# National Emission Standards for Hazardous Air Pollutants: Gold Mine Ore Processing and Production Area Source Category; and Addition to Source Category List for Standards > Briefs, arguments, decisions, and more. URL: https://www.frixlaw.com/law-library/documents/fr%3A2011-2608 ## Record - **Collection:** Federal Register - **Document type:** Rule - **Published:** February 17, 2011 - **Citation:** 76 FR 9450 ## Text ENVIRONMENTAL PROTECTION AGENCY 40 CFR Parts 9 and 63 [EPA-HQ-OAR-2010-0239; FRL-9242-3] RIN 2060-AP48 National Emission Standards for Hazardous Air Pollutants: Gold Mine Ore Processing and Production Area Source Category; and Addition to Source Category List for Standards AGENCY: Environmental Protection Agency (EPA). ACTION: Final rule. SUMMARY: EPA is adding the gold mine ore processing and production area source category to the list of source categories to be regulated under Section 112(c)(6) of the Clean Air Act due to its mercury emissions. EPA is also promulgating national emission standards for hazardous air pollutants to regulate mercury emissions from this source category. DATES: This final rule is effective on February 17, 2011. The incorporation by reference of certain publications listed in the final rule is approved by the Director of the Federal Register as of February 17, 2011. ADDRESSES: EPA has established a docket for this action under Docket ID No. EPA-HQ-OAR-2010-0239. All documents in the docket are listed on the http://www.regulations.gov Web site. Although listed in the index, some information is not publicly available, e.g., Confidential Business Information (CBI) 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 www.regulations.gov or in hard copy at the EPA Headquarters Library, Room Number 3334, EPA West Building, 1301 Constitution Ave., NW., Washington, DC. The EPA/DC Public Reading Room hours of operation are 8:30 a.m. to 4:30 p.m. Eastern Standard Time (EST), Monday through Friday. The telephone number for the Public Reading Room is (202) 566-1744, and the telephone number for the Air and Radiation Docket and Information Center is (202) 566-1742. FOR FURTHER INFORMATION CONTACT: Mr. Chuck French, Sector Policies and Program Division, Office of Air Quality Planning and Standards (D243-02), Environmental Protection Agency, Research Triangle Park, North Carolina 27711, telephone number (919) 541-7912; fax number (919) 541-3207, e-mail address: french.chuck@epa.gov. SUPPLEMENTARY INFORMATION: The information presented 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. Addition to Section 112(c)(6) Source Category List III. What is the statutory authority and regulatory approach for the proposed standards? IV. Summary of Significant Changes Since Proposal A. Applicability B. Final Emission Standards C. Compliance Dates D. Compliance Requirements E. Monitoring Requirements F. Definitions V. Summary of Responses to Major Comments A. Statutory Requirements B. Applicability C. MACT Floors D. Compliance Determinations E. Monitoring Requirements F. Definitions VI. Summary of Environmental, Economic and Health Benefits 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 this final rule include: Category NAICS code 1 Examples of regulated entities Industry: Gold Ore Mining 212221 Establishments primarily engaged in developing the mine site, mining, and/or beneficiating ( i.e., preparing) ores valued chiefly for their gold content. Establishments primarily engaged in transformation of the gold into bullion or dore bar in combination with mining activities are included in this industry. 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 63.11640 of subpart EEEEEEE (National Emission Standards for Hazardous Air Pollutants (NESHAP): Gold Mine Ore Processing and Production Area Source Category). 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 EPA 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 proposed 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 U.S. Court of Appeals for the District of Columbia Circuit by April 18, 2011. Under section 307(d)(7)(B) of the CAA, only an objection to this final rule that was raised with reasonable specificity during the period for public comment can be raised during judicial review. Moreover, 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) also provides a mechanism for us 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 the person listed in the preceding FOR FURTHER INFORMATION CONTACT section. II. Addition to Section 112(c)(6) Source Category List For reasons stated in the preamble to the proposed rule (75 FR 22470, April 28, 2010), we are adding the gold mine ore processing and production area source category to the list of source categories under section 112(c)(6) on the basis of its mercury emissions. The preamble for the proposed rule provides a description of this industry including the processes used and the typical control technologies applied. III. What is the statutory authority and regulatory approach for the proposed standards? As explained in the preamble to the proposed rule, CAA section 112(c)(6) requires that EPA set standards under section 112(d)(2) or (d)(4). The mercury standards for the gold mine ore processing and production area source category are being established under CAA section 112(d)(2), which requires maximum available control technology (MACT) level of control. Under CAA section 112(d), the MACT standards for existing sources must be at least as stringent as the average emissions limitation achieved by the best performing 12 percent of existing sources (for which the administrator has emissions information) for source categories and subcategories with 30 or more sources, or the best performing 5 sources for categories and subcategories with fewer than 30 sources (CAA section 112(d)(3)(A) and (B)). This level of minimum stringency is called the MACT floor. For new sources, MACT standards must be at least as stringent as the emission control that is achieved in practice by the best controlled similar source (CAA section 112(d)(3)). EPA also must consider more stringent “beyond-the-floor” control options. When considering beyond-the-floor options, EPA must consider not only the maximum degree of reduction in emissions of HAP, but must take into account costs, energy, and nonair quality health and environmental impacts when doing so. IV. Summary of Significant Changes Since Proposal This section summarizes the significant changes to the rule since proposal. Additional information on the basis for these changes and other changes can be found in the Summary of Responses to Major Comments in section V of this preamble and in the Summary of Comments and Responses document which is available in the docket for this action. A. Applicability We have clarified in § 63.11651 of the final rule that the term “gold mine ore processing and production facility” does not include individual prospectors and very small pilot scale mining operations. These types of operations are very small and were not included in the section 112(c)(6) inventory that was the basis for the listing of the gold mine ore processing and production source category. B. Final Emission Standards We have made changes to all of the proposed emission standards as the result of the following developments: (1) Inclusion of additional emissions test data received since proposal; 1 (2) additional analyses in response to public comments on the proposed rule; 2 and (3) further review of the data used to develop the standards for the proposed rule. The changes are summarized below and described in more detail in section V of this preamble. We estimate the final MACT standards will reduce mercury emissions from gold mine ore processing and production down to a level of about 1,180 pounds per year, which will be an estimated 77 percent reduction from the 2007 emissions level (5,000 lb/yr), a 95 percent reduction from year 2001 emissions level (about 23,000 lb/yr), and more than 97 percent reduction from uncontrolled emissions levels (more than 37,000 lb/yr). 1 The new test data used in final MACT standard calculations can be found in the docket as docket items: EPA-HQ-OAR-2010-0239-0359 and EPA-HQ-OAR-2010-0239-0360. 2 Analyses for the final MACT standards can be found in the docket in the document titled: “Development of the MACT Floors and MACT for the Final NESHAP for Gold Mine Ore Processing and Production” (also known as the “MACT Development Document”). Ore Pretreatment Processes In the proposed rule, the proposed mercury emission standards for both existing and new ore pretreatment processes were 149 pounds per million tons of ore processed (lb/million tons of ore). In the final rule, the emission standard for existing sources is 127 lb/million tons of ore; and for new sources the emission standard is 84 lb/million tons of ore. The final emission standards are based on several changes to the data set used in the MACT analysis. Since we issued the proposed rule, we collected emissions data from more recent tests that were not available at proposal. Further, we learned that two emissions tests that we used to develop the MACT floor in our proposed rule had been invalidated by the Nevada Division of Environmental Protection (NDEP), and we removed those test results from the database. Information on the specific tests invalidated and the rationale are available in the docket (docket item number EPA-HQ-OAR-2010-0239-0061). We also discovered that the test data for a unit within the ore pretreatment affected source at a facility should have been included as part of a different unit at the same facility. We have also dropped the data for one facility from the analysis because their autoclave was shut down in 2007 and dismantled, and the only test data we had for them was one test of the autoclave when it was operating in 2006. Moreover, we conducted additional beyond-the-floor analyses for the ore pretreatment affected source. The new information and analyses described above are discussed in more detail in section V.C of this preamble and in the MACT Development Document which is available in the docket for this rulemaking. The resulting data set included emissions data for four facilities that ranged from 45 to 165 lb/million tons of ore. Based on these data, and using the same upper prediction limit (UPL) approach used for proposal to account for variability, we determined the MACT floor to be 158 lb/million tons of ore for existing sources of ore pretreatment processes and 84 lb/ million tons of ore for new sources. As explained in the proposed rule (75 FR at 22482), the technologies that we estimate are needed to achieve the MACT floor level of performance for existing ore pretreatment processes include calomel-based mercury scrubbers on roasters and venturi scrubbers on autoclaves and ancillary roaster operations. The preamble to the proposed rule provides a description of the UPL and the approach and calculations used to derive the UPL. The UPL is also discussed further in section V. In our beyond-the-floor analysis, we evaluated the potential to add condensers and carbon adsorbers to control autoclaves, and the potential to add carbon adsorbers to control the ore pre-heaters. Based on this beyond-the-floor analysis, we concluded that it is feasible and cost-effective to establish the MACT standard for existing sources at a level lower than the MACT floor. Based on the analysis, we determined the MACT standard for existing sources to be 127 lb/million tons of ore. For new sources, we determined that it was not feasible and cost-effective to establish a standard lower than the new source MACT Floor (of 84 lb/million tons); therefore the MACT standard for new sources was determined to be 84 lb/million tons. The technologies needed to achieve the new source MACT floor will depend on the types of ore processed, amount of mercury in the ore, and specific process units used. Nevertheless, we conclude that, at a minimum, the controls that would be needed would include calomel-based mercury scrubbers on roasters and venturi scrubbers on autoclaves and ancillary roaster operations. Additional controls that will likely be needed to achieve emissions at or below the new source MACT floor level include condensers and carbon adsorbers on autoclaves, and carbon adsorbers on ore preheaters. Table 1 summarizes the MACT floor analysis for existing and new ore pretreatment processes. The beyond-the-floor analyses are explained further in section V of this preamble and in more detail in the MACT Development document. Table 1—MACT Floor Results for Ore Pretreatment Processes Facility Average performance (lb/million tons of ore) A 45 C 56 E 71 D 165 Average of the 4 facilities 84 99% UPL for existing sources ( i.e., the MACT Floor for existing sources) 158 99% UPL for new sources 1 ( i.e., the MACT Floor for new sources) 84 1 The MACT Floor for new sources is based on the average performance of Facility A (i.e., 45) plus an amount to account for variability (i.e., 45 + 39 = 84). Carbon Processes Under the proposed rule, all carbon processes were subject to the same proposed mercury emissions limits of 2.6 pounds per ton of concentrate (lb/ton of concentrate) for existing sources and, for new sources, either 0.14 lb/ton of concentrate or 97 percent reduction in uncontrolled mercury emissions. These limits would have applied to facilities that operate mercury retorts and facilities that do not operate mercury retorts. In the final rule, we distinguish between carbon processes with mercury retorts and carbon processes without mercury retorts because we believe there are unique differences in these two types of processes. Therefore, the final rule specifies separate emission standards for these two types of processes. Moreover, the final emission standards for carbon processes reflect inclusion of new test data that were not available at proposal. We also revised our data set based on new information that we received since proposal which impacted which sources were among the best performing sources. Based on the data that we have, there are 10 facilities that have carbon processes with mercury retorts, and we have mercury emissions data for all 10 of these facilities. There are approximately 7 facilities that have carbon processes without mercury retorts, and we have comprehensive and reliable mercury emissions data for 2 of these facilities. These 2 facilities are the best controlled facilities within that group based on the information we have. ( See section V for further details.) For carbon processes with mercury retorts, the