# National Emission Standards for Hazardous Air Pollutants for Area Sources: Asphalt Processing and Asphalt Roofing Manufacturing > Briefs, arguments, decisions, and more. URL: https://www.frixlaw.com/law-library/documents/fr%3AE9-27946 ## Record - **Collection:** Federal Register - **Document type:** Rule - **Published:** December 2, 2009 - **Citation:** 74 FR 63236 ## Text ENVIRONMENTAL PROTECTION AGENCY 40 CFR Part 63 [EPA-HQ-OAR-2009-0027; FRL-8983-6] RIN 2060-AO94 National Emission Standards for Hazardous Air Pollutants for Area Sources: Asphalt Processing and Asphalt Roofing Manufacturing AGENCY: Environmental Protection Agency (EPA). ACTION: Final rule. SUMMARY: EPA is promulgating national emissions standards for the control of emissions of hazardous air pollutants (HAP) from the asphalt processing and asphalt roofing manufacturing area source category. These final emissions standards for new and existing sources are based upon EPA's final determination as to what constitutes the generally available control technology or management practices (GACT) for the source category. DATES: This final rule is effective on December 2, 2009. ADDRESSES: EPA has established a docket for this action under Docket ID No. EPA-HQ-OAR-2009-0027. All documents in the docket are listed in the Federal Docket Management System index at http://www.regulations.gov . Although listed in the index, some information is not publicly available, e.g., confidential business information or other information whose disclosure is restricted by statute. Certain other material, such as copyrighted material, is not placed on the Internet and will be publicly available only in hard copy form. Publicly available docket materials are available either electronically through http://www.regulations.gov or in hard copy at the EPA Public Reading Room under the heading “Area Source National Emission Standards for Hazardous Air Pollutants (NESHAP) for Asphalt Processing and Asphalt Roofing Manufacturing.” The Public Reading Room is located at EPA/DC, EPA West, Room 3334, 1301 Constitution Ave., NW., Washington, DC and is open from 8:30 a.m. to 4:30 p.m., Monday through Friday, excluding legal holidays. The telephone number for the Public Reading Room is (202) 566-1744, and the telephone number for the Air Docket is (202) 566-1742. FOR FURTHER INFORMATION CONTACT: Mr. Warren Johnson, Outreach and Information Division, Office of Air Quality Planning and Standards (MC-C404-05), Environmental Protection Agency, Research Triangle Park, North Carolina 27711, telephone number: (919) 541-5124; fax number: (919) 541-0242; e-mail address: johnson.warren@epa.gov. SUPPLEMENTARY INFORMATION: Outline. The information in this preamble is organized as follows: I. General Information A. Does This Action Apply to Me? B. Where Can I Get a Copy of This Document? C. Judicial Review II. Background Information for This Final Rule III. Summary of Major Changes Since Proposal IV. Summary of Final Standards A. Do the Final Standards Apply to My Source? B. When Must I Comply With the Final Standards? C. What Are the Final Standards? D. What Are the Initial and Continuous Compliance Requirements? E. What are the Notification, Recordkeeping, and Reporting Requirements? F. What Are the Title V Permit Requirements? V. Summary of Comments and Responses A. Source Category Listing B. GACT Limits C. Initial Compliance Requirements D. Continuous Compliance Requirements E. Title V Permitting F. Definitions G. Cost Impacts H. Miscellaneous VI. Summary of Impacts of the Final Standards A. What Are the Air Impacts? B. What Are the Cost Impacts? C. What Are the Economic Impacts? D. What Are the Non-Air Health, Environmental, and Energy Impacts? VII. Statutory and Executive Order Reviews A. Executive Order 12866: Regulatory Planning and Review B. Paperwork Reduction Act C. Regulatory Flexibility Act D. Unfunded Mandates Reform Act E. Executive Order 13132: Federalism F. Executive Order 13175: Consultation and Coordination With Indian Tribal Governments G. Executive Order 13045: Protection of Children From Environmental Health and Safety Risks H. Executive Order 13211: Actions Concerning Regulations That Significantly Affect Energy Supply, Distribution, or Use I. National Technology Transfer and Advancement Act J. Executive Order 12898: Federal Actions To Address Environmental Justice in Minority Populations and Low-Income Populations K. Congressional Review Act I. General Information A. Does This Action Apply to Me? The regulated categories and entities potentially affected by the final standards include: Category NAICS code 1 Examples of regulated entities Petroleum Refineries 324110 Area source facilities that refine asphalt. Asphalt Shingle and Coating Materials Manufacturing 324122 Area source facilities that manufacture asphalt roofing materials. 1 North American Industry Classification System. This table is not intended to be exhaustive, but rather provides a guide for readers regarding entities likely to be affected by this action. To determine whether your facility would be regulated by this action, you should examine the applicability criteria in 40 CFR 63.11559 of subpart AAAAAAA (NESHAP for Area Sources: Asphalt Processing and Asphalt Roofing Manufacturing). If you have any questions regarding the applicability of this action to a particular entity, consult either the air permit authority for the entity or your EPA Regional representative as listed in 40 CFR 63.13 of subpart A (General Provisions). B. Where Can I Get a Copy of This Document? In addition to being available in the docket, an electronic copy of this final action will also be available on the Worldwide Web (WWW) through the Technology Transfer Network (TTN). Following signature, a copy of this final action will be posted on the TTN's policy and guidance page for newly final or promulgated rules at the following address: http://www.epa.gov/ttn/oarpg/. The TTN provides information and technology exchange in various areas of air pollution control. C. Judicial Review Under section 307(b)(1) of the Clean Air Act (CAA), judicial review of this final rule is available only by filing a petition for review in the United States Court of Appeals for the District of Columbia Circuit by February 1, 2010. Under section 307(b)(2) of the CAA, the requirements established by this final rule may not be challenged separately in any civil or criminal proceedings brought by EPA to enforce these requirements. Section 307(d)(7)(B) of the CAA further provides that “[o]nly an objection to a rule or procedure which was raised with reasonable specificity during the period for public comment (including any public hearing) may be raised during judicial review.” This section also provides a mechanism for EPA to convene a proceeding for reconsideration, “[i]f the person raising an objection can demonstrate to EPA that it was impracticable to raise such objection within [the period for public comment] or if the grounds for such objection arose after the period for public comment (but within the time specified for judicial review) and if such objection is of central relevance to the outcome of the rule.” Any person seeking to make such a demonstration to us should submit a Petition for Reconsideration to the Office of the Administrator, U.S. EPA, Room 3000, Ariel Rios Building, 1200 Pennsylvania Ave., NW., Washington, DC 20460, with a copy to both the person(s) listed in the preceding FOR FURTHER INFORMATION CONTACT section, and the Associate General Counsel for the Air and Radiation Law Office, Office of General Counsel (Mail Code 2344A), U.S. EPA, 1200 Pennsylvania Ave., NW., Washington, DC 20460. II. Background Information for This Final Rule Section 112(d) of the Clean Air Act (CAA) requires EPA to establish national emission standards for hazardous air pollutants (NESHAP) for both major and area sources of HAP that are listed for regulation under CAA section 112(c). A major source emits or has the potential to emit 10 tons per year (tpy) or more of any single HAP or 25 tpy or more of any combination of HAP. An area source is a stationary source that is not a major source. Section 112(k)(3)(B) of the CAA calls for EPA to identify at least 30 HAP which, as the result of emissions from area sources, pose the greatest threat to public health in the largest number of urban areas. The EPA implemented this provision in 1999 in the Integrated Urban Air Toxics Strategy (64 FR 38715, July 19, 1999). Specifically, in the Strategy, EPA identified 30 HAP that pose the greatest potential health threat in urban areas, and these HAP are referred to as the “30 urban HAP.” Section 112(c)(3) requires EPA to list sufficient categories or subcategories of area sources to ensure that area sources representing 90 percent of the emissions of the 30 urban HAP are subject to regulation. A primary goal of the Strategy is to achieve a 75 percent reduction in cancer incidence attributable to HAP emitted from stationary sources. Under CAA section 112(d)(5), we may elect to promulgate standards or requirements for area sources “which provide for the use of generally available control technologies or management practices (GACT) by such sources to reduce emissions of hazardous air pollutants.” Additional information on GACT is found in the Senate report on the legislation (Senate Report Number 101-228, December 20, 1989), which describes GACT as: * * * methods, practices and techniques which are commercially available and appropriate for application by the sources in the category considering economic impacts and the technical capabilities of the firms to operate and maintain the emissions control systems. Consistent with the legislative history, we can consider costs and economic impacts in determining GACT, which is particularly important when developing regulations for source categories, like this one, that have a number of small businesses. Determining what constitutes GACT initially involves considering the control technologies and management practices that are generally available to the area sources in the source category. We also consider the standards applicable to major sources in the same industrial sector to determine if the control technologies and management practices employed by those sources are transferable and generally available to area sources. In appropriate circumstances, we may also consider technologies and practices at area and major sources in similar categories to determine whether such technologies and practices could be considered generally available for the area source category at issue. Finally, as noted above, in determining GACT for a particular area source category, we consider the costs and economic impacts of available control technologies and management practices on that category. We are promulgating these national emission standards in response to a court-ordered deadline that requires EPA to issue standards for certain source categories listed pursuant to section 112(c)(3) and (k) by November 16, 2009 ( Sierra Club v. Johnson, no. 01-1537, D.D.C., March 2006). An additional rulemaking will be published in a separate Federal Register notice for the remaining source category due in November 2009. III. Summary of Major Changes Since Proposal The final rule contains several revisions and clarifications to the proposed rule made in response to public comments. We explain the reasons for the following changes in detail in the summary of comments and responses (section V of this preamble): • Revised the emission limits for asphalt roofing manufacturing lines using emissions data supplied by the industry; • Revised the initial compliance requirements to specify that compliance tests must be conducted while manufacturing the product with the greatest polycyclic aromatic hydrocarbons (PAH) and particulate matter (PM) emissions and to allow facilities to use process knowledge to demonstrate initial compliance for saturator-only lines; • Revised the initial compliance requirements to clarify procedures for using previously-conducted emission tests to demonstrate compliance; • Revised the equations for calculating asphalt charging rate and clarified the procedures for determining production rate; • Revised the continuous compliance requirements to allow for monitoring of parameter ranges (instead of maintaining the parameter below a maximum value) and use of equipment manufacturer specifications when establishing parameter values, and to remove the option to use a continuous emissions monitor (CEMS); • Revised the continuous compliance requirements to allow facilities to monitor the indicator light of electrostatic precipitators (ESPs) as an option to monitoring voltage; • Defined PM as the material collected using EPA Method 5A; and • Added definitions for “built-up roof operation” and “hot-mix asphalt operation” and clarified the definition of “saturator” with regard to impregnation vats. IV. Summary of Final Standards A. Do the Final Standards Apply to My Source? The final subpart AAAAAAA standards apply to each existing and new area source facility that processes asphalt and/or manufactures roofing products using saturation and/or coating processes that apply asphalt to a substrate. The standards do not apply to research or laboratory facilities, as defined in section 112(c)(7) of the CAA. B. When Must I Comply With the Final Standards? All existing area source facilities subject to this final rule are required to comply with the rule requirements no later than December 2, 2010. New sources are required to comply with the rule requirements by December 2, 2009 or upon startup of the facility, whichever is later. Because the majority of existing sources in this category are already well-controlled, we believe that one year is a reasonable amount of time to allow existing sources to conduct compliance testing and prepare compliance demonstrations showing compliance with the final rule. C. What Are the Final Standards? As discussed in section II.C of this preamble, the two production operations for which this category was listed are: (1) Asphalt processing (refining) operations; and (2) roofing product manufacturing operations. For asphalt processing, the final standards require the owner or operator to limit PAH emissions to 0.003 lb/ton of asphalt charged to the asphalt refining (blowing still) operation. Alternatively, owners or operators may comply with a PM emissions limit of 1.2 lb/ton of asphalt charged to the asphalt refining operation. The alternative PM limit ensures reductions in emissions of PAH that are at least equivalent to those achieved through compliance with the PAH emission limit. The final standards for new refining operations are the same as for existing sources. For asphalt roofing product manufacturing operations, we examined the process operations and other factors and determined that it was appropriate to establish subcategories that reflect the unique emission characteristic