                                   July 2023


Summary of Public Comments and Responses for New Source Performance Standards Review for Steel Plants: Electric Arc Furnaces Constructed After 10/21/74 & On or Before 8/17/83; Standards of Performance for Steel Plants: Electric Arc Furnaces & Argon-Oxygen Decarburization Constructed After 8/17/83And On Or Before May 16, 2022; Standards of Performance for Steel Plants: Electric Arc Furnaces and Argon-Oxygen Decarburization Vessels Constructed After May 16, 2022
                                       
                                       
                                       
                     U.S. Environmental Protection Agency
                 Office of Air Quality Planning and Standards
                Sector Policies and Programs Division (D243-02)
                 Research Triangle Park, North Carolina 27711

   The primary contact regarding questions or comments on this document is:

                       Donna Lee Jones jones.donnalee@epa.gov
                        U.S. Environmental Protection Agency
                OAQPS/Sector Policies and Programs Division (D243-02)
                    Research Triangle Park, North Carolina 27711
                               TABLE OF CONTENTS

ABBREVIATIONS AND ACRONYMS	iv
PREFACE	vi
LIST OF COMMENTERS	vii
1.0	40 CFR PART 60, SUBPART AAb LB/TON PM STANDARD	1
2.0	40 CFR PART 60, SUBPART AAb BSER FOR MELT SHOP OPACITY	28
3.0	REVISIONS TO 40 CFR PART 60, SUBPARTS AA AND AAa	46
4.0	MONITORING AND COMPLIANCE	59
4.1 Monitoring Requirements in Proposed 40 CFR Part 60, Subpart AAb	59
4.2 Bag Leak Detection System (BLDS)	68
4.3 Building Inspection Requirement in Proposed Subparts AA, AAa, and AAb	71
4.4 Calculation for Zero Percent Opacity Under Method 9	73
4.5 ASTM D7520-16 Digital Camera Opacity Technique	74
4.6 EPA Method 22 (Visible Emissions)	78
5.0	ECONOMIC IMPACTS	80
6.0	MISCELLANEOUS	82
6.1 Definitions	82
6.2 Startup, Shutdown, and Malfunction	83
6.3 Response to the EPA's Solicitation for Information on EAF NSPS Facility List for 2022	85
6.4 Environmental Justice	85
6.5 Integrated 15-minute Block Pressure Averages in 40 CFR Part 60, Subparts AA and AAa	87
6.6 Criteria for Exempting Capture System Operations Ducted to a Control Device	88
6.7 Procedural--Comment Period Extension	88
6.8 Adoption of Other Commenter's Comments	90
6.9 Regulation of Other NSPS Pollutants, Major Sources of HAP, and Greenhouse Gases	90
6.10 Reissuing of Rule	93





	
                          LIST OF FIGURES AND TABLES


Figure 1-1. Type of steel produced at facilities in three datasets: 13 facilities with 0 percent opacity EPA/EAF dataset, 30 total facilities EPA/EAF dataset, and 61 facilities in 2017 Industry (SMA) dataset.	8
Figure 1-2. EAF baghouse data: lb PM/ton steel, by baghouse type.	10
Figure 1-3. Corrected total facility gr/dscf vs. previous calculation	16
Figure 1-4. Total facility and individual baghouse PM, gr/dscf vs lb/ton.	18
Figure 1-5. Comparison of total facility gr/dscf vs lb/ton for 13 facilities with 0 percent opacity vs. all 30 facilities in EPA data set.	19
Figure 1-6. Range of EAF facility total baghouse lb PM/ton steel, by steel measurement type.	22


Table 1-1. Type of Emissions Controlled in the 50 Baghouses with lb/ton Values in the EPA EAF Data	9
Table 1-2. Total Facility Baghouse Emissions (lb/ton)	19
Table 1-3. Summary of Steel Production Measurement Points in the	21
Table 1-4. Melt Shop Baghouse Emissions and Production Rates at CMC in Cayce, SC	22
Table 1-5. Review of 18 Permits for Cost Information  -  Alphabetical By State	26




ABBREVIATIONS AND ACRONYMS

A/C		air-to-cloth ratio
AHSS		advanced high strength steel 
AISI		American Iron and Steel Institute
AOD		argon-oxygen decarburization
AVG		average 
BH		baghouse
BOPF		basic oxygen process furnace 
BACT		best available control technology
BLDS 		bag leak detection systems
BID 		background information document
BPT 		benefits per ton
BSER 		best system of emissions reduction
CO2e		carbon dioxide equivalent
CO		carbon monoxide
CAA 		Clean Air Act
CFD		complex computational fluid dynamic
CFR 		Code of Federal Regulations
CBI		confidential business information
COMS		continuous opacity monitoring systems
DCOT		digital camera opacity technique 
DRI		direct reduction ironmaking
DEC 		direct shell evacuation control
dscf 		dry standard cubic feet
dscm 		dry standard cubic meters
EAF 		electric arc furnace
EIA 		economic impact assessment
EJ 		environmental justice
EPA 		Environmental Protection Agency
EO		executive order
FR		federal register
ft		foot/feet
°F 		degrees Fahrenheit
g 		grams
GACT		generally available control technology
gr 		grains
gr/dscf		grains per dry standard cubic feet
GHG		greenhouse gas
HAP		hazardous air pollutants
II&S 		integrated iron and steel industry
LMF		ladle metallurgy furnace
LMS		ladle metallurgy station
LAER		Lowest Achievable Emission Rate
OSHA		Occupational Safety and Health Administration
lb 		pounds
lb/ton		pounds per ton
MACT		maximum achievable control technology
mg 		milligram
mg/dscm	milligrams per dry standard cubic meters
NAAQS	National Ambient Air Quality Standards
NESHAP	national emission standards for hazardous air pollutants
NGO		nongovernment organizations
NSPS 		new source performance standards
NOx		nitrogen oxides
O&M 		operating and maintenance
OAQPS 	Office of Air Quality Planning and Standards
OMB 		Office of Management and Budget
p. 		page
PM 		particulate matter
PM2.5 		particulate matter less than 2.5 micrometers
PRA 		Paperwork Reduction Act
PSD		Prevention of Significant Deterioration
RACT		Reasonably Available Control Technology
SO2		sulfur dioxide 
SSINA		Specialty Steel Industry of North America
SMA 		Steel Manufacturers Association
tpy 		tons per year
U.S. 		United States
U.S.C. 		United States Code
USD		United States Dollar 

                                    PREFACE

      On May 16, 2022, the U.S. Environmental Protection Agency (EPA) proposed new source performance standards (NSPS) for "Electric Arc Furnaces and Argon-Oxygen Decarburization Vessels Constructed After May 16, 2022" (40 CFR part 60, subpart AAb) from a review of the current NSPS for electric arc furnaces (EAF) and argon-oxygen decarburization vessels constructed after August 17, 1983 and on or before May 16, 2022 (40 CFR part 60, subpart AAa). The new subpart (40 CFR part 60, subpart AAb) being finalized reflects the comments received during the public comment period, which are summarized here along with the EPA responses. 
      
      The EPA also is finalizing amendments to update both the 1974 NSPS for EAF, titled "Standards of Performance for Steel Plants: Electric Arc Furnaces Constructed After October 21, 1974, and On or Before August 17, 1983" (40 CFR part 6, subpart AA) and the above-mentioned "Electric Arc Furnaces and Argon-Oxygen Decarburization Vessels Constructed After August 17, 1983 and On or Before May 16, 2022 (40 CFR part 60, subpart AAa)
      
      The EPA's responses to significant comments appear in the preamble to the final rule ONLY and are not repeated in this document although the comments themselves are included here. Copies of all comments submitted are available at the EPA Docket Center Public Reading Room. Comment letters are also available electronically through http://www.regulations.gov by searching Docket ID No. EPA-HQ-OAR-2002-0049.

                              LIST OF COMMENTERS

The EPA received 11 comment letters after May 16, 2022, on the proposed new source performance standards (NSPS) for "Electric Arc Furnaces and Argon-Oxygen Decarburization Vessels Constructed After May 16, 2022" (40 CFR part 60, subpart AAb) and proposed revisions to Standards of Performance for Steel Plants: Electric Arc Furnaces Constructed After October 21, 1974, and On or Before August 17, 1983; and Standards of Performance for Steel Plants: Electric Arc Furnaces and Argon-Oxygen Decarburization Vessels constructed after August 17, 1983 and on or before May 16, 2022. All comment letters are contained in Docket ID No. EPA-HQ-OAR-2002-0049. The docket item identification (ID) number and commenter name and affiliation are listed below. The comment letters are identified in this document by the last four digits of their docket ID item number for convenience in identification.

    Public Comments Received on Proposal by Commenter Name and Affiliation
                             Docket ID Item Number
                     Shorthand Commenter Affiliation Name
                                  Affiliation
EPA-HQ-OAR-2002-0049-0088
Citizen 
Citizen
EPA-HQ-OAR-2002-0049-0091
State 
Wisconsin Department of Natural Resources (WIDNR)
EPA-HQ-OAR-2002-0049-0093
Steel Co. 
Nucor, Inc
EPA-HQ-OAR-2002-0049-0094
Nongovernment organization (NGO)
GASP, Fairfield Environmental Justice Alliance, Group Against Smog and Pollution, Sierra Club, and Our Children's Earth Foundation
EPA-HQ-OAR-2002-0049-0095
Steel Co. 
U.S. Steel Corporation
EPA-HQ-OAR-2002-0049-0096
Steel Co. 
Cleveland Cliffs, Inc.
EPA-HQ-OAR-2002-0049-0097
Industry Trades 
American Iron and Steel Institute (AISI), the Steel Manufacturers Association (SMA), and the Specialty Steel Industry of North America (SSINA)
EPA-HQ-OAR-2002-0049-0098
Steel Co. 
SSAB, Inc.


        Public Comments Received Requesting Extension of Comment Period
                       by Commenter Name and Affiliation
                             Docket ID Item Number
                     Shorthand Commenter Affiliation Name
                                  Affiliation
EPA-HQ-OAR-2002-0049-0086
Industry Trades
AISI, SMA, and SSINA
EPA-HQ-OAR-2002-0049-0087
NGO
Greater Birmingham Alliance to Stop Pollution (GASP)
EPA-HQ-OAR-2002-0049-0089
Steel Co.
U.S. Steel Corporation

1.0	40 CFR PART 60, SUBPART AAb LB/TON PM STANDARD


Comment 1-1: The form of the standard should not be changed from the original Title 40 of the Code of Federal Regulations (CFR) part 60, subpart AAa rulemaking.