emission standard in the final rule is 2.2 lb/ton of concentrate for existing sources and 0.8 lb/ton of concentrate for new sources. For carbon processes without mercury retorts, the emission standard in the final rule is 0.17 lb/ton of concentrate for existing sources and 0.14 lb/ton of concentrate for new sources. For carbon processes, regardless of whether the facility operates a mercury retort, we estimate that to meet the MACT floor facilities would generally need to have mercury condensers and carbon adsorbers to control mercury emissions. We also considered beyond-the-floor options for both existing and new sources for these process groups, which were based on the addition of a second carbon adsorber; however, we rejected those options because they are not cost effective. Additional information on the analyses performed can be found in the MACT Development document in the docket for this rulemaking. We also eliminated in the final rule the compliance alternative of 97 percent reduction for new carbon processes. After reviewing the comments received on this proposed alternative standard and giving further consideration to the practicality of how it would be measured, we concluded that this option would be difficult to implement, particularly when multiple processes that are operated at different times vent to a single control device and stack. In addition, we have limited data supporting this compliance alternative. In proposing this alternative for comment, we had hoped to, but did not, receive additional data indicating that the 97 percent reduction option would be equivalent to the proposed new source limit of 0.14 pounds of mercury per ton of concentrate. For the reasons stated above, we eliminated the 97 percent control efficiency option for new carbon processes in the final rule. Table 2 summarizes the results of the MACT floor analysis for carbon processes with mercury retorts, and Table 3 summarizes the analysis for carbon processes without mercury retorts. Table 2—MACT Floor Results for Carbon Processes With Mercury Retorts Facility Average performance (lb/ton of concentrate) N 0.53 J 0.74 I 1.06 A 1.47 H 1.67 D 2.20 C 3.71 G 8.17 E 14.49 B 20.60 Average of top 5 1.1 99% UPL for existing sources ( i.e., MACT Floor for existing sources) 2.2 99% UPL for new sources ( i.e., MACT Floor for new sources) 0.8 Table 3—MACT Floor Results for Carbon Processes Without Mercury Retorts Facility Average performance (lb/ton of concentrate) M 0.058 F 0.098 Average of top 2 facilities 0.078 99% UPL for existing sources ( i.e., MACT Floor for existing sources) 0.17 99% UPL for new sources ( i.e., MACT Floor for new sources) 0.14 Non-Carbon Concentrate Processes Under the proposed rule, the mercury emission standards for non-carbon concentrate processes were 0.25 lb/ton of concentrate for existing sources and 0.2 lb/ton of concentrate for new sources. In the final rule, the emission standards for these sources are 0.2 lb/ton of concentrate for existing sources and 0.1 lb/ton of concentrate for new sources. These standards are based on using new emissions data that were not available when we developed the proposal, along with the data that were used for the proposal. For non-carbon concentrate processes, we estimate that to meet the MACT floors, for both existing and new sources, facilities would generally need to control mercury emissions using mercury condensers and carbon adsorbers. As explained in the proposed rule, we considered beyond-the-floor controls for these processes (which were based on adding a second carbon adsorber to the MACT floor level controls) but concluded those controls would not be a cost-effective option. There are approximately 3 facilities in the U.S. that use these types of processes. We have emissions tests data for 2 of these facilities. Table 4 summarizes the results of the MACT floor analysis for non-carbon concentrate processes. Table 4—MACT Floor Results for Non-Carbon Concentrate Processes Facility Average performance (lb/ton of concentrate) K 0.047 L 0.078 Average of 2 facilities 0.062 99% UPL for existing sources ( i.e., MACT Floor for existing sources) 0.2 99% UPL for new sources ( i.e., MACT Floor for new sources) 0.1 C. Compliance Dates In the final rule, we provide in § 63.11641 that the compliance date for existing sources is 3 years after promulgation of the final rule as opposed to 2 years as proposed. We reviewed the information provided in public comments on the challenges of installing new controls, especially for autoclaves, which, although the controls have not yet been demonstrated, have been proposed by facilities with autoclaves in their Nevada Mercury Control Program (NMCP) permit applications. We also considered the installation of new controls on the roaster preheaters, which also have not yet been demonstrated, but have been proposed by these facilities in their NMCP permit applications. We concluded that allowing 3 years for existing sources to comply is appropriate, given the complexity of the sources, the combinations of control devices that are needed in many cases, and the amount of time necessary for designing, installing, testing, and commissioning additional emission controls for mercury. D. Compliance Requirements Section 63.11646(a)(1) of the final rule does not include Method 30A, as was proposed, as an appropriate method for determining mercury concentration because it is not yet in general use. This paragraph further clarifies that the use of ASTM D6784-02 and Method 30B are allowed for compliance tests only if approved by the permit authority as opposed to automatically being allowed as in the proposal. The final rule also does not include the requirement to follow the acetone rinse procedures and the absence of cyclonic flow determination requirement, which were in subparagraphs (v) and (vi) respectively of our proposed § 63.11646(a)(1). Method 29 already includes requirements for the acetone rinse, so there is no need to specify those procedures in the rule; and Method 1, which is required by the rule, addresses the issue of cyclonic flow. In § 63.11646(a)(2), we changed the minimum sample volume when Method 29 is used to determine compliance from the proposed 60 dry standard cubic feet (dscf) to 30 dscf. We believe this volume is adequate for detecting mercury in the samples and determining mercury emissions for this industry. We have also expanded this section to address non-detect values. If the emission testing results for any of the emission points yield a non-detect value, the final rule requires that the minimum detection limit (MDL) be used to calculate the mass of emissions (in pounds of mercury) for that emission point that would subsequently be used in the calculations to determine if the source is in compliance with the MACT standard. If the resulting calculations indicate that mercury emissions are greater than the MACT emission standard, the owner or operator may repeat the mercury emissions testing one additional time for any emission point for which the measured result was below the MDL using procedures that produce lower MDL results. If this additional testing is performed, the results from that testing must be used to determine compliance. For sources with multiple emission units ( e.g., two roasters) ducted to a common control device and stack, we have clarified in § 63.11646(a)(3) that compliance testing must either be performed with all affected emissions units in operation, if this is possible, or units must be tested separately. We also clarified that the establishment of operating limits for units that share a common stack can be based on emissions when all process units are operating together, or based on testing units separately. However, this requirement does not affect the frequency and schedule for monitoring, which are specified in the rule. If facilities have batch type processes that cannot be operated simultaneously, then the facility can test some or all of the units individually. In § 63.11646(a)(6) and (7), we clarify that the production data used in compliance determinations are based on full calendar months. For the initial compliance test, data for all the full calendar months between publication of the final rule and the initial compliance test must be used. This initial compliance determination must include at least one full month of production data ( e.g., hours of operation, and million tons of ore processed or tons of concentrate processed) including the month the test was conducted. For subsequent annual compliance tests, data for the 12 full calendar months prior to the annual compliance test must be used to demonstrate compliance. In addition, we clarify in paragraphs § 63.11646(a)(5), (6) and (7) that compliance determinations are based on the number of 1-hour periods each process unit operates. By using the 1-hour period terminology, the final rule language is consistent with the terminology used in the General Provisions to part 63. Because the final rule does not include the 97 percent reduction option that was in § 63.11645(e)(2) of the proposed rule, we have removed from the final rule the compliance requirement for that option that was in § 63.11646(b) of the proposed rule, which addressed testing the inlets and outlets for sources choosing that proposed option. E. Monitoring Requirements Section 63.11647(a) of the final rule includes an additional option for monitoring mercury emissions from roasters. The proposed rule specified two options for monitoring mercury emissions: Paragraph (a)(1) specified weekly sampling using PS 12B; and paragraph (a)(2) specified continuous monitoring using a mercury continuous emissions monitoring system (CEMS). In the final rule, we added paragraph (a)(3) to provide a third option of continuous sampling using PS 12B. In addition, paragraph (a)(1) in the final rule was changed to require sampling at least twice per month using either PS 12B or Method 30B rather than weekly. We believe that Method 30B is an acceptable alternative method for monitoring purposes and allows owners and operators more flexibility in how they monitor roaster emissions. We also believe that sampling twice per month coupled with extensive parametric monitoring of control devices (as explained below) is sufficient for the monitoring option in paragraph (a)(1). Section 63.11647(a)(4)(iii) of the proposed rule would have required additional compliance testing if the mercury concentration in the ore fed to the roaster was higher than any concentration measured in the previous 12 months. We have removed this requirement from the final rule because it is not clear that the mercury content of the ore has a significant effect on the performance of mercury scrubbers applied to roasters, which are designed to handle and operate efficiently for a range of mercury inlet concentrations. In addition, condensers are used to recover liquid elemental mercury prior to the mercury scrubber, and any increase in mercury loading would likely result in an increase in the recovery of elemental mercury. The final rule incorporates several changes to § 63.11647(b), which addresses monitoring of calomel-based mercury scrubbers ( i.e., mercury scrubbers) that are used to control emissions from roasters. The proposed rule required monitoring of the scrubber liquid flow, liquid chemistry, scrubber pressure drop, and scrubber inlet gas temperature hourly. The final rule does not include the requirement to monitor pressure drop across calomel-based scrubbers because we conclude that pressure drop is not related to mercury emission control performance by this type of control device. In addition, the final rule allows hourly monitoring of the line pressure in the scrubber liquid supply line as an alternative to hourly monitoring of scrubber liquid flow rate. Line pressure monitoring is already in practice at some facilities and provides the same type of information as does liquid flow rate. As was proposed, the final rule allows the operating limit for scrubber liquid flow rate (or line pressure) and inlet gas temperature to be based on the minimum flow rate (or line pressure) or maximum inlet gas temperature established during the initial performance test. It also includes two additional options for setting these operating limits: (1) Based on the manufacturer's specifications if certain types of systems are designed to operate within a specified range of flow rates or temperatures; and (2) based on limits established by the permitting authority. If the facility chooses the option to establish the limits during initial compliance, the final rule requires the scrubber flow rate operating limit to be based on either the lowest value for any run of the initial compliance test or 10 percent less than the average value measured during the compliance test and the inlet gas temperature operating limit to be based on either the highest value for any run of the initial compliance test or 10 percent higher than the average value measured during the compliance test. This requirement takes into account the fact that, although initially the system may exhibit little variability from test run to test run, the short-term variability in performance may increase with time. Additional discussion of these changes can be found in section V.E of this preamble and in the Summary of Public Comments and Responses document in the docket for this rulemaking. In response to comments, we have revised the requirements for corrective action following control device monitoring parameter exceedances specified in § 63.11647(d). Under the final rule, if the corrective actions taken following an exceedance do not result in the parameter value ( e.g., liquid flow rate, line pressure, or inlet gas temperature) being returned to within the parameter range or limit within 48 hours, a mercury concentration measurement must be made to determine if the operating limit for mercury concentration is being exceeded. The measurement must be performed and the concentration determined within 48 hours after the initial 48 hours, or a total of 96 hours from the time the parameter was exceeded. If the measured mercury concentration meets the operating limit for mercury concentration, the corrective actions are deemed successful. In addition, the owner or operator may request approval from the permitting authority to change the parameter range or limit based on measurements of the parameter at the time the mercury concentration measurement was made. If, on the other hand, the measured mercury concentration indicates the operating limit