Date: December 13, 1999 To: John Spotila Administrator, Office of Information and Regulatory Affairs, Office of Management and Budget (OMB) From: Jere Glover, Chief Counsel for Advocacy, Small Business Administration (SBA) Subject: Secondary Aluminum We strongly urge that EPA provide appropriate regulatory relief to the aluminum die casters and foundries. To our knowledge, there is no legal requirement to include these small area sources within the scope of this rule. The agency has no data on dioxin/furans from these industries, and specifically chose not to include these facilities in the original testing to acquire data for this rulemaking. I am also particularly troubled by the lack of the opportunity to comment on critical issues, such as the failure to elicit public comment on dioxin/furans emissions data in the national emissions inventories for these industries, and the basis of emissions estimates, that are now being relied on by EPA after the comment period has long expired. If the agency has a concern about dioxin and furan emissions, but yet has no data from these facilities to support this regulation of upwards of 1800 facilities, another more cost-effective manner of addressing this concern should be adopted. It is Advocacy's recommendation that EPA should pursue one of the two options set out in the accompanying staff memorandum. These options exclude the appropriate die casters and foundries by use of a new definition of "secondary aluminum" operations, or by deferring action on these area sources pending additional testing and analysis. Enclosure Date: December 13, 1999 To: Art Fraas, Office of Management and Budget From: Kevin Bromberg Subject: Secondary Aluminum Air Toxics Standard (MACT) I. Die Casters and Foundries Generally Should be Excluded from Secondary Aluminum MACT This draft final rule requires aluminum die casters and foundries to spend between $25 to $120 million in testing costs to determine whether emissions of dioxins/furans (D/F) warrant control. Since at least July 1998, EPA declined to collect D/F emissions data from these industries despite the knowledge that customer returns would place up to 1850 facilities within the scope of this rule. The agency proposes to include these facilities within the secondary aluminum smelter category if they accept customer returns. However, the expected emissions reductions are very low, even by EPA estimates. In the December 1998 Prevention of Significant Deterioration guidance, EPA declared that "aluminum die casting facilities typically need not be considered secondary metal production plants." We suggest two options for addressing this issue: (1) excluding die casters and foundries with "clean customer returns," or (2) deferring die caster and foundry area sources from the final rule, pending further testing and examination of the emissions. II. EPA Has Not Collected Dioxin/Furans Data For Die Casters and Foundries; The EPA Emissions Inventory Does Not Identify Die Casters and Foundries as Sources of D/F Emissions EPA's only expressed justification for including these sources is the area source release of a significant amount of D/Fs. However, this rulemaking never discussed D/F releases from aluminum die casters and foundries, and such data is missing from all the EPA published air toxics inventories (where D/F emissions are included). The data now cited by EPA was not referenced in the preamble to this rulemaking (although the agency asserts that it is now found in this secondary aluminum MACT rulemaking record). Furthermore, the agency does not expect much D/F emissions reductions from these facilities, despite the imposition of significant testing costs. EPA's own draft final economic analysis reveals no emission reductions from these facilities, only substantial costs. The July 19, 1999 urban air toxics Federal Register notice includes all the industrial categories used to meet the 90% air toxics requirement: table 4-3 (in a related rulemaking document) indicates only 23 secondary aluminum foundries, which presumably includes only the secondary aluminum smelters. To our knowledge, EPA has not indicated that aluminum die casters or foundries were a significant source of D/Fs in any rulemaking document. No die casters or foundries were tested for emissions as part of this rulemaking. When NADCA requested D/F data from EPA in July, it was told only chip dryer data was available for their industry, and this data was never provided directly to NADCA (it was provided recently by SBA when it received the data.) [Industry now concedes that chip dryer operations (a fairly atypical process for these facilities) could be appropriately included in MACT.] Industry was not informed in the secondary aluminum proposal, or by EPA at a later date, that it had any other data relevant to D/F emissions for these two industries. Industry has not been able to respond to this new data now produced by EPA on Thursday from a secondary aluminum smelter. Although the agency staff now claims to have sent the emissions data relevant