Commenters 0097-Industry Trades and 0095-Steel Co. stated for the 40 CFR part 60, subpart AAa [original] rulemaking, the EPA rejected a production-based (or mass-based) standard in favor of the 40 CFR part 60, subpart AAa concentration-based limit: 
 
      A process weight format is based on a direct relationship between the quantity of pollutant emitted and the amount of input material consumed or product produced. Because of wide differences between EAF and AOD shops in operating procedures, such as the length of the steel production cycle, grade of steel produced, control technologies, vessel capacities, and other operating parameters, a simple direct relationship between mass emissions and steel production does not exist. Therefore, a process weight format was not selected for control devices regulated by the proposed standards. 
       
      Methodology to measure the concentration of emissions discharged to the atmosphere from control devices is readily available and well demonstrated. Concentration measurements are obtained directly from the stack emission test data. A concentration standard can be met equally well by a large or a small shop and by carbon and specialty steel shops. Consequently, a concentration format (i.e., mass emissions per unit volume of gas) was selected for control devices regulated by the proposed standards to ensure control of captured process and fugitive emissions. 

(48 FR at 37347)
 
The commenters continued that the EPA provides no explanation for the change in its position and fails to address the rationale the Agency provided in 1983 for adopting the current grain-loading standards in 40 CFR part 60, subparts AA, AAa. 
 
Commenter 0095-Steel Co. added that EPA's failure to justify this "depart[ure] from a prior policy" would render abandonment of the current concentration-based standard "arbitrary and capricious." (FCC v. Fox TV Stations, Inc., 556 U.S. 502, 515-16 (2009))

Response 1-1: For the EPA's response to issues raised in this comment in regard to the propose lb/ton particulate matter (PM) standard, see the EAF NSPS preamble to the final rule, section IV.A.2.


Comment 1-2: The EPA conducted evaluation on a concentration basis and not in the form of the proposed standard (lb/ton). 

Commenter 0093-Steel Co. stated the EPA in its proposal performed cost analyses based upon the air flowrates to the air pollution control device and rather than establishing a standard of performance for the air pollution control device (baghouse), the EPA set the PM emissions standard in terms of lb/ton steel produced on a facility-wide basis. The EPA analyzed the performance of emissions controls from EAFs on a concentration basis (milligram per dry standard cubic meters (mg/dscm)  -  grain per dry standard cubic feet (gr/dscf)) -- and not in the form of the proposed standard. The EPA must be consistent with the basis of their evaluation and establish a standard measuring compliance as a concentration exiting the control device. 
Response 1-2: For the EPA's response to issues raised in this comment in regard to the propose lb/ton PM standard, see the EAF NSPS preamble to the final rule, section IV.A.2.


Comment 1-3: There are applicable regulatory provisions to prevent inefficient operation of air pollution control devices so there is no need to change the format of standard to do so.

Commenter 0093-Steel Co. stated it appears that the EPA is concerned with a facility utilizing more than one air pollution control device, overcontrolling one device, and underutilizing another. The EPA's explanation disregards other applicable regulatory provisions to prevent inefficient operation of air pollution control devices. 

Commenters 0091-State and 0093-Steel Co. stated the general provisions of the NSPS in 40 CFR part 60, subpart A, 60.12, address any concerns of excess emissions by prohibiting the use of gaseous diluents to achieve compliance with a concentration-based emission standard. The commenters continued, if the EPA suspected a facility of introducing diluent air for the purposes of achieving compliance, then the EPA should exercise its enforcement authority under 40 CFR 60.12:

      40 CFR 60.12 Circumvention. No owner or operator subject to the provisions of this part shall build, erect, install, or use any article, machine, equipment or process, the use of which conceals an emission which would otherwise constitute a violation of an applicable standard. Such concealment includes, but is not limited to, the use of gaseous diluents to achieve compliance with an opacity standard or with a standard which is based on the concentration of a pollutant in the gases discharged to the atmosphere.

Commenter 0097-Industry Trades also noted the General Provisions of the NSPS regulations prohibit "circumvention" of an applicable standard, including banning "the use of gaseous diluents to achieve compliance with an opacity standard or with a standard which is based on the concentration of a pollutant in the gases discharged to the atmosphere." (40 CFR 60.12) 

Response 1-3: The EPA agrees with the commenters that the NSPS General Provisions prohibit the use of gaseous diluents to achieve compliance with a concentration-based emission standard and notes that the commenters also are aware that the use of gaseous diluents to achieve compliance with a concentration-based emission standard is prohibited. However, that prohibition does not address inefficiencies in operation and control systems that are not as obvious, and may not be done intentionally by the owner or operator. The new format of the EAF NSPS standard under 40 CFR part 60, subpart AAb, in units of lb/ton, when compared to the General Provisions, not only ensures that circumvention and inefficiencies will not occur, but also ensures that every facility is accountable for the total PM contribution to the environment from its control devices for every unit of steel produced as compared to a predetermined value developed from data from one-third of the industry. For example, a facility with a control device that might be on the brink of exceeding a concentration-based PM limit can no longer simply add another control device in parallel and effectively emit more total PM with two control devices than they previously did with one control device, all while making the same amount of steel. Another example is a facility operating their baghouses at higher air-to-cloth (A/C) ratios than that needed to achieve the optimum A/C ratio for control of PM from their EAF and AOD. 


Comment 1-4: The EPA must consider the fundamental fabric filter operations and use of dilution air.

Commenter 0097-Industry Trades said the EPA's statement "baghouse dilution air (added to EAF exhaust air) can be increased with the effect of lowering measured baghouse PM emission concentration and disguising the true performance of the baghouse" is unsupported by any evidence in the rulemaking record. The volume of air coming from the main evacuation system dwarfs the amount of possible dilution air. Dilution air is usually used for cooling intake air to the baghouse, which is essential to maintaining baghouse functionality. Moreover, attempting to manipulate dilution air to somehow improve compliance is not done for very good reasons: dilution air negatively affects the performance of the baghouse and the overall performance of the melting process along with possibly degrading operating components of the main melting equipment. Adding unnecessary dilution air would minimize baghouse efficiency and have the effect of short-circuiting the ventilation system, thereby causing the baghouse to perform poorly for maintaining the direct shell evacuation control (DEC) and melt shop fume control. This would result in a dustier melt shop, be detrimental to worker safety, and increase the need for secondary emission controls. Further, operating in this manner would increase the need for larger exhaust rates unnecessarily as well. In fact, all efforts are made to minimize dilution air getting into the melt shop. 

Commenter 0097-Industry Trades continued that the EPA's argument suggests that having high flow rates is inappropriate and is being done to cheat the system, when in fact higher flow rates show that the evacuation system is operating more efficiently and is removing PM as it is designed to do. If the control device meets a grain loading performance standard that is all the better; it means that the system is performing as it should. 

Commenter 0093-Steel Co. added grain loading of filter media is not linear to air flowrates. 

Response 1-4: The EPA's concern with dilution air is the volume of air added to the fabric filter intake, not into the furnace nor the DEC nor the room evacuation as the commenters suggest. The flowrate from the baghouse is measured at the baghouse outlet at the same time as the PM is collected and measured for compliance purposes. Added baghouse dilution air increases the denominator in the calculation of the mass PM per unit air flow, thereby lowering the ratio and lowering the test value as compared to the NSPS standard. Although the grain loading to filter media may not be linear with increased flowrate due to addition of dilution air, the increased air flow has an inverse impact on PM concentration, meaning that as total flowrate is increased due to [clean] dilution air, the PM concentration is decreased, which lowers the test value as compared to the NSPS standard making it easier to meet a standard in terms of volume of exhaust air.


Comment 1-5: The EPA must consider operations of secondary baghouse. 

Commenter 0097-Industry Trades stated the EPA assumption that "secondary baghouses can be operated inefficiently without exceeding the current NSPS limit" is false and fails to reflect the fact that secondary baghouses are operated to control fugitive emissions (such as from the EAF during charging and tapping) and emissions from other processes in the melt shop (such as from the ladle metallurgy station or (LMS)). Commenters 0097-Industry Trades and 0093-Steel Co. stated facilities operate these secondary emission controls as efficiently as possible to maintain visibility within the melt shop and a safe working environment. Commenter 0093-Steel Co. added operating the secondary baghouses inefficiently would be counterintuitive to the goals and objectives of the air pollution control investments. Commenter 0097-Industry Trades stated, in contrast, all baghouses that receive primary emissions from the EAF or AOD are operated as efficiently as possible, not only to meet the concentration-based grain loading standard but to minimize the amount of fugitive emissions for secondary baghouses to control. 

Commenter 0097-Industry Trades stated the EPA appears to believe that the second baghouse is utilized to dilute the level of PM that goes to the primary baghouse. Secondary baghouses are not employed to control furnace emissions; they are used to evacuate dust from the shop (during charging/tapping for worker safety, and to control fugitive emissions from a variety of potential sources). In some cases, a second baghouse is used primarily for the LMS or for other PM-emitting sources within the melt shop as well. The EPA also assumes that secondary emission controls are operated inefficiently because the NSPS limit is too high. Secondary baghouses are not intended to address compliance with the PM limit, which is based on performance of the primary emission control system/baghouse which operates during the melting/refining stages (40 to 60 minutes). Secondary baghouses collect the majority of PM that goes to such baghouses during charging (less than 1 minute to 3 minutes) and tapping (4 to 6 minutes), the period when the DEC is not engaged, as well as from other operations (LMS) to keep the melt shop clear. 

Commenter 0093-Steel Co. added the EPA's rationale for changing the form of the standard is not based on engineering analysis, but rather a mistrust of performance test evaluations. Importantly, the FR notice does not cite a specific scenario where the EPA found that a secondary baghouse was "operated inefficiently without exceeding the current NSPS limit."

Response 1-5: The EPA disagrees with the commenter. The EPA does not have a mistrust of performance test evaluations; nothing in the record alludes to this being the EPA's rationale in regard to performance tests. While inefficient secondary baghouse operation cannot be easily substantiated with engineering analyses, neither has it been refuted with engineering analyses by the commenter. Regardless, this information is not a prerequisite for changing the format of the standard. The format of the subpart AAb standard accounts for the total PM impact to the environment. Because the NSPS is a technology-based standard, in order to not penalize well-operated large facilities, the ratio of PM emitted to amount of steel produced is used rather than a tons per hour or other time-based limits. When all facilities are well-operated, state and local agencies can better balance the overall impact of industry to the surrounding communities.