for mercury concentration is exceeded, the exceedance must be reported as a deviation within 24 hours to the permitting authority, and the facility must perform a compliance test (pursuant to § 63.11647(d)) within 40 days to determine whether the source is in compliance with the MACT standard. We believe 40 days is appropriate because it may take 3 to 4 weeks to schedule and have the testing contractor on site, and, following completion of the test, another week or so to receive the final test results, and allows sufficient time to notify the permitting authority. We also removed the requirement that roasters must be shut down if a parameter is out of range. In § 63.11647(a)(1)(ii) of the final rule, we require these same corrective actions described above ( i.e., measuring mercury concentration within 48 hours, reporting a deviation if the data show the operating limit was exceeded within 24 hours, and conducting a compliance test within 40 days) for exceedances of mercury concentration operating limits indicated by the results of the twice monthly monitoring using PS 12B or Method 30B, CEMS, or continuous monitoring using PS 12B. In such cases, the owner or operator must use the results of the compliance test to determine if the ore pretreatment process affected source is in compliance with the emission standard. If the source is determined to be in compliance, the owner or operator may use this compliance test to establish a new operating limit for mercury concentration for the roaster. We also removed the requirement that roasters must be shut down if the mercury concentration is out of range. In the final rule, § 63.11647(f)(1) requires monthly sampling of the exhaust stream of carbon adsorbers using Method 30B. The duration of sampling must be at least the minimum sampling time specified in Method 30B and up to one week. The proposed rule required a full week of such sampling, but, as pointed out by one of the commenters, breakthrough of the sampling trap from exhaust streams with high mercury concentrations could occur before a week had elapsed. Section 63.11647(f)(2) of the final rule clarifies that sampling of the carbon bed must be collected from the inlet and outlet of the bed. This paragraph also specifies that, for carbon adsorbers with multiple carbon columns or beds, the sampling should be performed in the first and last column or bed rather than at the inlet or outlet. We have deleted § 63.11647(f)(3) in the proposed rule, which allowed the carbon bed change-out rate to be determined based on historical data and the estimated life of the carbon. We have concluded that this method would not be adequate to ensure that breakthrough does not occur earlier than expected. We have clarified § 63.11647(h) with respect to the monitoring of scrubbers (other than the calomel-based mercury scrubbers described above). Under the final rule, owners or operators are required to monitor and record water flow rate (or line pressure) and scrubber pressure drop once per shift; they also must record any occurrences when the water flow rate (line pressure) or pressure drop are outside the operating range, take corrective actions to return the water flow rate (line pressure) or pressure drop back in range, and record the corrective actions taken. At proposal, the water flow rate and pressure drop were to be monitored continuously. However, measuring the water flow rate (line pressure) and pressure drop once per shift will provide two to three measurements per day, and we believe that is sufficient to assure proper operations of the wet scrubber, and thus assure compliance with the emission standards. We have also added the option of monitoring the line pressure in the scrubber liquid supply line as an alternative to monitoring scrubber liquid flow rate because line pressure monitoring is already in practice at some facilities and provides the same type of information as does liquid flow rate. As was proposed, the final rule allows the operating limit for water flow rate and pressure drop to be based on the minimum value during the initial performance test. It also includes two additional options for setting the operating limit: (1) Based on the manufacturer's specifications; and (2) based on limits established by the permitting authority. We have also clarified that, for scrubbers on autoclaves, the pressure drop parameter range should be established from manufacturer's specifications only. F. Definitions We have added a definition of carbon adsorber to § 63.11651 to clarify that this term, as used in the final rule, includes control devices consisting of a single fixed carbon bed, multiple carbon beds or columns, carbon filter packs or modules, and other variations of carbon adsorber design. The definition of “gold mine ore processing and production facility” in § 63.11651 of the rule has been clarified to state that small operations, such as prospectors and very small pilot scale mining operations, that process or produce less than 100 pounds of concentrate per year are excluded from the source category. These prospectors and very small pilot-scale operations (that process at or below this level) were not included in the section 112(c)(6) inventory that was the basis for the listing of gold mine ore processing and production source category. These types of very small operations were not intended to be subject to the final rule, and we do not expect any significant emissions from them. We also clarified that the source category does not include facilities at which 95 percent or more of the metals produced are metals other than gold. For example, if other non-ferrous metals (such as copper, lead, nickel, or zinc) comprise 95 percent or more of the product, the facility is not part of the gold ore processing and production source category. V. Summary of Responses to Major Comments A. Statutory Requirements 1. Listing of the Gold Mine Ore Processing and Production Source Category Under Section 112(c)(6) Comment: One commenter stated that adding the gold mine ore processing and production category to the list of categories required by Clean Air Act (CAA) section 112(c)(6) was correct and required because gold mines accounted for a significant portion of the aggregate emissions of mercury in the baseline year (1990) and because they still do so today. Other commenters stated that EPA does not have the authority to list gold mining processing and production as a source category under section 112(c)(6) and noted that section 112(c)(6) requires EPA to list, by 1995, categories of sources that make up 90 percent of the 1990 emissions for a subset of hazardous air pollutants (HAP), including mercury. The commenters said that EPA concluded its statutory listing obligation for mercury in 1998 with the publication of a list of source categories constituting 90 percent of aggregate mercury emissions, and that gold mining was not included on that list in 1998. In addition, the commenters said that the CAA requires EPA to list all categories under section 112(c)(6) by 1995 and complete issuance of standards for all listed sources by 2000, a task that would be impossible if EPA had the authority to add source categories ad infinitum. Response: We appreciate the commenter's support in listing the gold mine processing and production area source category pursuant to section 112(c)(6). We disagree, however, with the commenters that assert that EPA is precluded from listing additional categories pursuant to section 112(c)(6). The commenters appear to be arguing that EPA is limited to a single listing opportunity under section 112(c)(6) and, having not listed gold mine ore processing and production in the initial 1998 listing effort, EPA is now foreclosed from doing so. There is nothing in the language of section 112(c)(6), however, that precludes EPA from listing additional source categories to the extent EPA determines that those categories are needed to meet the 90 percent requirement in section 112(c)(6). Indeed, the commenter's reading is contrary to the fundamental purpose of section 112(c)(6). The core requirement of section 112(c)(6) is that EPA “shall * * * list categories and subcategories of sources assuring that sources accounting for not less than 90 per centum of the aggregate emissions of each such pollutant” are subject to standards under either 11217FE0(d)(2) or (d)(4). EPA reasonably interprets section 112(c)(6) as allowing it to revise the list to add categories, where, as here, it determines that it needs the additional categories to meet the 90 percent requirement in section 112(c)(6). Indeed, EPA has previously revised the section 112(c)(6) list to add a source category, where EPA determined that category was needed to meet its 90 percent requirement for mercury. See 72 FR 74087 (Dec. 28, 2007) (adding area source electric arc furnaces to the section 112(c)(6) list). As explained in the proposed rule, we have a 1990 baseline emissions inventory, and it is against this baseline that we assess compliance with the 90 percent requirement for each of the pollutants specified in section 112(c)(6). EPA explained in the initial 1998 listing notice that it was using 1990 as the baseline year for assessing compliance with the 90 percent requirement. As EPA has developed emission standards for the sources included on the initial section 112(c)(6) list, it has acquired additional information on those sources and their emissions in 1990, which has resulted in some revisions to the 1990 baseline emissions inventory estimates. These revisions resulted in the need to regulate an additional source category. See 72 FR 74087 (setting standards for area source electric arc furnaces). In addition to obtaining additional information concerning the source categories on the initial list, EPA has obtained additional information concerning the 1990 emissions of other sources. As explained in the preamble to the proposed rule, at the time of the initial section 112(c)(6) listing, there was very little available information on mercury emissions from gold mine ore processing and production. See 75 FR 22471. Because EPA lacked emissions information on mercury emissions from this source category at the time of the listing decision, EPA was unable to estimate the 1990 baseline mercury emissions from the gold mine ore processing and production source category and include this category in the first listing effort. Based on information that became available after the initial listing, EPA now finds that regulation of the area source gold mine ore processing and production category is needed to meet the 90 percent requirement for mercury. 75 FR 22471. Under the commenters' view, EPA cannot add any additional categories to the section 112(c)(6) list following the initial listing. If true, EPA could not meet its section 112(c)(6) obligation—a result Congress could not have intended. EPA reasonably interprets section 112(c)(6) in a manner that allows the Agency to achieve that provision's core requirement. EPA repeats that it sees nothing in the language or purpose of section 112(c)(6) that precludes it from listing additional source categories as needed. Finally, Congress left to EPA's discretion which categories and subcategories of sources to include on the section 112(c)(6) list. We have determined that we need the gold mine ore processing and production source category to meet the 90 percent requirement in section 112(c)(6) for mercury and are therefore now setting standards for that category. We also reject the comment that the task of completing standards by 2000 would be impossible if EPA had the authority to add source categories. Nevertheless, EPA is under a court ordered deadline to complete section 112(c)(6) standards by January 16, 2011. ( Sierra Club v. EPA, Consolidated Case No. 01-1537, D.D.C). Comment: Some commenters claimed that EPA did not provide an adequate basis for its 1990 emissions estimate for gold mining processing and production. Specifically, they questioned EPA's estimated emissions of 4.4 tons from this source category in the 1990 baseline year. Response: Although the commenters question EPA's estimated emissions of 4.4 tons from this source category in the 1990 baseline year, they did not provide an alternative method for calculating such emissions or alternative data or assumptions that should be used. They also did not explain what they think the 1990 baseline emissions should have been. EPA continues to maintain that its baseline emissions estimate is reasonable. The methodology EPA used to derive that estimate is described in docket item EPA-HQ-OAR-2010-0239-0175. Comment: Several commenters stated that Phase 2 permits under the Nevada Mercury Control Program (NMCP), which are scheduled for issuance by the end of 2010, will result in MACT-level controls on all thermal units at Nevada gold mines. According to the commenters, these permits are the culmination of a 7-year collaborative effort between NDEP and the gold mining industry to substantially reduce mercury emissions from gold mine processes. The commenters said that the proposal does not address how the NESHAP will result in reductions in mercury at gold mines in areas of the country other than Nevada, where the mercury content of the ore in gold mines is non-existent or only a fraction of the amount found in Nevada, and Nevada accounted for 99 percent of mercury emissions associated with gold mining operations in the United States. According to the commenters, this shows that if Nevada has an equivalent mercury control program for the gold mining industry, then there is nothing to be gained from imposing a Federal program, and if EPA acknowledges that the mines in Nevada are already well controlled, then the listing of gold mining and the promulgation of an additional layer of regulation at substantial cost to industry, but with little environmental benefit, is both legally indefensible and practically unsupportable. Response: As explained above, we are regulating the gold mine ore processing and production source category to meet the 90 percent requirement in section 112(c)(6) for mercury and are therefore setting standards for that category. Based on our 1990 baseline inventory for section 112(c)(6) and other emissions information for subsequent years, we estimate that this industry was among the top ten highest emitting categories of mercury emissions in the U.S. in 1990 and has remained in the top 10 since that time. Moreover, even though most emissions are from facilities located in Nevada, several commenters expressed serious concerns about the potential for mercury emissions from new gold mines in other States ( e.g., Alaska). We share these concerns about potential emissions from new gold mine facilities. Finally, Congress left to EPA's discretion which categories and subcategories of sources to include on the section 112(c)(6) list. We are regulating the gold mine ore processing and production source category to meet the 90 percent requirement in section 112(c)(6) for mercury and are therefore now promulgating a Federal NESHAP for existing and new gold mine ore processing and production facilities. 