﻿
UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
REGION IX
75 Hawthorne Street
San Francisco, CA  94105

October 2020

Memorandum

Subject:Response to Comments Document for the EPA's Final Actions on the "Arizona State Implementation Plan Revision: Hayden Sulfur Dioxide Nonattainment Area for the 2010 SO2 NAAQS" and Rule R18-2-B1302, "Limits on SO2 Emissions from the Hayden Smelter"

From:Ashley Graham, Air Planning Office, Air and Radiation Division
  Kevin Gong, Rules Office, Air and Radiation Division

To: Docket # EPA-R09-OAR-2020-0109
Docket # EPA-R09-OAR-2020-0173

On May 22, 2020, the EPA proposed to partially approve and partially disapprove an Arizona state implementation plan (SIP) revision for attaining the 2010 1-hour primary sulfur dioxide (SO2) national ambient air quality standard (NAAQS) for the Hayden SO2 nonattainment area (Hayden SO2 Plan), and proposed a limited approval and limited disapproval of the Asarco-Hayden SO2 smelter rule (Arizona Rule R18-2-B1302, "Rule B1302"). The EPA's proposed actions provided a 30-day public comment period. During this period, we received comments from ADEQ, ASARCO LLC ("Asarco"), and Freeport-McMoRan Incorporated (FMI).,, The comments from FMI pertain to Rule B1302 and the comments from ADEQ and Asarco pertain primarily to our proposed partial approval and partial disapproval of the Hayden SO2 Plan. Summaries of all comments and the EPA's responses are included below.

Comments from ADEQ

Comment: ADEQ's comment letter expresses concern that the EPA's proposed action does not clearly acknowledge the work that ADEQ and Asarco have completed since identifying the modeling error that was part of the basis for the EPA's proposed disapproval of the modeled attainment demonstration and related elements. ADEQ describes the modeling error that was discovered in 2017 after the SIP revision was submitted to the EPA and discusses the extensive work that was conducted to develop a revised modeling methodology. These efforts include additional analyses, work to justify new assumptions and modeling parameters, and the development of new modeling files and a modeling technical support document (TSD), draft versions of which were shared with EPA staff for review. ADEQ does not dispute the modeling error and acknowledges that the EPA was required to take action on the SIP revision submitted in March 2017. However, ADEQ expresses concern that the language in the EPA's proposal could lead the reader to believe that it knowingly submitted a SIP revision containing a flawed attainment demonstration, that the error was a recent discovery, or that it has taken no action to resolve the modeling issue. ADEQ contends that a clarification regarding the additional modeling efforts would help avoid any misunderstanding. Finally, ADEQ asserts that the new modeling methodology shows attainment of the NAAQS and that it was approved by the EPA in 2018. 

Response: We agree that extensive work has been done by ADEQ and Asarco, in consultation with EPA staff, to correct the flawed modeling in the March 2017 submittal. While we noted in our proposal that ADEQ has been working with Asarco and the EPA on revised modeling, we acknowledge that the high level of effort that has gone into that work was not clearly presented in our proposed action and the sequence of ADEQ submitting the SIP revision in March 2017, identifying the error later in 2017, and subsequently working extensively with Asarco and the EPA to correct the error was not discussed.

In response to the statement that the new methodology was approved by the EPA in 2018, we would like to clarify that, while ADEQ and Asarco consulted with EPA staff to revise the modeling, and has shared new modeling files and a modeling TSD with EPA staff, these documents have not undergone ADEQ public notice and comment or been formally submitted to the EPA as a SIP revision. Therefore, the revised modeling has not been formally approved by the EPA and was not evaluated as part of our proposed action. Only upon such future submission, if it occurs, will the EPA be able to formally evaluate and make a determination regarding its adequacy to demonstrate attainment of the 2010 SO2 NAAQS. 

Comments from Asarco

Comment: Asarco notes that it has spent considerable time and resources since 2011, in collaboration with ADEQ and the EPA, to achieve attainment of the 2010 SO2 NAAQS in the Hayden NAA. The commenter states that Asarco's efforts, including improvements to the capture and control systems, retrofits and rebalancing of the converter aisle to enhance sulfur recovery at the acid plant, and installation of an improved preheater system to reduce startup emissions, have resulted in SO2 emission reductions of approximately 90 percent relative to pre-2010 levels.

Response: The EPA acknowledges the efforts that Asarco has undertaken to reduce SO2 emissions and improve air quality in the Hayden SO2 NAA. A summary of the equipment and process upgrades that have been implemented was included in our proposed action, and a more detailed discussion was included in the TSD accompanying our proposed action on Rule B1302. 