Comment 1-6: The EPA erroneously changed the limit for each control device to facility-wide PM limit. 

Commenter 0097-Industry Trades disagrees with the EPA's rationale that "there is no accounting for the total PM emissions from the facility" if a facility has separate baghouses. The commenter fails to see the relevance of "parity" between baghouses as long as emissions from the EAF are being properly captured and treated in (many) baghouses. If the EPA believes that the performance of one (say, the primary baghouse) is influenced by the presence or performance of the secondary baghouses or other emission controls, that has no technical basis. There is no incentive for facilities to operate the primary baghouse inefficiently. As a practical matter, facilities cannot, on-the-fly simply redirect exhaust gas streams from one baghouse to another. 

Commenter 0095-Steel Co. added the EPA should only amend the standard to require that the limit apply in the aggregate to all control devices through which a single EAF discharges, because that is the limit of the EPA's rationale. As worded, by aggregating a full "facility" rather than merely all control devices that capture a given EAF's emissions, the proposed rule would sweep beyond the rationale offered by the EPA and create compliance complications for other scenarios the EPA has provided no rationale for aggregating. For example, some facilities have multiple EAFs in the same melt shop, with each EAF venting to its own separate baghouse. Because the EAFs are generally operated independently and may have different production rates from each other in any given time period, it is not justified to subject the multiple independent EAFs to an aggregate production-based limit, and the EPA has not identified any rationale for doing so. The commenter continued that, likewise, some facilities have multiple melt shops, each containing one or more EAFs that each vent to separate control devices, and which function separately (i.e., production is not synchronized, and the various EAFs and melt shops may have different production rates). Because the EPA has provided no rationale for aggregating emissions for an entire melt shop if the melt shop has multiple EAFs, the EPA must retract that proposal, and either return to imposing PM limits on a per control device basis, or at most, only aggregate control devices that vent from the same EAF.

Response 1-6: The EPA contends in the new facility-wide PM limit per ton of steel produced, that a "new facility" for the purposes of the NSPS AAb, that is, each EAF or AOD and their associated control devices that begin construction, reconstruction or modification after May 16, 2022, should be accountable for controlling emissions consistent with the degree of emission limitation identified in this final action. This is true whether an EAF or AOD facility adds additional baghouses or additional EAF. If an EAF facility is in the business of making steel, it should be accountable for all steel made at the EAF facility, as defined, whether it is in 1, 2, or more EAF. This fact also is the basis for applicability of NSPS for modified facilities, where an existing facility is subject to NSPS if a physical change, or change in the method of operation, results in an increase in [total] emissions (40 CFR 60.2). 

For the situation where an existing EAF facility modifies or reconstructs so that 40 CFR part 60, subpart AAb applies, we have included a definition of "EAF facility" in 40 CFR part 60, subpart AAb as follows: "Electric arc furnace facility means the EAF(s) or AOD(s) subject to this rule [40 CFR part 60, subpart AAb] and the air pollution control equipment used to remove particulate matter from the effluent gas stream generated by the EAF(s) or AOD(s)." This means that only the equipment that qualifies as "new," e.g., a new or modified EAF and its control devices, would be subject to 40 CFR part 60, subpart AAb and the existing equipment, e.g., an EAF or AOD and its control devices, which is not included in the modification or reconstruction under 40 CFR part 60, would stay subject to its applicable rule, whether that be 40 CFR part 60, subpart AA or AAa.
The calculation of the new lb/ton standard is structured so that each EAF's control device lb/ton emissions are added to the emissions for other control devices for the same affected EAF under 40 CFR part 60, subpart AAb, or to the emissions from control devices for other affected EAF under 40 CFR part 60, subpart AAb, at the same facility. Because a ton of steel from one EAF is the same ton of steel from another EAF, under a facility-wide lb/ton standard, the production rates can be different for each EAF, or for same EAF with different control devices on a different test day. The lb/ton calculation is done for each control device to produce ratios which are, by definition, on the same basis (ton of steel) and then can be added together to produce a facility-wide lb/ton emission rate. Note that the test data acquired for calculating a gr/dscf standard is the same test data needed to calculate a lb/ton standard. The rule already requires determining the amount of steel produced (tapped) in the rule:

      40 CFR 60.274a Monitoring of operations 
      (h) During any performance test required under 40 CFR 60.8, and for any report thereof required by 40 CFR 60.276a(f) of this subpart, or to determine compliance with 40 CFR 60.272a(a)(3) of this subpart, the owner or operator shall monitor the following information for all heats covered by the test: 
      (1) Charge weights and materials, and tap weights and materials;

In addition, the EPA did not contend that steel operators can switch "on-the-fly" between primary and secondary baghouses. Rather, the issue we are addressing is a situation where secondary baghouses are added to a facility to "off load" PM from the primary baghouses (see the Cost Memorandum): "by dividing up the emissions into separate baghouses, with each baghouse falling under the same NSPS PM limit, there is no accounting for the total PM emissions from the facility").


Comment 1-7: Emissions not subject to EAF NSPS being ducted to melt shop baghouse system.

Commenter 0093-Steel Co. stated that most mills duct a number of other emissions units not subject to the NSPS through the melt shop baghouse system. Examples of emissions units not subject to the NSPS include ladle metallurgy furnace (LMF), casters, miscellaneous natural gas fired units and in some cases material handling. Separating these units out of a common melt shop exhaust system for a NSPS demonstration is not environmentally beneficial. Capturing and controlling the emissions units that are not part of the affected facilities is a "good air pollution control practice" and should be encouraged by the EPA. The EPA needs to clarify in 40 CFR 60.272b that the standard only applies to affected sources and not the entire facility. More baghouses are located at the "facility" for other operations that may not be subject to the EAF NSPS. 

Response 1-7: If facilities comply with the gr/dscf EAF NSPS standard when units not subject to the NSPS are ducted through the melt shop baghouse system, based on the comments, it is not clear how those circumstances differ with a lb/ton standard. However, the EPA understands the, what could be considered unique, circumstances raised by the commenter. In fact, this situation has already been addressed in the previous EAF NSPS at 40 CFR 60.275(g) and 40 CFR 60.275a(h) for 40 CFR part 60, subparts AA and AAa, respectively, and has been retained in the final rule for 40 CFR part 60, subpart AAb, as follows:
      40 CFR 60.275(h) Where emissions from any EAF(s) or AOD vessel(s) are combined with emissions from facilities not subject to the provisions of this subpart, determinations of compliance with 40 CFR 60.272b(a)(3) will only be based upon emissions originating from the affected facility(ies), except if the combined emissions are controlled by a common capture system and control device, in which case the owner or operator may use any of the following procedures during an opacity performance test and during shop opacity observations: 
         (1) Base compliance on control of the combined emissions; or
         (2) Utilize a method acceptable to the Administrator that compensates for the emissions from the facilities not subject to the provisions of this subpart.

While it may be convenient for a facility to include emissions from (auxiliary) units that are not in the affected EAF facility along with the emissions of the affected EAF facility control devices (under 40 CFR part 60, subpart AAb), this practice is not an exception that should need to be accommodated in the development of an EAF standard. Not only would it arbitrarily change the basis of the standard outside of the source category's definition, it would also make the magnitude of the standard source-specific because different facilities may have different auxiliary emission sources or intensities, and a facility may use a different combination or level of auxiliary units for different batches of steel. The facility could continue the practice of including auxiliary PM emissions with EAF PM control devices even with a lb/ton standard without consequence, so long as the overall PM emissions from the control device comply with the EAF NSPS total facility limit in lb/ton as in 40 CFR 60.275(b)(1), 40 CFR 60.275a(b)(1) 40 CFR 60.275b(b)(1). Or, the facility could determine the auxiliary emissions under 40 CFR 60.275(b)(2) or (3), 40 CFR 60.275a(b)(2) or (3), or 40 CFR 60.275b(b)(2) or (3). 


Comment 1-8: A lb/ton steel limit does not consider the different types of EAF mills. 

Commenter 0097-Industry Trades stated the EPA does not acknowledge nor address the fundamental fact that a "facility-wide lb/ton" production, or mass-based standard, ignores the substantial differences among EAF steel mills that directly bear on the particulate emissions per ton of steel produced. The commenter claims it is both unfair and inconsistent with best system of emissions reduction (BSER) to hold a small specialty steel EAF facility, with low tonnages and more time-intensive steel refining requirements, to the same production-based standard as a facility that produces 10-times or more steel with much shorter heat times (i.e., two facilities with vastly different production rates). 

Commenters 0097-Industry Trades and 0093-Steel Co. stated a compliance method based on PM per ton of steel produced does not take into consideration the various subcategories of EAF operations, differences in steel products, and variation in heat times and tonnages produced, which vary considerably depending on the product grade of steel and the mix of such products at various mills. Some carbon EAF mills produce high tonnages in relatively short heat times, while specialty EAF steel facilities produce much smaller tonnages over heat times that can be 2 to 3 times as long. 

Commenters 0097-Industry Trades and 0093-Steel Co continued, as the EPA noted in developing the 40 CFR part 60, subpart AAa standards: "The production of steel in an EAF is a batch process where `heats' or cycles range from 1 to 5 hours, depending upon the size and quality of the charge, the power input to the furnace, and the desired quality of the steel produced." Commenter 0093-Steel Co. added the EPA's statement, "The production of steel in an EAF is a batch process" (87 FR 29713, May 16, 2022), is not accurate and fails to acknowledge "Endless Charging Systems" and Consteel(R) continuous feed systems. 

Commenter 0093-Steel Co. added to determine appropriate standards of performance, the EPA should conduct a comprehensive evaluation of the different types of EAF mills (such as bar, sheet, and plate) and consider establishing different limitations and requirements for each subcategory. Commenter 0093-Steel Co. encourages the EPA to evaluate current designs and applications of baghouses for the control of PM. 

Response 1-8: For the EPA's response to issues raised in this comment in regard to the propose lb/ton PM standard, see the EAF NSPS preamble to the final rule, section IV.A.2.


Comment 1-9: A lb/ton steel limit does not consider the differences in inputs to steelmaking process. 

Commenter 0093-Steel Co. stated not all EAF facilities are using 100 percent scrap input to produce carbon and alloy steels, as indicated by the EPA (87 FR 29713, May 22, 2022). The commenter said that producing high quality steel and steel products, such as advanced high strength steel (AHSS) for the automobile and aviation industries, requires more alloys and processing, which then has the potential to increase the PM emission rate in terms of lb/ton of steel produced. The commenter indicated that these differences in material input require a more robust evaluation for establishing standards of performance than what the EPA has conducted. The EPA must conduct a more complete, comprehensive evaluation prior to finalizing the standards of performance.