2. Emission Standards for HAP Other Than Mercury Comment: One commenter stated that CAA section 112(c)(6) provides that EPA must “list categories and subcategories of sources assuring that sources accounting for not less than 90 percent of each [enumerated] pollutant are subject to standards under subsection (d)(2) or (d)(4) of this section.” The commenter also stated that the D.C. Circuit has held repeatedly that when EPA sets standards for a category or subcategory of sources under section 112(d)(2), EPA has a statutory duty to set emission standards for each HAP that the sources in that category or subcategory emit ( e.g., National Lime Ass'n v. EPA, 233 F.3d 625, 633-634 (D.C. Cir. 2000)). The commenter concluded that when EPA sets standards for gold mines under section 112(d)(2), as section 112(c)(6) requires it to do, EPA must set section 112(d)(2) emission standards for all the HAP that gold mines emit. The commenter said that EPA appears to believe that because gold mines are needed only to reach the section 112(c)(6) requirement of 90 percent for mercury and not for the other pollutants enumerated in section 112(c)(6), EPA's only obligation under section 112(c)(6) is to set section 112(d)(2) standards for mercury. The commenter said that section 112(c)(6) expressly requires EPA to issue section 112(d)(2) standards for the “sources” in the categories listed under section 112(c)(6), not some subset of the pollutants that those sources emit, and that section 112(d)(2) standards must include emission standards for each HAP that a source category emits. The commenter continued by stating that nothing in the CAA exempts EPA from this requirement. The commenter concluded that, had Congress wished to give EPA discretion to set standards for only some of the pollutants emitted by a category listed under section 112(c)(6), it would have done so expressly. Response: EPA disagrees with the comment that, even though EPA lists a category under section 112(c)(6) due to the emissions of one or more HAP specified in that section, EPA must issue emission standards for all HAP (including HAP not listed in section 112(c)(6)) that sources in that category emit. The commenter cited in support the opinion by the United States Court of Appeals for the DC Circuit in National Lime Ass'n v. EPA, 233 F.3d 625, 633-634 (D.C. Cir. 2000)). The part of the National Lime opinion referenced in the comment dealt with EPA's failure to set emission standards for certain HAPs emitted by major sources of cement manufacturing because the Agency found no sources using control technologies for those HAP. In rejecting EPA's argument, the court stated that EPA has “a statutory obligation to set emission standards for each listed HAP.” Id. at 634. The Court noted the list of HAP in section 112(b) and stated that section 112(d)(1) requires that EPA “promulgate regulations establishing emission standards for each category or subcategory of major sources * * * of hazardous air pollutants listed for regulation. * * *” Id. (Emphasis added). For the reasons stated below, we do not believe that today's final rule is controlled by or otherwise conflicts with the National Lime decision. National Lime did not involve section 112(c)(6). That provision is ambiguous as to whether standards for listed source categories must address all HAP or only the section 112(c)(6) HAP for which the source category was listed. Section 112(c)(6) requires that “sources accounting for not less than 90 per centum of the aggregate emissions of each such [specific] pollutant are subject to standards under subsection (d)(2) or (d)(4).” This language can reasonably be read to mean standards for the section 112(c)(6) HAP or standards for all HAP emitted by the source. Under either reading, the source would be subject to a section 112(d)(2) or (d)(4) standard. The commenter insists that once a section 112(d)(2) standard comes into play, all HAP must be controlled (per National Lime ). But this result is not compelled by the pertinent provision, section 112(c)(6). That provision is obviously intended to ensure controls for specific persistent, bioaccumulative HAP, and this purpose is served by a reading which compels regulation under section 112(d)(2) only of the HAP for which a source category is listed under section 112(c)(6), rather than for all HAP. The facts here support the reasonableness of EPA's approach. Gold mine ore processing is an area source category listed under section 112(c)(6) for regulation under section 112(d)(2) solely due to its mercury emissions. There is special statutory sensitivity to regulation of area source categories in section 112. For example, an area source category may be listed for regulation under section 112 if EPA makes an adverse effects finding pursuant to Section 112(c)(3) or if EPA determines that the area source category is needed to meet its section 112(c)(3) obligations to regulate urban HAP or its section 112(c)(6) obligations to regulate certain persistent bioaccumulative HAP. Therefore, unless an area source category emits a section 112(c)(3) urban HAP or a section 112(c)(6) HAP and EPA determines that such category is needed to meet the 90 percent requirement set forth in section 112(c)(3) and (c)(6), findings related to adverse human health or environmental effects are required before EPA can regulate that area source category—findings EPA is unable to make for non-mercury HAP emitted from the gold mine ore processing and production source category at this time. Moreover, to the extent EPA lists an area source category pursuant to section 112(c)(3) (whether that finding is based on adverse effects to human health or the environment or a finding that the source is needed to meet the 90 percent requirement in section 112(c)(3), the statute gives EPA discretion to set generally available control technology (“GACT”) standards for such sources. 42 U.S.C. 7412(d)(5). EPA does not interpret section 112(c)(6) to create a means of automatically compelling regulation of all HAP emitted by area sources unrelated to the core object of section 112(c)(6), which is control of the specific persistent, bioaccumulative HAP, and thereby bypassing these otherwise applicable preconditions to setting section 112(d) standards for area sources. Nor does National Lime address the issue, since the case dealt exclusively with major sources. 3 233 F. 3d at 633. Consequently, EPA disagrees with the comment that it is compelled to promulgate section 112(d)(2) MACT standards for all HAP emitted by gold mine ore processors. 3 EPA acknowledges that major sources regulated under section 112 must be subject to MACT standards for all HAP emitted from the source category consistent with National Lime. 3. Emission Standards for Fugitive Emissions Comment: One commenter stated that gold mines have significant fugitive emissions of mercury, but that EPA did not propose standards for these emissions or mention them in its proposal. The commenter said that EPA has a statutory obligation to set standards for gold mine mercury emissions under section 112(d)(2) and (3), and must set emission standards for all the mercury emissions from the listed category. Another commenter described a recent preliminary study at two facilities in Nevada that found fugitive mercury air emissions from various non-point sources at those two mining operations such as from leach pads and tailings ponds. One commenter stated that means to control fugitive emissions are available, such as enclosing their leaching operations. By enclosing the leaching process, the commenter believes that mines could eliminate this source of fugitive emissions. The commenter also stated that mines should not send tailings into open tailing ponds, but into closed treatment facilities that would remove mercury and other HAP from the tailings and prevent their release to the air. The commenter recommended that EPA evaluate the use of sulfur-based complexing agents for removing mercury during cyanidization of gold. According to the commenter, research indicates that these products appear useful for substantially reducing mercury in process solution during heap leaching. Response: Due to the lack of information, we have not included fugitive mercury emissions at gold mine facilities in our 1990 baseline emission estimate (or in our more recent emissions estimates) for the gold mine ore processing and production area source category. Accordingly, these fugitive emissions are not part of the source category we are listing and regulating in this final rule. Other than the recent preliminary research at two facilities, we have no data on fugitive mercury emissions at gold mine facilities. The recent preliminary research suggests that some fugitive emissions may be occurring at these facilities from large non-point sources such as tailings ponds, leach fields and waste rock piles. However, it is our understanding that this preliminary research has not yet been published or peer-reviewed. Thus, at this juncture, we do not have sufficient information on fugitive emissions. Furthermore, we have very little information on how these fugitive mercury emissions might be controlled. A few commenters suggested that certain compounds were available that may be useful for limiting these emissions. However, as far as we know, there has been no demonstration that these compounds would work effectively to limit the emissions, and we do not know the costs or potential adverse impacts of applying these chemicals. Therefore, we question the feasibility and practicality of applying these chemicals to limit fugitive mercury emissions from these non-point sources. We also question the feasibility and practicality of enclosing the leaching operations or the tailings ponds, as suggested by some commenters. As explained in the proposed rule, the gold mine ore processing and production area source category covers the thermal processes that occur after ore crushing, including roasting operations ( i.e., ore dry grinding, ore preheating, roasting, and quenching), autoclaves, carbon kilns, electrowinning, preg tanks, mercury retorts, and furnaces. The data and calculations used to derive the estimated 4.4 tons of mercury emissions for this source category for the 1990 baseline inventory for section 112(c)(6) reflect emissions from the thermal processes described above, and the final MACT standards address all of these processes. 4. Major Source Determination Comment: One commenter noted that the proposal stated that the gold mining processing and production source category consists of only area sources; however, the proposal indicated that actual emissions of hydrogen cyanide (HCN) at a few facilities were near the major source threshold. The commenter concluded that EPA violates both the CAA and its own regulations by basing its evaluation of whether gold mines are major sources on their actual emissions instead of their potential emissions. The commenter further noted that the proposal requested comment on a certification process that would allow gold mines to avoid major source status whereby companies could certify that they are area sources by implementing certain “management practices” and then certifying to EPA that they had done so. The commenter stated that such a certification process would be unlawful in calculating a sources “potential to emit” because the management practices are not “control equipment,” “restrictions on hours of operation or on the type or amount of material combusted, stored, or processed,” and would not be “federally enforceable.” Other commenters supported EPA's conclusion and determination that the gold mines are area sources of HAP. According to the commenters, EPA's methodology in making this determination was extremely conservative because EPA did not apply what the commenters believe to be a key correction factor. Application of this correction factor would have reduced the HCN emissions estimates from by approximately 40-50%. The commenters also stated that fence line testing at selected gold mine operations demonstrated that these levels of HCN were below all applicable public health standards. The commenters believe that, because the gold mines are area sources of HCN, they should not be subject to section 112 work practice standards or newly developed certification requirements. The commenters noted that it is not technically practical to set systematic work practice standards to reduce HCN emissions for every gold mining operation to follow because each mine is unique in its mineralogy and cyanide leaching processes, and different process solution pH values are necessary to enhance gold recovery. The commenters explained that for economic, health, and safety reasons, they already implement work practice standards designed to minimize HCN. The commenter concluded that the combination of these work practice standards and the annual TRI reporting more than adequately ensure that gold mining operations will remain area sources of HCN. Response: Contrary to the assertions of one of the commenter's, EPA did not state in the preamble to the proposed rule that the sources at issue had actual emissions ranging from 5 to 9 tons. By contrast, EPA stated that “a few facilities are close to the major source threshold due to hydrogen cyanide (HCN).” 