Document 916w5z6vnkGXpOG2aNz7Qg5K5

1. EPA's failure to sufficiently lower the Method 23 sampling volume for mobile hot cars will cause safety and operational issues for SunCoke. In the Proposed Rule, EPA proposed a minimum sample volume for Method 23 (PAI I) of 140 dscf (dry standard cubic foot). SunCoke explained that the proposed sample volume was too high for the pushing tests.8' While EPA agreed that a small volume allows a source to demonstrate compliance based upon the detection levels for pushing, EPA only lowered the sample volume to 105 dscf, which is not sufficient to address SunCoke's concerns associated with the length of time required for testing, specifically for the pushing process." EPA also disagreed with SunCoke's evaluation of the amount of time required for testing and stated that "the longest duration of any test run during the CAA section 114 request was 21 hours."87 But EPA clearly misunderstood SunCoke's comments because they state in their response "We also disagree with the commenter that a month would be required to complete three test runs "88 SunCoke did not state that three test runs will require a month of testing. A clarification of SunCoke's comments and additional context is provided below. Because pushing is a brief and intermittent activity that lasts between only two and four minutes for each oven, sampling must be conducted during the push of each oven on as many ovens as is necessary to achieve the specified sample volume. This means that to collect a sample volume of 105 dscf, each test run for PAI I would have to span two days, as demonstrated in the 2017 ICR testing where for each test run of PAII.s/DFs, approximately 90 dscf was collected on the first day of testing and the remaining volume (approximately 55 dscf) was collected on the second day of testing. Therefore, three test runs of PAH at the reduced volume of 105 dscf will still require six days of testing on the hot car. To this, we need to add nine days of testing for the other HAP (three days for Hg, three days for AG, and three days for I ICN). If there are no production delays or testing issues, this is 15 straight days of testing on each hot car. This doesn't even consider the existing ongoing permit requirement to test PM on the hot car (derived from the existing MAC-I. rule) every 2.5 years, which adds an extra three days of testing on each hot car. Therefore, it is very reasonable to expect that testing could take three weeks for each hot car based on the requirements in the Final Rule. Ultimately, extended testing periods not only raise personnel safety concerns, but also serve to disrupt critical aspects of production and maintenance. SunCoke requests that EPA lower the Method 23 sample volume to 90 ciscf or less to remedy these legitimate safety and operational concerns. J. EPA did not respond to SunCoke's comments on the definition of "acceptable makeup water." EPA did not respond to SunCoke's comments on the definition of "acceptable makeup water" in 40 C.F.R. 63.7352.89 Specifically, SunCoke requested that EPA also allow the use of "non-process wastewaters," as defined in 40 C.F.R. 420.02(r), as acceptable makeup water for quenching. SunCoke comment letter, p. 35. Response to Comments, Response 11-10, p. 185 f'i Id Icl SunCoke comment letter, Aft II, pp. l-2 - 23 - Sierra Club FOIA 2025-EPA-04883 ED_018388_00000109-00083 SC_EVERSPLIT0005757