refers to each 15.6 °C (288.7 °K, 520.0 °R, or 60.3 °F)
(c) Definitions. For the purpose of this clock hour in a calendar day. and one (1) atmosphere pressure (29.92
section, we are defining certain words (14) Hourly Average means an inches Hg or 760 mm Hg) for refinery
or initials as described in this arithmetic average of all valid and fuel gas emission calculations.
paragraph. Terms not defined below complete 15-minute data blocks in a (28) The initials SO2 mean sulfur
that are defined in the Clean Air Act or clock hour. Four (4) valid and complete dioxide.
regulations implementing the Clean Air 15-minute data blocks are required to (29) The initials SWS mean sour water
Act, shall have the meaning set forth in determine an hourly average for each stripper.
the Clean Air Act or such regulations. CEMS per clock hour. (30) The term 3-hour emissions means
(1) Aliquot means a fractional part of Exclusive of the above definition, an the amount of SO2 emitted in each of
a sample that is an exact divisor of the hourly CEMS average may be the eight (8) non-overlapping 3-hour
whole sample. determined with two (2) valid and periods in a calendar day, expressed in
(2) Annual Emissions means the complete 15-minute data blocks, for two pounds and rounded to the nearest
amount of SO2 emitted in a calendar (2) of the 24 hours in any calendar day. tenth (1⁄10) of a pound, where:
year, expressed in pounds per year A complete 15-minute data block for 3 hour emissions = Σ Hourly emissions
rounded to the nearest pound, where: each CEMS shall have a minimum of within the 3-hour period.
Annual emissions = S Daily emissions one (1) data point value; however, each (31) The term 3-hour period means
within the calendar year. CEMS shall be operated such that all any of the eight (8) non-overlapping 3-
(3) Calendar Day means a 24-hour valid data points acquired in any 15- hour periods in a calendar day:
period starting at 12 midnight and minute block shall be used to determine Midnight to 3 a.m., 3 a.m. to 6 a.m., 6
ending at 12 midnight, 24 hours later. the 15-minute block’s reported a.m. to 9 a.m., 9 a.m. to noon, noon to
(4) Clock Hour means a twenty-fourth concentration and flow rate. 3 p.m., 3 p.m. to 6 p.m., 6 p.m. to 9
(1⁄24) of a calendar day; specifically any (15) Hourly Emissions means the p.m., 9 p.m. to midnight.
of the standard 60-minute periods in a pounds per clock hour of SO2 emissions (32) Turnaround means a planned
day that are identified and separated on from a source (including, but not activity involving shutdown and startup
a clock by the whole numbers one (1) limited to, a flare, stack, fuel oil system, of one or several process units for the
through 12. sour water system, or fuel gas system) purpose of performing periodic
(5) Continuous Emission Monitoring determined using hourly averages and maintenance, repair, replacement of
System or CEMS means all continuous rounded to the nearest tenth (1⁄10) of a equipment, or installation of new
concentration and volumetric flow rate pound. equipment.
monitors, associated data acquisition (16) The initials H2S mean hydrogen (33) Valid means data that are
equipment, and all other equipment sulfide. obtained from a monitor or meter
necessary to meet the requirements of (17) Integrated sampling means an serving as a component of a CEMS
this section for continuous monitoring. automated method of obtaining a which meets the applicable
(6) Daily Emissions means the amount sample from the gas stream to the flare specifications, operating requirements,
of SO2 emitted in a calendar day, that produces a composite sample of and quality assurance and control
expressed in pounds per day rounded to individual aliquots taken over time. requirements of section 6 of
the nearest tenth (1⁄10) of a pound, (18) The initials MBER mean the ConocoPhillips’, CHS Inc.’s,
where: Montana Board of Environmental ExxonMobil’s, and MSCC’s 1998
Daily emissions = S 3-hour emissions Review. exhibits, respectively, and this section.
within a calendar day. (19) The initials MDEQ mean the (d) CHS Inc. emission limits and
(7) EPA means the United States Montana Department of Environmental compliance determining methods.
Environmental Protection Agency. Quality. (1) Introduction. The provisions for