profiles of the different process types (equipment configurations). We developed three subcategories based upon the various process types used in the industry: (1) Production lines that use a coater only, (2) production lines that use a saturator only, and (3) production lines that use both saturators and coaters. For existing coater-only production lines, the final standards require the owner or operator to limit PAH emissions from all coating mixers and coaters to 0.0002 lb/ton of product manufactured. Alternatively, owners or operators may choose to comply with a PM emission limit of 0.06 lb/ton of product manufactured. The alternative PM limit ensures reductions in emissions of PAH that are at least equivalent to those achieved through compliance with the GACT-based PAH emission limit. For existing saturator-only production lines, the final standards require the owner or operator to limit PAH emissions from all saturators (and wet loopers) to 0.0007 lb/ton of product manufactured. Alternatively, for saturator-only production lines, owners or operators can comply with a PM emissions limit of 0.30 lb/ton of product manufactured. The alternative PM limit ensures reductions in emissions of PAH that are at least equivalent to those achieved through compliance with the GACT-based PAH emission limit. For existing combined saturator and coater production lines, the final standards require the owner or operator to limit PAH emissions from all saturators, wet loopers, coating mixers, and coaters to 0.0009 lb/ton of product manufactured. The final standards for combined saturator and coater production lines alternatively allow owners or operators to comply with a PM emissions limit of 0.36 lb/ton of product manufactured. The alternative PM limit ensures reductions in emissions of PAH that are at least equivalent to those achieved through compliance with the GACT-based PAH emission limit. The final standards for new roofing product manufacturing operations for all subcategories are the same as those for existing sources. D. What Are the Initial and Continuous Compliance Requirements? The final standards require an initial compliance assessment of the process emissions or control device outlet concentration to demonstrate initial compliance with the applicable standard, and to establish monitoring parameter values ( e.g., temperature, pressure drop) for the process or control device that will be monitored to demonstrate continuous compliance. For PM control devices used on asphalt roofing lines, the final rule allows owners or operators to establish monitoring parameter operating ranges based upon equipment manufacturer guarantees. For existing sources, the final standards require owners or operators to conduct the initial compliance assessment by May 31, 2011. Owners or operators of new sources are required to conduct the initial compliance assessment by June 1, 2010 or within 180 days after startup, whichever is later. For existing and new blowing stills and asphalt roofing manufacturing lines, the final standards require owners or operators to demonstrate initial compliance by conducting emission tests or by using the results from an emission test conducted in the past five years that meets the specified criteria in the final rule. Specifically, owners or operators can use the results of the previously-conducted test only if the emission measurements were made using the test methods specified in Table 3 of the final rule. See 40 CFR 63.11562(d). Additionally, the owner or operator must be able to demonstrate that no process changes have been made since the date of the previous test, or that the results of the emissions test reliably demonstrate compliance despite any process changes. Id. For existing and new asphalt processing and asphalt roofing manufacturing lines that do not require a control device to comply with the emission limits, the final rule allows owners or operators to use process knowledge and engineering calculations, instead of compliance test results, to demonstrate initial compliance. For example, an owner or operator could use a mass-balance approach ( e.g., based upon asphalt throughput, asphalt content of the product manufactured) to demonstrate that the emission limits would not be exceeded. Continuous compliance with the final emission limits is demonstrated by monitoring parameters and process conditions established during the initial compliance assessment. The final standards require owners and operators to demonstrate continuous compliance based upon a 3-hour averaging period. If a thermal oxidizer is used to comply with the emission limits, the final standards require that the 3-hour average combustion zone temperature of each affected thermal oxidizer be maintained at or above the operating limit established during the initial compliance assessment. For PM control devices, the final standards require that the average 3-hour pressure drop and inlet gas temperature values be maintained within the range of established values. As an alternative to monitoring temperature and pressure drop, the final rule allows owners or operators to use a leak detection system for a filtration-based PM control device. If an ESP is used as the PM control device, the final standards require that the 3-hour average ESP voltage be maintained at or above the operating value established during the initial compliance test. As an alternative to monitoring the ESP voltage, the final rule allows owners or operators to monitor the device's indicator and warning lights on the device that signify when the ESP must be cleaned. For other types of control devices, the final standards allow the owner or operator to establish approved monitoring parameters and to maintain the value of those parameters within the operating values established during the initial compliance assessment. In cases where add-on control devices are not needed to comply with the final standards, owners or operators are required to establish a range of operating values for process parameters based upon written equipment manufacturer specifications, verify that the equipment is operating within that range during the initial compliance assessment, and maintain the 3-hour average of those parameters within the established values. During periods of startup and shutdown, the final standards require owners and operators to demonstrate compliance over a 24-hour averaging period. As is explained below, the final rule does not establish separate standards for malfunctions and the 3-hour averaging period applies during such events. Thus, consistent with Sierra Club v. EPA, 551 F.3d 1019 (DC Cir. 2008), the emission standards of this rule apply at all times. E. What Are the Notification, Recordkeeping, and Reporting Requirements? Affected new and existing sources are required to comply with certain requirements set forth in the General Provisions (40 CFR part 63, subpart A), as identified in Table 5 of this final rule. The General Provisions include specific requirements for notifications, recordkeeping, and reporting. Among other requirements, each facility is required to submit an initial notification that complies with the requirements in 40 CFR 63.9(b) of the General Provisions within 120 days of the effective date of the final rule and a notification of compliance status that complies with the requirements in 40 CFR 63.9(h) within 60 days after completion of the compliance assessment. Facilities are also required to submit semi-annual compliance summary reports. F. What Are the Title V Permitting Requirements? This final rule exempts the asphalt processing and asphalt roofing manufacturing area source category from title V permitting requirements unless the affected source is otherwise required by law to obtain a title V permit. For example, sources that have title V permits because they are major sources under the criteria pollutant program would maintain those permits. V. Summary of Comments and Responses We received a total of six comment letters from industry trade associations, an environmental advocacy group, State/local regulatory agency groups, and a control device equipment vendor on the proposed rule during the comment period. One commenter, an industry trade association, expressed support for the following provisions in the proposal package: • The roofing line subcategory designations; • The definition of the affected source for asphalt processing and asphalt roofing manufacturing operations; • The PAH and PM GACT emission standards for new and existing sources; • The definitions of “asphalt flux,” “asphalt processing operation,” and “blowing still;” • The use of PM emissions as a surrogate for PAH emissions; • The use of certain previously-conducted emission tests to demonstrate initial compliance with the emission limitations; and • The exemption from title V permitting requirements. We acknowledge the commenter's support for these provisions. Sections V.A. through V.H. contain summaries of the remaining comments that we received and our responses to those comments. A. Source Category Listing Comment. One commenter asserted that the Agency used inaccurate PAH emissions data for 1990 to list asphalt processing and asphalt roofing manufacturing area sources under CAA section 112(c)(3). The commenter asserted that urban area source PAH emissions in the industry in that baseline year were significantly lower than EPA's estimates and provided a copy of a report previously submitted to the Agency that the commenter contended supports that assertion. The commenter's report concludes that, by combining asphalt roofing manufacturing and asphalt processing into a single source category and using the outdated data, the EPA's PAH emissions estimate for the two categories is overstated by nearly two orders of magnitude. Based upon this information, the commenter stated that EPA should not be issuing GACT standards for asphalt processing and asphalt roofing manufacturing area sources under CAA section 112(c)(3). Response. We listed the asphalt processing and asphalt roofing manufacturing source category under CAA section 112(c)(3) in one of a series of amendments (November 22, 2002, 67 FR 70427) to the original source category list included in the 1999 Integrated Urban Air Toxics Strategy. As explained in more detail below, we included this source category on the section 112(c)(3) area source category list based upon emissions data for the 1990 baseline year. The asphalt processing and asphalt roofing manufacturing source category was listed for its contributions toward meeting the requirement that we list sufficient categories and subcategories of area sources to ensure that area sources representing 90 percent of area source emissions of PAH are subject to regulation under CAA section 112. While Congress required EPA to list sufficient categories or subcategories of area sources to ensure that areas sources representing 90 percent of the area source emissions of the 30 Urban HAP are subject to regulation under section 112 of the Clean Air Act, it left it to EPA's discretion to determine which categories and subcategories of sources to include on the list. As explained in the Integrated Urban Air Toxics Strategy, EPA based its listing decisions on the baseline National Toxics Inventory (NTI) that the Agency compiled for purposes of implementing its air toxics program after the 1990 CAA Amendments (64 FR 38706, 38711, n.10). The baseline NTI reflected HAP emissions from asphalt processing and asphalt roofing manufacturing area sources in 1990. EPA listed the asphalt processing and asphalt roofing area source category on the basis of that emissions data. EPA continues to believe that it was reasonable to rely on that data and that it acted appropriately in including the asphalt processing and asphalt roofing area source category on the list on the basis of that data. There is nothing in the comments that persuades EPA that the asphalt processing and asphalt roofing manufacturing area source category should not be included in the source category list. The report submitted along with the comments clearly reflects the Commenter's preference that a different source category, asphalt concrete manufacturing, be included on the list instead of asphalt processing and asphalt roofing manufacturing and that the inclusion of that source category would have also resulted in a cumulative percentage contribution in excess of 90 percent. This, however, misses the point. As stated above, Congress left it to EPA's discretion to determine which categories and subcategories to include on the list. Congress did not require EPA to establish a rank order of such categories and subcategories and then move from the highest ranking source category or subcategory to lower ranking categories or subcategories until a cumulative total of 90 percent was reached. Thus, as long as EPA had some basis for including a particular category or subcategory of area sources on the list, which is the case here, it can choose to include that category or subcategory even if there are other potential source categories or subcategories that arguably may contribute more to cumulative emissions. In this particular instance, EPA questions the accuracy of the emission factors used in the report submitted by the commenter. Specifically, the emissions factors in the commenter's report are based primarily on emissions data from 1998 and 1999 (with some reliance on 1994 data). The report takes these emission factors that are based on post-1990 data and applies them to 1990 production rates. As the commenter points out in its comments, PAH emissions in the asphalt processing and asphalt roofing manufacturing industry have declined since 1990. As a result, emission factors developed using emissions data from years after 1990 are likely to underestimate actual emissions in 1990. Moreover, even if EPA were to accept, for argument's sake, the revised emissions estimates set forth in the report submitted by the commenter, it would, for the reasons described below, continue to believe that the asphalt processing and asphalt roofing manufacturing category belongs on the 112(c)(3) source category list. First, EPA believes that it is most appropriate to consider asphalt processing and asphalt roofing manufacturing as a single source category rather than two separate source categories, as the commenter contends, because a single facility often includes both types of operations. Indeed, 90 percent of the facilities affected by the final rule conduct both asphalt processing and asphalt roofing manufacturing operations at the same site. We also believe that asphalt processing and asphalt roofing manufacturing operations are closely linked, regardless of co-location, because the purpose of blow stills at asphalt processing operations is to prepare asphalt flux, obtained from refineries, for use in manufacturing roofing products ( e.g., shingles, roll roofing). Second, while