Document K6VMdk1nzaXazeBJ54aKGDQqK

V t Tuesday December 29, 1992 Part II Environmental Protection Agency 40 CFR Part 63 National Emission Standards for Hazardous Air Pollutants; Compliance Extensions for Early Reductions; Final Rule SL 063368 61970 Federal Register / Vol. 57, No. 250 / Tuesday, December 29, 1992 / Rules and Regulations ENVIRONMENTAL PROTECTION AGENCY 40 CPR Pert 63 (AD-fRL-4534-3] National Emission Standards for Hazardous Air Pollutants; Compliance Extansions for Early Reductions agency: Environmental Protection Agency (EPA). ACTION: Final rule. SUMMARY: Regulations governing compliance extensions for early reductions of hazardous air pollutants (HAP's) were proposed in the Federal Register on June 13,1991. This action promulgates these regulations and implements the provisions of section 112(i)(5) of the Clean Air Act (CAA) (as amended in 1990). This rule establishes requirements and procedures for source owners or operators to obtain compliance extensions from section 112(d) standards and for reviewing agencies to follow in evaluating requests for extensions. dates: Effective Date. December 1992. Judicial Review. Under si 307(b)(1) of the CAA. judicial rjew of the actions taken by this notio available only by the filing of Aetition for review in the U.5. Court ^Rppeals for the District of Columbia (meuit within 60 days of today's ptAication of this rule. Under section 3oA)(2) of the CAA, the requirements thaAre the subject of today's notice am/ not be challenged later in civil qyriminal proceedings brought by Aa to enforce these requirements. ADDRESSES: Backgroun 'reformation Document. The bai d information document (BID) for romulgated standards may be ob ed from the U.S. EPA Library (MD-3 search Triangle Park, North Carolini 7711, telephone number (919) 541--j r77. Please refer to "National Emissio; Standards for Hazardous Air Po, itants: Compliance Extensions for Reductions-- Background Lnfi ation for Promulgated St lards" (EPA--450/3- 92-006b). The contains (1) a summary of al^ e public comments made on the posed standards and EPA's respoi s to the comments; and (2) a summ< of the changes made to the standai since proposal. Also available the EPA Library are three additional pporting documents. These documen (a) ling Document for Regulati, s Governing Compliance Extensii s for Early Reductions of Hazardous Air Pollutants" (EPA-450/ 112(d) standard by achieving early i 91r-013. July 1991); emission reductions of HAP's. Thai (b) "Questions and Answers about compliance extension would be Early Reductions Program" (EPA~45j obtained and authorized in acconnee 3-92-005, January 1992): and with section 112(i)(5) of the CAA^^ (c) "Procedures for Establishing Section 112(i)(5j allows an exiiig Emissions for Early Reduction source to be granted a 6-year extJslon Compliance Extensions" (EPA--45 of compliance with otherwise 91-012a, February 1992). applicable section 112(d) standi Docket Docket No, A-90-47, upon demonstration by the owft or containing supporting informatiqfused operator of the source that a 9oArcent in developing the promulgated reduction in HAP's (95 percenAr more standards, is available for publii in the case of particulates) haAeen inspection and copying * 3:30 achieved. An enforceable alterative a.m. and 3.30 p.m., Monday emission limitation reflectinAie Friday, at EPA's Air Docket reduction will be establishedKr the Waterside Mall, room M-15< source by a permit issued utgei Title V Floor, 401M Street SW,, Wi of the CAA. DC 20460. A reasonable fee The 90 (95) percent emisAn charged for copying. reduction must be achieveA>efore FOR FURTHER INFORMATION cUnkCV. Tot information concerning thiAarly Reductions Program regulAms, contact Mr. David Beck, telephonw919) 541- 5421, or Mr. Richard ColA, telephone (919) 541-5262. The addfta for both is Emission Standards DivAon (MD-13), U.S. Environmental PrnMiinn Agency, Research Triangle ParkAlorth Carolina, 27711. For informatioimonceming Method 301, contact kA Tony Wayne, telephone (919) 541-W78, Emission Measurement BranchAechnical Support Division (Kw-19), U.S. Environmental ProtAion Agency, Research Triangle North Carolina 27711. SUPPLEMENTARY I ATION: The information pr led in this preamble is organized as I bws: proposal of an applicable Ation 112(d) standard in most cases. Hwever, a source achieving the reduAon after proposal of a standard btAefore January 1,1994. may quAfy for a compliance extension fjAi section 112(d) standards by maAng an enforceable commitmeAbefore proposal of the standaAto achieve such a reduction. State or local air poMtion control agencies with Title VAnnitting authority will revieuAitle V operating permit applications A compliance extensions. They wiMssue permits containing altematiA emission limitations applicajA to sources achieving early redAtions in lieu of section 112(d) staAards in accordance with approved peAit programs under Title V of the CAMThe EPA Regional l. General Requirewits n. Summary of Clfiges and Impacts of the Final Rule m. Supporting DBuments IV. Public Puticflation V. Significant cAunants and Changes to the Proposed Me Offices will ravisA permit applications 'and issue Title lAermits for sources located in StateAiat do not have approved Title Awrmit programs. States will assufle the review and permitting actiAties upon approval of A Dennitlonf Source their Title V pAnit programs, or may be 8. Base YetAmiiiions C fiifim naMi Commitments D. AllowatAEmission Reductions E. ...........AiIIiiii of Early Reductions granted delegAon of the Early Reductions PAgram prior to approval of their Title V Armit program. F. High-RH Pollutants G. Stats Mhority H. Intertm With Tide V Permits I. InterfaB With Section 112(g) ModiAations ILSu #f Changes and Impact! of the Final Rd Moments received during : period on the proposed J. InterHs With Title I Provisions ificant changes have been K. TesMethods end Procedures L OtiAChange* to the Proposed Relation VI. AdiBiistrative Requirements a. r B. PBerwork Reductions Act C. Acutive Order 12291 D. flgulatory Flexibility Act [ regulation. The significant