(8) Exhibit means for a given facility (20) The initials mm mean CHS Inc. cover the following units:
named in paragraph (a) of this section, millimeters. (i) The flare.
exhibit A to the stipulation of the (21) The initials MSCC mean the (ii) Combustion sources, which
Montana Department of Environmental Montana Sulphur & Chemical Company. consist of those sources identified in the
combustion sources emission limit in (C) Hourly average volumetric flow recorded flow, yet any Flare Water Seal
section 3(A)(1)(d) of CHS Inc.’s 1998 rates in SCFH of the gas stream to the Monitoring Device indicated there was
exhibit. flare; no flow;
(2) Flare requirements. (D) Hourly average volumetric flow (G) For each 3-hour period in which
(i) Emission limit. The total emissions rates in SCFH of the flare pilot and the flare emission limit is exceeded, the
of SO2 from the flare shall not exceed purge gases; written report shall identify:
150.0 pounds per 3-hour period. (E) Hourly average temperature (in °F) (1) The date, start time, and end time
(ii) Compliance determining method. and pressure (in mm or inches of Hg) of of the excess emissions;
Compliance with the emission limit in the gas stream to the flare; (2) Total hours of operation with
paragraph (d)(2)(i) of this section shall (F) Hourly emissions from the flare in excess emissions, the hourly emissions,
be determined in accordance with pounds per clock hour; and and the 3-hour emissions;
(G) Daily calibration data for all flare, (3) All information regarding reasons
paragraph (h) of this section.
pilot gas, and purge gas CEMS. for operating with excess emissions; and
(3) Combustion sources. (iii) The quarterly written report shall
(i) Restrictions. Sour water stripper (4) Corrective actions taken to
contain the following information: mitigate excess emissions;
overheads (ammonia (NH3) and H2S (A) The 3-hour emissions in pounds
gases removed from the sour water in (H) The date and time of any
per 3-hour period from each flare; noncompliance with the requirements
the sour water stripper) shall not be (B) Periods in which only natural gas
burned in the main crude heater. At all of paragraph (d)(3)(i) of this section; and
or an inert gas was used as flare pilot (I) When no excess emissions have
times, CHS Inc. shall keep a chain and gas or purge gas or both; occurred or the continuous monitoring
lock on the valve that supplies sour (C) The results of all quarterly system(s) or manual system(s) have not
water stripper overheads from the old Cylinder Gas Audits (CGA), Relative been inoperative, repaired, or adjusted,
sour water stripper to the main crude Accuracy Audits (RAA), and annual such information shall be stated in the
heater and shall keep such valve closed. Relative Accuracy Test Audits (RATA) report.
(ii) Compliance determining method. for all total sulfur analyzer(s) and H2S (e) ConocoPhillips emission limits
CHS Inc. shall log and report any analyzer(s), and the results of all annual and compliance determining methods.
noncompliance with the requirements calibrations and verifications for the (1) Introduction. The provisions for
of paragraph (d)(3)(i) of this section. volumetric flow, temperature, and ConocoPhillips cover the following
(4) Data reporting requirements. pressure monitors; units:
(i) CHS Inc. shall submit quarterly (D) For all periods of flare volumetric (i) The main flare, which consists of
reports beginning with the first calendar flow rate monitoring system or total two flares—the north flare and the south
quarter following May 21, 2008. The sulfur analyzer system downtime, flare flare—that are operated on alternating
quarterly reports shall be submitted pilot gas or purge gas volumetric flow or schedules. These flares are referred to
within 30 days of the end of each H2S analyzer system downtime, or
herein as the north main flare and south
calendar quarter. The quarterly reports failure to obtain or analyze a grab or
main flare, or generically as the main
shall be submitted to EPA at the integrated sample, the written report
flare.
following address: Air Program Contact, shall identify: (ii) The Jupiter Sulfur SRU flare,
EPA Montana Operations Office, (1) Dates and times of downtime or
which is the flare at Jupiter Sulfur,
Federal Building, 10 West 15th Street, failure;
(2) Reasons for downtime or failure; ConocoPhillips’ sulfur recovery unit.
Suite 3200, Helena, MT 59626. (2) Flare requirements.
The quarterly report shall be certified (3) Corrective actions taken to
mitigate downtime or failure; and (i) Emission limits.
for accuracy in writing by a responsible (A) Combined emissions of SO2 from
(4) The other methods, approved by
CHS Inc. official. The quarterly report the main flare (which can be emitted
EPA in the flare monitoring plan
shall consist of both a comprehensive from either the north or south main
required by paragraph (h)(5) of this
electronic-magnetic report and a written flare, but not both at the same time)
section, used to determine flare
hard copy data summary report. shall not exceed 150.0 pounds per 3-
emissions;
(ii) The electronic report shall be on (E) For all periods that the range of hour period.
magnetic or optical media, and such the flare or any pilot or purge gas (B) Emissions of SO2 from the Jupiter
submittal shall follow the reporting volumetric flow rate monitor(s), any Sulfur SRU flare and the Jupiter Sulfur
format of electronic data being flare total sulfur analyzer(s), or any pilot SRU/ATS stack (also referred to as the
submitted to the MDEQ. EPA may or purge gas H2S analyzer(s) is Jupiter Sulfur SRU stack) shall not
modify the reporting format delineated exceeded, the written report shall exceed 75.0 pounds per 3-hour period,
in this section, and, thereafter, CHS Inc. identify: 600.0 pounds per calendar day, and
shall follow the revised format. In (1) Date and time when the range of 219,000 pounds per calendar year.
addition to submitting the electronic the volumetric flow monitor(s), total (ii) Compliance determining method.
quarterly reports to EPA, CHS Inc. shall sulfur analyzer(s), or H2S analyzer(s) (A) Compliance with the emission
also record, organize, and archive for at was exceeded; and limit in paragraph (e)(2)(i)(A) of this
least five (5) years the same data, and (2) The other methods, approved by section shall be determined in
upon request by EPA, CHS Inc. shall EPA in the flare monitoring plan accordance with paragraph (h) of this
provide EPA with any data archived in required by paragraph (h)(5) of this section. In the event that a single
accordance with this provision. The section, used to determine flare monitoring location cannot be used for
electronic report shall contain the emissions; both the north and south main flare,
following: (F) For all periods that the flare ConocoPhillips shall monitor the flow
(A) Hourly average total sulfur volumetric flow monitor or monitors are and measure the total sulfur
concentrations as H2S or SO2 in ppm in recording flow, yet any Flare Water Seal concentration at more than one location
the gas stream to the flare; Monitoring Device indicates there is no in order to determine compliance with
(B) Hourly average H2S concentrations flow, the written report shall identify: the main flare emission limit.
of the flare pilot and purge gases in (1) Date, time, and duration when the ConocoPhillips shall log and report any
ppm; flare volumetric flow monitor(s) instances when emissions are vented
from the north main flare and south (E) Hourly average temperature (in °F) Monitoring Device indicates there is no
main flare simultaneously. and pressure (in mm or inches of Hg) of flow, the written report shall identify:
(B) Compliance with the emission the gas streams to the ConocoPhillips (1) Date, time, and duration when the
limits and requirements in paragraph main flare and Jupiter Sulfur SRU flare; flare volumetric flow monitor(s)
(e)(2)(i)(B) of this section shall be (F) Hourly emissions in pounds per recorded flow, yet any Flare Water Seal
determined by summing the emissions clock hour from the ConocoPhillips Monitoring Device indicated there was
from the Jupiter Sulfur SRU flare and main flare and Jupiter Sulfur SRU flare; no flow;
SRU/ATS stack. Emissions from the and (G) Identification of dates, times, and
Jupiter Sulfur SRU flare shall be (G) Daily calibration data for all flare, duration of any instances when
determined in accordance with pilot gas, and purge gas CEMS. emissions were vented from the north
paragraph (h) of this section and the (iii) The quarterly written report shall and south main flares simultaneously;
emissions from the Jupiter Sulfur SRU/ contain the following information: (H) For each 3-hour period in which
ATS stack shall be determined pursuant (A) The 3-hour emissions in pounds a flare emission limit is exceeded, the
to ConocoPhillips’ 1998 exhibit (see per 3-hour period from the written report shall identify:
section 4(A) of the exhibit). ConocoPhillips main flare and the sum (1) The date, start time, and end time
(3) Data reporting requirements. of the combined 3-hour emissions from of the excess emissions;
(i) ConocoPhillips shall submit the Jupiter Sulfur SRU/ATS stack and (2) Total hours of operation with
quarterly reports on a calendar year Jupiter Sulfur SRU flare in pounds per excess emissions, the hourly emissions,
basis, beginning with the first calendar 3-hour period; and the 3-hour emissions;
quarter following May 21, 2008. The (B) Periods in which only natural gas (3) All information regarding reasons
quarterly reports shall be submitted or an inert gas was used as flare pilot for operating with excess emissions; and
within 30 days of the end of each gas or purge gas or both; (4) Corrective actions taken to
calendar quarter. The quarterly reports mitigate excess emissions; and