the commenter contends that asphalt concrete manufacturing should be included on the list instead of asphalt processing and asphalt roofing manufacturing, the fact is that, on a per facility basis, the asphalt processing and asphalt roofing manufacturing sources are larger PAH emissions sources than the asphalt concrete industry sources. As a result, EPA's regulation of the 75 sources in the asphalt processing and asphalt roofing manufacturing area source category is far more cost efficient and far more feasible from an implementation perspective than regulating the 3600 facilities engaged in asphalt concrete manufacturing. Finally, as explained above, Congress afforded EPA discretion in selecting the source categories to regulate to meet the 90 percent requirement in section 112(c)(3) and (k)(3)(B). Without the asphalt processing and asphalt roofing manufacturing source category, we will not meet this requirement. In conclusion, Congress required EPA to list sufficient categories and subcategories of sources of area sources to ensure that area sources representing 90 percent of the area source emissions of the 30 urban HAP are subject to regulation under CAA section 112. EPA has discretion to identify the categories and subcategories on the list and properly included asphalt processing and asphalt roofing manufacturing on the list. Nothing in the comments contradicts this. B. GACT Limits Comment. One commenter noted that EPA stated in the proposal notice that “[w]e believe that all asphalt processing and asphalt roofing manufacturing facilities will be able to meet the proposed standards using existing controls * * *” and that “* * * no additional air pollution control devices would be required.” The commenter was concerned that such proposals are merely paperwork exercises and are not responsive to Congress' intent in establishing the area source program under the Clean Air Act which the commenter believed should result in reductions in emissions from area sources of hazardous air pollution. Moreover, the commenter recommended that, “* * * in this rule and in future area source proposals, EPA incorporate provisions that will provide additional public health protection from the adverse effects of emissions of hazardous air pollutants from area sources.” Response. The commenter does not challenge any aspect of EPA's proposed GACT determination for this area source category. Instead, the commenter makes a blanket assertion that EPA is not acting consistently with the purposes of the area source provisions in the CAA ( i.e., sections 112(c)(3) and 112(k)(3)(B)), because it is not requiring emission reductions beyond the level that is currently being achieved from this well-controlled source category. In support of this assertion, the commenter compares the requirements in the proposed rule to the area source category's current emission and control status. Such a comparison is flawed. Congress promulgated the relevant CAA area source provisions in 1990 in light of the level of area source HAP emissions at that time. Congress directed EPA to identify not less than 30 HAP which, as a result of emissions from area sources, present the greatest threat to public health in the largest number of urban areas, and to list sufficient area source categories to ensure that sources representing 90 percent of the 30 HAP listed are subject to regulation. As explained in the Integrated Urban Air Toxics Strategy, EPA based its listing decisions on the baseline National Toxics Inventory (NTI) that the Agency compiled for purposes of implementing its air toxics program after the 1990 CAA Amendments (64 FR 38706, 38711, n.10). The baseline NTI reflected HAP emissions from asphalt processing and asphalt roofing manufacturing area sources in 1990. Thus, contrary to the commenter's suggestion, the relevant emission level for comparison is the emission level reflected in our baseline NTI, not the current emission level. Furthermore, in promulgating the area source provisions in the CAA, Congress did not require EPA to issue area source standards that must achieve a specific level of emission reduction. Rather, Congress authorized EPA to issue standards under section 112(d)(5) for area sources that reflect GACT for the source category. As Congress itself recognized, to qualify as being generally available, a GACT-based standard would most likely be based upon an existing control technology or management practice: “[A]n equipment standard would require neighborhood dry cleaning establishments to employ the commercially available systems associated with the lowest measured emissions * * * S. Rep. 101-128, at 171-172 (emphasis added). Thus, it is both reasonable and consistent with Congressional intent that the GACT-based standards being finalized today codify the use of the existing effective PAH control approach being used by sources in the category. For all of these reasons, this final rule is consistent with sections 112(c)(3), 112(k)(3)(B), and 112(d)(5). Comment. One commenter asserted that, although section 112(d)(5) does authorize EPA to issue GACT standards in lieu of MACT standards, the Agency's decision to do so is subject to familiar administrative law requirements. The commenter maintained that to be non-arbitrary, the decision must—at a minimum—be supported by a rational explanation. The commenter stated that EPA has provided no explanation whatsoever for its apparent decision to issue GACT standards pursuant to CAA section 112(d)(5), instead of MACT standards pursuant to section 112(d)(2) and (3) and, for this reason alone, its decision is arbitrary and capricious. The commenter also claimed that the proposed standards are based solely on cost and are thus unlawful and arbitrary. The commenter asserted that CAA section 112(d)(5) does not direct EPA to set standards based on what is cost effective; rather, according to the commenter EPA must establish GACT based on the “methods, practices and techniques which are commercially available and appropriate for application by the sources in the category considering economic impacts.” The commenter stated that because cost effectiveness is not relevant under CAA section 112(d)(5), the reliance on cost effectiveness as the sole determining factor in establishing GACT renders the proposed standards unlawful. Response. As the commenter acknowledged, in section 112(d)(5), Congress gave EPA explicit authority to issue alternative emission standards for area sources. Specifically, section 112(d)(5), which is titled “Alternative standard for area sources,” provides: With respect only to categories and subcategories of area sources listed pursuant to subsection (c) of this section, the Administrator may, in lieu of the authorities provided in paragraph (2) and subsection (f) of this section, elect to promulgate standards or requirements applicable to sources in such categories or subcategories which provide for the use of generally available control technologies or management practices by such sources to reduce emissions of hazardous air pollutants. See CAA section 112(d)(5). There are two critical aspects to section 112(d)(5). First, section 112(d)(5) applies only to those categories and subcategories of area sources listed pursuant to section 112(c). The commenter does not dispute that EPA listed the asphalt processing and asphalt roofing manufacturing area source category pursuant to section 112(c). Second, section 112(d)(5) provides that for area sources listed pursuant to section 112(c)(3), EPA “may, in lieu of” the authorities provided in section 112(d)(2) and 112(f), elect to promulgate standards pursuant to section 112(d)(5). Section 112(d)(2) provides that emission standards established under that provision “require the maximum degree of reduction in emissions” of HAP (also known as maximum available control technology (MACT)). Section 112(d)(3), in turn, defines what constitutes the “maximum degree of reduction in emissions” for new and existing sources. See section 112(d)(3). Webster's dictionary defines the phrase “in lieu of” to mean “in the place of” or “instead of.” See Webster's II New Riverside University (1994). Thus, section 112(d)(5) authorizes EPA to promulgate standards under section 112(d)(5) that provide for the use of GACT, instead of issuing MACT standards pursuant to section 112(d)(2) and (d)(3). The statute does not set any condition precedent for issuing standards under section 112(d)(5) other than that the area source category or subcategory at issue must be one that EPA listed pursuant to section 112(c), which is the case here. The commenter argues that EPA must provide a rationale for issuing GACT standards under section 112(d)(5), instead of MACT standards. The commenter is incorrect. Had Congress intended that EPA first conduct a MACT analysis for each area source category, Congress would have stated so expressly in section 112(d)(5). Congress did not require EPA to conduct any MACT analysis, floor analysis or beyond-the-floor analysis before the Agency could issue a section 112(d)(5) standard. Rather, Congress authorized EPA to issue GACT standards for area source categories listed under section 112(c), and that is precisely what EPA has done in this rulemaking. Although EPA need not justify its exercise of discretion in choosing to issue a GACT standard for an area source listed pursuant to section 112(c)(3), EPA still must have a reasoned basis for the GACT determination for the particular area source category. The legislative history supporting section 112(d)(5) provides that GACT is to encompass: * * * methods, practices and techniques which are commercially available and appropriate for application by the sources in the category considering economic impacts and the technical capabilities of the firms to operate and maintain the emissions control systems. See Senate Report on the 1990 Amendments to the Act (S. Rep. No. 101-228, 101st Cong. 1st session. 171-172). The discussion in the Senate report clearly provides that EPA may consider costs in determining what constitutes GACT for the area source category. Congress plainly recognized that area sources differ from major sources, which is why Congress allowed EPA to consider costs in setting GACT standards for area sources under section 112(d)(5), but did not allow that consideration in setting MACT floors for major sources pursuant to section 112(d)(3). This important dichotomy between section 112(d)(3) and section 112(d)(5) provides further evidence that Congress sought to do precisely what the title of section 112(d)(5) states—provide EPA the authority to issue “[a]lternative standards for area sources.” Notwithstanding the commenter's claim, EPA properly issued standards for the area source category at issue here under section 112(d)(5) and in doing so provided a reasoned basis for its selection of GACT for this area source category. As explained in the proposed rule and below, EPA evaluated the control technologies and management practices that reduce PAH emissions at asphalt processing and asphalt roofing manufacturing facilities. In its evaluation, EPA used information from an industry survey, discussed options for controlling PAH emissions with the industry trade associations, and reviewed operating permits to identify the emission controls and management practices that are currently used to control PM and PAH emissions. In our evaluation, we determined that all blow stills used to process asphalt are currently controlled using thermal oxidation. We also found that the majority of roofing manufacturing lines were controlled using some type of PM control device ( e.g., fiber-bed filters). Additionally, we determined that, due to market-driven process changes, the majority of roofing manufacturing facilities no longer use organic felt as the substrate for roofing materials. This process change significantly reduced the amount of asphalt used to manufacture a given quantity of roofing products. EPA disagrees with the commenter's assertions that EPA based its GACT determination solely on its estimate of cost effectiveness and that cost effectiveness is not relevant in determining what constitutes GACT. The Agency's consideration of cost effectiveness in establishing GACT and the Agency's views on what is a cost-effective requirement under section 112(d)(5) are relevant. The U.S. Court of Appeals for the DC Circuit has stated that cost effectiveness is a reasonable measure of cost as long as the statute does not mandate a specific method of determining cost. See Husqvarna AB v. EPA, 254 F.3d 195, 201 (DC Cir. 2001) (finding EPA's decision to consider costs on a per-ton-of-emissions removed basis is reasonable because CAA section 213 did not mandate a specific method of cost analysis). Further, we did not base our GACT determination solely on our estimate of cost effectiveness. Rather, we first carefully evaluated the methods, practices and techniques that are commercially available and appropriate for application by sources in the asphalt processing and asphalt roofing manufacturing area source category. Only then did we consider costs and economic impacts to determine what constitutes GACT for the source category. In doing so, we determined that, because sources in the asphalt processing and asphalt roofing manufacturing area source category currently have relatively low emissions of PAH based upon the use of existing controls, requiring additional controls would result in very high costs for only a modest incremental improvement in control. Finally, we believe the consideration of costs and economic impacts is especially important for determining GACT for the asphalt processing and asphalt roofing manufacturing area source category because of the number of existing sources that would need to retrofit controls on asphalt roofing manufacturing operations if the existing controls on those operations were determined inadequate. Even though we are not required to provide a specific rationale for why we chose to establish GACT-based standards, rather than MACT-based standards, EPA did in fact provide a rationale for doing so in the proposed rule. In the proposal, we explained that the facilities in the asphalt processing and asphalt roofing manufacturing area source category are already well controlled for PAH, the urban HAP for which the source category was listed pursuant to section 112(c)(3). See 74 FR 32826-32828. Consideration of costs and economic impacts is especially important when an area source category is comprised of sources that are already well-controlled. In such circumstances, a MACT floor determination, where costs cannot be considered, could result in very high costs for only a modest incremental improvement in control efficiency for sources in the area source category. EPA concluded that this would be the case were it to establish MACT-based emission standards for the asphalt processing and asphalt roofing manufacturing