i listed below, t significant change to the an to the high-risk list of Seventeen pollutants have ' 1 to the list and five i have been deleted. Weighting aeral Requirements seven pollutants have been A provision was added to day's rule applies to any existing at there are no increases in that seeks a 6-year compliance (elides as a result of the emission Knsion from an applicable section (on demonstration. Another SL 063369 61980 Federal Register / Vol. 57. No. 250 / Tuesday, December 29, 1992 / Rules and Regulations F. High-Risk Pollutants 1. Summary * Section ll2(i)(5)(E) specifies that with respect to pollutants for which high risks of adverse public health effects may be associated with exposure to small quantities, the Administrator shall limit the use of offsetting reductions in emissions of other HAP's from the source as counting towards the 90 (95) percent reduction in such high-risk pollutants. In the final rule, the number of pollutants on the high-risk list is 47, rather then 35 as proposed. The additional pollutants are listed in Table 1. Table 1. List of Pollutants Aooed to High-Risk List CAS No. Chemical 93863 *09974 ..... 334883..... 96128 -----70447 1220S7 151564 ..... Flhyim toiwtt (AlttMna) ... 77474 ____ 0............... UArtQftrueu mmfvuvvte .................... ..... 80344....... 0 ........ . 684935 ___ 62759 ....... 56362 7803512 .... D*vw*w* ............... .... ... ............................................... ,,............................ ,, 7723140 .... 6001352 .... ............. ......... .................... ................ ...... . .................................... tear 100 100 10 10 100 10 100 10 10 10 10 1000 100 10 10 10 100 These pollutants have been added to the list since proposal for a number of reasons including changes to the methodology EPA used to select the pollutants and updated health effects information. Five pollutants which were proposed for the high-risk list have been deleted. These are benzotrichloride; chloroprene; l,l,2,2-tetrachldroetbane; 2,4-toluene diisocyanate; and vinylidene chloride. The final list of hign-risk pollutants and their weighting factors are shown in Table 2. Table 2. List of High-Risk Pollutants CAS No. Chemical W*Mgnt-V` l-Q- facnr 53963 ....... 107028 _ 79061 79107 .... 197131 ..... 0 .......... 1332214 ... 71432 92875 0 ......... 542861 ..... 106990 ..... 0 ......... 57749 ....... 532274 ..... 0 ......... 107302 ... 0 ___ _ 334883 .. 132649 ... 96128 .... 111444 .... 79447 ...... 2-Acety<amlnonuotane AcnMm Acvic acid... - . _.. AcnSonitnie...........--..... Arterw compounds___ Aioaetos.........--...,,. BBsw-cymouomfomcoamtnoyot)uentdhesr_.. 1 3-Butedwne....... OCedommiumnc.o.m..p.o..u.n.d.s........ 2-0*o'oacMopfteflone ~ confounds ... Ch<o"3i*ietnyl PW. methyl CDoieieoonv'eetnnaenne.S.-w.-.-o-n-e--_- 1,2-l>tnomo-3cNoropfOpene. DeNotoetnyt (Bia(2- edw cMomethyOeSwi). Dimettycaftiamoyt cNo- noe. 100 100 10 10 10 100 100 10 1000 10 1000 10 10 100 100 100 10 10 10 10 10 10 100 Table 2. List of High-Risk Pollutants--Continued CAS No. Chemlcel taaor 122887 108934 ..... 151564 ..... 75218 -------76448 -------118741 _ 12-Diphaiyhydrazlne -- Ethylene dtoomtde........EttyanMne (Aaridtne) . Ethylene oNde_____ -- BeptacNw. -- HeeacMorabenisne___ 302012 -- 0 --, 0 101688 .... 60344 -------624838 __ 0 62759 _____ 684836 -- HWB _____ 75445 -------7803512 _ 7723140 73658 .-- 1748016 8001362 ... diene. HyUnune___________ Manganeee compounds. Mercury compounds___ Methylene dphenyl MPC)in>V /(MunUitt). Mesy hyosane-------------- MaShS Isocyanate Nfckei oomoounds_______ N-Naroaodewsthytemins . kl || -- in 4.. -- IvflWlWFTnnBnjWw .................................. Photfpnt ..*.*.**.***.... Photphint............................. Phoeohotue_______________ 1,2-Prwyanim<ne_______ 2,3,74-Tc*ehwo<*- barao-p-anm Tonphene (cModneted 75014 ...... Vky eriohde ................. 10 10 100 10 100 100 100 10 100 10 10 10 10 100 1000 10 10 10 10 100 100.000 100 10 The specific changes made to the methodology and the rationale for specific pollutants being added to or deleted from the list are described below, 2. Methodology for Selecting Pollutants for the High-Risk List The methodology used to formulate the list of high-risk pollutants, the list itself, and the weighted offsetting scheme are a direct response to the mandate in section U2(i)(5)(E) of the Act. The high-risk list and the weighted offsetting scheme may not be applicable to, or appropriate for, other sections in Title m of the Act Other provisions of section 112, such as establishing lesser quantity emission rates under section 112(a) and identifying the relative hazards to human health from emissions of each of the listed hazardous air pollutants under 112(g) may also require the ranking of pollutants. The approach for identifying the high-risk pollutants may or may not be found to be appropriate for these or other provisions. The selection of today's approach for the purposes of section 112(i)(5)(E) is not intended to establish a precedent for other provisions or preclude the consideration of other alternative methodologies. As presented in the preamble to the proposed rule, EPA used a number of criteria for selecting pollutants for the high-risk list. First, as specified by the CAA, chlorinated dioxins and fonns (as listed in section 112(b)) were included on the high-risk list Second, pollutants classified by EPA as Group A carcinogens (known human carcinogens) were judged to be of sufficient concern to be listed as highrisk. Third, as a way to systematically screen the remaining HAP's listed in section 112(b), EPA employed a three tiered screening analysis for selecting high-risk pollutants. In the first tier, health effects data were examined to rank pollutants based solely on health effects. The health SI* 063370 Federal Register / Vol. 57, No. 250 / Tuesday, December 29, 1992 / Rules and Regulations 61981 endpoints considered were The EPA agrees with the commenters' emission release and exposure scenarios carcinogenicity, reproductive and concerns regarding use of the TRI and across