(C) The results of all quarterly
shall be submitted to EPA at the (I) When no excess emissions have
Cylinder Gas Audits (CGA), Relative
following address: Air Program Contact, occurred or the continuous monitoring
Accuracy Audits (RAA), and annual
EPA Montana Operations Office, system(s) or manual system(s) have not
Relative Accuracy Test Audits (RATA)
Federal Building, 10 West 15th Street, been inoperative, repaired, or adjusted,
for all total sulfur analyzer(s) and H2S
Suite 3200, Helena, MT 59626. such information shall be stated in the
analyzer(s), and the results of all annual
The quarterly report shall be certified report.
calibrations and verifications for the (f) ExxonMobil emission limits and
for accuracy in writing by a responsible
volumetric flow, temperature, and compliance determining methods.
ConocoPhillips official. The quarterly
pressure monitors; (1) Introduction. The provisions for
report shall consist of both a
(D) For all periods of flare volumetric ExxonMobil cover the following units:
comprehensive electronic-magnetic
flow rate monitoring system or total (i) The Primary process flare and the
report and a written hard copy data
sulfur analyzer system downtime, flare Turnaround flare. The Primary process
summary report.
(ii) The electronic report shall be on pilot gas or purge gas volumetric flow or flare is the flare normally used by
magnetic or optical media, and such H2S analyzer system downtime, or ExxonMobil. The Turnaround flare is
submittal shall follow the reporting failure to obtain or analyze a grab or the flare ExxonMobil uses for about 30
format of electronic data being integrated sample, the written report to 40 days every 5 to 6 years when the
submitted to the MDEQ. EPA may shall identify: facility’s major SO2 source, the fluid
modify the reporting format delineated (1) Dates and times of downtime or catalytic cracking unit, is not normally
in this section, and, thereafter, failure; operating.
ConocoPhillips shall follow the revised (2) Reasons for downtime or failure; (ii) The following refinery fuel gas
format. In addition to submitting the (3) Corrective actions taken to combustion units: The FCC CO Boiler,
electronic quarterly reports to EPA, mitigate downtime or failure; and F–2 crude/vacuum heater, F–3 unit, F–
ConocoPhillips shall also record, (4) The other methods, approved by 3X unit, F–5 unit, F–700 unit, F–201
organize, and archive for at least five (5) EPA in the flare monitoring plan unit, F–202 unit, F–402 unit, F–551
years the same data, and upon request required by paragraph (h)(5) of this unit, F–651 unit, standby boiler house
by EPA, ConocoPhillips shall provide section, used to determine flare (B–8 boiler), and Coker CO Boiler (only
EPA with any data archived in emissions; when the Yellowstone Energy Limited
accordance with this provision. The (E) For all periods that the range of Partnership (YELP) facility is receiving
electronic report shall contain the the flare or any pilot or purge gas ExxonMobil Coker unit flue gas or
following: volumetric flow rate monitor(s), any whenever the ExxonMobil Coker is not
(A) Hourly average total sulfur flare total sulfur analyzer(s), or any pilot operating).
concentrations as H2S or SO2 in ppm in or purge gas H2S analyzer(s) is (iii) Coker CO Boiler stack.
the gas stream to the ConocoPhillips exceeded, the written report shall (2) Flare requirements.
main flare and Jupiter Sulfur SRU flare; identify: (i) Emission limit. The total combined
(B) Hourly average H2S concentrations (1) Date and time when the range of emissions of SO2 from the Primary
of the ConocoPhillips main flare and the volumetric flow monitor(s), total process and Turnaround refinery flares
Jupiter Sulfur SRU flare pilot and purge sulfur analyzer(s), or H2S analyzer(s) shall not exceed 150.0 pounds per 3-
gases in ppm; was exceeded, and hour period.
(C) Hourly average volumetric flow (2) The other methods, approved by (ii) Compliance determining method.
rates in SCFH of the gas streams to the EPA in the flare monitoring plan Compliance with the emission limit in
ConocoPhillips main flare and Jupiter required by paragraph (h)(5) of this paragraph (f)(2)(i) of this section shall be
Sulfur SRU flare; section, used to determine flare determined in accordance with
(D) Hourly average volumetric flow emissions; paragraph (h) of this section. If
rates in SCFH of the ConocoPhillips (F) For all periods that the flare volumetric flow monitoring device(s)
main flare and Jupiter Sulfur SRU flare volumetric flow monitor or monitors are installed and concentration monitoring
pilot and purge gases; recording flow, yet any Flare Water Seal methods used to measure the gas stream
to the Primary Process flare cannot ExxonMobil shall continue to use the sampling resolution of at least one (1)
measure the gas stream to the length-of-stain detector tube method at concentration measurement per minute,
Turnaround flare, ExxonMobil may this frequency until the H2S CEMS meet the requirements expressed in the
apply to EPA for alternative measures to measures an H2S concentration in the definition of ‘‘hourly average’’ in
determine the volumetric flow rate and refinery fuel gas stream equal to or less paragraph (c)(14) of this section, and
total sulfur concentration of the gas than 1200 ppmv continuously over a 3- meet the CEMS Performance
stream to the Turnaround flare. Before hour period. Specifications contained in section 6(C)
EPA will approve such alternative (C) When the length-of-stain detector of ExxonMobil’s 1998 exhibit, except
measures, ExxonMobil must agree that tube method is required, SO2 emissions that ExxonMobil shall perform a
the Turnaround flare will be used only from refinery fuel gas combustion shall Cylinder Gas Audit (CGA) or Relative
during refinery turnarounds of limited be calculated as follows: the Hourly Accuracy Audit (RAA) which meets the
duration and frequency—no more than emissions shall be calculated using requirements of 40 CFR part 60,
60 days once every five (5) years— equation 1, 3-hour emissions shall be Appendix F, within eight (8) hours of
which restriction shall be considered an calculated using equation 2, and the when the Coker unit flue gases begin
enforceable part of this FIP. Such Daily emissions shall be calculated exhausting through the Coker CO Boiler
alternative measures may consist of using equation 3. stack. ExxonMobil shall perform an
reliable flow estimation parameters to Equation 1: EH = K * CH * QH annual Relative Accuracy Test Audit
estimate volumetric flow rate and Where: (RATA) on the CEMS and notify EPA in
manual sampling of the gas stream to writing of each annual RATA a
the flare to determine total sulfur EH = Refinery fuel gas combustion hourly
emissions in pounds per hour, rounded minimum of 25 working days prior to
concentrations, or such other measures actual testing.
to the nearest tenth of a pound;
that EPA finds will provide accurate K= 1.688 × 10-7 in (pounds/standard cubic (C) Beginning on May 21, 2008,
estimations of SO2 emissions from the feet (SCF))/parts per million (ppm); ExxonMobil shall operate and maintain
Turnaround flare. CH = Hourly refinery fuel gas H2S a continuous stack flow rate monitor to
(3) Refinery fuel gas combustion concentration in ppm determined by the measure the stack gas flow rates in the
requirements. length-of-stain detector tube method as Coker CO Boiler stack. Whenever
(i) Emission limits. The applicable required by paragraph (f)(3)(ii)(B) of this ExxonMobil’s Coker unit is operating
emission limits are contained in section section; and
and Coker unit flue gases are exhausted
3(A)(1) of ExxonMobil’s 2000 exhibit QH = actual fuel gas firing rate in standard
cubic feet per hour (SCFH), as measured through the Coker CO Boiler stack, this
and section 3(B)(2) of ExxonMobil’s
by the monitor required by section CEMS shall be operational and shall
1998 exhibit.
(ii) Compliance determining method. 6(B)(8) of ExxonMobil’s 1998 exhibit. achieve a temporal sampling resolution
For the limits referenced in paragraph Equation 2: (Refinery fuel gas of at least one (1) flow rate measurement
(f)(3)(i) of this section, the compliance combustion 3-hour emissions) = Σ per minute, meet the requirements
determining methods specified in (Hourly emissions within the 3- expressed in the definition of ‘‘hourly
section 4(B) of ExxonMobil’s 1998 hour period as determined by average’’ in paragraph (c)(14) of this
exhibit shall be followed except when equation 1). section, and meet the Stack Gas Flow
the H2S concentration in the refinery Equation 3: (Refinery fuel gas Rate Monitor Performance
fuel gas stream exceeds 1200 ppmv as combustion daily emissions) = Σ (3- Specifications of section 6(D) of
measured by the H2S CEMS required by hour emissions within the day as ExxonMobil’s 1998 exhibit, except that
section 6(B)(3) of ExxonMobil’s 1998 determined by equation 2). ExxonMobil shall perform an annual
exhibit (the H2S CEMS.) When such Relative Accuracy Test Audit (RATA)
(4) Coker CO Boiler stack
value is exceeded, the following on the CEMS and notify EPA in writing
requirements.
compliance monitoring method shall be (i) Emission limits. When of each annual RATA a minimum of 25
employed: ExxonMobil’s Coker unit is operating working days prior to actual testing.
(A) ExxonMobil shall measure the and Coker unit flue gases are burned in (D) SO2 emissions from the Coker CO
H2S concentration in the refinery fuel the Coker CO Boiler, the applicable Boiler stack shall be determined in
gas according to the procedures in emission limits are contained in section accordance with the equations in
paragraph (f)(3)(ii)(B) of this section and 3(B)(1) of ExxonMobil’s 2000 exhibit. sections 2(A)(1), (8), (11)(a), and (16) of
calculate the emissions according to the (ii) Compliance determining method. ExxonMobil’s 1998 exhibit.