area source category. Comment. One commenter stated that EPA did not provide an explanation for its decision to narrowly focus the proposed rule on just PAH emissions. The commenter went on to make the following points. The commenter noted that in the 2003 NESHAP for the asphalt processing and asphalt roofing manufacturing major source category, the EPA stated that the major source category emits a variety of HAP. The commenter added that the preamble to the 2003 major source NESHAP (68 FR 22976, 22976 (Apr. 29, 2003)) stated that approximately 98 percent of emissions from the processing of asphalt and the manufacture of asphalt roofing consist of formaldehyde, hexane, hydrochloric acid (HCl), phenol and toluene. A combination of several different organic HAP comprise the remaining two percent of the total HAP emissions. The commenter said that in 2003, the EPA found that exposure to these HAP could result in both “chronic health disorders ( e.g., irritation of the lung, skin, and mucous membranes, effects on the central nervous system, and damage to the blood and liver) and acute health disorders ( e.g., respiratory irritation and central nervous system effects such as drowsiness, headache, and nausea).” Id. The commenter also noted that EPA classified two of the HAP (formaldehyde and polycyclic organic matter (POM)) as probable human carcinogens. The commenter stated that Section 112(d) requires that emission standards be developed for each HAP listed in section 112(b). Assuming arguendo that the Agency does not have to set separate standards for each HAP when issuing standards under section 112(d)(5), the commenter stated that the Agency still has an obligation to address all the HAP that a category emits when it sets GACT standards. Thus, the commenter asserted that EPA had an obligation to address the HAP emitted by asphalt processing and asphalt roofing manufacturing sources beyond PAH, especially in light of the fact that PAH is such a limited component of the HAP emitted by the source category. Further, the commenter added that the Agency's failure to even consider non-PAH HAP and to explain its failure to address these HAP is arbitrary and capricious. The commenter also noted that EPA failed to address all sources of HAP emissions in the asphalt processing and asphalt roofing manufacturing source category. The commenter pointed out that EPA noted in the 2003 major source NESHAP that, in addition to the blowing stills and roofing manufacturing operations addressed in the proposed rule, asphalt storage and process tanks, asphalt loading racks, sealant applicators, and adhesive applicators are also sources of HAP emissions. The commenter stated that the Agency's failure to acknowledge these emission sources and consider commercially available technology for reducing emissions from these sources was unlawful. Response. Section 112(k)(3)(B) of the CAA requires EPA to identify at least 30 HAP emitted from area sources that pose the greatest threat to public health in the largest number of urban areas (the “Urban HAP”) and identify the area source categories emitting such pollutants that are or will be listed pursuant to section 112(c)(3). Section 112(c)(3), in relevant part, provides: The Administrator shall * * *, pursuant to subsection (k)(3)(B) of this section, list, based on actual or estimated aggregate emissions of a listed pollutant or pollutants, sufficient categories or subcategories of area sources to ensure that area sources representing 90 percent of the area source emissions of the 30 hazardous air pollutants that present the greatest threat to public health in the largest number of urban areas are subject to regulation under this section. Thus, section 112(c)(3) requires EPA to list sufficient categories or subcategories of area sources to ensure that area sources representing 90 percent of the area source emissions of the 30 Urban HAP are subject to regulation. Section 112(d)(1) requires the Administrator to promulgate regulations establishing emissions standards for each area source category of HAP listed for regulation pursuant to section 112(c). EPA identified the 30 Urban HAP that posed the greatest threat to public health in the Integrated Urban Air Toxics Strategy (Strategy). In the Strategy and subsequent Federal Register notices, EPA listed the area source categories necessary to meet the 90 percent requirement in section 112(c)(3) and (k)(3)(B), and one of those categories was the Asphalt Processing and Asphalt Roofing Manufacturing area source category. We have interpreted sections 112(c)(3) and 112(k)(3)(B) together to require EPA to regulate only those Urban HAP emissions for which an area source category is listed pursuant to section 112(c)(3), not all urban HAP or all section 112(b) HAP emitted from a listed area source category. As stated above, section 112(k)(3)(B) addresses the strategy to control HAP from area sources in urban areas and the focus of the strategy as it relates to control of area sources is on the 30 HAP that pose the greatest threat to public health in the largest number of urban areas. Section 112(c)(3) specifically references section 112(k)(3)(B) as the basis for selecting area sources for listing to satisfy the Agency's responsibility for regulating urban HAP emissions from area sources. Under these provisions, area sources categories are listed because they emit one or more of the 30 listed Urban HAP and the Agency has identified the category as one that is necessary to satisfy the requirement to subject area sources representing 90 percent of the area source emissions of the 30 Urban HAP to regulation. EPA listed the Asphalt Processing and Asphalt Roofing Manufacturing area source category pursuant to sections 112(c)(3) and 112(k)(3)(B), based on the category's emissions of PAH, which is an urban HAP. Thus, consistent with the requirements of sections 112(c)(3) and 112(k)(3)(B), we must regulate the PAH emissions from the Asphalt Processing and Asphalt Roofing Manufacturing area source category, as these are the urban HAP emissions for which the category was listed to meet the 90 percent requirement in sections 112(c)(3) and (k)(3)(B). See 112(c)(3) (EPA must “ensure that area sources representing 90 percent of the area source emissions of the 30 hazardous air pollutants * * * are subject to regulation.”). We recognize that the source category emits other section 112(b) HAP, including other urban HAP; however, as stated above, sections 112(c)(3) and 112(k)(3)(B) do not require the Agency to regulate the area source category for any HAP other than those for which the category was listed. As to the other urban HAP emitted from this category, we have identified other area source categories that emit these urban HAP and subjecting those area source categories to regulation will satisfy the requirement to subject to regulation area sources that account for 90 percent of the area source emissions of those urban HAP. While the Agency is not required to regulate all section 112(b) HAP from area sources listed pursuant to section 112(c)(3) and 112(k)(3)(B), section 112 of the CAA does not preclude EPA from regulating other HAP from these area sources at our discretion and in appropriate circumstances. Section 112(d)(5) states that for area sources listed pursuant to section 112(c), the Administrator may, in lieu of section 112(d)(2) “MACT” standards, promulgate standards or requirements “applicable to sources” which provide for the use of GACT or management practices “to reduce emissions of hazardous air pollutants.” This provision does not limit EPA's authority to regulate only those urban HAP emissions for which the category is needed to achieve the 90 percent requirement in sections 112(k)(3)(B) and 112(c)(3). In fact, in two other area source rules, in addition to regulating the urban HAP that were necessary to satisfy the 90 percent requirement in sections 112(k)(3)(B) and 112(c)(3), we regulated additional section 112(b) HAP. Specifically, in the chemical manufacturing area source rule and the paint and allied products area source rule, although not required, we exercised our discretion to regulate other section 112(b) HAP beyond the urban HAP for which the categories were listed under section 112(c)(3) and (k)(3)(B), including non-urban section 112(b) HAP. The chemical manufacturing area source rule and the paints and allied products area source rule both involve specific circumstances which EPA believes justify regulating organic and metal section 112(b) HAP in addition to the specific urban HAP needed to meet the 90 percent requirement in section 112(c)(3) and (k)(3)(B), which served as the basis for the listing of the categories. In the chemical manufacturing area source rule, which establishes standards for 9 area source categories, we regulated such HAP because the emission standards designed to control the urban HAP for which the categories were listed were equally effective at removing other urban and non-urban metal and organic HAP, and demonstrating compliance for total HAP was less burdensome than demonstrating compliance for speciated HAP for those sources required to install add-on controls. In the paint and allied products area source rule, we included emission standards for HAP beyond the urban HAP for which the category was listed because the emission standards designed to control those urban HAP would also control other urban and non-urban metal and organic HAP. As noted above, the asphalt processing and asphalt roofing manufacturing area source category was listed solely due to emissions of PAH. By contrast, both the chemical manufacturing and the paint and allied products area source categories were listed for multiple urban HAP ( i.e., 1,3-butadiene; methylene chloride; 1,3-dichloropropene; hexachlorobenzene; acetaldehyde; hydrazine; chloroform; quinoline; ethylene dichloride; and HAP metal compounds (arsenic, cadmium, chromium, lead, manganese, and nickel) for chemical manufacturing, and benzene, methylene chloride, and compounds of cadmium, chromium, lead, and nickel for paint and allied products). For sources in these area source categories, it was reasonable to develop emission limits for non-urban HAP in part because the cost of estimating compliance for each urban HAP for which the categories were listed was overly burdensome. However, this same rationale is not appropriate in this rule because EPA listed the asphalt processing and asphalt roofing manufacturing source category based on the emissions of a single HAP (PAH). The co-control scenario also plays out differently in the context of the asphalt processing and asphalt roofing manufacturing area source category. Specifically, where an add-on control device like those used by facilities complying with the major source NESHAP ( e.g., a thermal oxidizer or a fiber-bed filter) is needed to comply with the final standards for the asphalt processing and asphalt roofing manufacturing area source category, the control device will achieve co-control of certain HAP other than PAH. For example, a thermal oxidizer will effectively control total HAP, total hydrocarbon (THC) and PM emissions and a fiber-bed filter will effectively control PM emissions. An emission limit based on the use of a thermal oxidizer ( e.g., a limit on total HAP or total THC) would, however, necessitate all emissions from regulated operations being routed to a thermal oxidizer or similar control device. At present, based on the available information, facilities only use thermal oxidizers to control emissions from asphalt processing operations. Thermal oxidizers are not currently used to control emissions from asphalt roofing manufacturing operations. As a result, such limits would require facilities to retrofit to route emissions from asphalt roofing manufacturing operations to a thermal oxidizer or similar control device. Such retrofits would increase the cost of complying with the standards to a level that is unacceptable for a GACT-based standard. We estimate that 29 existing facilities currently have a thermal oxidizer and the remaining 46 would need to install new controls. Even when assuming a best case scenario, whereby facilities would only need to install new ductwork to route emissions to an existing thermal oxidizer, we estimate that such facilities would have an estimated initial capital cost of $58,000 and annual maintenance costs adding up to $11,000. We believe that these estimates are unrealistically low, however, because the existing thermal oxidizers would also require supplemental fuel, and, in many cases, an upgrade of the control unit, in order to handle the increased emissions loading. We estimate that it would cost an average facility in excess of $1 million to install new thermal oxidation controls, with annual costs of just over $910,000 per year per facility for fuel and maintenance. In actuality, though, the costs could be much greater depending on the configuration of the facility. These cost concerns are further exacerbated by the fact that the benefits arising from co-control will be realized without EPA establishing specific emission limits for the co-controlled HAP. We therefore believe that we have appropriately exercised our discretion in regulating only the PAH emissions from the asphalt processing and asphalt roofing manufacturing area source category. The commenter further asserts that we failed to regulate all sources of HAP emissions. For the reasons described above, this rule establishes emissions standards for PAH only. To the extent the commenter is asserting that we failed to address all sources of PAH emissions, we disagree. We are required to regulate only those sources of PAH emissions that formed the basis of our listing decision. EPA based the listing of the asphalt processing and asphalt roofing manufacturing area source category solely on emissions from asphalt blowing (processing) and saturation of felt (using saturators, wet loopers, and coaters). Based on our review of the record supporting the listing decision, the record does not include emissions from asphalt loading racks, asphalt storage tanks, adhesive storage tanks, adhesive applicators, sealant storage tanks or sealant applicators. As a result, we did not establish PAH emission limits for those sources, as these emission sources were not part of the listed source category. Comment. One commenter stated that a significant problem with the proposal is that it would establish GACT standards that are actually more stringent—and significantly so—than the MACT standards for the industry. The commenter stated that they know of no other GACT standards that are more stringent than the corresponding MACT standards for the same industry. The commenter asserted that it makes no sense to have smaller area sources subject to more stringent standards than larger major sources. The commenter added that the very term “maximum achievable control technology” on its face indicates that the CAA section 112(d)(2) standards should be more stringent—they are the “maximum achievable” standards in contrast to the CAA section 112(d)(5) standards that are merely “generally available.” The commenter stated that for MACT, CAA section 112(d)(3) provides minimum levels of stringency, also known as the MACT “floor” levels. Thus, according to the commenter, the MACT standard for existing sources must be at least as stringent as the performance achieved by the average of the best performing 12 percent of sources in the category. The commenter stated that for new sources, the standard must be at least as stringent as that achieved by the best controlled similar source. In the subpart LLLLL asphalt processing and asphalt roofing manufacturing MACT rulemaking, the commenter noted that EPA concluded only six years ago that the average of the best performing 12 percent ( i.e., the 94th percentile of performance) was equivalent to the subpart UU NSPS limits. 