Comment: Asarco asserts that the statement in the EPA's proposal that an error in ADEQ's modeling "changed predicted SO2 concentrations such that the modeling no longer shows attainment of the 2010 SO2 NAAQS" is disingenuous because ADEQ's revised modeling demonstration shows attainment of the 2010 SO2 NAAQS. Asarco believes that the accompanying footnote in the proposed action suggests that the modeling error was discovered in 2020, rather than in 2017, and suggests that the EPA should have acknowledged that ADEQ's revised modeling shows attainment even if the EPA felt compelled to act only on the submitted version of the plan.
Response: As discussed in our response to ADEQ's comments in Section II.A of this notice, the EPA does not dispute that the modeling error was discovered in 2017. We referenced the 2020 email in our proposed action because we did not have contemporaneous documentation of the discovery of the modeling error to cite in our proposal. We did not intend for our proposal to suggest that the modeling error was identified in 2020 and acknowledge the extensive work that has been done by ADEQ and Asarco to revise the modeling in the March 2017 SIP revision.

We also note that ADEQ and Asarco have informally sent draft revised modeling to EPA staff, who have provided feedback on the draft revised modeling. However, as previously noted, ADEQ has not yet released the revised modeling for public notice and comment or formally submitted the modeling to the EPA as a SIP revision. Accordingly, the EPA has not yet reviewed the revised modeling for approvability under the applicable requirements of the CAA and EPA regulations. 

Comment: Asarco asserts that under CAA section 172(c)(6), "other control measures, means or techniques" may be sufficient to achieve and demonstrate attainment of the NAAQS, and therefore, it does not agree that the Hayden SO2 Plan cannot be approved without numeric fugitive emissions limits. Asarco contends that the EPA improperly relied upon selective citation of the CAA and EPA regulations and non-binding guidance to conclude that a numeric fugitive emissions limit is required. Asarco lists the "other control measures, means or techniques" provided for in the Hayden SO2 Plan, which it asserts are sufficient "to achieve and demonstrate attainment of the 2010 SO2 NAAQS," including new and upgraded capture and control equipment, operation and maintenance plans for process and control equipment, numeric emissions limits on the main stack, a new preheater system to reduce startup emissions, work practice controls for fugitive emissions, and fugitive emissions studies to evaluate the efficacy of the improved gas capture and control equipment.

Response: We disagree with this comment. Section 172(c)(6) of the CAA requires attainment plans to include "enforceable emission limitations, and such other control measures, means or techniques" as necessary or appropriate to provide for attainment. The guidance documents we cited in our proposal (i.e., the General Preamble and the 2014 SO2 Guidance) describe and interpret CAA section 172(c)(6) and other binding statutory and regulatory requirements. While the guidance documents are not themselves binding, they guide the EPA's review of SIP submittals for compliance with the relevant requirements. In any case, the text of section 172(c)(6) is clear that the EPA must determine whether a submitted SIP includes all enforceable emission limitations and other measures that are necessary to provide for attainment. While measures other than emission limits might be sufficient by themselves in some circumstances (for example, where a particular source contributes little to the attainment problem or is not susceptible to a numeric limit due to technological limitations), such circumstances do not exist in this case, given that fugitive SO2 emissions at the Hayden facility have the potential to cause or contribute to NAAQS violations and are capable of being continuously monitored. 

The measures listed in Asarco's comment, while important components of the control strategy, do not ensure that fugitive emissions will remain at the level that was assumed in the attainment modeling. In particular, the installation of new and improved capture and control equipment was expected to reduce fugitive emissions, but, in the absence of ongoing monitoring, it is not known whether these changes were sufficient to reduce emissions to the level necessary to achieve attainment. Similarly, operation and maintenance requirements and work practice controls are helpful for ensuring that process and control equipment are properly operated, but they do not correspond to or assure achievement of any particular level of emissions. 

The fugitive emissions studies, the first of which began last year, will provide better information regarding the actual level of fugitive emissions from the facility. However, these studies will last for only one year each and do not correspond to any numeric emission limit. Therefore, if one of the studies were to show that fugitive emissions exceeded the levels assumed in the attainment modeling, this would not constitute a violation of an emissions limit that could give rise to an enforcement action. Rather, it would simply trigger a requirement for Asarco to conduct new modeling to assess whether the NAAQS would still be attained at the higher emissions levels. If that modeling shows an increased likelihood of a NAAQS exceedance, then Asarco would have to submit to ADEQ a proposed revision to its operations and maintenance plan and associated modeling to demonstrate attainment of the NAAQS. ADEQ would then submit revisions to the operational limits and volumetric flow monitoring provisions, and a revised attainment demonstration to the EPA as a SIP revision. 