equations in paragraph (f)(3)(ii)(C) of (A) Compliance with the emission (5) Data reporting requirements.
this section. limits referenced in paragraph (f)(4)(i) of (i) ExxonMobil shall submit quarterly
(B) Within four (4) hours after the H2S this section shall be determined by reports beginning with the first calendar
CEMS measures an H2S concentration in measuring the SO2 concentration and quarter following May 21, 2008. The
the refinery fuel gas stream greater than flow rate in the Coker CO Boiler stack quarterly reports shall be submitted
1200 ppmv, ExxonMobil shall initiate according to the procedures in within 30 days of the end of each
sampling of the refinery fuel gas stream paragraphs (f)(4)(ii)(B) and (C) of this calendar quarter. The quarterly reports
at the fuel header on a once-per-hour section and calculating emissions shall be submitted to EPA at the
frequency using length-of-stain detector according to the equations in paragraph following address: Air Program Contact,
tubes pursuant to ASTM Method (f)(4)(ii)(D) of this section. EPA Montana Operations Office,
D4810–06, ‘‘Standard Test Method for (B) Beginning on May 21, 2008, Federal Building, 10 West 15th Street,
Hydrogen Sulfide in Natural Gas Using ExxonMobil shall operate and maintain Suite 3200, Helena, MT 59626.
Length-of-Stain Detector Tubes’’ a CEMS to measure sulfur dioxide The quarterly report shall be certified
(incorporated by reference, see concentrations in the Coker CO Boiler for accuracy in writing by a responsible
paragraph (j) of this section) with the stack. Whenever ExxonMobil’s Coker ExxonMobil official. The quarterly
appropriate sample tube range. If the unit is operating and Coker unit flue report shall consist of both a
results exceed the tube’s range, another gases are exhausted through the Coker comprehensive electronic-magnetic
tube of a higher range must be used CO Boiler stack, the CEMS shall be report and a written hard copy data
until results are in the tube’s range. operational and shall achieve a temporal summary report.
(ii) The electronic report shall be on this section, and the results of all annual (3) All information regarding reasons
magnetic or optical media, and such calibrations and verifications for the for operating with excess emissions; and
submittal shall follow the reporting volumetric flow, temperature, and (4) Corrective actions taken to
format of electronic data being pressure monitors; mitigate excess emissions; and
submitted to the MDEQ. EPA may (E) For all periods of flare volumetric (J) When no excess emissions have
modify the reporting format delineated flow rate monitoring system or total occurred or the continuous monitoring
in this section, and, thereafter, sulfur analyzer system downtime, Coker system(s) or manual system(s) have not
ExxonMobil shall follow the revised CO Boiler stack CEMS downtime, been inoperative, repaired, or adjusted,
format. In addition to submitting the refinery fuel gas combustion system such information shall be stated in the
electronic quarterly reports to EPA, CEMS downtime, flare pilot gas or purge report.
ExxonMobil shall also record, organize, gas volumetric flow or H2S analyzer (g) Montana Sulphur & Chemical
and archive for at least five (5) years the system downtime, or failure to obtain or Company (MSCC) emission limits and
same data, and upon request by EPA, analyze a grab or integrated sample, the compliance determining methods.
ExxonMobil shall provide EPA with any written report shall identify: (1) Introduction. The provisions for
data archived in accordance with this (1) Dates and times of downtime or MSCC cover the following units:
provision. The electronic report shall failure; (i) The flares, which consist of the 80-
contain the following: (2) Reasons for downtime or failure; foot west flare, 125-foot east flare, and
(A) Hourly average total sulfur (3) Corrective actions taken to 100-meter flare.
concentrations as H2S or SO2 in ppm in mitigate downtime or failure; and (ii) The SRU 100-meter stack.
the gas stream to the flare(s); (4) The other methods, approved by (iii) The auxiliary vent stacks and the
(B) Hourly average H2S concentrations EPA in the flare monitoring plan units that can exhaust through the
of the flare pilot and purge gases in required by paragraph (h)(5) of this auxiliary vent stacks, which consist of
ppm; section, used to determine flare the Railroad Boiler, the H–1 Unit, the
(C) Hourly average SO2 concentrations emissions; H1–A unit, the H1–1 unit and the H1–
in ppm from the Coker CO Boiler stack; (F) For all periods that the range of 2 unit.
(D) Hourly average volumetric flow the flare or any pilot or purge gas (iv) The SRU 30-meter stack and the
rates in SCFH of the flare pilot and volumetric flow rate monitor(s), any units that can exhaust through the SRU
purge gases; flare total sulfur analyzer(s), or any pilot 30-meter stack. The units that can
(E) Hourly average volumetric flow or purge gas H2S analyzer(s) is exhaust through the SRU 30-meter stack
rates in SCFH in the gas stream to the exceeded, the written report shall are identified in section 3(A)(2)(d) and
flare(s) and in the Coker CO Boiler identify: (e) of MSCC’s 1998 exhibit.
stack; (1) Date and time when the range of (2) Flare requirements.
(F) Hourly average H2S concentrations the volumetric flow monitor(s), total (i) Emission limit. Total combined
in ppm from the refinery fuel gas sulfur analyzer(s), or H2S analyzer(s) emissions of SO2 from the 80-foot west
system; was exceeded, and flare, 125-foot east flare, and 100-meter
(G) Hourly average refinery fuel gas (2) The other methods, approved by flare shall not exceed 150.0 pounds per
combustion units’ actual fuel firing rate EPA in the flare monitoring plan 3-hour period.
in SCFH; required by paragraph (h)(5) of this (ii) Compliance determining method.
(H) Hourly average temperature (in °F) section, used to determine flare Compliance with the emission limit in
and pressure (in mm or inches of Hg) of emissions; paragraph (g)(2)(i) of this section shall
the gas stream to the flare(s); (G) For all periods that the range of be determined in accordance with
(I) Hourly emissions in pounds per the refinery fuel gas CEMS is exceeded, paragraph (h) of this section. In the
clock hour from the flare(s), Coker CO the written report shall identify: event MSCC cannot monitor all three
Boiler stack, and refinery fuel gas (1) Date, time, and duration when the flares from a single location, MSCC shall
combustion system; and range of the refinery fuel gas CEMS was establish multiple monitoring locations.
(J) Daily calibration data for the CEMS exceeded; (3) SRU 100-meter stack
described in paragraphs (f)(2)(ii), (H) For all periods that the flare requirements.
(f)(3)(ii) and (f)(4)(ii) of this section. volumetric flow monitor or monitors are (i) Emission limits. Emissions of SO2
(iii) The quarterly written report shall recording flow, yet any Flare Water Seal from the SRU 100-meter stack shall not
contain the following information: Monitoring Device indicates there is no exceed:
(A) The 3-hour emissions in pounds flow, the written report shall identify: (A) 2,981.7 pounds per 3-hour period;
per 3-hour period from the flare(s), (1) Date, time, and duration when the (B) 23,853.6 pounds per calendar day;
Coker CO Boiler stack, and refinery fuel flare volumetric flow monitor(s) and
gas combustion system; recorded flow, yet any Flare Water Seal (C) 9,088,000 pounds per calendar
(B) Periods in which only natural gas Monitoring Device indicated there was year.
or an inert gas was used as flare pilot no flow; (ii) Compliance determining method.
gas or purge gas or both; (I) For each 3-hour period and (A) Compliance with the emission
(C) Daily emissions in pounds per calendar day in which the flare limits contained in paragraph (g)(3)(i) of
calendar day from the Coker CO Boiler emission limits, the Coker CO Boiler this section shall be determined by the
stack and refinery fuel gas combustion stack emission limits, or the fuel gas CEMS and emission testing methods
system; combustion system emission limits are required by sections 6(B)(1) and (2) and
(D) The results of all quarterly or exceeded, the written report shall section 5, respectively, of MSCC’s 1998
other Cylinder Gas Audits (CGA), identify: exhibit.
Relative Accuracy Audits (RAA), and (1) The date, start time, and end time (B) MSCC shall notify EPA in writing
annual Relative Accuracy Test Audits of the excess emissions; of each annual source test a minimum
(RATA) for the CEMS described in (2) Total hours of operation with of 25 working days prior to actual
paragraphs (f)(2)(ii) (flare total sulfur excess emissions, the hourly emissions, testing.
analyzer(s); pilot gas or purge gas H2S the 3-hour emissions, and the daily (C) The CEMS referenced in
analyzer(s)), (f)(3)(ii), and (f)(4)(ii) of emissions; paragraph (g)(3)(ii)(A) of this section
shall achieve a temporal sampling shall use length-of-stain detector tubes Method for Hydrogen Sulfide in Natural
resolution of at least one (1) pursuant to ASTM Method D4810–06, Gas Using Length-of-Stain Detector
concentration and flow rate ‘‘Standard Test Method for Hydrogen Tubes’’ (incorporated by reference, see
measurement per minute, meet the Sulfide in Natural Gas Using Length-of- paragraph (j) of this section) with the
requirements expressed in the definition Stain Detector Tubes’’ (incorporated by appropriate sample tube range. If the
of ‘‘hourly average’’ in paragraph (c)(14) reference, see paragraph (j) of this results exceed the tube’s range, another
of this section, and meet the ‘‘CEM section) with the appropriate sample tube of a higher range must be used
Performance Specifications’’ in sections tube range. If the results exceed the until results are in the tube’s range.