75 FR 22479. EPA failed to clarify in the preamble to the proposed rule that the range of 5 to 9 tons represented potential to emit calculations for the largest-emitting sources. Specifically, as explained in the document “Estimated Emissions of HCN from Gold Mine Facilities in the U.S.” (which is available in the docket for this rulemaking), EPA estimated the potential to emit for the five largest sources assuming that these sources would be operating every day of the year, 24 hours a day, at 100 percent of its current capacity. These assumptions and calculations resulted in a potential to emit estimate of 5 tons of HCN per year for the largest source. EPA then completed a second set of calculations, using the same assumptions ( i.e., operating every day of the year, 24 hours a day, at full capacity), but without applying the surface area correction factor, and those calculations resulted in a conservative potential to emit estimate of 9 tons of HCN per year for the largest source. The emission estimates for the remaining large facilities were all below 9 tons. The commenters correctly point out that in determining whether a source is a “major source” under CAA section 112, we must consider the source's potential to emit, as well as its actual emissions. See CAA section 112(a)(1) and 40 CFR 63.2. As noted above, we specifically examined the sources' potential to emit and concluded that all sources' potential to emit were below the major source thresholds. Some commenters allege that EPA significantly overestimated HCN emissions from the larger sources by not accounting for certain correction factors. They assert that if one were to account for the appropriate correction factors in developing the potential to emit values, HCN emissions would “range from 3.7-4.5 tpy for the larger mines compared to the 5-9 tpy estimate” ( See document titled “PTE Emission Estimates for HCN” by the Nevada Mining Association, which is available in the docket for this action). Other commenters make a blanket, unsupported assertion that the Agency has underestimated HCN emissions from the source category because they believe that without the management practices currently employed by sources in the category, HCN emissions would exceed the major source thresholds at the larger sources. These latter commenters, however, made only conclusory statements and did not demonstrate that HCN emissions from the larger sources would exceed the major source thresholds if the management practices were not employed. In sum, EPA has developed conservative estimates of the sources' potential to emit HCN. At one end of the range EPA estimates potential emissions of 5 tons per year of HCN for the largest source, which is well below the major source threshold of 10 tons per year of a single HAP. At the other end of the range EPA estimates potential emissions of 9 tons per year for that same largest source, which is a conservative estimate and is still below the major source threshold. The emission estimates for the remaining large facilities were all below 9 tons. We understand that the sources at issue implement various management practices as part of their operations to minimize the use and emissions of cyanide to protect workers, to comply with Mine Safety and Health Administration (MSHA) standards, to comply with their agreements to the International Cyanide Code, and for economic reasons (to reduce operational and supply costs). We currently do not have sufficient information to explicitly quantify emissions reductions achieved through these management practices, but nothing in the record suggests that the facilities would be major sources if they failed to employ the management practices. Accordingly, we are taking final action today to list the gold mine ore processing and production area source category and regulate its mercury emissions pursuant to CAA section 112(c)(6). Although not required, we intend to send letters to various Gold Mining Processing and Production companies pursuant to Section 114 of the Clean Air Act to confirm our conclusion that the sources' potential to emit remain below major source thresholds. 5. Title V Permit Exemption Comment: In the proposal preamble, EPA solicited comment on whether a title V exemption “is appropriate under section 502(a) for any particular sources in this category.” One commenter offered the following reasons for not exempting gold mines from title V permitting requirements: • EPA did not properly determine whether some or all sources in the category are major sources by determining each source's potential to emit. • The CAA allows EPA to exempt area sources from title V permitting only if it establishes that compliance with the title V permitting requirements would be “impracticable, infeasible or unnecessarily burdensome.” However, EPA does not claim that such requirements are “impracticable,” “infeasible,” or “unnecessarily burdensome” for gold mines. • It is feasible and within the gold mining companies' financial means to comply with title V permitting requirements. The commenter believes that the text and legislative history of the CAA make plain that Congress intended ordinary citizens to be able to get emissions and compliance information about air toxics sources and to be able to use that information in enforcement actions and in public policy decisions on a State and local level. According to the commenter, Congress did not think that enforcement by States or other government entities was enough; if it had, Congress would not have enacted the citizen suit provisions. The commenter said that, if a source does not have a title V permit, it is difficult or impossible for a member of the public to obtain relevant information about its emissions and compliance status or to bring enforcement actions. The commenter stated that to the extent the informational and enforcement benefits provided by title V permits can be considered a burden, these benefits far outweigh that burden. The commenter also noted that title V provides important monitoring benefits and that title V permits are necessary to provide adequate monitoring. The commenter concluded by stating that the legislative history of the CAA shows that Congress did not intend EPA to exempt source categories from compliance with title V unless doing so would not adversely affect public health, welfare, or the environment; however, exempting gold mines from title V would adversely affect public health, welfare and the environment by depriving the public of important informational and enforcement benefits. One State agency commented that additional title V permitting would subject both the source and the State agency to additional resource burdens. The commenter points out that major sources of criteria pollutant emissions are currently subject to title V permit requirements in Nevada and that sources not subject to major source permitting requirements are subject to Nevada's minor source permitting program. In addition, the NMCP requires all mining sources to obtain mercury-specific operating permits to construct. The commenter believes that these permit programs would provide a strong basis for implementing and enforcing any Federal MACT requirements for the gold mining industry, and there would be nothing gained by subjecting these sources to title V permitting. Several commenters stated that EPA should exercise its discretion and exempt the gold mine ore processing and production industry from the title V requirements as impracticable, infeasible, and unnecessarily burdensome. The commenters said that, in light of EPA's findings in other similar rulemakings for area sources, the four factors set forth in the Exemption Rule support a finding that title V permitting is “unnecessarily burdensome” for the gold mine ore processing and production area source category. In discussing the first factor of the Exemption Rule, whether title V would result in significant improvements to the compliance requirements, the commenters said that the proposed NESHAP for the gold mine ore processing and production area source category includes extensive monitoring, recordkeeping, and reporting requirements that are more comprehensive than title V requirements. The commenters believe that Nevada regulations and permits provide an additional layer of compliance assurance on the Federal NESHAP that obviates the need for title V permitting. The commenters claimed that the additional layering of title V does not “significantly improve” upon the proposed and existing compliance requirements. Regarding the second factor in the Exemption Rule, whether title V permitting would impose significant burdens on the area source category and whether the burdens would be aggravated by any difficulty the sources may have in obtaining assistance from permitting agencies, the commenters said that there are extensive administrative burdens and costs associated with the title V permitting process, including mandatory activities that have been previously identified by EPA. The commenters claimed that many of the area source gold mines are owned and operated by small entities that are already required to comply with comprehensive State permitting requirements for mercury emissions and that requiring title V permits for them would result in resources being redirected away from more useful and necessary efforts. The commenters explained that the third factor in the Exemption Rule examines whether the costs of title V permitting for the area source category would be justified, taking into consideration any potential gains in compliance likely to occur for such sources. The commenters claim that there do not appear to be any gains in compliance to justify the additional costs that would be imposed on these area sources from title V permitting based on the lack of significant improvements in compliance requirements and the substantial additional costs and burdens associated with title V compliance. The commenters noted that the fourth factor in the Exemption Rule analysis is whether there are implementation and enforcement programs in place that are sufficient to assure compliance with the NESHAP for the area source category, without relying on title V permits. The commenters claimed that the proposed rule includes all necessary monitoring to effectively implement its requirements, and the area sources for the gold mine ore processing and production are already permitted under State permit programs. According to the commenters, all non-title V sources in Nevada are required to hold “Class II” operating permits that must contain, among other things, all applicable emission limitations and standards. The commenters said that other States where gold mine ore processing and production area source are located either would be covered by a comparable delegated State air program or by EPA. The commenters stated that EPA regularly provides title V exemptions for area sources similar to gold mine ore processing and production area sources and cited examples from the past year. The commenters claim that the existing and proposed compliance and monitoring requirements for the gold mines are generally more stringent than those found in the other NESHAPs for which EPA has granted a title V permit exemption. The commenters stated that exempting the gold mine ore processing and production area source category from title V permitting will not adversely affect public health, welfare, or the environment because title V permits do not generally impose substantive air quality control requirements. According to the commenters, requiring title V permits also carries the potential of adversely affecting public health, welfare, or the environment by shifting State agency resources away from ensuring compliance with a program that is reducing mercury emissions from gold mines. The commenters stated that EPA should exempt the gold mine ore processing and production area source category from title V permitting requirements, and at a minimum, should exempt area source gold mines that are subject to Nevada's comprehensive mercury control program. Response: After reviewing the comments, we continue to believe that it is appropriate that all gold mine ore processing and production facilities be required to obtain title V permits. Most of the other area source categories for which we have provided title V permit exemptions have hundreds or thousands of facilities that are mostly owned by small businesses. In contrast, there are an estimated 21 facilities that are subject to this final rule, and, based on our research and analyses, none of the facilities are owned by small businesses; most of these facilities are owned by large, and in some cases, multi-national, corporations. Therefore, we conclude that the argument of financial burden, which has supported title V exemption for other source categories, does not apply to the gold mining industry (see Economic and Small Business Analysis, which is available in the docket). Currently, it is our understanding that 7 of the 21 facilities that will be subject to the final rule already have title V permits (5 in Nevada and 2 in other states). Further, there are approximately 5 facilities in all other States ( i.e., except Nevada) that do not currently have title V permits that will be subject to this final rule, so title V permitting will apply to no more than a few facilities in any one of these other States. Therefore, we do not believe the requirement for title V permitting will be overly burdensome to the permitting authorities in those States. Although there are more facilities in Nevada that will be subject to the final rule, as the commenters point out, Nevada already has an effective permitting system in place. Five of the 14 gold mine facilities in Nevada already have title V permits. Because of Nevada's existing permitting system and experience with title V permitting, we do not think that it is an undue burden on the State of Nevada to require title V permits from the other gold mine facilities located within the State. We also think it is important for the public in States where these facilities are located to have access to emissions and monitoring data and the opportunity for public involvement in the permitting of these facilities that is provided by title V permitting. 