66 FR 58617-20 (Nov. 21, 2001) (subpart LLLLL MACT proposal). The commenter stated that there have not been changes in the industry since publication of the final MACT standards in 2003 that would be expected to have rendered the assumptions for the MACT standards invalid. Thus, the commenter asserted that there is no basis for determining that any standards more stringent than the NSPS or MACT standards are “generally available.” The commenter stated that “The legislative history is replete with support for the proposition that GACT standards are to be less stringent than MACT standards. The Senate Report for the 1990 CAA Amendments states that “[t]he Administrator may require area sources to install MACT, but also has the option to impose less stringent emissions limitations reflecting generally available control technology.” Senate Report 101-228, in Congressional Research Service, A Legislative History of the Clean Air Act Amendments of 1990 (“A Legislative History”) 8338, 8490 (emphasis added). See also floor statement of Sen. Moynahan (“Clearly, this [GACT] requirement is less demanding than the maximum achievable control technology required for major point sources”) (April 3, 1990 Senate floor debate on S. 1630, in A Legislative History 6946, 7083); House Energy and Commerce Committee Markup of H.R. 3030 (The Waxman amendment requires EPA to regulate 90 percent of the area source emissions of each hazardous air pollutant. EPA may elect to establish controls based on “generally available control technology” in lieu of the more stringent controls based on “maximum achievable control technology” that would apply to major sources.” (Apr. 12, 1990 Clean Air Facts description of committee markup, in A Legislative History 2446, 2561). Another commenter added that the preamble did not contain any explanation for EPA's decision to impose more stringent requirements on smaller, lower-emitting facilities than on major sources. The commenter also cited rationale in Senate Report 101-228 that indicates the Congress intended GACT standards for area sources to be less stringent than MACT standards for major sources. The commenter also noted that EPA has taken the position that GACT is a less stringent standard in the preamble to the area source rulemaking for perchloroethylene dry cleaning facilities (58 FR 49354, 49356). Response. As described in detail below, we disagree with the commenters' basic premise that a GACT-based standard will always be less stringent than a previously-promulgated MACT-based standard, particularly in circumstances such as those here where the relevant MACT-based standard is more than 6 years old. Further, in this particular instance, the major source MACT-based NESHAP and the area source GACT-based standards are not directly comparable because they regulate different pollutants and different collections of process equipment. The MACT standards regulate total HAP with no speciation. The MACT also covers additional process equipment ( i.e., asphalt, adhesive, and sealant storage tanks, and adhesive and sealant applicators) that are not covered under the GACT-based standards. In assessing what constitutes GACT for the asphalt processing and asphalt roofing manufacturing area source category, we evaluated the control technologies and management practices that reduce PAH emissions at the asphalt processing and asphalt roofing manufacturing facilities that compose the source category. In our evaluation, we used information from an industry survey, discussed options for controlling PAH emissions with the industry trade association, and reviewed operating permits to identify the emission controls and management practices that are currently used to control PM and PAH emissions. In our evaluation, we determined that all of the blow stills used by facilities in the source category to process asphalt are currently controlled using thermal oxidation. We also found that the majority of roofing manufacturing lines was controlled using some type of PM control devices ( e.g., fiber-bed filters). Additionally, we determined that due to market-driven process changes, the majority of roofing manufacturing facilities no longer use organic felt as the substrate for roofing materials. The process change of no longer using organic felt as a substrate has significantly reduced the amount of asphalt used to manufacture a given quantity of roofing products. For all of these reasons, it is understandable that the GACT standard for this category is different than the MACT standard. After considering all of this information, we then considered costs and economic impacts in order to determine what actually constitutes GACT for the asphalt processing and asphalt roofing manufacturing area source category. While MACT-based standards for a given source category would most likely be more stringent than GACT-based standards for the same sources if the standards were developed at the same point in time, that is not the case here. Here, the GACT standards are based upon more recent process equipment, control device, and emissions data that were analyzed to support development of these standards, specifically. In contrast, the MACT standards were based upon data collected in 1995. Additionally, the GACT-based standards focus on the HAP (PAH) and processes (blowing stills and saturators, wet loopers, coaters, and coating mixers) for which this area source category was listed. The MACT-based standards were developed using a floor analysis for total HAP over a wider span of process equipment. Under such circumstances, the previously established MACT standard cannot reasonably be considered dispositive of the question of what constitutes GACT. Rather, as with any GACT determination, in determining what constitutes GACT for the asphalt processing and asphalt roofing manufacturing area source category, we first carefully evaluated the methods, practices and techniques that are commercially available and appropriate for application by sources in the asphalt processing and asphalt roofing manufacturing area source category. We then considered costs and economic impacts to determine what constitutes GACT. The GACT-based standards in this final rule reflect the Agency's determination, based on this evaluation, of GACT for the asphalt processing and asphalt roofing manufacturing area source category. Comment. One commenter did not believe that the proposed standards represent a GACT level of control because EPA used unrepresentative data, did not account for variability in establishing the emission limits, and determined the emission limits using the average. In developing the proposed GACT standards, the commenter noted that EPA used data from only one source in each source category. The commenter also stated that not only is the data too sparse, but it is not representative of GACT because the data were collected to support a MACT rulemaking ( i.e., the data were collected at the best-controlled sources in the industry). The commenter submitted PM emissions data from member companies for coater-only lines, saturator-only lines, and lines containing coaters and saturators. The commenter noted that there are numerous subpart UU NSPS compliance tests available documenting PM emissions from industry sources. The commenter added that, because the PM data have been collected to demonstrate compliance with air permits and the subpart UU NSPS, the data would meet the quality assurance and quality control standards required by State air pollution control agencies. The commenter stated that the standards should consider the variability in emissions due to: operational distinctions between different facilities or units ( i.e., roofing lines); between-test variability ( i.e., variability in measurements made at the same facility or unit at different times); and within-test variability ( i.e., measurement variations in individual test runs). The commenter stated that EPA and the courts have recognized the importance of using representative data and accounting for such variability between facilities, processes, and test results. In Sierra Club v. EPA, 167 F.3d 658, 665 (DC Cir. 1999), the U.S. Court of Appeals for the DC Circuit stated in a MACT case (under CAA section 129): “It is reasonable to suppose that if an emissions standard is as stringent as ‘the emissions control that is achieved in practice’ by a particular unit, then that particular unit will not violate the standard. This only results if ‘achieved in practice’ is interpreted to mean ‘achieved under the worst foreseeable circumstances.' ” The commenter stated that, in approving EPA's decision to account for variability in a CAA section 112 case by not setting the standards based upon the lowest emission limits, the court correctly pointed out that “even the best performing sources occasionally have spikes.” Mossville Environmental Action Now v. EPA, 372 F.3d 1232, 1242 (DC Cir. 2004). Similarly, the commenter noted that, under the technology-based NSPS, the DC Circuit's decisions “evince a concern that variables be accounted for, that the representativeness of test conditions by [sic] ascertained, that the validity of tests be assured and the statistical significance of results determined.” National Lime Ass'n v. EPA, 627 F.2d 416, 452-53 (DC Cir. 1980). See also Portland Cement Ass'n v. Ruckelshaus, 486 F.2d 375, 396 (DC Cir. 1973), cert. denied, 417 U.S. 921 (1974) (remanding NSPS in part due to “the lack of any indication of statistical reliability” in test results used to set standards). Moreover, the commenter asserted that a single test almost by definition cannot be representative of conditions found throughout an industry. The commenter said that the DC Circuit has held under CAA section 111, “a uniform standard must be capable of being met under most adverse conditions which can reasonably be expected to recur * * *” National Lime Ass'n, 627 F.2d at 431 n.46. See also Portland Cement Ass'n, 486 F.2d at 396 (noting industry point that “a single test offered a weak basis” for inferring that plants could meet the standards). Without accounting for variation among different emissions tests, the commenter stated that it cannot be determined with a significant degree of statistical confidence that even a single unit will not be able to meet the standard over a reasonable period of time, when one can expect adverse conditions to be present. The commenter noted that the courts have recognized this same basic principle in reviewing technology-based effluent standards under the Clean Water Act. As the Fifth Circuit stressed in reviewing “best practicable technology” or “BPT” standards under Clean Water Act section 304(b)(1): The same plant using the same treatment method to remove the same toxic does not always achieve the same result. Tests conducted one day may show a different concentration of the same toxic than are shown by the same test on the next day. This variability may be due to the inherent inaccuracy of analytical testing, i.e., “analytical variability,” or to routine fluctuations in a plant's treatment performance. Chemical Manufacturers Ass'n v. EPA, 870 F.2d 177, 228 (5th Cir. 1989). The commenter said that the Fifth Circuit upheld the standards because EPA expressly stated that they should be achievable “at all times apart from instances of upsets,” and because the Clean Water Act contains an “upset defense.” Id. at 230. See also American Petroleum Institute v. EPA, 540 F.2d 1023, 1035-36 (10th Cir. 1976) (“Even in the best treatment systems, changes occur in ability to treat wastes * * * [V]ariability factors present[] a practical effort to accommodate for variations in plant operations”); FMC Corp. v. Train, 539 F.2d 973, 985 (4th Cir. 1976) (variability factors account for “the fact that even in the best treatment systems changes continually occur in the treatability of wastes”). See also 47 FR 24534, 24546 (1982) (in setting general pH effluent limitation under the Clean Water Act, EPA pointed out that it “traditionally has recognized that it must take variability into account in establishing effluent limitations, and in recognition that 100 percent compliance is theoretically impossible, the Agency has generally set daily effluent limitations which would be met approximately 99 percent of the time”). The commenter noted that EPA pointed out in its brief in the Sierra Club v. EPA MACT case under CAA section 129 (discussed above), that simply trying to set a technology-based emission standard by considering a very limited dataset “ignores the critical distinction between an emission level that is ‘observed' on a particular occasion versus an emission level the Administrator determines is `achieved in practice' through performance because it is capable of being met continuously under the range of operating conditions that can reasonably be expected.” EPA brief at 35. Limited test results—the “observed” emissions levels—bear no relationship at all to what a variety of differently configured plants (or even a single unit) can achieve on a continuous basis. This is because each test produces a very limited sample of data. It does not provide a full enumeration of the available data for the unit's performance over a long period of time. See Natrella, Environmental Statistics, supra, chapter 1. The commenter stated that EPA inappropriately ignored basic statistical principles for environmental standard-setting. The commenter said that in any normally distributed set of data, 50 percent of the data points will be higher than the mean. Even assuming that the data were representative, a standard that 50 percent of sources do not meet would lead to a level of control more stringent than that generally available. The commenter stated that the use of the average uncontrolled emissions derived from a single test at a saturator/wet looper and a single test at a coater/coating mixer at one facility (the Tamko Frederick, MD facility) is inappropriate