6(C) and (D) of MSCC’s 1998 exhibit, tube’s range, another tube of a higher (6) Data reporting requirements:
except that MSCC shall also notify EPA range must be used until results are in (i) MSCC shall submit quarterly
in writing of each annual Relative the tube’s range. reports beginning with the first calendar
Accuracy Test Audit at least 25 working (5) SRU 30-meter stack. quarter following May 21, 2008. The
days prior to actual testing. (i) Emission limits. quarterly reports shall be submitted
(4) Auxiliary vent stacks. (A) Emissions of SO2 from the SRU within 30 days of the end of each
(i) Emission limits. 30-meter stack shall not exceed 12.0 calendar quarter. The quarterly reports
(A) Total combined emissions of SO2 pounds per 3-hour period; shall be submitted to EPA at the
from the auxiliary vent stacks shall not (B) Emissions of SO2 from the SRU following address: Air Program Contact,
exceed 12.0 pounds per 3-hour period; 30-meter stack shall not exceed 96.0 EPA Montana Operations Office,
(B) Total combined emissions of SO2 pounds per calendar day; Federal Building, 10 West 15th Street,
from the auxiliary vent stacks shall not (C) Emissions of SO2 from the SRU Suite 3200, Helena, MT 59626.
exceed 96.0 pounds per calendar day; 30-meter stack shall not exceed 35,040 The quarterly report shall be certified
(C) Total combined emissions of SO2 pounds per calendar year; and for accuracy in writing by a responsible
from the auxiliary vent stacks shall not (D) The H2S concentration in the fuel MSCC official. The quarterly report
exceed 35,040 pounds per calendar burned in the heaters and boilers shall consist of both a comprehensive
year; and described in paragraph (g)(1)(iv) of this electronic-magnetic report and a written
(D) The H2S concentration in the fuel section, while any of these units is hard copy data summary report.
burned in the Railroad Boiler, the H–1 exhausting to the SRU 30-meter stack, (ii) The electronic report shall be on
Unit, the H1–A unit, the H1–1 unit, and shall not exceed 160 ppm per 3-hour magnetic or optical media, and such
the H1–2 unit, while any of these units period and 100 ppm per calendar day. submittal shall follow the reporting
is exhausting to the auxiliary vent (ii) Compliance determining method. format of electronic data being
stacks, shall not exceed 160 ppm per 3- (A) Compliance with the emission submitted to the MDEQ. EPA may
hour period and 100 ppm per calendar limits in paragraph (g)(5)(i) of this modify the reporting format delineated
day. section shall be determined by in this section, and, thereafter, MSCC
(ii) Compliance determining method. measuring the H2S concentration of the shall follow the revised format. In
(A) Compliance with the emission fuel burned in the heaters and boilers addition to submitting the electronic
limits in paragraph (g)(4)(i) of this described in paragraph (g)(1)(iv) of this quarterly reports to EPA, MSCC shall
section shall be determined by section (when fuel other than natural also record, organize, and archive for at
measuring the H2S concentration of the gas is burned in one or more of these least five (5) years the same data, and
fuel burned in the Railroad Boiler, the heaters or boilers) according to the upon request by EPA, MSCC shall
H–1 Unit, the H1–A unit, the H1–1 unit, procedures in paragraph (g)(5)(ii)(C) of provide EPA with any data archived in
and the H1–2 unit (when fuel other than this section. accordance with this provision. The
natural gas is burned in one or more of (B) Beginning June 20, 2008, MSCC electronic report shall contain the
these units) according to the procedures shall maintain logs of: following:
in paragraph (g)(4)(ii)(C) of this section. (1) The dates and time periods that (A) Hourly average total sulfur
(B) Beginning June 20, 2008, MSCC emissions are exhausted through the concentrations as H2S or SO2 in ppm, in
shall maintain logs of: SRU 30-meter stack, the gas stream to the flare(s);
(1) The dates and time periods that (2) The heaters and boilers that are (B) Hourly average H2S concentrations
emissions are exhausted through the exhausting to the SRU 30-meter stack of the flare pilot and purge gases in
auxiliary vent stacks, during such time periods, and ppm;
(2) The heaters and boilers that are (3) The type of fuel burned in the (C) Hourly average SO2 concentrations
exhausting to the auxiliary vent stacks heaters and boilers during such time in ppm from the SRU 100-meter stack;
during such time periods, and periods. (D) Hourly average volumetric flow
(3) The type of fuel burned in the (C) Beginning June 20, 2008, MSCC rates in SCFH in the gas stream to the
heaters and boilers during such time shall measure the H2S content of the flare(s) and in the SRU 100-meter stack;
periods. fuel burned when fuel other than (E) Hourly average volumetric flow
(C) Beginning June 20, 2008, MSCC natural gas is burned in a heater or rates in SCFH of the flare pilot and
shall measure the H2S content of the boiler that is exhausting to the SRU 30- purge gases;
fuel burned when fuel other than meter stack. MSCC shall begin (F) Hourly average temperature (in (F)
natural gas is burned in a heater or measuring the H2S content of the fuel at and pressure (in mm or inches of Hg) in
boiler that is exhausting to an auxiliary the fuel header within one (1) hour from the gas stream to the flare(s);
vent stack. MSCC shall begin measuring when any heater or boiler begins (G) Hourly emissions in pounds per
the H2S content of the fuel at the fuel exhausting to the SRU 30-meter stack clock hour from the flare(s) and SRU
header within one (1) hour from when and on a once-per-3-hour period 100-meter stack;
a heater or boiler begins exhausting to frequency until no heater or boiler is (H) Daily calibration data for all flare
an auxiliary vent stack and on a once- exhausting to the SRU 30-meter stack. CEMS, all pilot gas and purge gas
per-3-hour period frequency until no To determine the H2S content of the fuel CEMS, and the SRU 100-meter stack
heater or boiler is exhausting to an burned, MSCC shall use length-of-stain CEMS;
auxiliary vent stack. To determine the detector tubes pursuant to ASTM (iii) The quarterly written report shall
H2S content of the fuel burned, MSCC Method D4810–06, ‘‘Standard Test contain the following information:
(A) The 3-hour emissions in pounds recording flow, yet any Flare Water Seal (i) Within 365 days after receiving
per 3-hour period from the flare(s) and Monitoring Device indicates there is no EPA approval of the flare monitoring
SRU 100-meter stack, and 3-hour H2S flow, the written report shall identify: plan required by paragraph (h)(5) of this
concentrations in the fuel burned in the (1) Date, time, and duration when the section, each facility named in
heaters and boilers described in flare volumetric flow monitor(s) paragraph (a) of this section shall install
paragraphs (g)(1)(iii) and (iv) of this recorded flow, yet any Flare Water Seal and calibrate, and, thereafter, calibrate,
section while any of these units is Monitoring Device indicated there was maintain and operate, a continuous flow
exhausting to the SRU 30-meter stack or no flow; monitoring system capable of measuring
auxiliary vent stacks and burning fuel (I) For each 3-hour period and the volumetric flow of the gas stream to
other than natural gas; calendar day in which the flare the flare(s) in accordance with the
(B) Periods in which only natural gas emission limit, the SRU 100-meter stack specifications contained in paragraphs
or an inert gas was used as flare pilot emission limits, the SRU 30-meter stack (h)(2)(iii) through (vi) of this section.
gas or purge gas or both; emission limits, or auxiliary vent stack The flow monitoring system shall
(C) Daily emissions in pounds per emission limits are exceeded, the require more than one flow monitoring
calendar day from the SRU 100-meter written report shall identify: device or flow measurements at more
stack; (1) The date, start time, and end time than one location if one monitor cannot
(D) Annual emissions of SO2 in of the excess emissions; measure the total volumetric flow to
pounds per calendar year from the SRU (2) Total hours of operation with each flare.
100-meter stack; excess emissions, the hourly emissions, (ii) Volumetric flow monitors meeting
(E) The results of all quarterly the 3-hour emissions, and the daily the proposed volumetric flow
Cylinder Gas Audits (CGA), Relative emissions; monitoring specifications below should
Accuracy Audits (RAA) and annual (3) All information regarding reasons be able to measure the majority of
Relative Accuracy Test Audits (RATA) for operating with excess emissions; and volumetric flow in the gas streams to the
for all total sulfur analyzer(s), all H2S (4) Corrective actions taken to flare. However, in rare events (e.g.,
analyzer(s), and the SRU 100-meter mitigate excess emissions; upset conditions) the flow to the flare
stack CEMS, and the results of all may exceed the range of the monitor. In
(J) For instances in which emissions
annual calibrations and verifications for such cases, or when the volumetric flow
are exhausted through the auxiliary vent
the volumetric flow, temperature, and monitor or monitors are not working,
stacks or 30-meter stack, the quarterly
pressure monitors; other methods approved by EPA in the
written report shall identify:
(F) For all periods of flare volumetric flare monitoring plan required by
(1) The dates and time periods that
flow rate monitoring system or total paragraph (h)(5) of this section shall be
emissions were exhausted through the
sulfur analyzer system downtime, SRU used to determine the volumetric flow
auxiliary vent stacks or the 30-meter
100-meter CEMS downtime, flare pilot rate to the flare, which shall then be
stack;