Commenter 0097-Industry Trades added the proposed EAF/AOD 0.16 lb/ton PM limit unnecessarily complicates compliance testing by over-measuring results, especially when many EAF/AOD baghouses also control non-EAF process units, including LMS whose emissions are included in the baghouse compliance tests as the LMS are part of the EAF steelmaking sequential steelmaking process (LMS are not turned off during compliance testing as they are a necessary part of the sequential process). 


Response 1-9: The data used for the lb/ton analyses as well as the entire EPA dataset are similar to the distribution of product types in the industry in terms of carbon steel vs. high alloy vs. stainless steel. See Figure 1-1 which shows the various types of EAF baghouses in the 13 facilities with 0.000 percent shop opacity compared to the 30 EAF in the EPA EAF data with lb/ton data, compared to the 61 EAF facilities in the industry data out of the total of 88 that have product type that is known. If there is any trend, the lb/ton data for 13 facilities with 0.000 percent shop opacity have a higher representation of high alloy and stainless steel production than the national EAF population, which would slant the data towards higher emissions according to the commenter. Facilities that produce high alloy and stainless steel use a higher grade of steel input and not 100 percent scrap, as the commenter points out. However, it may be the case that emissions are similar between high alloy EAF processes and carbon steel. A higher grade of input to the EAF with high alloy and stainless steel production that has lower PM emissions than lower grade carbon steel, may be offset by the extra processing required after the high alloy and stainless steel leaves the EAF, which generates more PM than with production of the typical carbon steel. 

      Note: C = carbon steel; H = high alloy steel; S = stainless. 
      For SMA data: C,H = C,H only and C,H,S; H = H only and H,S.
      
      Figure 1-1. Type of steel produced at facilities in three datasets: 13 facilities with 0 percent opacity EPA/EAF dataset, 30 total facilities EPA/EAF dataset, and 61 facilities in 2017 Industry (SMA) dataset.


As discussed above under response to Comment #1-7, while it is convenient for a facility to include PM emissions from (auxiliary) units that are not in the EAF NSPS source category along with PM emissions from EAF source category processes into the EAF facility's control devices, this practice is not an exception that should need to be accommodated in the development of an EAF standard. Not only because it would arbitrarily change the basis of the standard outside of the source category's official definition, but it would make the magnitude of the standard source-specific because different facilities may have different auxiliary emission sources or intensities, and any one facility may use a different combination or level of auxiliary units for different batches of steel. The facility could continue without consequence the practice of including their auxiliary PM emissions sources with their EAF PM if the overall PM emissions from the control device comply with the EAF NSPS total facility limit in lb/ton, the same as has been done in the past with the EAF standard in gr/dscf format under 40 CFR part 60, subparts AA and AAa when compensating for emissions from sources not subject to the EAF NSPS standard. Or, the facility can separate out the emissions from the auxiliary units as per the instructions in 40 CFR 60.275(b)(2) or (3), and 40 CFR 60.275a(b)(2) or (3), or 40 CFR 60.275b(b)(2) or (3).

The PM emissions from secondary baghouses that control LMS and other alloy operations can be low; therefore, their lb/ton contribution to the total facility lb/ton would also be low. The benefit of using the lb/ton factor is that it makes the low PM sources less significant to the total for the same amount of steel produced. In many cases, the secondary emissions, which also may include LMS and other alloy operations, are combined with primary emissions and are controlled in the same baghouse. Some of these primary and secondary control combinations are still at the low end in the list of ranked baghouses in lb/ton. See Figure 1-2 (which corresponds to the 50 individual baghouses with lb/ton data listed in Table 3 of the Emission Memorandum for the proposed rule, Docket ID Item No. EPA-HQ-OAR-2002-0049-0061) and which illustrates the data from various baghouse types. While the primary and secondary combined emissions baghouses (grey line in Figure 1-2) cover a wide range of data and predominate at the higher emission levels, the secondary-only and primary-only baghouses are in the lower 2/3[rd] of the combined primary and secondary lb/ton values. Combination primary and secondary baghouses are the majority in number (30 out of 80) and predominate at higher PM emissions, as seen in Figure 1-2. Table 1-1 summarizes the type of emissions controlled in the 50 baghouses with lb/ton values in the EPA EAF data.

Table 1-1. Type of Emissions Controlled in the 50 Baghouses with lb/ton Values in the EPA EAF Data
                                       
                 Type of Emissions Controlled in EAF Baghouses
                                 No. Baghouses
                                       

Primary and secondary
                                      30
                                                                               

Primary only
                                       8
                                                                               

Secondary only
                                       8
                                                                               

AOD
                                       3
                                                                               

AOD and secondary
                                       1
                                                                               

Total baghouses
                                      50
                                                                               

Note: Total of 50 baghouses for which lb/ton values could be developed.

                                                                               
                                       
Figure 1-2. EAF baghouse data: lb PM/ton steel, by baghouse type. 


Comment 1-10: The proposed "facility-wide" PM limit fails to meet the EPA's Compliance Monitoring Strategy Goals. 

Commenter 0093-Steel Co. stated the EPA released in October 2016 its latest version of "CAA Stationary Source Compliance Monitoring Strategy" and stated its goals of the compliance monitoring strategy: Goal (1) of the compliance monitoring strategy to "Provide national consistency in developing stationary source air compliance monitoring programs, while at the same time provide states/locals/tribes/territories with flexibility to address local air pollution and compliance concerns" will be difficult to achieve. The commenter continued, in evaluating Goal (3) of the compliance monitoring strategy, to "Provide a framework for developing stationary source air compliance monitoring programs that focuses on achieving measurable environmental results," affected facilities already monitor and demonstrate compliance with emission limitations in terms of gr/dscf. Since the 1974 standards of performance, the EPA has established and provided a framework for demonstrating compliance that focuses on achieving measurable environmental results. By revising the form of the standard to a particulate limit of lb/ton of steel produced, the EPA will create national inconsistencies between EPA Regions and air pollution control agencies when determining compliance with the standard. In doing so, the EPA will not meet the goals of its compliance monitoring strategy.

Commenters 0093-Steel Co. and 0091-State stated the emission limitation should be in terms that are clearly defined and practically enforceable. A PM emission limit in terms of lb/ton of steel produced will create inconsistency between EPA Regions and state, local, and tribal air pollution control agencies and is difficult to measure in practice. The EPA should maintain an emission limit in terms of gr/dscf, which is consistent with the emission guarantees provided by baghouse vendors and bag leak detection systems (BLDS) certification requirements, as well as the test method used to demonstrate compliance with the emission limit. Facilities already monitor and demonstrate compliance with emission limitations in terms of gr/dscf. 

Response 1-10: We disagree with the commenter. The same test methods used to measure grains of PM for a gr/dscf standard are required to be used to measure pounds of PM for a lb/ton limit, i.e., EPA Method 5. The measurement of steel produced is something EAF facilities already determine as part of performance tests using EPA Method 5 the EAF rules, for both the NSPS (40 CFR 60.274a(h)(1)) and NESHAP (40 CFR 63.10686(d)(5)), for Title V permitting, and likely also track for their own business purposes. The EAF NESHAP (40 CFR part 63, subpart YYYYY) describes procedures to develop a lb/ton emissions format from the EAF test data required for the rule in 40 CFR 63.10786(d)(5): "To determine compliance with the PM emissions limit in paragraph (c) of this section for an EAF or AOD vessel in a lb/ton of steel format, compute the process-weighted mass emissions (Ep) for each test run using Equation 1 of this section."

The facilities that provided EAF test data to the EPA in 2010 for the purpose of presenting mercury emission data also provided the production rate as part of their PM and mercury testing. These data were requested by the EPA under the authority of CAA section 114 (hereafter called the EPA/EAF data).

While change can be a challenge in any regulatory area, EPA determined that requiring new sources to track total facility emissions rather than just baghouse exhaust concentrations would result in a better form of the standard that will ensure the emission reductions from the BSER are achieved. Vendors and state and regional inspectors can continue to use gr/dscf to evaluate operation of the baghouse and BLDS. However, facilities need also to be responsible for their total PM impact in terms of mass of PM emitted per ton of steel produced as they increase production. 

Facilities and states only can assess whether increases in total PM emissions are warranted by determining the rate that PM is emitted per ton of steel produced for the same facility both before and after a planned change. It is therefore easier to obtain information on total PM and steel production rates rather than determining the PM emitted per airflow rate for every individual baghouse at a facility. As stated in other responses in this document, where a total facility lb/ton limit is in effect, expansions that plan to increase emissions, where the increase is also an increase relative to the amount of steel produced will not be allowed if the final total PM emitted and total steel produced ratio is not equal to or less than 0.16 lb/ton.


Comment 1-11: The change from a concentration to lb/ton limit complicates compliance and does not result in better control and greater assurance of compliance. 

Commenter 0097-Industry Trades and 0095-Steel Co. stated the EPA's assertion that switching to a lb/ton standard will "result in better control nor greater assurance of compliance" is incorrect. Under the current standards in 40 CFR part 60, subparts AA, AAa, compliance is readily demonstrated through EPA Method 5 monitoring of the stack on the primary control device/baghouse. This is a direct measurement of the filtering ability of the baghouse and evidence of compliance with concentration limits without all of the unnecessary variables the new rule introduces which are not directly related to emissions. Under the proposal, facilities would be required to track tonnages produced during stack tests and match those to emissions data. 

Response 1-11: For the EPA's response to issues raised in this comment in regard to the propose lb/ton PM standard, see the EAF NSPS preamble to the final rule, section IV.A.2.


Comment 1-12: The lb/ton limit does not consider vendor guarantees on control systems. 