6. Reconstruction Comment: Several commenters believe it is appropriate to group under each of the umbrella “affected sources” all the equipment associated with each particular process in order to ensure a reasonable application of the reconstruction provisions found in the General Provisions. The commenters asked that EPA reaffirm that the 50 percent fixed capital cost trigger for determining reconstruction would be measured against all equipment components needed for the defined processes, and that reconstruction at one affected source as defined in the standard will not affect or result in reconstruction at another affected source. The commenters also noted that the definition of “reconstruction” authorizes EPA to establish special provisions in a particular standard for the application of the reconstruction criteria to the affected source. The commenters said that the “carbon processes” affected source illustrates that the affected source can consist of several pieces of interconnected equipment that together constitute the process line, and it can be anticipated that production needs will give rise to the need to add more pieces of equipment to an existing carbon process line or even to install a whole new carbon process line. The commenters provided three examples: Adding a new component to an existing carbon processes group; construction of a new carbon group due to expansion at a facility that has an existing carbon group; and installation of new pollution control equipment. The commenter said that consideration of whether or where new MACT requirements should apply in these examples warrants the development of special reconstruction provisions in this standard, or EPA should clarify that the three examples would not be considered reconstruction under the proposed rule. The commenters asked that EPA either clarify that the three examples would not be considered reconstruction, or alternatively, add the following provisions to the proposed rule: (1) An addition of a new piece of equipment to address production requirements is not considered a reconstruction, (2) the expansion of a facility by the construction of a completely new process line will not be considered a reconstruction of an existing process line, and (3) the installation of air pollution control equipment to comply with this standard is not considered a reconstruction. Response: The determination of what constitutes a reconstruction is directly tied to the definition of the affected source and the definition of reconstruction in the part 63 General Provisions: Reconstruction, unless otherwise defined in a relevant standard, means the replacement of components of an affected or a previously nonaffected source to such an extent that: (1) The fixed capital cost of the new components exceeds 50 percent of the fixed capital cost that would be required to construct a comparable new [affected] source; and (2) It is technologically and economically feasible for the reconstructed source to meet the relevant standard(s) established by the Administrator (or a State) pursuant to section 112 of the Act. Upon reconstruction, an affected source, or a stationary source that becomes an affected source, is subject to relevant standards for new sources, including compliance dates, irrespective of any change in emissions of hazardous air pollutants from that source. For each of the four affected sources in the final rule, we have defined the affected source as the collection of processes associated within each affected source. Consequently, if one process within the affected source is upgraded or replaced with a new process, the 50 percent fixed capital cost criterion would be based on the fixed capital cost of replacing all processes in the affected source, not just the capital cost of the process being upgraded or replaced. For example, if a new carbon kiln is added to an existing group of carbon processes with mercury retorts, the capital cost of the new carbon kiln would be divided by the fixed capital cost of constructing a comparable new affected source containing all of the processes within the existing affected source of carbon processes with mercury retorts to calculate the percent for comparison to the 50 percent criterion. With regard to the scenario where a new carbon process with a mercury retort is installed, the affected source is defined as the collection of all applicable processes within the affected source, and because of this, a facility could not have two carbon processes with mercury retorts affected sources, such as the commenter suggested, where one group is new and the other is existing. For example, if a new group of carbon processes with mercury retorts is installed at a facility in addition to an existing group of carbon processes with mercury retorts, the two groups (all carbon processes with mercury retorts at the facility) collectively would be a single affected source. In this case, the fixed capital cost criterion would be based on the fixed capital cost of replacing the existing affected source with a comparable new affected source, and if the new processes exceed 50 percent of that cost, all of the carbon processes with mercury retorts would be subject to the new source limit for carbon processes. There would not be separate and different emission standards for the two sets of carbon sources with mercury retorts (the older group and newer group) because the collection of all of these processes is the affected source. We do not see a necessity to provide criteria for this final rule that are different from the requirements in the General Provisions for determining what constitutes a reconstruction. We also think it is appropriate to exclude the cost of emission control equipment from the cost calculation for reconstruction determinations. B. Applicability Comment: Several commenters stated the rule should exempt individuals (prospectors), laboratories, small mining operations, and non-leaching operations. The commenters urged EPA to include in the final rule all of the following exemptions to avoid the problem of unintended regulation of sources that were not meant to be included in the source category: Gold mining operations that produce less than 100 pounds of concentrate per year, which would exempt analytical labs that perform small bench scale processing tests on gold ores; gold mining operations that do not leach or dissolve gold, which would exempt placer and other non-leaching operators, including both small commercial efforts as well as individual recreation-type prospectors; and gold mining operations that process less than 1,000 tons per year of gold ore, which would exempt certain small scale pilot plants and related testing operations. The commenters said that the exemptions suggested above will not reduce in any way the effectiveness of the proposed rule in controlling mercury emissions from the targeted larger mines, nor will they lead to increased mercury emissions, but they will exclude regulation of a large number of small operators who do not emit any significant mercury. Response: Section 63.11640(c) of the proposed rule provides that the emissions standards for this area source category do not apply to research and development facilities, as that term is defined under CAA section 112(c)(7). We did not receive any adverse comments concerning this provision, and are finalizing the provision in this rule. Further, as mentioned above in section IV, we are clarifying in this final rule that this area source category does not include individual prospectors and very small pilot scale mining operations. Prospectors and other very small pilot-scale operations ( e.g., operations that produce or process less than 100 pounds of concentrate per year) are very small and were not included in the section 112(c)(6) inventory that was the basis for the listing of gold mine ore processing and production source category. We believe that emissions from the very small scale operations described above to be very minimal. By contrast, the commenter's suggested 1,000 tons/yr ore threshold may include operations beyond the very small scale pilot operations discussed above. We believe that the 100 pounds of concentrate per year more appropriately reflect these very small scale operations. We are not making the suggested change of excluding operations that do not leach or dissolve gold because certain gold mine facilities in the source category use flotation or gravity flotation processes and perform thermal processing of concentrate in melt furnaces, which can have significant emissions of mercury. However, as mentioned above we are clarifying that this final rule does not apply to these very small scale operations. C. MACT Floors 1. Consideration of Variability in Determining Floors Comment: One commenter acknowledged that EPA may consider variability in calculating the best sources' performance, but stated that EPA's method of considering variability seeks to assure that none of the sources among those identified as best performers would ever exceed the floor level. The commenter claims that such an approach ignores the reality that sources' emission levels are largely within their control, and although a great deal of variability may be statistically conceivable if EPA chooses a high enough prediction limit (in this case the 99th percentile) that does not mean that a well-operated source actually would experience such variability. The commenter said that one of the main points of having emission standards is to ensure that sources not only deploy the appropriate control measures, but also use those control measures consistently to minimize emissions. The commenter said that using an upper prediction limit to set standards reflecting the statistical worst performance these sources could have in a purely statistical sense does not yield an accurate picture of the best sources actual performance, and it is especially arbitrary in the absence of any explanation of why EPA thinks that the relevant best sources' performance would ever be so bad, other than the fact that it is statistically possible. Response: As described previously, the MACT floor limits are calculated based on the performance of the lowest emitting sources in each of the MACT floor pools. We ranked all of the sources for which we had data based on their emissions and identified the lowest emitting sources. As the commenter concedes, EPA can consider variability in assessing sources' performance when setting MACT standards. See Brick MACT, 479 F.3d at 881-82; and Mossville Envt'l Action Now v. EPA, 370 F.3d 1232, 1241-42 (D.C. Cir 2004) (reaffirming that EPA can assess variability in determining the level of emissions control achieved by the best performing sources). Variability in facilities' performance has various causes. One source of variability for these facilities is the differing mercury concentrations in the input materials. Another source of variability is due to normal variations in performance of the control devices for which both run-to-run and test-to-test variability must be accounted. 4 A review of the run-by-run emissions data in the record shows that emission rates from one run to the next for well-operated sources can vary by as much as a factor of 8. We need to account for sources' variability (both due to control device performance and variability in inputs) in assessing sources' performance when developing technology-based standards. Accordingly, EPA accounts for variance in test data, between units, and among facilities when developing the MACT standard. 4 Run-to-run variability is essentially within-test variability, and encompasses variability in individual runs comprising the compliance test, and includes uncertainties in correlation of monitoring parameters and emissions, and imprecision of stack test methods and laboratory analysis. 72 FR at 54877 (Sept. 27, 2007). Test-to-test variability results from variability in pollution device control efficiencies over time. Test-to-test variability can be termed long-term variability. 72 FR at 54878. In determining the MACT floor limits, we first determine the average emissions of the top performers based on available data. We then assess variability of the best performers by using a statistical formula designed to estimate a MACT floor level that is equivalent to the average of the best performing sources based on future compliance tests. Specifically, the MACT floor limit is an upper prediction limit (UPL) calculated with the Student's t-test. The Student's t-test has also been used in other EPA rulemakings ( e.g., NESHAP for Cement Manufacturing, NSPS for Hospital/Medical/Infectious Waste Incinerators, and NESHAP for Industrial, Commercial, and Institutional Boilers and Process Heaters) in accounting for variability. A prediction interval for a future observation is an interval that will, with a specified degree of confidence, contain the next (or some other pre-specified) randomly selected observation from a population. In other words, the prediction interval estimates what the upper bound of future values will be, based upon present or past background samples taken. The UPL consequently represents the value which we can expect the mean of future observations ( i.e., emission test runs) to fall below within a specified level of confidence, based upon the results of an independent sample from the same population. In other words, if we were to randomly select a future test condition from any of these sources ( e.g., average of 3 runs) we can be 99 percent confident that the reported level will fall at or below the 99 percent UPL value. We note that the methodology accounts for both short-term and long-term variability and encompasses run-to-run and test-to-test variability. For this rule, we used the 99 percent UPL analysis on the emissions data for the top performing sources to account for the variance. In the context of determining the MACT floor, the 99 percent UPL represents the value below which the mean of future compliance tests (based on, for example, a 3-run average) would fall 99 percent of the time. A 99 percent level of confidence means that a facility, whose emissions are consistent with the best performing sources, has one chance in 100 of exceeding the emission standard. We believe that using the 99 percent UPL is appropriate for this rule. As noted above, this approach is consistent with several other previous rulemakings. It also makes sense from a practical standpoint. If we selected a lower number ( e.g., 95 percent UPL) this would mean that a best performing source that is performing at the MACT level of control would potentially exceed the limit 5 percent of the time—which we do not believe is a reasonable approach for this rule. See Mossville, 379 F.3d at 1241-42); see also 70 FR at 59438 (Oct. 12, 2005) (explaining use of 99th percentile). With regard to the commenter's statement that no sources among the best performers would ever exceed the MACT standard, we believe this is incorrect. The commenter provided no basis for this statement, and we do not believe the commenter based this statement on an analysis of the variability in the data. We do not believe that the UPL analysis reflects the statistical worst performance the top five performing sources could have. The UPL calculation is dependent on the data that we have, and reflects the actual variability in the test data for the best performing sources. It does not reflect worst-case performance. We continue to believe that