gas or purge gas volumetric flow or H2S used to calculate SO2 emissions. In
(2) The heaters and boilers that were quarterly reports, sources shall indicate
analyzer system downtime, failure to
exhausting to the auxiliary vent stacks when these other methods are used.
obtain or analyze a grab or integrated
or 30-meter stack during such time (iii) The flare gas stream volumetric
sample, or failure to obtain an H2S
periods; and flow rate shall be measured on an actual
concentration sample as required by
(3) The type of fuel burned in the wet basis, converted to Standard
paragraphs (g)(4)(ii)(C) and (g)(5)(ii)(C)
heaters and boilers during such time Conditions, and reported in SCFH. The
of this section, the written report shall
periods; and minimum detectable velocity of the flow
identify:
(K) When no excess emissions have monitoring device(s) shall be 0.1 feet
(1) Dates and times of downtime or
occurred or the continuous monitoring per second (fps). The flow monitoring
failure;
(2) Reasons for downtime or failure; system(s) or manual system(s) have not device(s) shall continuously measure
(3) Corrective actions taken to been inoperative, repaired, or adjusted, the range of flow rates corresponding to
mitigate downtime or failure; and such information shall be stated in the velocities from 0.5 to 275 fps and have
(4) The other methods, approved by report. a manufacturer’s specified accuracy of
EPA in the flare monitoring plan (h) Flare compliance determining ±5% of the measured flow over the
required by paragraph (h)(5) of this method. range of 1.0 to 275 fps and ±20% of the
section, used to determine flare (1) Compliance with the emission measured flow over the range of 0.1 to
emissions; limits in paragraphs (d)(2)(i), (e)(2)(i), 1.0 fps. The volumetric flow monitor(s)
(G) For all periods that the range of (f)(2)(i) and (g)(2)(i) of this section shall shall feature automated daily
the flare or any pilot or purge gas be determined by measuring the total calibrations at low and high ranges. The
volumetric flow rate monitor(s), any sulfur concentration and volumetric volumetric flow monitor(s) shall be
flare total sulfur analyzer(s), or any pilot flow rate of the gas stream to the flare(s) calibrated annually according to
or purge gas H2S analyzer(s), is (corrected to one (1) atmosphere manufacturer’s specifications.
exceeded, the written report shall pressure and 68° F) and using the (iv) For correcting flow rate to
identify: methods contained in the flare standard conditions (defined as 68°F
(1) Date and time when the range of monitoring plan required by paragraph and 760 mm, or 29.92 inches, of Hg),
the volumetric flow monitor(s), total (h)(5) of this section. The volumetric temperature and pressure shall be
sulfur analyzer(s), or H2S analyzer(s) flow rate of the gas stream to the flare(s) monitored continuously. Temperature
was exceeded; and shall be determined in accordance with and pressure shall be monitored in the
(2) The other methods, approved by the requirements in paragraph (h)(2) of same location as volumetric flow, and
EPA in the flare monitoring plan this section and the total sulfur the temperature and pressure monitors
required by paragraph (h)(5) of this concentration of the gas stream to the shall be calibrated prior to installation
section, used to determine flare flare(s) shall be determined in according to manufacturer’s
emissions; accordance with paragraph (h)(3) of this specifications and, thereafter, annually
(H) For all periods that the flare section. to meet accuracy specifications as
volumetric flow monitor or monitors are (2) Flare flow monitoring: follows: The temperature monitor shall
be calibrated to within ± 2.0% at temporal sampling resolution of at least analyzer(s), other methods, approved by
absolute temperature and the pressure one (1) flow rate measurement per EPA in the flare monitoring plan
monitor shall be calibrated to within ± minute, meet the requirements required by paragraph (h)(5) of this
5.0 mmHg; expressed in the definition of ‘‘hourly section, shall be used to determine total
(v) The flow monitoring device(s) average’’ in paragraph (c)(14) of this sulfur concentrations, which shall then
shall be calibrated prior to installation section, and be installed in a manner be used to calculate SO2 emissions. In
to demonstrate accuracy of the and at a location that will allow for quarterly reports, sources shall indicate
measured flow to within 5.0% at flow accurate measurements of the total when these other methods are used.
rates equivalent to 30%, 60%, and 90% volume of the gas. Gas flow rate monitor (2 ) The total sulfur analyzer(s) shall
of monitor full scale. accuracy determinations shall be achieve a temporal sampling resolution
(vi) Each volumetric flow device shall required at least once every 48 months of at least one (1) concentration
achieve a temporal sampling resolution or more frequently at routine refinery measurement per 15 minutes, meet the
of at least one (1) flow rate measurement turn-around. In cases when the flow requirements expressed in the definition
per minute, meet the requirements monitoring device or devices are not of ‘‘hourly average’’ in paragraph (c)(14)
expressed in the definition of ‘‘hourly working or the range of the monitoring of this section, be installed, certified (on
average’’ in paragraph (c)(14) of this device(s) is exceeded, other methods a concentration basis), and operated in
section, and be installed in a manner approved by EPA in the flare monitoring accordance with 40 CFR part 60,
and at a location that will allow for plan required by paragraph (h)(5) of this Appendix B, Performance Specification
accurate measurements of the total section shall be used to determine 5, and be subject to and meet the quality
volume of the gas stream going to each volumetric flow of the gas which shall assurance and quality control
flare. Each temperature and pressure then be used to calculate SO2 emissions. requirements (on a concentration basis)
monitoring device shall achieve a In quarterly reports, sources shall of 40 CFR part 60, Appendix F.
temporal sampling resolution of at least indicate when other methods are used; (3) Each affected facility named in
one (1) measurement per minute, meet or paragraph (a) of this section shall notify
the requirements expressed in the (B) Use parameters and methods the Air Program Contact at EPA’s
definition of ‘‘hourly average’’ in approved by EPA in the flare monitoring Montana Operations Office, Federal
paragraph (c)(14) of this section, and be plan required by paragraph (h)(5) of this Building, 10 West 15th Street, Suite
installed in a manner that will allow for section to calculate the volumetric flows 3200, Helena, MT 59626, in writing of
accurate measurements. of the gas, in SCFH. each Relative Accuracy Test Audit a
(vii) In addition to the continuous (3) Flare concentration monitoring: minimum of 25 working days prior to
flow monitors, facilities may use flare (i) Within 365 days after receiving the actual testing.
water seal monitoring devices to EPA approval of the flare monitoring (B) Grab or integrated total sulfur
determine whether there is flow going to plan required by paragraph (h)(5) of this concentration monitoring: If a facility
the flare. If used, owners or operators section, each facility named in chooses grab or integrated sampling
shall install, calibrate, operate, and paragraph (a) of this section shall instead of continuous total sulfur
maintain these devices according to determine the total sulfur concentration concentration monitoring, the facility
manufacturer’s specifications. The of the gas stream to the flare(s) using shall comply with the methods
devices shall include a continuous either continuous total sulfur analyzers specified in either paragraph
monitoring system that: or grab or integrated sampling with lab (h)(3)(i)(B)(1) (‘‘Grab Sampling’’) or
(A) Monitors the status of the water analysis, as described in the following (h)(3)(B)(i)(B)(2 ) (‘‘Integrated
seal to indicate when flow is going to paragraphs: Sampling’’), and the requirements of
the flare; (A) Continuous total sulfur paragraphs (h)(3)(i)(B)(3) (‘‘Sample
(B) Automatically records the time concentration monitoring. If a facility Analysis’’), (h)(3)(i)(B)(4)
and duration when flow is going to the chooses to use continuous total sulfur (‘‘Exemptions’’), and (h)(3)(i)(B)(5)
flare; and concentration monitoring, the following (‘‘Missing or Unanalyzed Sample’’) of
(C) Verifies that the physical seal has requirements apply: this section, as follows:
been restored after flow has been sent to (1 ) The facility shall install and (1) Grab Sampling. Each facility that
the flare. calibrate, and, thereafter, calibrate, chooses to use grab sampling shall meet
If the water seal monitoring devices maintain and operate, a continuous total the following requirements: if the flow
indicate that there is no flow going to sulfur concentration monitoring system rate of the gas stream to the flare in any
the flare, yet the continuous flow capable of measuring the total sulfur consecutive 15-minute period
monitor is indicating flow, the concentration of the gas stream to each continuously exceeds 0.5 feet per
presumption will be that no flow is flare. Continuous monitoring shall occur second (fps) and the water seal
going to the flare. at a location or locations that are monitoring device, if any, indicates that
(viii) Each facility named in representative of the gas combusted in flow is going to the flare, a grab sample
paragraph (a) of this section, that does the flare and be capable of measuring shall be collected within 15 minutes.
not certify that only natural gas or an the normally expected range of total The grab sample shall be collected at a
inert gas is used for both the pilot gas sulfur in the gas stream to the flare. The location that is representative of the gas