Commenters 0095-Steel Co. and 0097-Industry Trades stated it is critical to obtain vendor guarantees from suppliers when constructing new facilities or EAFs and associated control systems, to ensure that the purchased equipment can comply with applicable standards. Vendors can guarantee that the filters/control device have a specific removal rate (i.e., vendors can only guarantee the difference between the clean and dirty side of the bag). Obtaining such guarantees is what gives facilities comfort that the equipment they purchase will perform such that compliance is assured. Commenters continued, such comfort is not possible with the Agency's proposed "facility-wide" PM limit, [because] vendors do not offer lb/ton guarantees for specific equipment, and certainly not on a facility-wide basis. This is understandable given a supplier's ability to design equipment to a given concentration or control specification, but lack of ability to control the many factors that influence lb/ton efficiency, especially where a vendor may not be the sole facility-wide designer. Vendors have no control over the tonnage of steel produced or how the steel tonnage estimate comports with the duration of the PM measurement. The commenters concluded the EPA should take this into account by setting a concentration-based emission standard and noted, the EPA has previously acknowledged the importance of being able to obtain vendor guarantees when setting the 40 CFR part 60, subpart AAa limits in 1984 (49 FR at 43840), and thus it would be arbitrary and impermissible for the EPA to ignore that consideration here. 

Response 1-12: For the EPA's response to issues raised in this comment in regard to the propose lb/ton PM standard, see the EAF NSPS preamble to the final rule, section IV.A.2.


Comment 1-13: lb/ton limits are not good for a short-term basis. 

Commenter 0095-Steel Co. stated lb/ton limits were not intended and are not adequately demonstrated to be reliably or representatively evaluated on an hourly basis. Rather, lb/ton limits are generally used for long term averages, such as monthly or annual production, because inventory is verified over longer terms than an hourly basis, and hourly production and conditions are more variable on an hourly basis than if averaged over a longer period. In fact, as the EPA recognized in the record for the proposed rule, many facilities do not track production on an hourly basis, and instead track production based on a per heat or batch basis.

From the NSPS proposal Emissions Memorandum: 

      Not all facilities and test data were able to be used to create emissions factors in mass units. In some cases, no production data were available for some or all the runs. One reason for fewer or no run data was because the production rate was reported as confidential business information; another reason was because the facility reported production in weight of steel per heat and not per hour.

Accordingly, verifying a lb/ton limit based on performance tests taken over the course of a few hour-long measurements are less consistent and generalizable than concentration-based limits.

Response 1-13: We disagree with the commenter that hourly lb/ton values are needed in the short term for the new 40 CFR part 60, subpart AAb rule. Compliance tests to determine PM emissions require at least three runs, where each run can last 4 hours or more. Therefore, the total compliance test time is on the order of a total of 12 hours or more. This period would not be considered short term, nor an hourly value. As discussed in the EPA response to a previous comment, (See EPA Response #1-6, earlier in this document), the amount of steel tapped is already required in the EAF rule during compliance tests (40 CFR 60.274(i)(1), 60.274a(h)(1)). 

Although the commenter's quote from the Emissions Memorandum[6] from the proposal is correct, stating that steel production was not reported in the EPA test data for some facilities, only a few facilities (3 out of 33) did not have complete data. This fact is also stated in the proposed Emissions Memorandum[6]:

      However, at least three runs of data were available, minimally, for each baghouse in the data that was used. Table 3 lists the 30 facilities with data available to be expressed in units of lb/ton, and the average baghouse emission data for their 50 baghouse in lb/ton (the units reported).

For one facility (3 test reports), the PM data submitted that was not used in the NSPS analysis was from an older test report that the facility submitted as part of a 2010 CAA section 114 request to submit previous test reports and included total heat (i.e., batch) production in tons rather than hourly production. Consequently, the 2010 request by the EPA to report production rates could not be retroactively added to the older test reports submitted along with test reports from 2010. For two other facilities (10 test reports), the production rates were reported as confidential business information, likely because these facilities were concerned about the public knowing the size and production capability of their facility. Due to numerous tests, and in some cases multiple test reports, submitted by each facility (see Section 3.0 PM Data, Appendix 5, and Table 5 in proposal Emissions Memorandum[6]), and despite the lack of data in a few cases, the EPA had sufficient data to analyze for the proposed and final lb/ton limit. ]


Comment 1-14: The EPA established a conversion ratio based on emissions from a single facility. 

Commenter 0097-Industry Trades stated the EPA ignored the vast majority of the available EAF test reports for 50 baghouses at 30 facilities with production-based and concentration-based emission data and instead established a conversion ratio based on emissions from a single facility. Commenters 0097-Industry Trades and 0095-Steel Co. stated the EPA's approach to rely solely on NAS Ghent (KY) (Model Plant E) to derive an equivalence between the concentration and production-based limits is flawed. Conceptually, it makes no sense for the EPA to rely on data from just a single facility, neglecting relevant data from the other 29 facilities for which the EPA has pair-wise concentration and production data. 

Response 1-14: The EPA is not quite clear on the point made by the commenter about the analysis based on a "single facility." The cost analysis used a model plant approach, which by definition required each model plant to have one emission value. The model plants represent the range in the EAF data in terms of lb/ton. See the EPA response to Comment #1-17 below for a discussion of the distribution of 13 and 30 EAF facilities in terms of their facility-wide total lb/ton emissions. The reason that the model plant analysis was based on 13 facilities with 0.000 percent shop opacity is explained in response to Comment #1-13 and in the Cost Memorandum.[2][,] However, it is also explained in response to Comment #1-13 that the lb/ton distribution is nearly identical between 13 facilities with 0.000 percent shop opacity and all 30 facilities with lb/ton data. The fact that the standard selected to reflect BSER for PM from control devices was based on the lb/ton value for one of the model plants is part of a typical model plant analysis. There must be only one value to be an emission limit.

If the commenter is referring to the fact that the highest-emitting facility, NAS-KY, was used in the proposal to estimate the lb/ton equivalent to the current standard in grains per dry standard cubic feet of exhaust (gr/dscf) (i.e., the "baseline"), the EPA used the highest-emitting facility in gr/dscf because the standard in gr/dscf was closer to this facility than any other in the data set; therefore, the lb/ton equivalent was expected to be closer as well. And because there was not a strong correlation of facility gr/dscf emissions to facility lb/ton emissions in the data, using a facility value in gr/dscf close to the gr/dscf standard was considered to be a reasonable approach that would incorporate less potential error. However, for the final rule, we did not use a baseline estimate. Rather, we used the highest-emitting facility, NAS-KY, for the highest-emitting model plant to which lower emitting model plants are compared rather than starting at a baseline. Similar results were obtained with this approach. Therefore, the 0.16 lb/ton value also was determined to be the limit under BSER, which is a capture system and fabric filter (See Comment #1-17 below).

The determination of the limit for NSPS is based on the "best" system of emission reduction, which in a model plant approach is one facility, but would be based on one number even under any other approach. And because costs are required to be considered in determining the appropriate level of control for the NSPS, the EPA did not select the lowest emitting facility or lowest emitting model plant as the basis for the limit for BSER. 


Comment 1-15: Total facility PM emissions in gr/dscf cannot be obtained by adding each baghouses' concentration.

Commenters 0097-Industry Trades and 0095-Steel Co. stated that as part of the lb/ton analysis, the EPA needs to verify that the plant-wide concentration values are properly calculated (i.e., not added, when multiple baghouses are present). The commenter also stated the EPA erroneously adds the four separate baghouse concentration values from NAS Ghent (KY). Concentrations are not additive.

Commenter 0097-Industry Trades stated that the ability of the control device (i.e., the baghouse) to remove PM is related more to the air flow, filter surface area (i.e., the A/C ratio noted prior), and filter composition of the baghouse, rendering grain loading a better metric than the tons or production basis for the form of the standard. 

Response 1-15: The EPA agrees that method used to obtained total facility PM emissions in gr/dscf was incorrect in the proposal. The previous analysis erroneously assumed each baghouse had the same flowrate so that the individual baghouse gr/dscf PM concentrations could be added for a facility-wide sum. However, EPA has corrected this for this final action. The correction used the following formula and steps to obtain total facility baghouse PM emissions in gr/dscf from total facility gr/min PM emissions and total baghouse flowrates, as follows:

SUMfacility BH PM gr/min {Average [SUMeach run (BH gr/dscf x BH dscf/min)]}/SUMfacility AVG BH flows =
                           Total Facility PM gr/dscf

Step 1: Determine individual PM emission rates for each run from individual baghouses in gr/min from individual baghouse PM concentrations in gr/dscf multiplied by flowrate for each run: 

                      RUN gr/dscf x dscf/min = RUN gr/min

Step 2: Average the run PM emission rates in gr/min for each baghouse to obtain overall test PM emission rate for each baghouse:
             Average RUNS = gr/min for each TEST for each BAGHOUSE

Step 3: Sum all baghouse PM emission rates in gr/min for all baghouses at a facility to obtain facility PM baghouse emission rate in gr/min:
                                       
Sum of average TEST gr/min for all BAGHOUSES = Total FACILITY gr/min for all baghouses

Step 4: Determine the average flowrate for each test at each baghouse at the facility from each run, as for emissions, and sum up all the flowrates for a facility-wide total BH flowrate:
                                       
Sum of individual baghouse FLOWRATES dscf/min = 
                                    Total FACILITY baghouse FLOWRATES, dscf/min
                                       
Step 5: Determine the facility-wide PM concentration in gr/dscf by dividing the total facility emission rate in gr/min by total facility baghouse flowrate in dscf/min:
                                       
Total FACILITY gr/dscf = Total FACILITY gr/min / Total FACILITY flowrate dscf/min

The EPA is aware that the correction resulted in higher emissions for some facilities in our analysis. This was likely due to baghouses that had both higher PM emissions and higher flowrates and, therefore contributed more to the total facility PM emissions than the baghouses with lower, flowrates. The following graph (Figure 1-3) illustrates the total facility gr/dscf PM values with the previous (proposal) method (blue) vs. the corrected values (orange). At higher emission rates, the difference between the corrected (finalized) values and previous (proposal) values, are greater. However, the EPA determined that this correction has no bearing on the lb/ton standard developed in the proposed and final rules because the total facility gr/dscf analysis was done only for informational purposes.
These corrections to the facility-wide gr/dscf calculations included with the documentation for the final rule do not change the level of the standard in lb/ton because the EPA did not derive lb/ton values from gr/dscf data. See the Emissions Memorandum[6] prepared for both the proposed and final rules for the procedures used to develop the lb/ton ratios. The EPA used reported tons/hr steel production and lb/hour PM test run emissions rates in the submitted test reports to develop lb PM per ton of steel. Note, a comparison of the (corrected) total facility gr/dscf data to the lb/ton data is shown in Figure 1 of the preamble to the final rule.
                                       
Figure 1-3. Corrected total facility gr/dscf vs. previous calculation


Comment 1-16: EPA's conversion of gr/dscf to the lb/ton of steel produced in the melt shop is incorrect.