many varied source categories are developmental toxicity, acute lethality, has consequently dropped tier 3 from anticipated under the Early Reductions and systemic effects other than acute the analysis. The EPA also agrees that Program. lethality (e.g., neurologic disorders). Consideration of health effects alone, chemicals not currently in common use The HEM was used to estimate a should still be listed as high risk in case theoretical downwind concentration of however, was judged Insufficient to production of that chemical is resumed a typical HAP. Because the pollutants address section 112(i)(5)(E) which at a later date. The EPA now believes are released from various types of describes high-risk pollutants as those that tiers 1 and 2 adequately identify sources with variable stack parameters for which high risks of adverse public those pollutants that could reasonably (stack height, emission velocity, health effects mey be associated with be expected to adversely affect public distance to nearest residence, etc.) there exposure to small quantities. health and thus limited the offsetting of was no attempt made to model site- Accordingly, tier 2 of the analysis was these pollutants in the Early Reductions specific conditions. Meteorological data exposure modeling using EPA's Human Program. The impact of dropping tier 3 used were representative of between 2 Exposure Model (HEM). The HEM is that the following pollutants have and 10 years of data from the selected provided an estimate of the ambient concentration 500 meters from a source emitting 10 tons per year at an average height of 10 meters under median meteorological conditions. As described in the preamble to the proposed rule, the ambient concentration predicted by the HEM was compared to selected risk levels to screen for pollutants likely to have an adverse effect at the exposure level that was modeled. For carcinogens, the modeled ambient concentration was compared to the concentration of each carcinogen that would result in an increase in cancer risk of one in ten thousand to the most exposed individual For noncarcinogens, the modeled ambient concentration was compared to either (1) a level one order of magnitude greater than a verified inhalation reference concentration or oral reference dose, (2) a lowest observed effect level (LOEL) divided by an uncertainty factor of loo, or (3) the dose or concentration of a chemical that causes death in 50 percent of the exposed population (LD*> or LC*o) divided by an uncertainty factor of 1000. Tier 3, which considered actual pollutant emissions as reported to the TR1, was intended to determine whether the pollutants were actually emitted at levels that could reasonably be expected to adversely affect public health. Pollutants with a low exposure potential were consequently eliminated from the high-risk list. Several commenters stated that it was inappropriate to eliminate pollutants been added to the high-risk list: 2acetylaminofluorene; 1,2 diphenyl hydrazine; toxaphene; nickel compounds; N-nitrosodimethylamine; ethylene imine (aziridine); dimethylcarbamoyl chloride; phosphorus and diazomethane. Many commenters recommended changes to the parameters chosen for the tier 2 exposure analysis. Several commenters felt that the exposure assumptions were not conservative for several reasons including the meteorological data, the stack parameters, and the limitations of the HEM. One commenter stated that one year of meteorological data is not representative and that downwash was not considered. Two commenters suggested a shorter stack height of 3.5 meters and a 20 meter distance to the nearest residence. Another commenter recommended that the modeling use more realistic assumptions about the pollutants and emission sources. One commenter contended that the HEM is too simplistic because it does not consider actual exposure, short-term exposures or population activity patterns. Other commenters made suggestions which would make the analysis less conservative. Two commenters fait that the 500 meter distance to the nearest residence was too close. One of these commenters suggested a distance of 3000 meters and a 10 year population residence time (rather than the 70 year lifetime exposure associated with the cancer potency factor). The second tier of the screening meteorological station. A10 meter stack height was chosen to represent a chemical plant. For the proposed rule, the modeling parameters also assumed a 70-year continuous exposure and a 500 meter distance to the nearest residence. Upon re-evaluation, EPA has made changes to these two parameters. First, the assumption that exposure to a pollutant occurs continuously over a 70-year period has been changed to 33 years for the purposes of this rulemaking. Thirty-three years is a duration representative of the 95th percentile for the number of years an individual would remain at the same residence, based on data on population mobility and mortality. The 33-year duration was selected for use in this analysis as representative of a reasonable worst case approach to assessing exposure duration. Second, EPA reconsidered the choice of 500 meters to the nearest residence and has revised the analysis using a 200 meter distance. The 200 meter distance to the nearest residence is the distance that has been most often used by EPA for analyses estimating exposure to emissions from point sources. This distance represents a reasonable worst case for point source and is judged to represent a reasonable worst case for facilities that are anticipated to participate in the Early Reductions Program. The net impact of these changes was the addition of hexachlorocyclopantadiene. phosphine, panthion, and manganese compounds to the list. Four commenters disagreed with the from the high-risk list based on analysis to select the high-risk one in ten thousand presumptive risk emissions data. The commenters pollutants was a generic exposure benchmark for carcinogens, stating that contended that use of the TRI is modeling exercise. The EPA intended it was not conservative enough. One