and purge gas, shall determine the concentration monitoring system shall combusted in the flare. Thereafter, the
volumetric flow of each pilot gas and require more than one concentration sampling frequency shall be one (1) grab
purge gas stream for which natural gas monitoring device or concentration sample every three (3) hours, which
or inert gas is not used by one of the measurements at more than one location shall continue until the velocity of the
following methods: if one monitor cannot measure the total gas stream going to the flare in any
(A) Measure the volumetric flow of sulfur concentration to each flare. Total consecutive 15-minute period is
the gas using continuous flow sulfur concentration shall be reported as continuously 0.5 fps or less. Samples
monitoring devices on an actual wet H2S or SO2 in ppm. In cases when the shall be analyzed according to
basis, converted to Standard Conditions, total sulfur analyzer or analyzers are not paragraph (h)(3)(i)(B)(3) of this section.
and reported in SCFH. Each flow working or the concentration of the total The requirements of this paragraph
monitoring device shall achieve a sulfur exceeds the range of the (h)(3)(i)(B)(1) shall apply to each flare at
a facility for which the sampling representative of the flaring event. The for which natural gas or inert gas is not
threshold is exceeded. owner or operator shall demonstrate to used by one of the following methods:
(2) Integrated Sampling. Each facility EPA that it was infeasible or unsafe to (A) Measure the H2S concentration of
that chooses to use integrated sampling collect a sample or to collect a sample the gas by continuous H2S analyzer. The
shall meet the following requirements: if at the sampling location approved by H2S concentration analyzer(s) shall
the flow rate of the gas stream to the EPA in the flare monitoring plan achieve a temporal sampling resolution
flare in any consecutive 15-minute required by paragraph (h)(5) of this of at least one (1) concentration
period continuously exceeds 0.5 feet per section. The owner or operator shall measurement per three (3) minutes,
second (fps) and the water seal also demonstrate to EPA that any meet the requirements expressed in the
monitoring device, if any, indicates that sample collected at an alternative definition of ‘‘hourly average’’ in
flow is going to the flare, a sample shall location is representative of the flaring paragraph (c)(14) of this section, be
be collected within 15 minutes. The incident. If a facility experiences installed, certified (on a concentration
sample shall be collected at a location ongoing difficulties collecting grab or basis), and operated in accordance with
that is representative of the gas integrated samples in accordance with 40 CFR part 60, Appendix B,
combusted in the flare. The sampling its flare monitoring plan approved by Performance Specification 2, and be
frequency, thereafter, shall be a EPA pursuant to paragraph (h)(5) of this subject to and meet the quality
minimum of one (1) aliquot for each 15- section, EPA may require the facility to assurance and quality control
minute period until the sample revise its flare monitoring plan and use requirements (on a concentration basis)
container is full, or until the end of a 3- continuous total sulfur concentration of 40 CFR part 60, Appendix F. In cases
hour period is reached, whichever monitoring as described in paragraph where the H2S analyzer or analyzers are
comes sooner. Within 30 minutes (h)(3)(i)(A) of this section or other not working or the H2S concentration
thereafter, a new sample container shall reliable method to determine total sulfur exceeds the range of the analyzer(s),
be placed in service, and sampling on concentrations of the gas stream to the other methods approved by EPA in the
this frequency, and in this manner, shall flare. flare monitoring plan required by
continue until the velocity of the gas (5) Missing or Unanalyzed Samples. paragraph (h)(5) of this section shall be
stream going to the flare in any For facilities using a sampling method used to determine the H2S concentration
consecutive 15-minute period is specified in either paragraph of the gas, which shall then be used to
continuously 0.5 fps or less. Samples (h)(3)(i)(B)(1) (‘‘Grab Sampling’’) or calculate SO2 emissions. In quarterly
shall be analyzed according to (h)(3)(i)(B)(2) (‘‘Integrated Sampling’’) of reports, sources shall indicate when
paragraph (h)(3)(i)(B)(3) of this section. this section, if a required sample is not other methods are used; or
The requirements of this paragraph obtained or analyzed for any reason, (B) Use methods approved by EPA as
(h)(3)(i)(B)(2) shall apply to each flare at other methods approved by EPA in the part of the facility’s flare monitoring
a facility for which the sampling flare monitoring plan required by
plan required by paragraph (h)(5) of this
threshold is exceeded. paragraph (h)(5) of this section shall be
section to estimate the H2S
(3) Samples shall be analyzed using used to determine total sulfur
concentration of the gas.
ASTM Method D4468–85 (Reapproved concentrations, which shall then be
2000) ‘‘Standard Test Method for Total (4) Calculation of SO2 emissions from
used to calculate SO2 emissions. In
Sulfur in Gaseous Fuels by flares. Methods for calculating hourly
quarterly reports, sources shall indicate
Hydrogenolysis and Rateometric and 3-hour SO2 emissions from flares
when these other methods are used.
Colorimetry,’’ (incorporated by (6) Reporting. For facilities using a shall be submitted to EPA as part of the
reference, see paragraph (j) of this sampling method specified in either flare monitoring plan required by
section) ASTM Method D5504–01 paragraph (h)(3)(i)(B)(1 ) (‘‘Grab paragraph (h)(5) of this section.
(Reapproved 2006) ‘‘Standard Test Sampling’’) or (h)(3)(i)(B)(2 ) Following approval by EPA, such
Method for Determination of Sulfur (‘‘Integrated Sampling’’) of this section, methods shall be followed for
Compounds in Natural Gas and Gaseous since normally only one (1) sample per calculating hourly and 3-hour SO2
Fuels by Gas Chromatography and flare will be analyzed for a 3-hour emissions from a facility’s flare(s).
Chemiluminescence,’’ (incorporated by period, the total sulfur concentration of (5) By October 20, 2008, each facility
reference, see paragraph (j) of this a sample obtained during a given 3-hour named in paragraph (a) of this section
section) or 40 CFR part 60, Appendix period shall be substituted for each hour shall submit a flare monitoring plan.
A–5, Method 15A ‘‘Determination of of such 3-hour period. If integrated Each flare monitoring plan shall
Total Reduced Sulfur Emissions From sampling for a flare produces more than include, at a minimum, the following:
the Sulfur Recovery Plants in Petroleum one (1) sample container during a 3- (i) A facility plot plan showing the
Refineries.’’ Total sulfur concentration hour period, and the gas in each location of each flare in relation to the
shall be reported as H2S or SO2 in ppm. container is analyzed separately, the general plant layout;
(4) Exemptions. For facilities using a concentrations for the containers shall (ii) Drawing(s) with dimensions,
sampling method specified in either be averaged. For that flare, the resulting preferably to scale, and an as-built
paragraph (h)(3)(i)(B)(1) (‘‘Grab average shall be substituted for each process flow diagram of the flare(s)
Sampling’’) or (h)(3)(i)(B)(2) (‘‘Integrated hour of the 3-hour period during which identifying major components, such as
Sampling’’) of this section, obtaining a the sampling occurred. The substituted flare header, flare stack, flare tip(s) or
sample is not required if flaring is a hourly total sulfur concentrations burner(s), purge gas system, pilot gas
result of a catastrophic or other unusual determined per this paragraph shall be system, water seal, knockout drum, and
event, including a major fire or an used to determine hourly emissions molecular seal;
explosion at the facility, such that from the flare. (iii) A representative flow diagram
collecting a sample at the EPA-approved (ii) Each facility named in paragraph showing the interconnections of the
location during the relevant period is (a) of this section that does not certify flare system(s) with vapor recovery
infeasible or constitutes a safety hazard, that only natural gas or an inert gas is system(s), process units, and other
provided that the owner or operator used for both the pilot gas and purge gas equipment as applicable;
shall collect a sample at an alternative shall determine the H2S concentration (iv) A complete description of the gas
location if feasible, safe, and of each pilot gas and purge gas stream flaring process for an integrated gas
flaring system that describes the method (B) A detailed description of this section may assert an affirmative
of operation of the flares; manufacturer’s specifications, defense to a claim for civil penalties for
(v) A complete description of the including, but not limited to, make, exceedances of such limits during
vapor recovery system(s) which have model, type, maintenance, and quality periods of malfunction, startup, or
interconnection to a flare, such as assurance procedures for the integrated shutdown. To establish the affirmative
compressor description(s); design sampling device, if used; and defense and to be relieved of a civil
capacities of each compressor and the (C) A complete description of the penalty in any action to enforce such a
vapor recovery system; and the method proposed method to alert personnel limit, the owner or operator of the
currently used to determine and record designated to collect samples that the facility must meet the notification