There is substantial risk that fugitive emissions from the facility could cause or contribute to violations of the 2010 SO2 NAAQS. Consequently, the Plan must assure that these emissions are limited in an enforceable manner. A process for future evaluation of fugitive emissions and potential future SIP revisions contingent on the results of that evaluation cannot substitute for enforceable limitations on fugitive emissions. Moreover, if fugitive emissions were to increase during the period between the two studies or after the second study, there would be no mechanism to address those increased emissions. In contrast, if the Plan were to rely on enforceable numeric fugitive emissions limits corresponding to the modeled fugitive emissions levels, with ongoing monitoring, recordkeeping and reporting requirements, then an exceedance of any of these emissions levels would be a violation of the SIP that could result in an immediate enforcement action by ADEQ, the EPA, or a third party. Such an approach would satisfy the requirement of CAA section 172(c)(6) for enforceable limits and other measures that provide for attainment of the 2010 SO2 NAAQS.

Finally, Asarco lists the stack emission limits among the control measures that it believes are sufficient to demonstrate attainment. As discussed in our proposal, the stack emission limits would be enforceable were it not for the flaws in monitoring, recordkeeping and reporting requirements. In any case, the stack limits have no bearing on the SIP's flaw in not imposing an enforceable limit for fugitive SO2 emissions. 

For the foregoing reasons, we conclude that the requirements for enforceable limits and other measures that provide for attainment of the SO2 NAAQS under CAA section 172(c)(6) have not been satisfied.

Comment: Asarco reiterates its view that the EPA's proposal is dismissive of the progress that Asarco has made in reducing total SO2 emissions at the Hayden smelter, and that it implies that fugitive emissions controls at the smelter are inadequate. Asarco cites emissions reductions observed based on the initial data collected during the first fugitive emissions study to assert that fugitive emissions are well below what is needed to ensure attainment of the 2010 SO2 NAAQS.

Response: The EPA acknowledges the progress that has been made to reduce SO2 emissions at the Hayden smelter. As discussed in Asarco's comments and in the TSD accompanying our proposed action on Rule B1302, Asarco's SO2 control strategy includes several equipment and process upgrades, including replacement of the electrostatic precipitator and flash furnace with a new vent gas baghouse system; replacement of five 13-foot diameter converters with new 15-foot diameter units that operate more efficiently; installation of extended secondary and tertiary hooding in the converter aisle to maximize ventilation gas capture during charging, transfer, and tapping operations; and improvements to the acid plant with an upgraded pre-heater system. ADEQ has estimated that the converter retrofit project would reduce SO2 emissions from the smelter by 90 percent between 2011 and 2019.

With regards to the adequacy of the fugitive emissions controls, the EPA disagrees that there are sufficient data to conclude that fugitive emissions are below the level needed to ensure attainment. Asarco references emissions reductions based on initial data collected during the first fugitive emissions study, stating that "[u]nder the Plan, fugitive emissions fall from a maximum annual average of 295 pounds/hour to an average range between 4.3 and 39.8 pounds/hour." However, Asarco has not provided the hourly emissions data from specific roofline sources over an extended period that would be necessary to assess whether the recently monitored levels of fugitive emissions have been consistently at or below the levels necessary for attainment. Moreover, even if recent fugitive emissions have been below the modeled level, there is no assurance that these levels will be maintained over the long-term because, as described in the previous response, the Plan and Rule B1302 do not include any ongoing requirements to measure fugitive emissions or assure that these emissions remain low.

Comment: Regarding the EPA's position that Rule B1302 subsection (E)(4) "provides an option for alternative sampling points that could undermine the enforceability of the stack emission limit by providing undue flexibility to change sampling points without undergoing a SIP revision," the commenter states that the EPA's concern is not justified and lacks merit because the provision requires Asarco to demonstrate to ADEQ's satisfaction that the measurement "would yield inaccurate results or would be technologically infeasible" prior to using an alternative sampling point. Asarco asserts that it would be indefensible for the EPA to require inaccurate results be used to demonstrate attainment. Lastly, Asarco notes that it has recommended that ADEQ withdraw subsection (E)(4) because Asarco and ADEQ have agreed that the monitoring points are yielding acceptable results so this issue should be resolved upon ADEQ's submittal of a revised plan.