inappropriate because not all pollutants that the modeling results be used to commenter recommended that EPA are reported, only large facilities in the determine which pollutants merited include every carcinogen with a potency manufacturing sector report to the further analysis based on exposure factor greater than 1.35 x 10-4 (pg/m3)~` inventory, and not all types of source potential. The EPA did not use the based on modeling a 3.5 meter stack. categories are covered. The commenters modeling results to estimate actual risk Another commenter argued that felt that EPA should not eliminate levels, but did use the modeling results Congress intended that sources with risk substances because they are not to differentiate between pollutants. The levels of one in one million be currently in common use and that the EPA still considers the use of generic controlled. In contrast, one commenter tier 3 screen should be dropped from modeling parameters appropriate for recommended the benchmark be the analysis. this analysis because a wide range of lowered to one in one thousand. This SL 063371 f 61982 Federal Register / Vol. 57, No. 250 / Tuesday, December 29, 1992 / Rules and Regulations commenter believes that the one in ten source's emissions pose a risk greater in developing an inhalation reference thousand risk benchmark is inconsistent than one in one million after application concentration. For this analysis, EPA with the "Savings Provision" in section of control technology (as stipulated in made a judgment that if a source's 112(q)(l), where standards promulgated section 112(d)). The commenter is emissions wera such that a LOEL or under section 112 in effect prior to. the reminded that the Early Reductions LDjo could potentially be exceeded 1990 amendments are to remain in lores Program provides an extension of the (after application of the uncertainty unless revised by EPA. compliance time for applying control factors and under the modeling scenario The exposure modeling exercise technology that is stipulatedoy used) then the pollutant could be conducted as pari of tier 2 of the standards promulgated under section considered high risk. In this application, screening analysis was used as a tool to 112(d). After the compliance time EPA believes uncertainty factors of 100 identify pollutants that could extension, facilities participating in the and 1000 to be appropriate. Similarly, potentially cause a high risk to public Early Reductions Program must still with respect to using a value one order health (under the exposure scenario that apply the required tedmolosy if their of magnitude above the reference was modeled). The EPA has decided emissions exceed specified levels. In concentration, EPA believes this to he that for carcinogens a one in ten addition, facilities that participate in the an appropriate estimate of a level at thousand presumptive risk level is Early Reductions Program are not which health effects could potentially appropriate for this analysis. This exempted burn a residual risk test under be significant enough to be regarded as decision is based primarily on guidance section 112(f), so the commenter's high risk. concerning acceptable risk discussed in concern that the intent of Congress is Two commentera strongly urged EPA the benzene NESHAP promulgated on not being fulfilled is unwarranted. to confine the selection ofcarcinogens September 14.1969 (54 FR 38044). In With respect to the Saving* Provision for the high-risk list to consideration of the preamble to the benzene decision, of section H2(q)(l) the commeuter is carcinogenic potency and not make use the use of the one in ten thousand reminded that today's rule is not an of the weight-of-evidence classification. benchmark is described--"EPA will emission standard, but procedures for Three commenters objected to the consider the extent of the estimated risk obtaining an Early Reductions addition of four Group A carcinogens to were an individual exposed to the compliance extension. The use of the the list after the pollutants had been maximum level of pollutant for a presumptive risk benchmark of one in excluded by the 3-tiered process. On the lifetime. The EPA will generally ten thousand in the screening analysis other hand, two other commenters presume that if the risk to that for selecting high-risk pollutants has no recommended that EPA prioritize the individual la no higher than bearing on NESHAP that have already high-risk list using a qualitative rather approximately one in ten thousand, that been promulgated. than quantitative approach (i.e., use risk level is considered acceptable and Two commenters questioned the weight-of-evidence reviews for EPA then considers the other health and benchmarks chosen for haalth effects Identifying substances rather than risk factors to complete an overall other than cancer. These conunentere potency factors). judgment on acceptability. The felt that neither the LOEL divided by a The EPA believes it appropriate for presumptive level provides e safety factor of 100 nor the LD divided identification ofhigh-risk pollutants to benchmark for judging the acceptability by a safety factor of 1000 were consider both the weight of evidence of maximum individual risk, but does conservative enough. The commenters and carcinogenic potency ofa pollutant not constitute a rigid line for "^ng also thought that using a value one The EPA stands by the decision to list that determination." In the case of the order ofmagnitude above the reference Group A carcinogens as high risk based screening analysis for high-risk concentration was not conservative on the feet that these are known human pollutants under the Early Reductions enough. carcinogens. The EPA judged that provisions, EPA believes the one in tan As stated in the preamble to the emissions ofknown human carcinogens thousand presumptive risk level has proposed rule, uncertainty factors were are of sufficient concern that use of been used appropriately. The applied to the health affects benchmarks offsetting reductions in other less conunentere are reminded