the amount of vapors recovered; trigger for collecting a sample has requirements of paragraph (i)(2) of this
(vi) A complete description of the occurred; section in a timely manner and prove by
proposed method to monitor, determine, (xi) A complete description of the a preponderance of evidence that:
and record the total volume and total methods to be used to estimate flare (i) For claims of malfunction:
sulfur concentration of gases combusted emissions when any flare, pilot gas, or (A) The excess emissions were caused
in the flare, including drawing(s) with purge gas volumetric flow monitoring by a sudden, unavoidable breakdown of
dimensions, preferably to scale, devices, total sulfur analyzers, or grab or equipment, or a sudden, unavoidable
showing the following information for integrated sampling methods, or pilot failure of a process to operate in the
the proposed flare gas stream gas or purge gas H2S analyzers are not normal or usual manner, beyond the
monitoring systems: working or available, or the operating control of the owner or operator;
(A) The locations to be used for all range of the monitors or analyzers is (B) The excess emissions:
monitoring and sampling, including, but exceeded; (1) Did not stem from any activity or
not limited to: Flare flow monitors, total (xii) A complete description of the event that could have been foreseen and
sulfur analyzers, concentration proposed data recording, collection, and avoided, or planned for; and
integrated sampling, concentration grab management system and any other (2) Could not have been avoided by
sampling, water seal monitoring relevant specifications and information better operation and maintenance
devices, pilot and purge gas flow referenced in paragraphs (h)(2) and (3) practices;
monitors, and pilot and purge gas of this section for each flare monitoring (C) Repairs were made as
concentration monitors; system; expeditiously as possible when the
(vii) A description of the method(s) (xiii) The following information for applicable emission limitations were
used to determine, and reasoning each flare using a water seal monitoring being exceeded. Off-shift and overtime
behind, all monitoring and sampling device: labor were used, to the extent
locations; (A) A detailed description of practicable;
(viii) The following information manufacturer’s specifications, (D) The amount and duration of the
regarding pilot gas and purge gas for including, but not limited to, make, excess emissions (including any bypass)
each flare: model, type, maintenance, and quality were minimized to the maximum extent
(A) Type(s) of gas used; assurance procedures; practicable during periods of such
(B) A complete description of the (B) A complete description of the emissions;
monitor(s) to be used, or the other proposed methods to determine that the (ii) For claims of startup or shutdown:
parameters that will be used and water seal is no longer intact and flow (A) All or a portion of the facility was
monitored, to determine volumetric is going to the flare, and the data used in startup or shutdown mode, resulting
flows of the pilot gas and purge gas to establish, and reasoning behind, these in the need to route gases to the flare;
streams for which natural gas or inert methods; (B) The periods of excess emissions
gas is not used; and (xiv) A schedule for the installation that occurred during startup and
(C) A complete description of the and operation of each flare monitoring shutdown were short and infrequent
analyzer(s) to be used to determine, or system consistent with the deadline in and could not have been prevented
other methods that will be used to paragraphs (h)(2) and (h)(3) of this through careful planning and design or
estimate, the H2S concentrations in the section; and better operation and maintenance
pilot gas and purge gas streams for (xv) A complete description of the practices; and
which natural gas or inert gas is not methods to be used for calculating (C) The frequency and duration of
used; hourly and 3-hour SO2 emissions from operation in startup or shutdown mode
(ix) A detailed description of flares. were minimized to the maximum extent
manufacturer’s specifications, (6) Thirty (30) days prior to installing practicable;
including, but not limited to, make, any continuous monitor or integrated (iii) For claims of malfunction,
model, type, range, precision, accuracy, sampler pursuant to paragraphs (h)(2) startup, or shutdown:
calibration, maintenance, quality and (3) of this section, each facility (A) If the excess emissions resulted
assurance procedure, and any other named in paragraph (a) of this section from a bypass of control equipment or
relevant specifications and information shall submit for EPA review a quality a process, then the bypass was
referenced in paragraphs (h)(2) and (3) assurance/quality control (QA/QC) plan unavoidable to prevent loss of life,
of this section for all existing and for each monitor or sampler being personal injury, or severe property
proposed flow monitoring devices and installed. damage;
total sulfur analyzers; (i) Affirmative defense provisions for (B) All possible steps were taken to
(x) The following information if grab exceedances of flare emission limits minimize the impact of the excess
or integrated sampling is used: during malfunctions, startups, and emissions on ambient air quality;
(A) A complete description of shutdowns. (C) All emissions monitoring systems
proposed analytical and sampling (1) In response to an action to enforce were kept in operation if at all possible;
methods if grab or integrated sampling the emission limits in paragraphs (D) The owner or operator’s actions in
methods will be used for determining (d)(2)(i), (e)(2)(i), (f)(2)(i), and (g)(2)(i) of response to the excess emissions were
the total sulfur concentration of the gas this section, owners and/or operators of documented by properly signed,
stream going to the flare; the facilities named in paragraph (a) of contemporaneous operating logs;
(E) The excess emissions were not be available to claims for injunctive following address: American Society for
part of a recurring pattern indicative of relief. Testing and Materials (ASTM), 100 Barr
inadequate design, operation, or (j) Incorporation by reference. (1) The Harbor Drive, Post Office Box C700,
maintenance; materials listed in this paragraph are West Conshohocken, PA 19428–2959,
(F) At all times, the facility was incorporated by reference in the www.astm.org, or by calling (610) 832–
operated in a manner consistent with corresponding paragraphs noted. These 9585.
good practices for minimizing incorporations by reference are (i) ASTM Method D4468–85
emissions; and approved by the Director of the Federal (Reapproved 2000), Standard Test
(G) During the period of excess Register in accordance with 5 U.S.C. Method for Total Sulfur in Gaseous
emissions, there were no exceedances of 552(a) and 1 CFR part 51. These Fuels by Hydrogenolysis and
the SO2 NAAQS that could be attributed materials are incorporated as they exist Rateometric Colorimetry, IBR approved
to the emitting source. on the date of the approval, and notice for paragraph (h)(3)(i)(B)(3) of this
(2) Notification. The owner or of any change in these materials will be section.
operator of the facility experiencing an published in the Federal Register. The (ii) ASTM Method D4810–06,
exceedance of its flare emission limit(s) materials are available for purchase at Standard Test Method for Hydrogen
during startup, shutdown, or the corresponding address noted below, Sulfide in Natural Gas Using Length-of-
malfunction shall notify EPA verbally as and all are available for inspection at Stain Detector Tubes, IBR approved for
soon as possible, but no later than noon the National Archives and Records paragraphs (f)(3)(ii)(B), (g)(4)(ii)(C), and
of EPA’s next working day, and shall Administration (NARA) and at the Air (g)(5)(ii)(C) of this section.
submit written notification to EPA Program, EPA, Region 8, 1595 Wynkoop (ii) ASTM Method D5504–01
within 30 days of the initial occurrence Street, Denver, CO. For information on (Reapproved 2006), Standard Test
of the exceedance. The written the availability of this material at Method for Determination of Sulfur
notification shall explain whether and NARA, call 202–741–6030, or go to: Compounds in Natural Gas and Gaseous
how the elements set forth in paragraph http://www.archives.gov/ Fuels by Gas Chromatography IBR
(i)(1) of this section were met, and federal_register/ approved for paragraph (h)(3)(i)(B)(3) of
include all supporting documentation. code_of_federal_regulations/ this section.
(3) Injunctive relief. The Affirmative ibr_locations.html.
Defense Provisions contained in (2) The following materials are [FR Doc. E8–7868 Filed 4–18–08; 8:45 am]
paragraph (i)(1) of this section shall not available for purchase from the BILLING CODE 6560–50–P

Rule: Federal Implementation Plan For The Billings/Laurel, MT, Sulfur Dioxide Area

Încărcat de

Informații document

Titlu original

Drepturi de autor

Formate disponibile

Partajați acest document

Partajați sau inserați document

Opțiuni de partajare

Vi se pare util acest document?

Este necorespunzător acest conținut?

Drepturi de autor:

Formate disponibile

Rule: Federal Implementation Plan For The Billings/Laurel, MT, Sulfur Dioxide Area

Încărcat de

Drepturi de autor:

Formate disponibile

Monday,

April 21, 2008

(b) Comment (ExxonMobil, WSPA, instances they had seen copious

provide for attainment and maintenance as to prevent attainment or interfere

technique served the modeling been ignored. The settlement documents

Because we will have primary

with applicable law or otherwise § 52.1392 Federal Implementation Plan for

Environmental Protection Agency. Quality. (1) Introduction. The provisions for

S-ar putea să vă placă și