the UPL does yield an accurate picture of the best sources' performance as best as possible with taking into account variance between the facilities, units at the facilities, and between test runs for the different units (including variability in input materials). Furthermore, although the average of several data sets may show a top performing source meeting the emission standard by a significant margin, the variability in emissions inherent in any one compliance test could easily indicate much higher emissions, and, in some cases, an exceedance of the emission standard. We continue to believe that the UPL analysis evaluated at 99 percent confidence is appropriate for this source category. Moreover, we believe the data we used to calculate the MACT standards are representative of the normal performance of the best performing sources for several reasons. First, the test results that we are using in our MACT database are tests conducted under Nevada's mercury emission control program, and are conducted to determine whether a facility is in compliance with State requirements. Facilities typically try to perform as well as they can during such tests. State (and often EPA) permitting authority staff are notified before a performance test is conducted to provide an opportunity to attend and observe the test, and they often attend to ensure the source is operating properly and that the testing is performed according to the strict requirements in the codified test methods. Test reports are carefully reviewed by the permitting authority, and any failure to follow the test method or abnormal operation of a source is flagged. These data are usually invalidated, and invalidated tests are not used in our MACT standard calculations. For example, several tests from these facilities were invalidated by the NDEP because the specified testing procedures were not followed or the emission control device was not operating properly, and we have not used those results in our analysis for those reasons. We have collected additional data from test reports not available at the time of proposal, and one of those tests was invalidated because NDEP representatives discovered that the emission control device was not operating properly during the test. Therefore, we also did not use those test data. The commenter believes that floors must be set at the average emission level achieved by the best performers when they are operating properly. We agree that the performance data characterizing the emission level achieved by the top performers must be data obtained when they are operating properly, and we believe that is the case for our current database for this source category. As described above, the MACT floor is based on the average performance of the top performers plus an amount to account for variability. We have appropriately developed a MACT standard based on emissions from the top 5 best performing sources that accounts for variability because, over an extended period of time, the emissions from each of these best performing facilities (even the best controlled) will vary above and below the facility average. For example, we expect that about half of the duration of the year the emissions from a best performing facility would be somewhat below their average and that about half of the duration of the year their emissions would be somewhat higher than their average. If we set the MACT limit exactly equal to the average emissions level achieved by the best performers (without accounting for variability), and we had a source that was performing at exactly the MACT level over the course of the year, the measured emissions level on roughly half the days of the year would suggest that the source is emitting at levels above the MACT limit, and on about half of the other days of the year the measured emissions level would suggest that the source is emitting at levels less than the MACT limit. We reasonably and appropriately accounted for variability in the data consistent with established statistical theory and practice and judicial precedent. Finally, ignoring variability of the best performing sources and using only the average performance would virtually guarantee that some of even the best performers would exceed the floor limit at least some of the time. Thus, we developed a MACT standard based on the average of the best performing sources that accounts for variability. We accomplished this by calculating the MACT standard from this average performance and accounting for variability by using the 99 percent UPL. The specific calculations are presented in the MACT floor document in the docket for this rulemaking. Furthermore, we agree with the comment that one of the points of having emission standards is to ensure that sources not only deploy the appropriate control measures, but also use those control measures consistently to minimize emissions. We believe that the MACT standards established in this rule along with the requirements to monitor and maintain control device parameters within certain ranges will ensure control measures are applied consistently to minimize emissions. Comment: Another commenter stated that consideration should be given to defining the inherent range of measurement error and requiring more test runs in order to reduce variability due to process variation. The commenter said that this would also better clarify when variability was due to operational controls, which could be addressed, rather than due to factors that cannot be controlled, such as mercury content in the ore. The commenter asked for clarification on how inconsistent runs should be treated, what defines an acceptable set of runs, and at what point more runs would be required to provide reliable data. The commenter also stated that the degree of variability allowed in the development of the new source limit for ore pretreatment appears to be out of line with the new source limits for carbon processes and non-carbon processes. The commenter believes that ore pretreatment variability for new sources is higher than existing sources because low thermal units were included in the same category, high emissions were allowed in the data set, and variable emissions were allowed in data set. The commenter recommended that, if EPA continues to use Goldstrike as the best performing source for new source MACT, then they should re-evaluate and reduce the variability to be equal to or less than the variability for existing sources. Response: We agree with the commenter that the testing process would be more accurate if the number of test runs was increased. However, we balance several factors in determining the minimum number of runs required, and because the compliance testing is supplemented by various types of continuous or periodic parametric monitoring, we have concluded that three test runs are appropriate for this final rule. Although we have not proposed a formal procedure to assess the consistency of test runs, the permitting authority performs routine reviews of compliance test data to identify potential outliers and results that suggest further investigation is needed. For example, a routine review tracks trends in performance, and in particular, flags any trends in deteriorating performance over time. An unusually high run among the three runs also attracts attention and would be examined to determine if it might have been caused by a problem with the process, control device, sampling, or analysis. If the permitting authority identifies inconsistent runs, they have the authority to invalidate any or all runs. A source would be required to perform more runs to provide reliable data if two to three runs were invalidated. We agree with the commenter that the degree of variability used in the development of the proposed new source MACT standard for the ore pretreatment group appeared to be inconsistent with the degree of variability used in the development of the proposed new source MACT standard for carbon processes and non-carbon concentrate processes. We agree with the commenter that the ore pretreatment degree of variability at proposal for new sources was higher than the degree of variability for existing sources. We do not believe that the variability was higher because low thermal units ( i.e., autoclaves) were included in the same category, but because two tests of the ore preheater/dry grinding processes at Goldstrike were allowed in the data set. These tests had, as the commenter identified, inconsistently high emissions (as compared to other tests at other times for the same units) and inconsistent variability between the runs. We have determined that the tests the commenter is referring to are not representative of normal operation, and those tests have been removed from our database because the NDEP invalidated the tests due to possible sample contamination. ( See the MACT Floor Document in the docket for the final rulemaking for more details). We continue to use Goldstrike as the best performing source for the ore pretreatment new source MACT, and the variability for new source MACT is now less than that of the variability for existing source MACT, and is less than the variability calculated at the time of proposal. 2. General Comments on MACT Comment: Some commenters stated that the MACT floor already represents installation and operation of MACT controls, and the use of emissions data from facilities that are already controlling their mercury emissions creates an artificially low MACT floor. The commenters said that the low MACT floor penalizes facilities that voluntarily invested in pollution control technology and creates a substantial disincentive for industry and States to move ahead of EPA in reducing emissions of HAP. Response: We acknowledged at proposal that many gold mine facilities are already well controlled for many reasons, including participation in the NMCP. We also acknowledge that the top performing facilities that are the basis for the MACT floor calculation are the top performers because they have installed controls. CAA section 112(d)(3)(B) requires that, for a category with fewer than 30 sources, the MACT floor not be less stringent than “the average emission limitation achieved by the best performing 5 sources ( for which the Administrator has or could reasonably obtain emission information ).” (Emphasis added). EPA has information on the well-controlled facilities and used the information to conduct MACT floor analysis, as required by the CAA. Although the MACT floor may be considered more stringent in comparison to floors that would have been established if no facilities had mercury emission controls, we do not consider the floor to be “artificially low” because consistent with the statute, it reflects the level achieved in practice by the best performing sources. See 112(d)(3). We do not believe that the MACT floor penalizes facilities that invested in pollution control technology because those facilities will be able to meet the MACT standards. We do not consider that this final rule creates a disincentive for industry and States to move ahead of EPA in reducing HAP emissions because as facilities reduce mercury emissions by adding controls required by State programs, they will be able to meet the NESHAP. Most of the facilities that will not meet the current final standards have already proposed to add controls to their units in their Phase 2 applications for the NMCP. 3. MACT for the Ore Pretreatment Group Comment: Several commenters supported EPA's general approach to establish three groups of affected sources in the proposal. On the other hand, several commenters suggested that EPA develop separate emission standards for roasters and autoclaves for existing and new sources. One commenter stated that roaster and autoclave processes are different from each other based on the mercury species released, controls utilized, and their rates of mercury emissions. The commenter said that roasters commonly reach temperatures of 400° to 700°C, releasing gaseous elemental mercury, whereas autoclaves commonly reach temperatures of 175° to 230°C producing reactive gaseous mercury and sulfate and forming mercury sulfate. According to the commenter, autoclaves are expected to be able to improve efficiency over time. The commenter noted that roasters produce one to two orders of magnitude higher emissions than do autoclaves. The commenter believes that facilities that only use autoclaves should not be allowed the leeway to emit at the rate that facilities employing roasters are allowed. The commenter recommends that the ore pretreatment group be divided into high temperature pretreatment processes (roasters) and low temperature pretreatment processes (autoclaves and ancillary roaster processes, such as dry grinding, pre-heating, and quenching). Response: We discussed in section V.A. of the preamble to the proposed rule our rationale for establishing the different affected sources, including the ore pretreatment processes affected source. We believe it is appropriate to maintain the ore pretreatment group affected source, as we had proposed. We do not agree with the comment that roasters necessarily have higher emissions that are one to two orders of magnitude higher than emission from autoclaves. The available data show a wide range in emissions from autoclaves (from 0.4 to 115 lb/million tons of ore). This range overlaps the range for roasters and their ancillary equipment, which have combined emissions between 42 to 71 lb/million tons of ore. Regardless of the mercury species released, controls utilized, operating temperatures, or control efficiency over time, autoclaves and roasters process the same input material ( i.e., ore) and are intended for the same purpose ( i.e., to oxidize the ore). Therefore, we believe that it is appropriate to maintain the ore pretreatment affected source as we had proposed, keeping roasting operations and autoclaves together. Comment: One commenter stated that EPA failed to consider beyond-the-floor standards for roasters and that if additional reductions are achievable at roasters, then EPA must set additional beyond-the-floor standards for roasters. A commenter also stated that although EPA's standard for new ore pretreatment facilities is as high as its standard for existing facilities, EPA does not propose or discuss setting beyond-the-floor standards for new sources. The commenter claims that EPA has a statutory obligation to ensure that its new source standards reflect the maximum achievable reduction in emissions. Response: Following proposal, we continued to investigate