that the since the LOEL and LD values hazardous pollutants under an Early generic exposure analysis was not represent levels at which health affects Reduction demonstration should be meant to estimate the actual risk bom are known to occur. Thee* uncertainty limited. Three of the pollutants any one type of source category, but factors were meant to account for (benzidine. his(chlonMn*tbyi)sther. and rather was used to differentiate between variables such as intenpadea variations chloromethyl methyl ether) ere pollutants baaed mi relative toxidtie*. and sensitive subpopulations, but do extremely potent carcinogens and ware The EPA recognizes that other not incorporate all the factors typically excluded from tbs 3-tiered process at modeling assumptions could have been used in developing reference tier 3. This means that, according to the used that would nave resulted in either concentrations. The uncertainty factors 1989 TR1, these pollutants wets not more or fewer pollutants exceeding the used to develop an inhalation reference emitted by any reporting facility in presumptive risk level. However, EPA concentration are typically applied to e quantities sufficient to exceed the risk does not agree with the couunentar's lowest observed adverse effect level benchmark level proposed by EPA. in suggestion that every carcinogen with a (LOAEL) or no observed adverse effects spite of this, EPA felt it prudent to list potency factor greater than 1.35 x icr* level (NOAEL). The resulting reference them as high risk in case a source's (jjg/m3)-1 should be considered high concentration represents a level at emissions changed in the future. As risk. which the potential far advene public discussed above, since proposal of the With respect to the comment that health effects is considered to be high-risk list, EPA has decided to Congress intended that sources posing negligible. However, the purpose of this eliminate tier 3 from the screening risks of one in one million be analysis was to identify pollutants analysis. As a result, even if EPA nad controlled, the commenter is referring to whose emissions could potentially decided not to use weight of evidence section H2(f)(A), the so-called "residual present a high risk, not a de minimis as a criterion, these carcinogens would risk" provisions. In this provision it is risk. Therefore, the uncertainty factors still be listed as high-risk pollutants. stipulated that additional emission used in this analysis were not meant to Benzene was also included on the list standards must be promulgated if a account for all the factors typically used even though the risk benchmark level SL 063372 Federal Register / Vol. 57, No. 250 / Tuesday, December 29, 1992 / Rules and Regulations 61983 was not exceeded when a 10 tons per year emission rate was modeled. The rationale for this decision was that in addition to being a known human carcinogen, benzene is a very high volume chemical and emissions in excess of 10 tons per year are not uncommon. The EPA believes that offsetting emissions of this known human carcinogen with less hazardous pollutants should also be limited. Two commenters suggested that EPA revise the selection criteria to include only known human carcinogens (Group A), probable human carcinogens (Group Bl), and noncarcinogens for which chronic human health effects can be expected to occur at extremely low levels of exposure. In selecting pollutants for the highrisk list, EPA considered weight-of evidence, potency and exposure potential of the pollutant. As described above, a weight-of-evidence classification of Group A (known human carcinogen) was deemed sufficient to list a pollutant as high-risk. The EPA does not agree with tire commenters suggestion that Group B2 and Group C carcinogens should not be considered for or included on the highrisk list In order to be listed, however, carcinogens other than Group A had to pass tier 2 of the analysis which considered potency and exposure potential. The EPA also disagrees with the commenters suggestion that only chronic human health effects should be considered. Section 112(b) of the CAA describes HAPs as thorn pollutants which present, or may present, a threat of adverse human health effects including substances which are carcinogenic, mutagenic, teratogenic, neurotoxic, which cause reproductive dysfunction, or which are acutely or chronically toxic. Thus, EPA believed it appropriate to assess (and include on the high-risk list where appropriate) pollutants that are acutely as well as chronically toxic. 3. Weighted Index for High-Risk Pollutants In the proposed rule, EPA described an indexed offsetting system based on the toxicity of the high-risk pollutants relative to each other. Several commenters believe that the weighted index system conflicts with the CAA because it allows offsetting of high-risk pollutants with non-high risk pollutants. The commenters state that high-risk pollutants should not be allowed to be offset by non-high risk pollutants, or only be allowed when high-risk pollutants are emitted in trace amounts. Other commenters agreed with the weighting factor concept and the with different health endpoints was ability to offset high-risk pollutants. restricted, this could potentially exclude In the final rule, a number of some facilities from participating in the weighting factors have been adjusted as Early Reductions Program. The EPA described below, but the weighting feels that such restrictions would not factor system as s whole is essentially benefit the goals of the Program as a unchanged. Contrary to several whole. commenters contentions, the high-risk Several comment letters made pollutant strategy does not conflict with reference to the development of section 112(i)(5)fE). That provision weighting factors for the high-risk requires the Administrator to limit, by pollutants. One commenter regulation, the use of offsetting recommended the weighting factors be reductions of other HAP*t in counting based on weight of evidence towards the 90 (93) per cent reduction classification. On the other hand, three