for setting standards. The commenter further stated that even assuming this is actually a median data point, 50 percent of the emission sources will have emissions higher than this source. The commenter noted that a paper published in a peer-reviewed journal showed that the emissions from uncontrolled coaters are variable (the standard deviation was 169 percent of the mean). The commenter stated that if the assumption is made that the data are distributed according to the t-Density function, this means that more than 33 percent of sources would be expected to have uncontrolled emissions of greater than 0.83 pounds/ton of product. To meet the 0.03 pound PM/ton of product standard, the commenter said that the cleanest of these sources (at 0.83 lbs/ton) would have to have unvarying emissions, and continuous control efficiencies of greater than 96 percent efficiency. The commenter also stated that EPA has inappropriately used average values in converting the emissions data to pounds of PM emitted per ton of product manufactured and in assessing the removal performance of high-efficiency air filter (HEAF) in calculating the proposed standards. The commenter suggested that a valid and reasonable approach to calculate representative emissions for such a small data set is to add two standard deviations to the mean (x) of the 3 stack testing runs. Assuming data are normally distributed, the commenter said that approximately 97.8 percent of sources in a normally distributed population would fall below this x + 2 standard deviations envelope. The commenter stated that because of EPA's flawed analysis, the proposed PAH and PM GACT emission standards for asphalt roofing manufacturing are too stringent and that EPA's assertion that the GACT standards can be met is incorrect. Response. We agree with the commenter that, as a general matter, it is desirable to have as robust a data set as possible when establishing emission limits. We also note, however, that EPA must often work with the data it has even though we might prefer to have additional data. We had a reasonable set of data upon which to base the proposed rule and it is within our discretion to determine whether it is appropriate to seek additional data before proposing to take a particular action. See, Natural Resources Defense Council v. EPA, 529 F.3d 1077 (D.C. Cir. 2008) (Recognizing that it is within EPA's discretion to determine when it is appropriate to rely on existing data rather than exercising its authority under section 114 of the Clean Air Act to obtain additional or new data.) In addition to actually having sufficient data upon which to base the proposed rule, we faced time constraints that precluded obtaining even more data due to the fact that we were trying to meet a court-ordered deadline for issuing the proposed rule. Finally, the rulemaking process itself is one of the primary ways in which EPA obtains relevant information. We agree with the commenter that additional roofing line emissions data would be helpful in establishing the GACT-based limits for this area source category. We also agree that variability in emissions is one of several important factors that need to be considered in establishing the GACT limits and that we had a less than desirable amount of data with which to consider statistical variability at proposal. The additional data provided with the industry comments, in combination with the data EPA relied on in developing the proposed rule, provides a robust data set for use in assessing both the actual performance of sources and the variability in that performance with the result that the final emission limits will be more statistically sound than those contained in the proposed rule. Consequently, the final GACT-based limits have been revised to take into account the additional data submitted by the commenter for asphalt roofing lines. Additionally, we considered the standard deviation of the data in establishing the revised emission limits. We are adding one standard deviation to the average of the data to account for variability. We considered adding two standard deviations to the average but we did not believe this approach was representative of GACT because the resulting emission limits were above the limits that most facilities already achieve. For the combined coater/saturator roofing lines, we are establishing the emission limits as the sum of the emissions limits for the coater-only and saturator-only lines. We used this approach for the combined coater/saturator roofing lines because the emissions are additive ( i.e., the process units are in series). The revised GACT limits for new and existing coater-only production lines are 0.0002 lb PAH/ton of product manufactured (or 0.06 lb PM/ton of product manufactured). For new and existing saturator-only production lines, the revised GACT limit is 0.0007 lb PAH/ton of product manufactured (or 0.30 lb PM/ton of product manufactured). For new and existing combined saturator and coater production lines, the revised GACT limit is 0.0009 lb PAH/ton of product manufactured (or 0.36 lb PM/ton of product manufactured). C. Initial Compliance Requirements Comment. One commenter contended that EPA proposed a very short compliance deadline for existing sources—only one year from issuance of the final rule. See section 63.11560(a). The commenter noted that the proposed one-year compliance deadline is premised upon EPA's assumption that sources will not have to install or modify air pollution control equipment to meet the standards. The commenter stated that this assertion is not true; however, as shown by the subpart UU NSPS test data in a report submitted by the commenter, a number of facilities have been operating above the proposed PM standards in the GACT proposal. Thus, according to the commenter, contrary to the proposal's justification, if the final standards are anywhere near the level of the proposed standards, the commenter stated that a number of facilities will need to make significant improvements to and/or reconstruct existing PM control equipment or install new equipment altogether to meet the proposed GACT limits. The commenter stated that NSPS subpart UU and MACT Method 5A testing data show that 20—50 percent of the potential GACT regulated sources surveyed by EME Solutions would be in non-compliance with the proposed GACT limits. Given that these sources will have to perform engineering testing(s) to assess compliance status, analyze results, design/develop solutions to the reason(s) for potential noncompliance, fabricate and install the solutions, and then perform compliance testing; eighteen months is much too short a time period. The commenter noted that the proposal also recognizes that there are uncontrolled sources in the industry. For example, many coating mixers are not currently controlled. Even if a facility has existing PM control equipment, the commenter contended that it will be necessary to install ducting to vent the currently-uncontrolled affected sources to the controls. The commenter also noted that many States require a construction permit to make modification to emissions control technology already in place. The permitting alone can take 9 months or longer. In addition, the commenter stated that the subpart LLLLL MACT standards provided a 3-year compliance date for existing sources, even though they were less stringent than the proposed GACT standards. The commenter said that there is no logical rationale for having a three-year compliance date for the MACT standards yet only a one-year compliance date for more stringent GACT standards. The commenter stated that for all these reasons, the final rule should provide that a facility has three years from the date of issuance of that rule to comply with the GACT standards. For all these reasons, the commenter believed that a three-year compliance deadline is appropriate, and that the proposed section 63.11560(a) should be amended by substituting the term “three years” where “one year” is currently found in the bracketed language. Response. We disagree with both the commenter's basic premise that existing sources will need three years to comply with the final standards and the assumptions underlying that premise. The commenter assumes that either new control devices will need to be installed, or existing controls upgraded, to comply with the PAH or PM emission limits. We believe that this assumption is incorrect. In this final rule, we revised the emission limits based on our assessment of additional data and to account for variability. As a result, we believe that no new add-on controls will be needed to comply with the final GACT standards. Consequently, we believe that the proposed compliance deadline of one year is adequate. If an owner or operator believes that additional time beyond the one year compliance period is needed to install controls, the owner or operator can request a compliance extension from the Administrator (or a State with an approved title V permit program), as authorized by CAA section 112(i)(3)(B) and specified in section 63.6(i)(4)(i)of the NESHAP General Provisions. Comment. One commenter noted that the deadline for conducting performance tests for existing sources stated in the proposal preamble was incorrect because it said that the performance test must be conducted within 180 days after publication of the final rule in the Federal Register , rather than 180 days after the compliance date as specified in the regulatory text. The commenter said that the preamble to the final rule should clarify that the preamble to the proposal was in error because the rule language specifies that existing facilities must demonstrate initial compliance within 180 calendar days after the compliance date. The commenter also noted that EPA uses multiple terms for the same requirement ( i.e., “performance testing,” “compliance testing”). The commenter asserted that the use of multiple terms for the same requirement can cause confusion when interpreting the regulatory requirements. The commenter recommended that EPA refer to this testing as “compliance testing” throughout the final GACT rule. Response. We agree with the commenter and have corrected the inconsistencies in the final rule. Comment. One commenter stated that either one or both of the asphalt density calculations have been improperly derived. The commenter said that either the calculations in English units or in metric units are inaccurate; as they do not give the same answer after the unit conversions are made. The commenter requested that EPA revise these equations as appropriate. Response. We agree with the commenter and we have corrected the English-unit values for the constants K 1 and K 2 in the asphalt density equations of the final rule. Comment. One commenter believed that the requirement in the proposed rule (section 63.11562(h)(1)) to conduct the compliance tests under conditions that represent normal operation and not during periods of startup, shutdown, or malfunction is overly broad. The commenter stated that there can be a significant range of “normal operation,” and the requirement as stated can lead to confusion among regulators and the regulated community. The commenter added that some asphalt roofing manufacturing facilities would find it impossible to meet the proposed requirement to manufacture a certain product during compliance testing because they do not manufacture such products. The commenter noted that the proposal also differs from the approach taken in the subpart LLLLL MACT rule. The commenter suggested that the final rule require that the test be performed while manufacturing the roofing product that is expected to result in the greatest amount of HAP emissions. Response. We agree with the commenter's suggestion that compliance tests be performed while manufacturing the roofing product that is expected to result in the greatest amount of PAH emissions. As a result, the final rule specifies that initial and subsequent compliance tests must be conducted while manufacturing the product that has the highest PAH and PM emissions. We have also eliminated the requirement that compliance tests be conducted under conditions that represent normal operation and not during periods of startup, shutdown or malfunction. We believe that this change addresses both aspects of the comment. Requiring that the compliance test be conducted while manufacturing the product that has the highest PAH and PM emissions eliminates the need to specifically reference normal operating conditions. We are appropriately requiring compliance testing during those periods when the facility is manufacturing the product that has the highest PAH and PM emissions. Comment. One commenter stated that it would be helpful if EPA explained how the production rate is determined. The commenter questioned if the production rate was based on actual daily production, monthly production, the daily average of monthly production or some other calculation. The commenter also questioned how the production rate would be determined in plants that run continuously, so that production spans more than one calendar day. Response. The production rate to be used in determining compliance with the asphalt roofing manufacturing emission limits is the production rate at which the roofing line was operating during the compliance test. If a facility is demonstrating initial compliance with the emission limits using the average of three 1-hour emission tests, the production rate used for the compliance demonstration would be the average rate over the 3-hour period (in terms of pounds of product manufactured). The final rule clarifies that the production rate used for determining compliance must be the average production rate utilized during the compliance test. Comment. One commenter supported EPA's decision to set the PM standards based upon filterable PM emissions, as is clear from the choice of Method 5A to measure PM emissions. The commenter noted that the data upon which the standards were based were of filterable PM emissions, so it would be inappropriate to include condensable particulate for compliance purposes. The commenter asserted that doing so would be inconsistent with the basis of the standards. The commenter believed that the preamble to the final rule should make it clear that in measuring PM emissions, the rule contemplates only filterable PM (the “front half”), and that it would be inappropriate to also require measurement of condensable PM (the “back half”). The commenter also recommended adding a definition for PM to section 63.11566. The commenter said that the definition should state that “Particulate matter (PM) means the filterable particulate matter as measured using the front half of Method 5A.” Should States require that the front half and back half meet these stringent standards, this would result in a regulation far stricter than that mandated by the CAA. The commenter stated that facilities