Response: The EPA disagrees that this issue lacks merit. The EPA is not suggesting that inaccurate sampling points be required to be used to demonstrate attainment, but rather that any change to sampling points should be the subject of EPA and public review through a SIP revision. As noted in our proposal, one of four basic principles that apply to all SIPs and control strategies is replicability, which means that "where a rule contains procedures for changing the rule, interpreting the rule, or determining compliance with the rule, the procedures are sufficiently specific and non-subjective such that two independent entities applying the procedures would obtain the same result." We find that the language in Rule B1302 subsection (E)(4) allowing for "measurement of the flow rate at an alternative sampling point" where the measurement in the outlet of the control equipment "would yield inaccurate results or would be technologically infeasible" is too general and subjective to ensure that two independent entities applying this standard would reach the same conclusion. For example, ADEQ might find that measurement of stack gas volumetric flow rate in the outlet of a particular piece of SO2 control equipment is technologically infeasible in a situation where the EPA might conclude that such measurement is feasible. Moreover, the rule does not specify any procedures or criteria for determining whether measurement at the alternative sampling point would yield accurate and representative results. Therefore, this provision of the rule is inconsistent with the principle of replicability.

As stated in the April 2019 Comment Letter conveying the EPA's comments to ADEQ regarding Rule B1302, the EPA agrees that withdrawal of subsection (E)(4) is appropriate and will resolve this issue, if such withdrawal occurs.

Comment: Asarco objects to the EPA's position that Rule B1302 subsection (E)(6) "allows for nearly 10 percent of total facility SO2 emissions annually to be exempt from continuous emissions monitoring systems; this deficiency could compromise the enforceability of the main stack emission limit." The commenter asserts that there is no deficiency and the basis for disapproval lacks merit because the provision to allow Asarco to petition ADEQ to replace the continuous emissions monitoring system (CEMS) with annual stack testing and report emissions rates as a pounds per hour (lb/hr) or pounds per ton production factor would still allow calculation of the emissions rates. Asarco states that there were legitimate concerns that it would not be able to perform a relative accuracy test audit (RATA) of the CEMS due to the low concentrations of SO2 present, but that it has now determined that it can perform a RATA of the relevant CEMS and has requested that ADEQ withdraw subsection (E)(6) in ADEQ's submittal of a revised plan to resolve this issue.

Response: The EPA disagrees that this issue lacks merit. While the rule language does provide for an emissions value that can allow for the calculation of an overall stack emissions rate, we do not consider this sufficient to ensure the enforceability of the one-hour main stack emissions limit given the large variability in hourly emissions from the Asarco facility. The commenter asserts that units encompassed by the provision typically emit less than 75 lb/hr SO2; however, we note that Asarco's emissions estimate for these units forecasts a maximum emission rate as high as 417 lb/hr SO2 (out of a total 1069.1 lb/hr or 1518 lb/hr main stack limit). In addition, we note that source test results represent a "snapshot" of unit emissions (and of corresponding unit operations) at the time of the source test. Generally, source tests must be performed at approximately 80 to 100 percent of maximum operating levels, and emissions limits relying upon a source test for demonstrating compliance typically require continuous monitoring of one or more parameters of unit operation. This allows for the determination that unit operations are representative of source test conditions and ensures the validity of the source test result. Rule B1302 subsection (E)(6), however, relies solely on source test results for demonstrating compliance, which we do not consider sufficient to ensure enforceability of the main stack emissions limit. As stated in our April 2019 Comment Letter, the EPA agrees that withdrawal of subsection (E)(6) is appropriate and will resolve this issue, if such withdrawal occurs.

Comment: Asarco objects to the EPA's position that Rule B1302 "lacks a method for measuring or calculating emissions from a shutdown ventilation flue; this omission could compromise the enforceability of the main stack emission limit." Asarco asserts that the concern is unfounded and lacks merit. Asarco explains the purpose of the shutdown ventilation flue and describes the procedure for calculating emissions for planned and unplanned shutdowns. Asarco notes that the procedure and resulting values are included in the SIP documentation but that to resolve the issue, it has requested that ADEQ revise the operation and maintenance plan requirements in the SIP to document the SO2 emitted during planned and unplanned use of the shutdown ventilation flue and require the use of the operation and maintenance plan value in compliance calculations.

Response: The EPA disagrees that the concern is unfounded and lacks merit. While the procedure for calculating emissions for planned and unplanned shutdowns and the value are included in supporting documentation for the Plan, they are not included in Rule B1302 or elsewhere in the SIP; therefore, they are not currently enforceable. 