of high-risk pollutants. The statute does commenters recommended the not say or even imply that the regulation weighting factors be based only on must prohibit such offsetting potency as described by the cancer Kreductions. In response to the comment that these icy factor. Another commenter er suggested that the cancer offsets should only be allowed whan potency factor be multiplied by one these pollutants are emitted in trace million to arrive at the weighting factor. amounts, EPA notes that the weighting Finally, one commenter suggested a system effectively restricts the offsetting combination approach where both of emissions of the high-risk pollutants. weight-of-evidence and potency would Such a large reduction of a non-high be considered. In this approach, risk pollutant is needed to offset anlgh- pollutants with the amn weight-of- riak pollutant (up to 100,000 to 1) that evidence classification and having only relatively trace amounts could be potency factors within an order of traded. Even in situations where the magnitude would be grouped together. pollutant concentrations are relatively The weighting factor system has been large (i.e., not trace amounts), it la revised slightly since the high-risk list appropriate to allow offsetting as long as was proposed, hi the final rule, the weighting factor system is followed. carcinogens, with a Group C If one ton of a HAP with a weighting classification (possible human factor of 100 is offset with a 10 ton carcinogens) have been assigned a lower reduction in a pollutant with a weight. Greater uncertainty exists weighting factor of 10. the offset regarding the evidence for a Group C generally should provide a similar classification than that exiating for degree of risk reduction. Consequently, chemicals classified es Group A (known EPA believes such offsets should not be human carcinogen) or B (prooably prohibited. Note that these relative carcinogenic to humans). A Group C trades deal with decreases in the high- classification is defined by positive risk pollutants. It is not envisioned that carcinogenicity in a single experiment, any successful Early Reduction a tumor response of marginal statistical submittal will include any increase in significance, or finding benign tumors high-risk pollutants other than only. A Group B classification, whereas, incidental trace amounts. is defined by carcinogenicity in two or Several commenters stated that more animal species, strains, or trading among pollutants with different experiments, either with or without types of health effects is inappropriate. limited human evidence. Group A is The commenters recommend using reserved for known human carcinogens. critical health effects and toxic potency The greater uncertainty regarding a information to place the listed Group C classification is reflected in s pollutants into categories, with trading lower weighting factor. As a result, two allowed within a category. Trading Group C carcinogens (l ,1,2,2- across categories would be allowed only tetrachloroethane and vinylidene for categories baaed on similar health chloride) have been assigned weighting effects. Weighting factors based on factors of 1 rather than 10 as proposed. relative potencies would be used in The EPA does not feel, however, that such tradeoffs. chemicals classified in Group B should The development of the weighting necessarily be weighted lower than system and the decision to allow trading those classified as Group A. With between carcinogens and respect to Group B chemicals, a sound noncardnogens is intended to provide foundation exists regarding flexibility to the participating facilities carcinogenicity in animals, but the in achieving their emission reduction human data are either inconclusive or of goals. As such, it is a policy decision limited value, most likely reflecting the and not one based purely on scientific difficulty of obtaining quality grounds. If tradingW.ween pollutants epidemiologic data. For this weighting SL 063373 61984 Federal Register / Vol. 57, No. 250 / Tuesday, December 29, 1992 / Rules and Regulations system then, chemicals classified as Group A or B are combined and ranked by potency. For carcinogens, the actual weighting factors are based on the potency of the high-risk carcinogens relative to carcinogens not on the high-risk list. This approach is the same as was proposal Multiplying the potency factors by one million as one commenter suggested to calculate weighting factors would result in the same relative ranking of the pollutants, but higher weighting factors. The EPA believes that basing the weighting factors on the differences between the potency of the high-risk carcinogens and the geometric mean potency of the group of carcinogens not on the high-risk list is more justifiable. Three commenters questioned EPA's policy decision concerning the weighting factors for noncardnogens. The commenters believe that EPA must assess many additional factors when ranking the hazards of noncardnogens. These factors include the severity of effect, its reversibility, and chemical properties such as the halMifa of volatile compounds in air, vapor ressure, persistence, and ioaccumulation potential. In assigning weighting factors to the high-risk pollutants, EPA attempted to bam the factors on the relative toxidty of the compounds. For the cardnogens, the cancer potency factor is a straightforward measure of relative toxidty and was used in conjunction with the weight of evidence classification to develop the weighting fadors. For the noncardnogens, however, there is not a comparable measure of toxidty that can be used consistently for pollutants with different health effects. In the absence of such a measure, EPA proposed to assign a weighting fador of 10 to the noncardnogens on the high-risk list. Many of the cardnogens were also assigned a weighting fador of 10. The EPA chose a weighting fador of 10 for the noncardnongens in recognition that noncancer health effects canhe as seriously debilitating as cancer. Another alternative would have been to assign these pollutants a