Commenter 0097-Industry Trades stated in order to convert gr/dscf to the lb/ton of steel produced in the melt shop, one has to multiply the gr/dscf by the dscf/ton or exhaust flow per unit of steel production. Unfortunately, this dscf/ton factor is highly dependent on many variables such as the size of the EAF, the total energy input into the EAF and the source of the energy, including the non-electrical energy inputs such as the amounts of natural gas and oxygen used, the grade of steel being made, the total heat time and the durations of the various modes such as melting and refining during the heat, the quantity of "hot heel" practice for the shop, and many other variables  -  all of which can vary from heat to heat. As such, unlike the relationship between exhaust flow and combustion variables in other source types (such as fossil fuel boilers) where the relationship between exhaust flow and heat input is stable and depends solely on the fuel type (called an F-factor), this is not the case in an EAF melt shop. The commenter stated, given the significant variability associated with the dscf/ton parameter in any melt shop, the EPA should not set a limit using the lb/ton form of the standard without analyzing the variability of this factor and then properly accounting for it in setting the stringency of the standard at the same level as the gr/dscf standard. 

Response 1-16: The EPA did not derive lb/ton values directly from gr/dscf data. See the proposal's Emissions Memorandum[6] for the procedures used to develop the lb/ton ratios. The EPA used reported tons/hr steel production and lb/hour PM test run emissions rates in the test reports (as reported by the testing contractor) to develop lb PM per ton of steel. See the Emission Memorandum's[6] Appendix B: EAF NSPS PM Test Data for Emission Factor Development. 

We would expect the air flow rate in the baghouse to be relatively consistent, given that the induced draft fans are unlikely to change. While the PM input could vary with steel production, we would not expect the air flow rate to change. Therefore, flow rate is expected to be relatively independent of steel production.

The calculations are as follows:

Data
Run steel production rate during testing (tons/hr)
Run PM filterable (lb/hr)

Calculations
Run PM Filterable (lb/ton) = run PM lb/hr / run steel ton/hr
Test lb/ton = AVG run PM lb/ton
Baghouse lb/ton = AVG all test lb/ton for baghouse
Facility lb/ton = Sum of all baghouse test lb/ton for facility


Comment 1-17: Data from facilities that achieved 0.000 percent shop opacity do not support the EPA's presumption that facilities with higher baghouse emissions capture shop PM more effectively.

Commenter 0097-Industry Trades stated, regarding EPA's proposed 0 percent opacity standard, that the data do not support EPA's simplistic presumption that facilities with higher baghouse emissions capture shop PM more effectively and therefore have 0 percent shop opacity. Even if the data supported EPA's speculation that facilities with higher baghouse emissions have lower shop opacity, that would not justify EPA discarding the majority of the performance test reports that contained both performance-based and concentration-based emissions data. The test reports could have been used to test whether there was a relationship between performance-based and concentration-based limits  -  i.e., whether these two metrics of emissions at the baghouse exhausts had a robust relationship. Reasonably discerning that relationship requires EPA to use as much data as possible, even though EPA surmised (baselessly) that some of those facilities had lower capture efficiency. When discerning whether two variables are related, more data are better than less data. As such, EPA's decision to ignore the majority of its available performance test data was arbitrary and quite wrong from an analytical perspective.

Response 1-17: The EPA disagrees with this comment. The commenter appears to misunderstand the relationship between melt shop opacity and baghouse emissions, the similarity in the 30 vs 13 facility datasets, and the regulatory process used to develop the lb/ton standard. First, no data was "discarded" in the analysis as the commenter suggests. Rather, in the first step in the development of the standard, the EPA looked at baghouse PM data in units of gr/dscf and lb/ton on both the baghouse level and total facility level for all baghouses and facilities for which there were data to develop a lb/ton emission factor. This step included 30 of the 33 facilities in the dataset, where only three of the 33 facilities did not have production data that allowed calculation of the lb/ton values. See the graphs in Figure 1-4 that illustrate these relationships. (Note that the gr/dscf data have been corrected as a result of comments in Comment #1-15). 

As shown in the graphs below, the facility-level lb/ton did not correspond as well to the facility-level gr/dscf as the baghouse gr/dscf vs lb/ton values. This is likely due to the fact that the facilities with facility-wide data above the line of best fit had baghouses that had mostly higher lb/ton to gr/dscf ratios, so the summing process over all baghouses at the facility exaggerated the difference between these facilities and the facilities that are better-corelated between total facility PM in gr/dscf and lb/ton. However, when data for lb/ton and gr/dscf are each plotted on a facility-wide total PM basis, while the numeric values are different between gr/dscf and lb/ton, the trends are the same between gr/dscf and lb/ton, as also with 13 vs. 30 facilities. See graphs in Figure 1-5. 

The EPA is not saying, as the commenter asserts, that higher baghouse emissions mean better melt shop control (the converse of EPA's statement), only that, all things being equal, higher capture of melt shop emissions represent the highest PM capture and control by a baghouse and then possibly the highest baghouse emissions that the facility would emit as compared to less capture and control of melt shop emissions, all other parameters being the same. 

The purpose of finalizing a standard in lb/ton and on a facility-wide total PM basis is, respectively: (1) to obtain a standard as an emission rate, such as lb/ton, rather than a non-dilution corrected concentration in order to eliminate the effect of fluctuations in baghouse dilution air on the measured concentration; and (2) to present total facility PM emissions that take into account all control devices and that more easily can be used to determine the total facility PM impact on surrounding environment and communities. 

                                       
                                       
Figure 1-4. Total facility and individual baghouse PM, gr/dscf vs lb/ton.



                                       
Figure 1-5. Comparison of total facility gr/dscf vs lb/ton for 13 facilities with 0 percent opacity vs. all 30 facilities in EPA data set.


Once the decision was made to develop a facility-wide lb/ton standard, the next step in the analysis was to determine the numeric value of the standard. All data for 30 facilities with lb/ton values could have been used instead of only 13 facilities with 0.000 percent opacity. It is shown in Figure 1-5 that the range in lb/ton emission data on a facility-wide lb/ton based on 30 facilities vs the 13 facilities with 0.000 percent opacity are almost exactly the same. The high value of 0.16 lb/ton was the same in each dataset, whereas the low values were close. (See Table 1-2). But in the BSER analysis for the PM limit from fabric filters based on model plants, the high values for lb/ton turned out to be the most important. Therefore, the use of 13 facilities vs. 30 facilities would have had the same limit under the BSER analysis for PM from control devices.

             Table 1-2. Total Facility Baghouse Emissions (lb/ton)
                                     Data
                                 13 Facilities
                                 30 Facilities
Low
                                    0.0130
                                    0.0049
High
                                    0.1600
                                    0.1600


The original reason for the use of only 13 facilities rather than 30 (in the proposal) was that during the calculations to estimate the emissions at a hypothetical facilities achieving 0 percent vs 6 percent melt shop opacity, it was evident that the estimate of total emissions at 0.000 percent shop opacity needed to account for the emissions based on 6 percent opacity that were no longer emitted when 0 percent opacity was being achieved. It was obvious from a mass balance perspective that the controlled PM emissions in a 0 percent shop opacity scenario would be collected by melt shop capture systems and then mostly controlled, i.e., collected in facility baghouses, but with a higher amount of PM being emitted from the baghouse than what was collected with 6 percent opacity. In order to not penalize facilities that collect all their shop emissions (i.e., achieve 0.000 percent shop opacity) with an overly stringent facility baghouse limit in lb/ton, and also avoid estimating the redirected PM by calculations involving numerous assumptions, the dataset used to calculate the limit for BSER for a facility-wide control device PM limit was reduced to include only the facilities that achieved 0.000 percent opacity. This decision resulted in a dataset of only 13 facilities for proposal. But, as seen in Table 1-2, the range in the total facility data are almost identical between the 13 and 30 facilities, and the shape of the two curves (13 facilities vs. 30 facilities) are nearly identical, as seen Figure 1-5. 


Comment 1-18: The EPA's proposed production-based data were not correlated with the concentration-based data.

Commenter 0097-Industry Trades stated the test reports could have been used to test whether there was a relationship between performance-based and concentration-based limits  -  i.e., whether these two metrics of emissions at the baghouse exhausts had a robust relationship. 

The commenter further stated if the EPA were to persist with this approach (lb/ton value), the EPA would need to include all EAF test data for which it has pairs of concentration and mass-based values for the same test.

Response 1-18: The EPA did not find any correlation between gr/dscf and lb/ton values. The basis of these terms are entirely different, which is why EPA is revising the format of the PM limit. The gr/dscf limit assesses baghouse performance in relation to exhaust flowrate, whereas a lb/ton limit assesses facility performance in controlling PM emissions as a function of steel produced. While proper operation of baghouses is essential for a well-controlled facility, one facility can operate multiple baghouses within the gr/dscf limit while emitting more PM per ton of steel produced than other facilities also within the gr/dscf limit. This fact is established by the lb/ton data in Figure 1-5, where the various levels of lb/ton total facility PM emissions all are from facilities whose baghouses meet the gr/dscf limits of the current NSPS.


Comment 1-19: The correlation to determine the production-based value corresponding to the 0.0052 gr/dscf concentration limit is incorrect.

Commenter 0097-Industry Trades stated the proper correlation between concentration and production values is needed in order to use that correlation to determine the production-based value corresponding to the 0.0052 gr/dscf concentration limit. Or EPA needs to conclude, if that data so indicate, that there is not a strong correlation.

Commenters 0097-Industry Trades and 0095-Steel Co. state that because the gr/dscf total facility calculation was incorrect, the starting point of the EPA's scaling attempt (i.e., NAS Ghent's (KY) 0.0040 gr/dscf value) is wrong. 

Response 1-19: As stated earlier in Response 1-14, there is no correlation between the gr/dscf and lb/ton data on an industry level. However, the paired relationship between one facility's lb/ton and gr/dscf emission factors at one point can be used to estimate the missing value in a pair at another close points if one value is known. This is the reason why, in the proposal, the highest lb/ton value in the data was used to estimate the baseline lb/ton equivalent to the current NSPS gr/dscf standard of 0.0052 gr/dscf. The results of the analysis to determine the limit for BSER would have been the same if the highest emitting facility was used as the starting point in the cost analysis, rather than the baseline. Note that the facility total gr/dscf analysis has been corrected as per comments received and is addressed in more detail in another comment response (Comment #1-15). In addition, a baseline model plant was not used in the model plant analysis for the final rule, so an estimate of production-based value corresponding to the 0.0052 gr/dscf concentration limit was not needed.