to die casters and foundries to the EPA staff involved in compiling the emission inventory records, three Federal Register notices, between 1997 and 1999, showed no data or emissions estimates relating to these facilities - only the 23 secondary aluminum smelters, accounting for approximately 4% of the D/F emissions. The public had two opportunities to review and comment on the draft inventories, and no one, to our knowledge indicated that die casters and foundries were missing from the D/F inventories. Considering the large size of the affected universe, it is appropriate to seek emissions data before imposing regulations. III. EPA Has The Discretion to Defer Regulation of Die Caster/Foundry Area Sources Under the Air Toxics Area Source Strategy Furthermore, even if die casters and foundries were appropriately part of the original definition of the secondary aluminum category, the July 1999 urban air toxics strategy Federal Register notice makes it quite clear that EPA has the authority to redefine the category to conform to the facilities "subject to regulation." 64 Fed. Reg. 38720. "Under section 112(e)(4), listing of a particular source category isn't considered final agency action until EPA issues emissions standards for that source category." Id. at 38721. For example, if EPA found that die caster and foundry and sweat furnaces area sources, under certain conditions, do not have significant D/F emissions, it could delete them from the regulation and the category of secondary aluminum operations. In the July 1999 notice, EPA indicates that it has only selected at this time new area sources for regulation that might clearly be needed to achieve the requirement to address 90% of the area sources of air toxics including D/F, specifically using the figure of 15% of total area source emissions. In this case, EPA's latest estimate is that these facilities would account for approximately 1% of total emissions (not now in the draft economic analysis). Thus, consistent with the new announced strategy, if these facilities were being examined for the first time now, EPA would not even be recommending this category for regulation, but would await future analysis with new and more refined emissions inventories. Given the published categories being used to reach the 90% standard in July, the low emissions estimate, the uncertainty surrounding this estimate, no direct emissions data, and other factors such as cost, EPA has a strong basis for concluding that such sources may be excluded from this category at this time, or permanently, if the agency deems it appropriate. IV. Proposed Solutions: Based on the most current economic impact analysis, EPA's proposed resolution is to require one time testing for the affected facilities (EPA estimate $26,000; industry estimate $50,000 - $120,000) for up to 1850 facilities to determine if the D/F standard is exceeded, with the expectation that all such facilities would pass the test. We suggest that an expenditure that could exceed $100 million is not the best approach, when no testing of even a sample of such facilities has yet occurred, and EPA, until one week ago, expected that all those facilities would remain unregulated after testing. (EPA more recently speculated that 30% of these facilities may require to do some pollution prevention or make other adjustments, in response to testing, but the basis for this estimate is unknown.) We propose two constructive ways to handle this issue. EPA could defer regulation of the die caster and foundry area sources, pending additional testing and analyses. It is not appropriate to regulate hundreds of facilities without any emissions testing. The industry commenters support the option of testing their facilities, if EPA chooses to further examine this industry. At that time, EPA can determine whether the die caster and foundry D/F emissions should constitute part of the 90% regulation of air toxics. The agency could also conclude that die casters and foundries would become part of the secondary aluminum category with its own standards. EPA can announce that it will defer a decision on this subcategory, pending the development of new data, and permitting the industry to opportunity to comment on this data. Secondly, EPA could redefine "clean charge" to include customer returns (which are generally very clean) to exclude customer returns with visible contamination or without trace contamination with specific numerical standards for excluding operations with excessive fluxes and amounts of returns. NADCA/AFS has suggested a 1% flux restriction, which is copied from the December 1998 EPA Prevention of Significant Deterioration guidance. Although EPA claims it was unable to make such a distinction for post-consumer waste, there is no evidence that this problem exists for customer returns. My understanding that such returns have only been machined, and are very lightly contaminated, if at all. This is a straight-forward solution that would resolve the issue without requiring the expenditure of future EPA resources.