Comment: Regarding the EPA's position that Rule B1302 "lacks a method for calculating hourly SO2 emissions," Asarco asserts that the calculation method is presented in subsections (F)(1) and (F)(2) and acknowledges that there was a typographical omission of the "valid hour" definition that was included in Arizona's submission. Asarco notes that it has submitted to ADEQ the same definition included in the EPA-approved plan for the 2010 SO2 NAAQS for the Miami, Arizona area and that Asarco has requested that ADEQ include it in a revised submittal to resolve the issue.

Response: The omission of the "valid hour" definition leads to ambiguity in how hourly emissions are calculated, thus undermining enforceability. However, the EPA agrees that inclusion of a "valid hour" definition will clarify the method for calculating hourly SO2 emissions for the Hayden facility and will resolve this issue, if submitted to the EPA in a future SIP revision.

Comment: The commenter states that Asarco is disappointed that the EPA has not evaluated a fundamental part of the Hayden SO2 control strategy  -  i.e., the "dual limit." Asarco discusses its rationale for the dual limit, states that there is no basis for the EPA to question it, and states that it is presumptively approvable under the EPA's SO2 Guidance.

Response: As noted in our proposal on Rule B1302, we are approving the main stack emission limit because it is more stringent than the existing requirements in state law, as well as new operational standards and monitoring, recordkeeping, and reporting requirements for the smelter.  However, as noted in our proposed action on the Hayden SO2 Plan, we are not evaluating its adequacy to ensure attainment of the 2010 SO2 NAAQS because (1) ADEQ has not demonstrated that the emission limits in Rule B1302 are sufficient to provide for attainment, and (2) the stack emission limit is not fully enforceable due to various deficiencies in Rule B1302.

Comment: Asarco states that it disagrees with the EPA's conclusion that the modeling in the Hayden SO2 Plan is flawed. It notes that the revised modeling that was informally submitted to EPA staff indicates that the Converter Retrofit Project meets the RACM/RACT requirements and that Asarco's understands that the revised modeling will be submitted to the EPA as a SIP revision.

Response: As discussed above, the EPA has not reviewed the revised modeling because, as Asarco acknowledges, it has not been formally submitted to the EPA as a SIP revision. The EPA's proposal to disapprove the RACM/RACT demonstration is based on the modeling that was submitted as part of the March 2017 SIP submittal. Both ADEQ and Asarco acknowledge the error in the modeling in the March 2017 submittal. The EPA will review any revised modeling upon formal submission of such modeling to the EPA as a SIP revision.

Comment: Asarco states that ADEQ intends to submit a SIP revision that includes updated modeling that shows attainment; removal of Rule B1302, Section (E)(4); removal of Rule B1302, Section (E)(6); a provision in the operation and maintenance plan to demonstrate the quantity of SO2 present during planned and unplanned use of the shutdown ventilation flue; and a "valid hour" definition that is the same as the definition in the approved Miami SO2 SIP. Asarco reiterates its position that the CAA does not require the Hayden SO2 SIP to include numeric fugitive emissions limits but notes that it is working with ADEQ to establish workable emissions limits and monitoring provisions for demonstrating compliance with such limits. Asarco also states that the submission of the SIP revision is imminent and recommends that the EPA prioritize action on the pending revised submittal rather than development of a new plan.

Response: As discussed above, the EPA disagrees with the commenter's assertion that the CAA does not require enforceable emissions limitations for fugitive emissions. Section 172(c)(6) of the Act requires attainment plans to include "enforceable emission limitations, and such other control measures, means or techniques" as necessary and appropriate to provide for attainment. With regards to the SIP revision that ADEQ and Asarco have been working on, the EPA will review the submittal for approvability under the applicable requirements of the CAA and EPA regulations once it has undergone ADEQ public notice and comment and been formally submitted to the EPA. While the EPA looks forward to reviewing the prospective submittal, the EPA must also fulfill its obligation under section 110(k) of the CAA to act on ADEQ's 2017 submittal.

Comments from FMI

Comment: FMI commented on transitional provisions in R18-2-715(I), R18-2-715.01(V), and R18-2-715.02(F). The commenter stated that these provisions are intended to clarify the applicability of current SIP-approved rules for the 1971 SO2 NAAQS in both the Hayden and Miami SO2 nonattainment areas, until the effective date of the rules for the 2010 SO2 NAAQS.