might be required to install thermal oxidizers to comply, a decision that would result in increased emissions of greenhouse gases to reduce already low emissions of PAH. Response. The data upon which the alternative PM emission limits are based were collected using EPA Method 5A of Appendix A of 40 CFR 60 (Determination of Particulate Matter Emissions from the Asphalt Processing and Asphalt Roofing Industry). Using Method 5A, PM in vent gas samples taken from the source is collected on a glass fiber filter maintained at a temperature of 42 ± 10 °C (108 ± 18 °F). The PM mass, which includes any material that condenses at or above the filtration temperature, is determined gravimetrically after the removal of uncombined water. Consequently, we agree with the commenter that it would be inappropriate to establish emission limits that include contributions from PM that is captured in the sampling train downstream of the Method 5A filter since we do not have data that reflect those contributions. Therefore, for purposes of this final rule, we are defining PM to include any material determined gravimetrically using EPA Method 5A—Determination of Particulate Matter Emissions From the Asphalt Processing And Asphalt Roofing Industry (40 CFR 60, Appendix A). Comment. One commenter noted that the proposal allows the use of the results of performance testing conducted during the past five years to show compliance and indicates that a source must be able to demonstrate that “the results of the performance test, with or without adjustments, reliably demonstrate compliance despite any process changes.” The commenter requested further explanation of this provision, because it is likely that most process adjustments would trigger a re-test. Another commenter stated that the rule should specify that only emission increases resulting from a process change that is above a de minimis level would prevent a previous test from being used. Response. We clarified the final rule preamble by removing the term “with or without adjustment” because that language was unclear. While we agree that there are many types of process changes that could increase PAH and PM emissions such that the previously-conducted test would not be valid, we believe that some changes would not invalidate the results of the previously-conducted test. We included the option to use existing tests to provide flexibility to the affected facilities. We intend that it is the responsibility of the owner or operator to demonstrate that the process adjustment or change did not invalidate the results of the previously-conducted test. Consequently, we are not including de minimis emissions levels in the final rule. Comment. One commenter noted that some facilities have conducted required PM compliance testing under various state-managed air permit programs. The commenter said that, in some cases, the methodologies used in these tests are somewhat different than Method 5A. However, the commenter noted that in all cases the methods are approved by a State agency prior to use and typically are carefully evaluated by state experts. The commenter asserted that preventing a facility from using a legitimate, accepted test previously used to establish compliance will result in unnecessary costs and potential conflicts with existing, state-issued, air permit terms and conditions. The commenter asserted that in this scenario requiring the prior test to conform exactly to Method 5A does not provide any additional benefit to the environment, and it merely adds cost, uncertainty and confusion. Response. We disagree with the commenter that the final rule should provide a blanket allowance for the use of state-approved test methods in lieu of EPA Method 5A. The final rule, through reference to the NESHAP General Provisions, allows owners or operators to petition the Administrator to use alternative test methods and procedures. The EPA retains the authority to approve alternative test methods based on site-specific information. This mechanism can be used to obtain approval to use the results of a previously conducted test, as well as to obtain approval to use an alternative test method in the future. Comment. One commenter supported EPA's decision to allow facilities to use “process knowledge and engineering calculations” in lieu of a performance test to demonstrate initial compliance at a roofing line that does not include a saturator. The commenter noted that companies often have the necessary information and data to show that they will be in compliance with the emission standards if they operate their plants in such a way as to meet specified parameters. However, the commenter questioned why the option was limited to roofing lines that do not include a saturator. The commenter noted that the proposal offers no explanation for this limitation. The commenter asserted that the same principles apply to roofing lines with saturators and asphalt processing operations. Response. In the proposal, we limited the option to use process knowledge and engineering calculations because we believed that a coater-only line was the only equipment configuration that could potentially demonstrate compliance without using an add-on control device. However, we agree with the commenter that the technical basis for allowing the option does not preclude application of the option to lines containing saturators. Therefore, the final rule does not limit to coater-only lines the use of process knowledge and engineering calculations, in lieu of an emissions test, to demonstrate initial compliance. However, we are clarifying that the option is applicable only to roofing lines that do not need a control device to comply with the GACT limits. D. Continuous Compliance Requirements Comment. Two commenters stated that the pressure drop monitoring requirement for control devices in the final rule should specify that the pressure drop must be maintained in the range established during the initial compliance test, rather than below a maximum limit. The commenters noted that if the filter develops a tear or it is removed after the initial test, the pressure drop would decrease. In this scenario, the commenters said that the filter removal or tear would not cause a violation of the operating limit but the air pollution control device would not be operating properly. A third commenter noted that filters become more efficient and remove more particulates as their differential pressure increases. Another commenter stated that as long as the ability of the blower to move air is not impeded ( i.e., as long as the operating limit of the technology is not exceeded), increased pressure drop actually improves PM removal efficiency. The commenter said that the key to PM filtration technology is not the pressure drop but the velocity of air moving through the capture and control system. The commenter said that pressure drop is actually a surrogate for air flow measurement. The commenter stated that the design maximum pressure drop is based on the ability of the blower providing air flow for capture of the emissions at the source (the air flow captures the PM emissions and transports the PM to the filtration). The commenter noted that the proposed approach of maintaining the pressure drop below a maximum level is contrary to the way filtration-based PM control technology used in asphalt roofing lines works. Response. We agree with the commenters that requiring that the pressure drop be maintained within a predetermined range and monitored to ensure that this is the case is a better indicator of control system performance than requiring the pressure drop be maintained below a maximum level. The final rule, therefore, specifies that the pressure drop and temperature must be maintained within the range established by the initial compliance assessment. Comment. One commenter recommended that the pressure drop temperature compliance parameters be based upon the specifications of the manufacturer of the filtration technology. The commenter said that many years of Method 5A compliance testing has demonstrated that as long as the inlet emissions stream does not exceed the manufacturer's temperature and pressure drop limits, and the control technology is operated as specified by the manufacturer, the technology will remove the PM from the stream as guaranteed. The commenter stated that many States have recognized the validity of this approach to deliver compliance with PM emissions limits by requiring that, in both construction and operating permits, emissions sources operate control technologies as per manufacturing requirements. The commenter said that language in the permit either incorporates or references the manufacturer's written operating requirements as compliance parameters. The commenter stated that limiting the allowable pressure drop to levels below manufacturer's guaranteed performance limits will force facilities to replace and dispose of expensive filtration media well before the end of its guaranteed performance which would result in the increased generation and disposal of solid wastes, with no net increase in reduction of PM and PAH emissions. Also, the commenter said that if the compliance test did not occur late in the expected life of the filter media, the pressure drop measured will be low because the pressure drop is lower for new filtration media than for old filtration media. The commenter added that the inlet temperature to the filtration technology is dominated by ambient conditions ( e.g., when outside temperatures are high, the inlet temperatures of emissions stream to the filtration technology will be high). Thus, the commenter said that if a facility cannot time the compliance test to occur during the hottest time of the year, the source will surely experience higher inlet temperatures during high temperature time periods. The commenter stated that member companies have already experienced this problem in operating under the subpart LLLLL asphalt processing and asphalt roofing manufacturing MACT. The commenter noted that facilities in the industry have received notices of violations for inlet temperatures that exceeded those measured during the performance test, then re-tested at the elevated temperature. The commenter said that these re-tests showed that they still did not exceed the MACT PM emission limits. The commenter also provided a graphical figure that shows a consistent correlation between temperature and emissions does not exist. The commenter recommended that facilities be allowed two options for establishing and monitoring pressure drop and temperature in the final rule. Under the commenter's first option, the parameters would be based upon manufacturer's specifications. The source would conduct an initial compliance test. The PM emissions from the control device would need to be shown to be below the final GACT limits. As long as the pressure drop was below the manufacturer's requirements, the source would be considered to be in compliance with the pressure drop compliance parameter. Under the commenter's second option, the parameter values would be established as under the proposal, but a measurement that did not exceed that value by a certain percent would not be considered to be a deviation (the commenter suggested 30 percent for pressure drop and 10 percent for temperature). The commenter stated that EPA has allowed a similar buffer over parameters measured during the performance test in existing MACT standards, including Subpart N for Chromium Electroplating, at section 63.343, allowing a buffer on differential pressure, and Subpart NNN for Wool Fiberglass, at section 63.1382, allowing production rate to exceed 20 percent above the tested rate for up to 10 percent of the operating time in a semiannual period. Another commenter, a control device equipment vendor, asserted that filters will perform adequately when operated within the design and pressure limits imposed by the manufacturer. The commenter added that filtration equipment will operate adequately at temperatures within the limits specified by the equipment manufacturer. Response. We agree with the commenters that equipment manufacturer specifications for filter media performance are appropriate for use in establishing monitoring parameter ranges, particularly considering the difficulty in conducting emission tests that capture the performance of the control device at the high and low end of its operating range. Consequently, we are adopting the commenter's first option in that the final rule allows owners or operators to use equipment manufacturer performance specifications for filter media in establishing monitoring parameters. Comment. One commenter was very concerned about the way the proposal would have facilities set their compliance parameter limits for pressure drop and temperature through an initial compliance test. The commenter believed that EPA's proposed approaches lack a technical basis and would result in numerous potential violations of the operating limits even when PM and PAH emissions are well below the emission standards. The commenter suggested alternative methodologies that are more appropriate for establishing parameter limits. The commenter noted that the proposal would treat all “deviations” from the operating parameter limits ( i.e., all exceedances of parameter limits) as potential violations of the emission standards. The commenter thought that this approach was excessively harsh, particularly because several factors make it almost certain that established operating parameter limits will be exceeded at times even when a facility is not exceeding the GACT emission standards, and is operating its processes and control equipment well. For example, the commenter stated that an exceedance of a temperature parameter limit does not mean that a facility is exceeding the emission standard; the ambient temperature has a significant effect on the temperature monitored and the amount of emissions is actually controlled by the temperature of the asphalt in the coating mixer, coater, and/or saturator. For that reason, the commenter noted that the preamble to the Subpart UU (NSPS for asphalt processing and asphalt roofing manufacturing) states that “periods of temperature excursions * * * would not, of themselves, constitute a violation of the numerical emission limits. The commenter noted that even if the temperature is measured at the coater or saturator, an exceedance of the temperature parameter limit does not mean that the source is exceeding the standards. The commenter asserts that the same is true for deviations from a set pressure drop parameter limit. As discussed above, it would not be at all surprising for a roofing line to exceed its pressure drop limit but still emit fewer PM or PAH emissions than the actual emission standard allows. Consequently, the commenter stated that EPA should follow an approach similar in some ways to one that EPA established in its