Comment 1-20: The mass-based limits need to be estimated on a consistent basis.

Commenter 0097-Industry Trades stated that EPA should ensure that the mass-based limits for each plant are estimated on a consistent basis (i.e., using the same form of production tons in the denominator) since plants can use a mix of metrics for production tons  -  such as melted tons, cast tons, etc.

Commenter 0091-State stated if the EPA will not retain the concentration-based limit, the EPA should add a definition of "tons of steel produced" and define the emission limit in terms of lb/ton of charge materials or lb/ton of tap materials, which is consistent with the records required in the proposed rule. 

Response 1-20: To address this comment, we analyzed the EPA EAF data used in the lb/ton analysis based on the point where steel produced was reported as being measured. Table 1-3 shows that in over half of the test reports, the point of steel production measurement was not specified. However, as seen in Figure 1-6, when comparing the lb/ton values for the unspecified production data, to data where the measurement location was specified, the magnitude of the data values (e.g., all high lb/ton rates vs. all low lb/ton rates) were not correlated to any one type of measurement point. Therefore, there was not a correlation between emission rate, as lb/ton, and any one steel production measurement point. In addition, the lb/ton data for unspecified locations covered the entire range in the EPA EAF data. Our conclusion is that the lb/ton limit is not dependent upon the type of measurement used for the amount of steel produced. Details of the analysis follow below.
 
       Table 1-3. Summary of Steel Production Measurement Points in the 
           Responses to the 2010 CAA Section 114 Information Request
                                     Total
                              Number Test Reports
          Was the point specified where steel produced was measured?
                            Type of Steel Measured
                              Number Test Reports
                         percent of Total Test Reports
                                      120
                                      No
Not specified
                                      51
                                      43%
                                       

                                      Yes
Specified 
                                      69
                                      57%
                                       

                                      Yes
 
Raw steel/liquid/hot metal/tapped 
                                      29
                                       
                                      24%

                                      Yes

Cast/billet 
                                      40
                                       
                                      33%


                                       
Figure 1-6. Range of EAF facility total baghouse lb PM/ton steel, by steel measurement type.


We looked at the reported tons of steel melted (raw) vs. cast at one facility (CMC, Cayce, SC) taken from the previous test reports submitted along with their 2010 test data. In 2006, compliance tests were performed on the EAF melt shop baghouse at the CMC facility. The compliance report included both melted (raw) and cast steel for two (#2 and #3) of the three runs performed at the baghouse, with only cast steel reported for run #1. The ratio of melted steel to cast steel for runs #2 and #3 was 1.0 for both runs, as shown in Table 1-4. The PM emissions rate in lb PM/ton steel for the two runs using both melted and cast steel was 0.044 lb PM/ton melted steel and 0.045 lb PM/ton cast steel, also shown in Table 1-4. These values are almost identical between the two measurement points and well below the proposed limit of 0.16 lb/ton. Therefore, the analysis here shows that the type of steel did not impact the value determined as the lb/ton emission rate. 

Table 1-4. Melt Shop Baghouse Emissions and Production Rates at CMC in Cayce, SC
                          April 26, 2006 Test Report
                                 Run PM lb/hr
                                lb PM/ton steel
                                 Tons per Run
                             Ratio Melt/Cast Steel


                                 Melted Steel
                                  Cast Steel
                                 Melted Steel
                                  Cast Steel
                                       
Run 2
                                     4.64
                                     0.045
                                     0.045
                                      493
                                      492
                                      1.0
Run 3
                                     4.59
                                     0.043
                                     0.044
                                     1,486
                                     1,473
                                      1.0
Average Two Runs
                                     0.044
                                     0.045
 


The amount of steel tapped already is required to be recorded in the current EAF NSPS, 40 CFR 60.274(i)(1) and 60.274a(h)(1), and was also proposed in 40 CFR subpart AAb in 60.274b(h)(1): 
      
      During any performance test required under 40 CFR 60.8 or 40 CFR 60.272b(d), and for any report thereof required by 40 CFR 60.276b(f) of this subpart, or to determine compliance with 40 CFR 60.272b(a)(3) of this subpart, the owner or operator shall monitor the following information for all heats covered by the test:
      
      (1) Charge weights and materials, and tap weights and materials;

For the new 40 CFR part 60, subpart AAb subpart, the lb/ton calculations require the value for the amount of steel tapped during the PM compliance tests, as above in 40 CFR 60.274b(h)(1), in order to show compliance with 40 CFR 60.272b(a)(1). 


Comment 1-21: EAF National Emission Standards for Hazardous Air Pollutants (NESHAP) 40 CFR part 63 subpart YYYYY requires concentration limits. 

Commenter 0095-Steel Co. stated the maximum achievable control technology (MACT) standard in the NESHAP for EAFs (40 CFR part 63, subpart YYYYY) independently limits PM emissions from each EAF to 0.0052 gr/dscf. The proposed rule would thus have the result of subjecting facilities to both a lb/ton limit (via NSPS) and a concentration limit (via NESHAP). Since facilities will still have to track gr/dscf anyway to comply with NESHAP limits, it would be inefficient and unreasonable to also require a lb/ton limit.

Response 1-21: For the EPA's response to issues raised in this comment in regard to the propose lb/ton PM standard, see the EAF NSPS preamble to the final rule, section IV.A.2.


Comment 1-22: The EPA must define how to calculate lb/ton steel during periods of zero production.

Commenter 0093-Steel Co. asserted that baghouses are operated even when steel is not produced. The EPA must clearly identify and explain how an emission rate is calculated for periods of collection, capture, and control of melt shop emissions during periods of zero production.

Response 1-22: Compliance tests for all EAF baghouses, not just those collecting melt shop emissions, only should be performed when steel is being produced. The requirement to record steel production during compliance testing is already included in 40 CFR 60.274b(h)(1) as well as in the previous NSPS in 40 CFR 60.274(i)(1) and 40 CRF 60.274a(h)(1)--a requirement that indirectly requires steel production to be occuring. To make this fact clear, we have added this requirement explicitly to the final rule. However, if a compliance test is being performed during a batch of steel production, if the batch is completed and no more steel is being produced (i.e., tapped), if PM is still being emitted from the EAF and, consequently, the baghouse, then the compliance test should continue to collect PM until all PM emissions cease. All the PM emissions should be included in the test run total because these emissions result from the steel produced, regardless of whether the emissions are entirely concurrent with the steel tapped.


Comment 1-23: Air-to-cloth (A/C) ratios from integrated iron and steel industry should not be used for EAFs.

Commenter 0097-Industry Trades stated that the EPA offers no explanation why using integrated iron and steel industry (II&S) baghouse data was relevant to EAF baghouse controls in the first place or why the EPA presumed that relative rank placement of five facilities along a ranking of II&S baghouse A/C ratios allowed the EPA to presume those facilities' PM emissions were based on control through a baghouse with the same A/C ratios. Moreover, the use of II&S data is inexplicable because the EPA has in its possession the A/C ratios for many plants with EAFs. This information was available to the EPA in the docket for the 40 CFR part 63, subpart YYYYY NESHAP for EAFs and the EPA even summarized the A/C ratios for the EAF baghouses that were operated during these performance tests. 

Commenter 0097-Industry Trades continued the EPA's derived average, median, minimum, and maximum A/C ratios are all incorrect. The derived A/C ratios misstate the actual A/C ratios reported by the three model facilities for which the EPA had actual performance test data (Model Plants A, B, and E). For instance, Model Plant E is the North American Stainless facility in Ghent, Kentucky (NAS-KY), which operates four baghouses. For those four baghouses, the facility reported to the EPA A/C ratios of 4.1, 4.5, 4.5, and 5.0 ft/min -none of which are close to the EPA's erroneously derived A/C ratio of 7.2. 

Response 1-23: For the EPA's response to issues raised in this comment in regard to the propose lb/ton PM standard, see the EAF NSPS preamble to the final rule, section IV.A.2.


Comment 1-24: The EPA must conduct a more robust cost analysis. 

Commenter 0093-Steel Co. questioned whether the EPA researched more recent air quality permit applications and relevant costs associated with the application of baghouses, including the capture systems. It appears that the EPA did not contact baghouse manufacturers and the suppliers of fabric filter materials to determine current costs of control equipment. The information in the docket does not include such information. The EPA admittedly did not have the data necessary to assess the baghouse air-to-cloth (A/C) ratio and compare to the actual particulate control performance used in its overall determination and instead the EPA applied A/C ratio data from unrelated performance tests. The EPA should re-evaluate its proposal and conduct a more robust cost analysis prior to finalizing the standard of performance for EAFs.
 
Response 1-24: The baghouse cost data used in the analysis for the limit based on BSER was taken from the sixth edition of the EPA Cost Manual, which is the latest version that is available. While we expect the emissions and operation of EAF to be specific to the industry, we do not expect the baghouse cost parameters to change with each industry, unless noted. All EAF baghouse cost parameters were updated to 2021 values and values specific to EAF were used for flowrates and A/C ratio. The detailed cost information used in the lb/ton analyses for the EAF NSPS proposal were described in the "Cost Memorandum."[2] 

We disagree with the commenter that operating air permits or permits of any kind include cost information. The EPA has searched and reviewed available permits for 15 Nucor facilities and 3 recently-built EAF facilities by other companies (see Table 1-5) and found no relevant costs associated with baghouses or capture systems. A search was done for an additional nine facilities but no permits were found in state webpages. In the preamble for the EAF NSPS proposal, the EPA requested detailed information on EAF source category-specific costs to further inform the development of the final rule (87 FR 29717: "The EPA seeks comment regarding this cost analysis and seeks detailed information on EAF source category-specific costs to further inform the development of the final rule.") While we appreciate the suggestions submitted by the industry for sources of cost data, the commenters did not identify any more detailed facility-specific costs and cost parameters or, minimally, case studies. 

Since proposal, we have determined that a facility in Chicago, Finkl Steel, has an operating permit from 2018 and construction permit from 2008 that included a 0.19 lb/ton PM limit on melt shop baghouse emissions that is very similar to the PM lb/ton limit for BSER in the proposed and final rules. So, despite the limited amount of cost information, the EPA analysis appears to be similar to an independent source. 

Ultimately, the EPA proposed and is finalizing the lb/ton value represented by the highest emitting facility in the data set. This value for lb/ton was determined to be cost-effective and is a limit that is demonstrated by all the sources in the EPA/EAF data. This emission limit is intended to reflect the degree of emission limitation achievable through application of the BSER.