the performance of facilities with ore pretreatment processes and opportunities for additional control. We collected data from more recent tests that were not available at proposal, and these new data show that emission control performance at these facilities has continued to improve. We identified two previous tests in the proposal database that were suspect, and we confirmed with NDEP that these tests should be invalidated and not used in the analysis because of possible sample contamination. We have also dropped the data for one facility from the analysis because their autoclave was shutdown in 2007 and dismantled, and we only had one test of the autoclave when it was operating in 2006. For these reasons, we did not include data for that facility in the analysis, which is now based on the only four facilities currently operating. Based on the addition and change described above with respect to our available data, we revised the MACT floor analysis for the ore pretreatment processes. The revised MACT floor for existing sources decreased from 175 lb/million tons at proposal to 158 lb/million tons, and the new source MACT floor dropped from 163 lb/million tons to 84 lb/million tons. The MACT floor limit for existing ore pretreatment processes is based on the use of calomel-based mercury scrubbers on roasters and wet scrubbers on autoclaves and ancillary roaster operations. We conducted a beyond-the-floor analysis during the development of the proposed rule. The roasters were already equipped with very good mercury controls (condensers and calomel-based mercury scrubbers), and we did not identify any beyond-the-floor options for the roasters. However, we identified as a beyond-the-floor control for autoclaves the installation of both a refrigeration unit (or condenser) and a carbon adsorber. We continue to believe that the roasters stacks are well controlled, but since our proposal, we have identified a beyond-the-floor control option (carbon adsorption) for the ore pre-heaters/dryers (ancillary roaster operation) that could achieve additional emissions reductions of approximately 70 percent (or more) for those units. Two of the three facilities with roasters have already proposed in their NMCP Phase 2 permit applications to apply controls to their preheaters/ore dryers, and these two companies have submitted cost estimates for applying a carbon adsorption system. Using the cost estimates submitted by the affected facilities, we estimate the capital costs for control of roaster preheaters/dryers for the three facilities with roasters as $3 million with a total annualized cost of $1.6 million per year. We also estimate a reduction of 118 lb/yr of mercury emissions would be achieved at an overall cost effectiveness of about $13,800 per pound of mercury. We believe that these costs and cost effectiveness are reasonable. As required under CAA section 112(d)(2), we have also considered non-air quality health and environmental impacts and energy requirements of this additional control. We conclude that this is an acceptable beyond-the-floor control technology for existing roaster preheaters/ore dryers. Therefore, we included the beyond-the-floor control for ore preheaters/dryers, as well as the beyond-the-floor control for autoclaves, in determining the MACT standard in this final rule for existing sources of ore pre-treatment processes. After applying the appropriate variability analyses to the data, we determined that the MACT standard for existing sources is 127 lb/million tons of ore. As mentioned above, we have revised the new source MACT floor. We also did a beyond-the-floor analysis for new sources in the ore pre-treatment processes group. However, we did not establish the MACT standard for new sources based on this beyond-the-floor analysis because we did not identify a feasible and cost-effective option to achieve reductions greater than the new source MACT floor. Therefore, for new sources of ore pretreatment processes, the MACT “floor” is the MACT standard for the affected source. The final new source MACT standard is 84 lb/million tons of ore, which is considerably more stringent compared to the proposed standard of 149 lb/million tons of ore and reflects the maximum achievable reduction in emissions. Comment: One commenter stated that the proposed estimated capital costs of $890,000 and total annualized cost of $720,000 for beyond-the-floor autoclave controls are not representative of actual costs of installing a refrigeration unit (or condenser) and a carbon adsorber on autoclaves. The commenter estimates that capital costs for autoclave controls will range from $18 million to at least $30 million, and annual operating costs could range from $2 million to $60 million, depending on which controls, if any, are determined to be technically feasible. The commenter believes that based on these cost estimates, beyond-the-floor MACT controls would be cost prohibitive and are not justified for the ore pretreatment affected source group. Another commenter estimated that for the installation of carbon adsorbers on their autoclaves to control mercury emissions, the capital costs would range from $30 million to $35 million, annual operating costs would be $2 million per year, and the annual energy requirements would be 11,400 megawatt-hours per year with an annual energy cost of $900,000. Response: After reviewing the new cost estimates provided by the commenters, we agree that capital and total annualized cost estimates of the beyond-the-floor controls on autoclaves in the proposal were underestimated. We evaluated the detailed cost estimate based on an engineering study for a carbon adsorption system provided by one of the commenters (see details in the comment above on capital, operating, and energy costs), and our review of these details indicates it to be a reasonable cost estimate and more representative. Therefore, we have used this estimate as the basis for our estimate of the costs of the beyond-the-floor mercury emission controls for autoclaves. Our revised estimates are that the capital cost for installing carbon adsorbers on autoclaves would be $29.3 million, with a total annualized cost of $4.9 million per year, which would result in an estimated reduction of 431 lb/yr of mercury emissions per year and an overall cost effectiveness of about $11,000 per pound of mercury. Based on these new costs and estimated reductions we conclude that the beyond the floor controls are affordable and justified for the ore pretreatment affected source. Comment: Several commenters noted that, at the proposed new source MACT limit of 149 pounds/million tons of ore, the proposed new source Donlin Creek Mine, located in Alaska, would be allowed to emit 3,200 lb/yr of mercury based on a projected production rate of 22 million tons/yr of ore. Response: With respect to this proposed new gold mine in Alaska, the commenters' estimate of 3,200 lb/yr of mercury emissions is inaccurate and a significant overestimate for a number of reasons. The two primary reasons are that, based on available information, if the facility is built, only an estimated 15 percent of the ore mined will be processed in autoclaves (not 100 percent as assumed by the commenters), and that the commenters' estimate is based on assuming that the average emissions level for the facility throughout the year would be at the maximum allowed at the proposed new source limit (149 lb/million tons of ore), which has been significantly reduced since proposal. With the new source MACT standard in the final rule that is about two times more stringent ( i.e., lower) than the proposed MACT standard, along with corrections described above, we estimate that far less than 3,200 lb/yr would be emitted from this new source if it is ever built. Assuming continuous operation for 365 days per year, an estimated 21.5 million tons/yr of ore mined, about 3.2 million tons/yr processed in autoclaves (15 percent), and assuming the source would emit at the average emission level used to calculate the revised new source MACT (45 lb/million tons of ore), we calculate that mercury emissions would be about 144 lb/yr, which is about 5 percent of the estimate provided by the commenters. Considering that the facility has yet to go through the permitting process and that, if it is built, it will likely include emissions controls that would reduce the emissions below 45 lb/million tons of ore, we believe that, if the facility is built, emissions would quite likely be lower than 144 lb/yr. 4. MACT for Carbon Processes Comment: Several commenters objected to including Facility M in the MACT floor determination for new and existing sources in the carbon processes affected source because it is not representative of, or similar to, other sources, because it has unusually low mercury concentrations in its ore, and no need for a retort to remove and recover mercury. They noted that, because the mercury content of the gold ore is fixed, the only way for other facilities to reduce emissions of mercury is to apply mercury emission controls, but, for many facilities, emission controls will not be enough to meet the proposed MACT standard. The commenters stated they were aware that the DC Circuit Court had constrained EPA's discretion to set floors that fail to consider material inputs, but they said gold mines were different from the remanded source categories (brick kilns and cement kilns) because gold mining operations process very large quantities of ore, and the ore is the only material input that results in mercury emissions. The commenters stated that, in adopting section 112, Congress expressly cautioned EPA against setting standards that would require mining operations to change the ore used as essential feedstock. The commenters said that, by ignoring the mercury content in the ore being mined and processed at the facilities in the MACT floor determination, EPA is requiring facilities to consider the substitution of, or changes in, the ore that is processed because there is no other way to achieve the standard. The commenters recommended that EPA address, as a threshold matter, the differences in processing and emissions across facilities that result from the variable concentration of mercury in ore. The commenters recommended that Facility M not be considered the “best controlled similar source” for purposes of setting the new source MACT floor because the facility is not similar to other sources. The commenters stated that, if EPA does not exclude from the source category facilities that do not use retorts to process concentrate, then they should subcategorize them. Response: After consideration of comments and a re-examination of the design of the facilities at issue, the emission controls, and other factors affecting emissions from the carbon processes at Facility M, we agree that this facility is quite different and unique compared to most other gold mine ore processing and production facilities, including other facilities in Nevada, in its carbon process. The difference is manifested in the processing train in that mercury retorts are not needed or used at Facility M to recover mercury. As the commenter notes, the CAA allows EPA to “distinguish among classes, types and sizes of sources within a category” in developing MACT emission standards, and gold mine facilities without mercury retorts are different in both class and type from those with mercury retorts. Accordingly, in the final rule, we identify and set separate MACT standards for these two different types of carbon processes: those that use mercury retorts; and those, such as the carbon process at Facility M, that do not use mercury retorts. As part of our re-analysis of the MACT floor and the MACT for sources that are in the carbon processes with mercury retorts group and sources that are in the carbon processes without mercury retorts group, we considered new data that were not available at the time of proposal. Over the past one to two years since our data collection effort for the proposal, facilities in Nevada have continued to add controls and improve emission control as part of the NMCP. The new data indicate there were two facilities with carbon processes without mercury retorts operating in 2009. Using the data from these two facilities, we determined that the MACT floor limits for carbon processes without mercury retorts are 0.17 lb/ton of concentrate for existing sources and 0.14 lb/ton of concentrate for new sources (based on the best performing facility, Facility M). In our beyond-the-floor analysis, we considered the addition of a carbon adsorber on an uncontrolled emission unit within an existing affected source. We estimate the capital cost as $210,000 with a total annualized cost of $72,000 per year, an emission reduction of 1.63 lb/yr of mercury, and a cost effectiveness of $44,000/lb of mercury. We do not believe that the small emission reduction that this control option would achieve is justified in light of its cost. We therefore decided not to go beyond-the-floor. We also considered possible beyond-the-floor options for new carbon processes without mercury retorts, but concluded these options were not cost-effective or feasible. Therefore, for new and existing sources of carbon processes without mercury retorts, the MACT floor limit is the MACT standard for this affected source. As part of our re-analysis for the carbon group processes with mercury retorts, we collected and evaluated additional data. As discussed above, several of the facilities have improved emission control over the levels observed in the database we used at proposal. Two facilities with newly-installed controls replaced two higher-emitting facilities that were in the top 5 at proposal, and all three of the other facilities that remained in the top 5 had lower levels of emissions after considering the new data. The results are that the MACT floor limits for carbon processes with mercury retorts are 2.2 lb/ton of concentrate for existing sources and 0.8 lb/ton of concentrate for new sources (based on the best performing facility, Facility N). In the beyond-the-floor analysis, we evaluated the impacts of adding a second carbon adsorber in series with the controls applied to achieve the MACT floor level of control. We estimate the capital cost would be $3 million with a total annualized cost of $1.3 million per year, an emission reduction of 9 lb/yr of mercury, and a cost effectiveness of $150,000/lb of mercury. B [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%3A2011-2608. Public record. Not legal advice.