subpart NNN fiberglass MACT standards. The subpart NNN wool fiberglass standards consider whether an affected source is operating outside of its parameter limits for more than 5 percent of the time during a 6-month block reporting period. The commenter believes that EPA should borrow from this approach, and require that the facility conduct a new compliance test if a roofing line has operated outside of the established parametric limits, as we have proposed them, for more than 5 percent of the time in any semiannual reporting period. The commenter said that this would essentially be a combination of the approaches taken by the wool fiberglass MACT standards and the subpart UU NSPS for asphalt roofing manufacturing. If the re-test shows the line to be emitting more PAH or PM than the standard allows, commenter said that the facility could be judged to be in violation of the GACT standard. If the re-test shows that emissions do not exceed the standard, commenter said that there would be no violation. Response. We acknowledge the difficulty in establishing appropriate monitoring parameter ranges for filtration-based PM control devices. As noted in earlier responses to comments above, the final rule allows owners or operators to establish a range of parameter values for monitoring using manufacturer performance specifications. The EPA believes that allowing the use of manufacturer specifications provides owners or operators sufficient flexibility in establishing appropriate parameter ranges. Consequently, we are not including a re-test provision in the final rule. The parameter ranges established by the facility and approved by the delegated authority are not-to-exceed values. A parameter exceedance would be a violation of the monitoring requirements but not necessarily a violation of the emission limits. Additionally, we are not including the re-test provision because we do not believe it is possible in all cases to replicate the conditions that caused the exceedance during a re-test. Comment. One commenter noted that some of the ESP units currently in operation in the industry are not provided with voltage meters, nor are they easily modified to add meters for the voltage reading. The commenter said that such ESPs are typically provided with a green indicating light. The commenter said that this light is used to assess the operation of the unit and determine when cleaning is needed. The commenter added that the light burns a solid green during normal operation and the light flashes as the cells gradually become dirty; the dirty cells are then replaced with clean spares. The commenter stated that contractors have been contacted to provide proposals to modify the existing units to add the required voltage indicators. The commenter said that current estimates are around $50,000 to modify the exiting units to add voltage meters and another $25,000 to $50,000 to add controls to automatically provide the 3-hour average voltage (cost varies depending upon the current automation capability of a facility). The commenter said that the high cost of these modifications is not reasonable, given that the use of the indicating light ensures that the ESP will operate properly. The commenter therefore believed that routine monitoring and logging of the ESP monitoring light is the only reasonable method to verify the operation of an ESP that does not have voltage meters and that EPA should allow this method of compliance. Response. We agree with the commenter that requiring retrofits for voltage monitors is not cost efficient. We also believe that monitoring the ESP instrumentation ( e.g., indicator light) provides sufficient monitoring of the ESP performance. Therefore, the final rule allows owners or operators to monitor the ESP instrumentation as an option to monitoring voltage. Additionally, the final rule specifies that failure to service the ESP within one hour of the potential problem is an exceedance of the monitoring standards, which is consistent with previously promulgated area source rules ( e.g., area source NESHAP for iron and steel foundries, and area source NESHAP for aluminum, copper and other nonferrous foundries). Comment. One commenter stated that CEMs are not suitable for asphalt fumes for continuous sampling of PM. The commenter noted that EPA Method 5A is used for stack PM sampling of asphalt fumes and Method 5A requires that the emission stream be cooled to allow the fume aerosols to condense and this PM portion is then recovered from the sample train with an after test solvent wash. The commenter stated that a continuous analyzer does not exist that will perform this PM sampling. Response. We agree with the commenter and the CEMS option has been removed from the final rule. Comment. One commenter supported the proposed provision that, for periods of startup and shutdown, would allow owners and operators to demonstrate compliance with the emission standard over a 24-hour averaging period. The commenter advocated, however, that EPA adopt a similar 24 hour averaging approach for determining compliance with the temperature requirements of the rule. Another commenter expressed concerns with the proposed provision that, for periods of startup and shutdown, allows owners and operators to demonstrate compliance with the emission standard over a 24-hour averaging period. Specifically, the commenter expressed concern regarding the public health impacts of excess emissions during SSM episodes. Response. We appreciate the one commenter's support of the provision that, for periods of startup and shutdown, allows owners and operators to demonstrate compliance with the emission standard over a 24-hour averaging period. However, we reject the commenter's suggestion that the 24-hour averaging period be extended to temperature. As stated elsewhere in this preamble, we have modified the rule to require that the owner/operator establish a temperature range for the inlet gas temperature to the PM control device during the initial compliance assessment and to then maintain the 3-hour average inlet gas temperature within that range during operations. We believe that these changes, which allow the owner/operator to establish a temperature range, obviate any need for a longer averaging time for temperature. We proposed the use of a 24-hour averaging period for determining compliance with the emission standards to account for emissions generated during periods of startup and shutdown based on the format we chose for the emission standards, i.e., lbs of emissions per ton of product produced. During periods of startup and shutdown, the process will continue to produce emissions. Even though emissions during such periods will be less than those that occur during normal operations when measured on an hourly basis, i.e., pounds of emissions per hour of operation, production during such periods will be very limited. As a result, it will be very difficult, if not impossible, to demonstrate compliance with a standard stated in terms of pounds of emissions per ton of product produced if a 3-hour averaging period is used. Specifically, emissions generated during periods of startup and shutdown will be less on an hourly basis than those generated during normal operations for a number of reasons. First, during periods of startup, the temperature of the asphalt is raised until it reaches the optimal temperature for use when producing product. Similarly, during periods of shutdown, the temperature of the asphalt is being reduced from the temperature which is optimal for production. As the temperature of the asphalt increases, the rate of volatilization also increases, resulting in increased PAH emissions as measured on a pounds per hour basis. As a result, during startup, PAH emissions, as measured on a pounds per hour basis, increase until the temperature of the asphalt reaches the optimal temperature for production after which the temperature is maintained at a steady state. During shutdown, the reverse process occurs, i.e., as the process is shut down, the asphalt cools, the rate of volatilization decreases and hourly PAH emissions decrease. Second, during startup and shutdown, the asphalt usage rate, and hence the hourly PAH emission rate, fluctuates. During startup, the asphalt usage rate gradually increases until it reaches the rate present during normal production. As a result, during startup, the hourly PAH emission rate gradually increases until it reaches the rate that exists during periods of normal production. During shutdown, the reverse occurs, i.e., the hourly asphalt usage rate gradually decreases from the rate present during normal production. Thus, except for the very start of the shut-down period, the hourly PAH emission rate is lower than during periods of normal production. The rate of production, i.e., the amount of product produced on an hourly basis, also fluctuates during periods of startup and shutdown. At the commencement of startup, no product is being produced as the asphalt is being brought up to the proper temperature for normal production. The rate of production then gradually increases until the process reaches, and is maintained at, the rate of normal production. During shutdown, the rate of production is gradually reduced from its normal rate to zero. Thus, in light of the production-based format of the standard and the emission characteristics described above that occur during startup and shutdown at asphalt processing and asphalt roofing manufacturing facilities, we concluded that it was appropriate to provide a longer averaging period for determining compliance during periods of startup and shutdown. We chose a 24-hour averaging period because, based on the exercise of our best engineering judgment, we determined that this was an appropriate period since the record indicates that the startup and shutdown processes can take up to 9 hours to complete. We also considered establishing a 16-hour averaging period as this represents two normal 8-hour shifts, but concluded that this would not provide adequate time for conditions to normalize. The final rule, therefore, allows sources to determine compliance with the emission standard based on a 24-hour averaging period, as opposed to a 3 hour period. We acknowledge the one comment regarding the health concerns associated with emissions that are generated during start-up and shut-down events; however, the GACT standards are technology-based standards as opposed to health- or risk-based standards. For the reasons described above, we think a 24-hour averaging period during periods of startup and shutdown is reasonable and the commenter has provided no evidence to the contrary. In the proposed rule, we proposed to also apply the 24-hour period for measuring compliance to malfunction events. We are not adopting this approach in the final rule. Rather, the final rule requires compliance with the standard based on a 3-hour average at all times, except as explained above, for periods of startup and shutdown, in which case the rule provides that owners and operators demonstrate compliance with the standard over a 24-hour averaging period. In re-examining the record for this rulemaking, we recognized that the data in the record supporting a longer averaging period related solely to startup and shutdown events. Moreover, in contrast to startup and shutdown events which are routine and distinct operating modes, a malfunction is defined as a “sudden, infrequent, and not reasonably preventable failure of air pollution control and monitoring equipment, process equipment or a process to operate in a normal or usual manner * * *” 40 CFR 63.2. As discussed above, EPA has properly accounted for different periods of operation, including periods of startup and shutdown, in establishing the standards in this rule. Since a malfunction is not a distinct operating mode, malfunction emissions do not need to be factored into the development of CAA section 112(d) standards, which, once promulgated, apply at all times. Sierra Club v. EPA, 551 F.3d 1019 (DC Cir. 2008). Thus, the final rule does not establish a different averaging period for use in measuring compliance during malfunction events. Further, even if malfunctions were considered a distinct operating mode, we believe it would be impracticable to take into account malfunctions in setting CAA section 112(d) standards. Because, by definition, malfunctions are sudden and unexpected events, it would be difficult to set a standard that would account for the myriad of different emissions that could occur during malfunctions. In addition, the type, frequency, and duration of the malfunctions may differ significantly between sources. Finally, setting an emissions standard that accounts for all different potential types of malfunctions would allow a source to emit excessive quantities of uncontrolled pollution and would not provide an incentive for sources to minimize the occurrence of malfunctions. E. Title V Permitting Comment. One commenter argued that the Agency's proposal to exempt the asphalt processing and asphalt roofing manufacturing area source category from title V requirements is unlawful and arbitrary. The commenter stated that section 502(a) of the CAA authorizes EPA to exempt area source categories from title V permitting requirements if the Administrator finds that compliance with such requirements is “impracticable, infeasible or unnecessarily burdensome.” 42 U.S.C. section 7661a(a). The commenter noted that EPA did not claim that title V requirements are impracticable or infeasible for any of the source categories it proposes to exempt, but that EPA instead relied entirely on its claim that title V would be “unnecessarily burdensome.” Response. Section 502(a) of the CAA states, in relevant part, that: * * * [t]he Administrator may, in the Administrator's discretion and consistent with the applicable provisions of this chapter, promulgate regulations to exempt one or more source categories (in whole or in part) from the requirements of this subsection if the Administrator finds that compliance with such requirements is impracticable, infeasible, or unnecessarily burdensome on such categories, except that the Administrator may not exempt any major source from such regulations. See 42 U.S.C. section 7661a(a). The statute plainly vests the Administrator with discretion to determine when it is appropriate to exempt non-major ( i.e., area) sources of air pollution from the requirements of title V. The commenter correctly noted that EPA based the proposed exemptions solely on a determination that title V is “unnecessarily burdensome,” and did not rely on whether the requirements of title V are “impracticable” or “infeasible”, which are alternative bases for exempting area [Text truncated at 120,000 characters. The full text is on the page linked above.] --- Source: Frix Law Library, https://www.frixlaw.com/law-library/documents/fr%3AE9-27946. Public record. Not legal advice.