weighting fador of 1 because other indices were not available. The EPA rejeded this option because only by assigning a weighting factor greater than 1 is offsetting limited between these noncarcinogens and other pollutants not on the high-risk list. In the final rule, three noncardnogens were assigned weighting fadors greater than 10. The rationale for increasing the weighting fador from 10 to 100 for mercury is based on consideration of persistence in the environment and bioaccumulation, as discussed below. After a review of the sdentific data, two other noncardnogens (acrolein and 2* chloroacetophenone) have been assigned weighting fadors of 100 rather than 10 in the final rule to provide an adequate margin of safety from adverse health effects. This decision is explained fully in a memorandum to the docket entitled "Need for Additional Weighting of Two Chemicals". Four commenters believe EPA should establish more weighting fador categories to refled more accurately the different toxidty levels of the various pollutants. On the other hand, two commenters believe more pollutants should be grouped together so that there is not such a wide range in weighting fadors. The EPA has reviewed the weighting fador categories and still finds four separate categories to be appropriate. As described in the comment above concerning weighting fadors there were some adjustments made to the placement of certain pollutants. However, the general categories with weighting fadors of 100,000; 1000; 100; and 10 are still believed to be appropriate. The difference between the most toxic carcinogen and the least toxic carcinogen with a weighting fador of 10 is about a fador of 35 (with the exception of one pollutant). Given the uncertainties inherent in the health data, EPA does not feel that this is enough of a difference to warrant another weighting fador category. Another commenter suggested that the weighting fador couldbe customized for each facility by relating the weighting fador to the degree to which a reference concentration is exceeded. The commenter believed that this approach would emphasize reductions in cases where reference concentrations may be exceeded and assigns less weight to high-risk pollutants in cases where the reference concentrations are not likely to be exceeded. There are several reasons why EPA prefers that weighting fadors not be customized for each participating facility. The primary reason is that the weighting fadors reflect the differences in the toxidty of the pollutants regardless of emission rates. For example, if fadlity X emits 50 tons of benzidine, a 1000 to 1 trade is necessary with a HAP not on the high-risk list. If fadlity Y emits 1 ton of benzidine, a 1000 to 1 offset is also required. The EPA believes this will emphasize reductions in the high-risk pollutants even if these emissions are small. In addition, participation in the Early Reductions Program is available to all fadlities throughout the Nation. The EPA prefers that for consistency in program implementation the weighting fadors for the high-risk pollutants not be customized for each participating fadlity. In addition, sucn customizing can greatly complicate the application preparation and review process: fadlities and reviewing agendas may not have the expertise or resources to either perform or review site-specific modeling analyses and determine appropriate weighting fadors. Four commenters discussed pollutants that persist in the environment and/or bioaccumulate. These commenters recommended that any pollutants identified as persistent or bioaccumulative be added to the list The commenters further suggested that an additional weighting fador of 10 be applied to these pollutants. After consideration of public comments, EPA has dedded to adjust the weighting fadors for chlordane, heptachlor, hexschlorobenzene, mercury compounds, and toxapbene from 10 to 100 based on persistence in the environment and bioaccumulation. The weighting fadors for these particular pollutants were increased based on persistence and bioaccumulation data compiled for EPA in support of the CAA Great Waters Study. 4. Additiona/Deletians to the High-Risk List In addition to the changes to the highrisk list described above, EPA added four pollutants to the list and deleted three pollutants on the basis of health effects data. Methyl hydrazine, 2chloroacetophenone, and 1,2-dibromo- 3-chloropropane were added baaed qn newly available or revised RICs. Nnitroso-N-methlyurea was added because a unit risk estimate is available in a health assessment document Benzotrichloride was deleted from the list because an inhalation potency factor was unavailable. Further review of the studies supporting the listing of chloroprane led to the determination that the primary study was not acceptable and, therefore, chloroprane was removed from the list The RfC for 2,4-toluene diisocyanate is undergoing review based on new data and, therefore, 2,4-toluene diisocyanate was removed from the list Two commenters stated that when pollutants are added to the list, a source should not be exempt from further reducing those pollutants, but the source should be given up to three years to adjust its emissions to continue to qualify for the Early Reduction extension or else comply with the emissions standard under section SL 063374  Federal Register / Vol. 57, No. 250 / Tuesday, December 29, 1992 / Rules and Regulations 61985 112(d). Other commenter* believe the health efleets. As stated above, three reduction as a condition for grantin changBs to the high-risk pollutant list chemicals were deleted from the list extension. As long as EPA (adminis should not affect the status of a because of health effects data and two through the Region) remains the previously submitted commitment or were assigned weighting factors of one permitting authority, the CAA sp< application. because of their Group C cancer 90 (95) percent reduction. The EPA has decided that sources classification. All other requests to Finally, two commenters conle with an approved enforceable delete pollutants from the