Document N22ypEbM21n59x6aOoNonaKpQ

4. EPA's Action Level Is Technically Flawed and, Therefore, Arbitrary If the Proposed Rule was adopted in its current form, regulated entities would be required to conduct root cause analysis and corrective action upon exceeding an entirely arbitrary annual average concentration "action level" for benzene of 3ug/m3. As EPA explains, the action level does not correlate to any particular metric related to risk. 88 Fed. Reg. at 55886. Rather, the action level is based on modeling fenceline benzene concentrations using benzene emissions inventories in the facility wide risk assessment, and assuming that the reported emissions represented full compliance with all standards, adjusted for additional proposed control requirements. Id. at 55885. For purposes of its modeling analysis, EPA assumed that the nearest cuff--silt polar grid receptor was the best representation of each facility's fenceline concentration, unless there was a census block centroid nearer to the fenceline than the nearest off-site polar grid receptor, or an actual receptor was identified from review of the site map. Id. at 55887. In other words, the action level was derived based on modeled offsite concentrations of benzene rather than the benzene concentrations actually detected at the facility fenceline. Id. Only receptors that were estimated to be outside the facility fenceline were considered in determining the maximum benzene level for each facility, which also corresponded with the maximum benzene concentration modeled at the fenceline. And, although EPA acknowledged that "[d]uc to differences in short-term meteorological conditions, short-terni (i.e., 2-week average) concentrations at the fenceline can vary greatly," EPA did not adjust the action level to account for this "great" variability. Id. EPA, without any support, assumed that a 2-week measurement period would take care of the "great" variability. Id. More importantly, EPA's "modeling exercise" appears to entirely ignore the actual sampling data which showed that the "[a]verage benzene Ac concentrations ranged from 0.1 to 40 ugim3 . . . at the fencelinc."H As a result, some facilities would have to perform root cause investigations and take corrective actions as soon as the rule goes into effect, even though EPA concluded that monitoring is unnecessary based on the technology review. Coke Facilities Method DTE/EES USS Clairton Fencehne Concentration, Hg/m3" ABC SunCoke-Haverhill I 325A&B Naphthalene) Benzene 2 Naphthalene Benzene eo Naphthalenes Benzene 15 Naphthalene Benzene o: 325A (six ror'hsj lb - O :5A 0 2. 28 uW D 2 06 0O1 -o 5A - O :3A intedor Concenrrntion,lag/m3`" 25 25 78 339 3 34 NA 12 67 t. sDi"erenie the between r-easureu value arc hatkground torcentration C' uelta C' ) s* Average duals all time periods CC Burns Harbor Naphthalene Benzene 3 o 4 142 1% Notably, the lowest benzene emissions were from SunCoke's Haverhill facility, an HNR at 0.1 u2/m3. In contrast, the benzene emissions detected at ByP facilities were 90 to 4,000% higher than the levels detected at SunCoke's I Iaverhill facility. Yet EPA did not remark upon this discrepancy nor attempt to distinguish between 13y1) and HNR facilities in determining the need for monitoring. EPA-IIQ-OAR-2003-0051, EPA, ltemorrtmitem, Fugitive Monitoring et, Coke Oven Feteililies, at 11 (Jul 1, 2023) 42 Sierra Club FOIA 2025-EPA-04883 ED_018388_00000109-00043 SC_EVERSPLIT0005717