The commenter also noted that, when the EPA approved Arizona's attainment plan and new rules for the Miami SO2 nonattainment area, it did not act on all of the transitional provisions. Accordingly, the commenter explained that there is an inconsistency between the EPA's SIP-approved rules and ADEQ's rules (i.e., a "SIP gap"). Therefore, the Miami copper smelter must comply with both the old SIP rules for attaining the 1971 SO2 NAAQS and the new SIP-approved rules for attaining the revised 2010 SO2 NAAQS. The commenter asserted that "[t]his unintended consequence therefore subjects the copper smelter to an array of duplicative regulatory requirements that no longer serve any purpose." The commenter also noted that ADEQ has sought to remedy the SIP gap issue by submitting a request to withdraw A.A.C. R18-2-715(F)(2) and R18-2-715(H), which apply only to the copper smelter in the Miami SO2 nonattainment area, from the Arizona SIP.

FMI therefore requested that the EPA either amend its proposed action on Rule B1302 to include a proposed approval of Arizona's revisions to A.A.C. R18-2-715 and R18-2-715.01, or to propose such approval in a separate, but concurrent action. The commenter stated that, by doing so, the EPA could "avoid having any period with a SIP gap by taking simultaneous final action on A.A.C R18-2-B1302 and R18-2-715 and R18-2-715.01" and "allow the existing SIP rules for attaining the 1971 SO2 NAAQS to be properly subsumed by the newly approved SIP rules for attaining the revised 2010 SO2 NAAQS." The commenter asserted that such an approach would be consistent with the EPA's efforts to implement Executive Order 13771, "Reducing Regulation and Controlling Regulatory Costs."

Response: The commenter is correct that the EPA has not yet proposed to act on the transitional provisions in A.A.C. R18-2-715(I) and R18-2-715.01(V). As we noted in the TSD for our proposed action on Rule B1302 for the Hayden area, in order to act on the revisions to 715 and 715.01, "we need to evaluate the effect of sunsetting the existing SIP-approved requirements of those rules in conjunction with the new requirements for the Hayden smelter established in Rule B1302." In conducting this evaluation, we explained that:
      
      . . . Rule B1302 does not include a numeric fugitive emission limit, whereas Rule 715 subsection (G) includes an annual average fugitive limit of 295 lb/hr. 
      
      In order to ensure that the existing fugitive limit of 295 lb/hr remains in the SIP, we are not acting on the revisions to Rule 715 at this time. Similarly, we are not acting on Rule 715.01, which includes requirements for SO2 compliance determination and monitoring that support the enforceability of the emission limits and requirements in Rule 715.
      
In other words, approval of R18-2-715(I) and R18-2-715.01(V) at this time would result in the removal of the existing SIP-approved fugitive emission limit and associated compliance requirements for the Hayden Smelter without a new fugitive emissions limit to replace it. For the reasons described in our proposed action on Rule B1302, as well as our proposed and final actions on the Hayden SO2 Plan and the associated responses to comments, in the absence of a fugitive emissions limit, the Plan does not provide for the attainment of the 2010 SO2 NAAQS. Therefore, an action to approve R18-2-715(I) and R18-2-715.01(V) -- and thereby remove the existing fugitive emissions limit from the SIP without replacement -- would interfere with attainment of the 1971 and 2010 SO2 NAAQS. Such an action would be impermissible under CAA section 110(l), which prohibits the EPA from approving any SIP revision that would interfere with applicable requirements concerning attainment and reasonable further progress or any other applicable CAA requirement. Therefore, we have not proposed to approve the transitional provisions in R18-2-715(I) and R18-2-715.01(V). 

We acknowledge that our inability to approve these provisions has resulted in a SIP gap and that that the requirements in Rule 715 that apply to the Miami smelter are now duplicative of the requirements in SIP-approved rule A.A.C. R18-2-C1302. However, because the transitional provisions that apply to Hayden and Miami are inseverable from one another (i.e., both are contained within a single paragraph within R18-2-715(I) and R18-2-715.01(V)), we cannot separately approve the transitional provisions for Miami without also approving the provisions for Hayden, which is prohibited by CAA section 110(l). 

On March 10, 2020, the EPA received a submittal from ADEQ seeking to withdraw A.A.C. R18-2-715(F)(2) and R18-2-715(H), which apply only to the Miami SO2 nonattainment area, from the Arizona SIP. As noted by the commenter, approval of this SIP revision would remedy the SIP gap issue for the Miami area. We intend to act on this submittal in a separate rulemaking, as it is outside of the scope of this action, which concerns only Rule B1302. 