list were that States should have the optioi commitment, or an approved permit denied. The rationale supporting the allowing fatilities in their States specifying an alternative emission limit, specific determinations are addressed in partidpate in the Program or de ing will not be affected when a HAP is the BID. either added to the list in section 112(b), or is newly designated as a high-risk C. State Authority alternative amission limits. The Early Reductions Progra national program mandated byj pollutant. The EPA expects to update A number of commenters through the CAA and as such, the list of high-risk pollutants as the recommended that States be deli not have the authority to disa science requires. It is conceivable that a authority for implementing the partidpation. However, Stab source would have to revise a post- Reductions Program even bea!foreJ the power to impose additi reduction demonstration with each are given permitting authority ler stricter State standards or revision of the list. This requirement title V of the CAA. than 90 (95) percent reducti could add significantly to The EPA specifically solidti condition of granting an Ea administrative burden for both the comment on whether this Reductions compliance e> sion under source and the reviewing agency. Given should be delegated to the Based their title V State permitti rogram that the source will be in full compliance with the emission standards under section 112(d) at the end of the six year extension, EPA feels this additional administrative burden is unnecessary. 5. Comments on Specific High-Risk Pollutants on the comments received, proceeding to establish the delegating the Early Reducti to thoee States that seek ' advance of having a title V permit program. Section CAA authorizes the Adm approve a State program implementation ana enfi t of the H. Interface With Tide V its One commenter was imed that the Early Reductions nil ired compliance extensions t in the form of a title V permit. Sim re is a chance that the regulai *.for operating permits will not be ` until after the first set of section Id) standards The EPA received one comment omission standards and i are proposed, the coi iter believed asking why lead was not listed as a requirements of the sectii including that it may be too lati apply for a high-risk pollutant. partial delegation of the That compliance extensii ail sources Airborne lead emissions are currently is, EPA anticipates that , if a state covered by the seed 112(d) standards, regulated by a National Ambient Air so desires, delegate to itethe The regulation alii s sources covered Quality Standard (NAAQS). The EPA is authority to develop implement the by standards proj prior to January currently reviewing the data the Early Reductions The EPA I.1994, to submit enforceable NAAQS is based on and expects to anticipates that it blish commitment prior proposal of the make a decision on revising the delegation guidance ie near future standard. The is then allowed standard in the near future. Given this, that will be useful States in until January 1,1 , to achieve the and other policy considerations developing programi ir submittal, reductions and submit a permit regarding identifying a criteria pollutant Of course, until time as a State application aa la is December 1,1993. as high risk, at this time EPA believes has an approved St; program under Sources which' already achieved it is appropriate to leave lead title V, it cannot ii a title V permit. the required ion prior to proposal compounds off the high-risk list. See CAA section: Therefore, of an applicabli on 112(d) standard One commenter urged EPA to add the Administrator tot delegate his must submit a lit application radionuclides to the high-risk list. authority to issue it establishing containing the ;uction demonstration The EPA did not addradionuclides to an alternative on limit under prior to such iposal; or if a Federal or the high-risk list because radionuclide emissions are measured in terms of activity rather than masa and it would section 112(i)(5) il such time as the state has an appi ed title V permit program. State permitti] program is not yet in place, the it application for the Early Redui is source may be be extremely difficult to equate the two Various com bts were received for the purpose of offsetting. The EPA pertaining to a te's authority to submitted u o 120 days after the permitting ' ority has established a recognizeshowever, that radionuclides impose stride could potentially be present in trace compliance e: intents for i than those permitting m under title V. In the latter situat >n, even though the permit amounts from some combustion sources. To account for this, language specified in The CAA rule. applicatioi ould have to be submitted s in section 112(i)(5)(A) after prop] 1. the source owner or has been added to the final rule that stipulates that if a radionuclide source that "Nothim this paragraph shall preclude a from requiring operator st still document that the required iy reductions were achieved is included in the emissions pool, EPA will not allow increases in radionuclide emissions under any post-reduction scenario. Numerous commenters requested that specific pollutants be removed from the high-risk list. Most of the commenters took issue with the scientific basis of reductions xcess of those spedfied in this at sgraph as a condition of granting ixtension * *. Although not sped Illy stated, it is implied that the exce: ductions can only be required a condition of the State granting extension. The CAA, thereto mplies that when the State is prior to iposal of the applicable section (d) standard. In the event that sui 'documentation cannot be providi r, the source's Early Reductions dem ition would be disallowed and the so' would have to meet the sectioi 12(d) standard, It become apparent recently that EPA potency factors or other specifics of the pe: ting authority, the State may prehensive title V permitting the scientific studies that document the require ater than 90 (95) percent Inism, either a federal rule for SL 063375