Table 1-5. Review of 18 Permits for Cost Information  -  Alphabetical By State
                      Facility Name - Subject to the NSPS
                                   City Name
                                     State
                             Permit Document Type
      Costs Associated With Application of Baghouses or Capture Systems?
ArcelorMittal/Nippon Steel 
Calvert
                                      AL
Title V renewal
                                      No
Nucor Steel Decatur
Trinity
                                      AL
Title V renewal
                                      No
Nucor Steel Tuscaloosa, Inc.
Tuscaloosa
                                      AL
Title V renewal
                                      No
Big River Steel (U.S. Steel)
Osceola
                                      AR
Final Title V
                                      No
Nucor Steel  -  Arkansas
Blytheville
                                      AR
Final Title V
                                      No
Nucor-Yamato Steel Co.
Blytheville
                                      AR
Final Title V
                                      No
Nucor Steel Florida
Frostproof
                                      FL
Final Title V
                                      No
Nucor Steel Kankakee, Inc.
Bourbonnais
                                      IL
Final Title V
                                      No
Nucor Steel Indiana
Crawfordsville
                                      IN
Title V renewal
                                      No
Nucor Steel Gallatin, LLC
Ghent
                                      KY
Final Title V
                                      No
Nucor Steel Brandenburg
Brandenburg
                                      KY
Title V (proposed)
                                      No
Nucor Steel Hertford County
Cofield
                                      NC
Final Title V
                                      No
Nucor Steel Auburn, Inc.
Auburn
                                      NY
Title V renewal
                                      No
Nucor Steel Marion, Inc.
Marion
                                      OH
Final Title V
                                      No
CMC Steel Oklahoma
Durant
                                      OK
Final Title V
                                      No
Nucor Steel Berkeley/Yamato Steel
Huger
                                      SC
Final Title V
                                      No
Nucor Steel Longview, LLC. 
Longview
                                      TX
Final Title V
                                      No
Nucor Steel Seattle, Inc.
Seattle
                                      WA
Final Title V
                                      No





2.0	40 CFR PART 60, SUBPART AAb BSER FOR MELT SHOP OPACITY


Comment 2-1: The proposed 0 percent melt shop opacity limit disregards workers' safety by requiring the closure of roof and buildings.

Commenters 0093 and 0098-Steel Co.'s, and 0097-Industry Trades stated the proposed 0 percent melt shop opacity limit disregarded workers' safety related to heat stress and material handling activities and that, therefore, the EPA should reconsider the 0 percent opacity limit. Commenter 0095-Steel Co stated the proposal did not include an analysis of impacts for closure of building openings. A review of the impacts on worker heat stress would be necessary and that the EPA had provided no justification for requiring melt shops to close all openings. 

Commenters 0093 and 0098-Steel Co.'s, and 0097-Industry Trades noted the current proposed rule did not address heat stress concerns, which was in conflict with OSHA's current Heat Stress Initiative and National Enforcement Program that identified "iron and steel mills" specifically as a high hazard industry for heat stress. Commenters stated that safe melt shop operation required air flow to minimize potential heat stress on workers and equipment. The commenters claimed that negative pressure alone was not sufficient to maintain proper airflow through the melt shop and that cross drafts were necessary and doors and other access points needed to be open. One commenter had significant concerns about employee health impacts from the proposed totally enclosed melt shop, particularly for its Mobile, Alabama, which is facility located in an extreme climate area, as the proposed changes could cause greater heat stress on employees and would necessitate design and structural changes that the EPA failed to consider in its proposal. Commenter 0091-State stated that 100 percent capture and 0 percent opacity may not be safe. Commenter 0097-Industry Trade noted when evaluating two different control systems, the EPA may not simply choose the most cost-effective air pollution control system if it potentially has adverse impacts on the health and safety of workers within the melt shop. Commenter 0095-Steel Co stated that facilities need to allow air changes to protect worker health and safety.

Commenters 0093 and 0098-Steel Co.'s, and 0097-Industry Trades referenced the 1984 amendments, which dismissed the option for a closed roof configuration to achieve 0 percent opacity due to the impacts of heat stress on worker safety and equipment functioning. Commenter 0095-Steel Co cited statements made by the EPA in the 1984 40 CFR part 60, subpart AAa rulemaking that "the visible emission limits were selected-based on the performance of the capture and control technologies that served as the basis for Regulatory Alternative B (partially open roof monitor)." and that "Regulatory Alternative C (closed roof) was not considered suitable as the basis for national standards of performance because it is based on a closed roof configuration which may aggravate worker and equipment heat stress problems." 

Response 2-1: For the EPA's response to issues raised in this comment in regard to the proposed melt shop opacity standard, standard, see the EAF NSPS preamble to the final rule, section IV.B.2.


Comment 2-2: Zero percent shop opacity is functionally equivalent to total building enclosure.
Commenter 0097-Industry Trades stated that although the EPA did not specifically propose "total building enclosure", the effect of a combined 0 percent shop opacity and a requirement for no holes or other openings in the melt shop was functionally equivalent, which the Agency previously considered unviable and unsafe for workers. The commenter referenced the EPA's past citing of the Secondary Lead Smelting NESHAP as a basis for asserting the achievability of 0 percent opacity. The commenter noted the following differences between lead smelters and EAF steel mills: lead smelters did not need large bay doors, as required by EAF steel mills for movement of large ladles, stacks of billets, dump trucks full of slag, and other equipment/materials; lead smelters operated at much lower temperatures (lead melts at 621.5[o]F and steel melts at 3000[o]F); and due to the high heat, EAF steel mills required larger makeup air and greater air flow. The commenter stated that such air flow needs rendered maintaining negative pressure within and throughout the melt shop impossible in any practical sense. Commenter 0095-Steel Co stated to comply with the 0 percent opacity standard for shop opacity and the requirement to eliminate any openings, plants would need to enclose their melt shop EAF operation. Commenter 0091-State stated that 100 percent capture and 0 percent opacity was difficult to achieve in practice.

Response 2-2: The proposed rule 40 CFR part 60, subpart AAb did not require a closed roof nor a totally enclosed melt shop and neither does this final action. The 0 percent shop opacity limit only necessitates that no particles be emitted from the shop; it does not restrict air flow from exiting or entering the shop. In order to comply with the finalized regulations, facilities would need to clean the air of particles before allowing hot air to exit the shop opening. This is the basis for the addition of a partial roof canopy used in the proposed cost estimates.

The 0 percent shop opacity limit in proposed 40 CFR part 60, subpart AAb does not include a specific requirement to plug holes in buildings nor does it require that buildings not have doors or, if the buildings have doors, that they can't use them. The 0 percent shop opacity limit in 40 CFR part 60, subpart AAb only requires that no particles be present in air that leaves the shop. Holes in walls and doors left open for long periods, other than the moving in and out of equipment, or open at all times should not be part of a facility's ventilation strategy or building management plan in order to achieve the proper ventilation for worker safety and operation of equipment. When particles exit melt shops through holes in the walls, the result illustrates a way to circumvent the shop limit and is evidence of a poor ventilation strategy. Even if it were clean air exiting melt shops through holes in the walls or doors being left open, the result isn't only a compliance issue, but is a sign of poor ventilation and could also lead to reducing the structural integrity of the building.


Comment 2-3: The opacity limit for BSER causes adverse environmental impacts.

Commenter 0097-Industry Trades stated the EPA must consider other non-air-quality environmental impacts of a standard, and thus the [opacity] limit based on BSER may not always reflect the lowest air emission standard achievable if the lower standard could result in other adverse environmental impacts. The commenter noted when evaluating two different control systems, the EPA may not simply choose the most cost-effective air pollution control system if it would create adverse environmental impacts on other media. 

Response 2-3: The commenter does not mention what adverse environmental impacts in particular could result by finalizing the proposed rule. There are no known impacts to water and no large impacts to solid waste due to any of the control technologies that are BSER in the proposed rule. The additional PM collected when 6 percent opacity is lowered to 0 percent is a small amount compared to the total PM collected facility-wide and is easily accommodated in the baghouse waste handling system, especially considering that most facilities already are operating near 0 percent opacity for their shop. Additionally, most EAF facilities recycle their baghouse dust. For an additional discussion on possible adverse environmental impacts, see "Section IV. Summary of Cost, Environmental, and Economic Impacts, A. What are the air quality and other environmental impacts?" in the proposed EAF NSPS rule:

      Solid wastes would increase slightly, approximately 17 tons per facility, on average, with the additional PM collected to meet 0 percent shop opacity limit under proposed 40 CFR part 60, subpart AAb as compared to current facilities meeting opacity limits under 40 CFR part 60, subparts AA and AAa [i.e., 6 percent opacity]. The small increase in solid wastes would be the same for both the carbon and specialty steel shops. However, most EAF dust is recycled to reclaim zinc. 

(87 FR 29722.) The EPA has concluded that this modest increase in the solid wastes produced does not outweigh the emission reductions gained by the more stringent limit.


Comment 2-4: Top-charged EAFs must have building openings for bringing in materials to EAF and for other equipment issues.

Commenter 0095-Steel Co. noted that top-charged EAFs must have building openings for bringing in materials to the EAFs (e.g., crane transfer of scrap and other material transfer equipment for process inputs such as lime and carbon). Commenters 0093 and 0098-Steel Co.'s, and 0097-Industry Trades Commenters also noted the movement of scrap and finished products required large openings to avoid pinch points and other material handling hazards.

Commenters 0093 and 0098-Steel Co.'s, and 0097-Industry Trades stated elevated wind conditions and ambient pressures outside of the shop will cause pressure shifts within the melt shop and reduce capture efficiency. 
Commenter 0095-Steel Co commented that openings in melt shop are needed to protect machinery from heat stress, and to allow efficient combustion. In addition, eliminating openings could affect the capture efficiency of the existing canopy and require plants to redesign melt shops and/or existing canopies. 

Commenter 0095-Steel Co stated if a stronger vacuum resulted from changes to the building or capture system, then structural reevaluation of equipment and structures would be required to ensure they can withstand the additional pressure changes. A stronger vacuum could also affect the temperature in the ducts to the baghouse, which could impact efficiency and damage bags in the baghouse. Any additional power requirements for the fans in the capture system would increase power consumption and indirectly increase emission from electricity generation. Small changes to the building structure may impact airflow and pressure of the canopy and could result in significant cost to the facility. Small changes in blower sizing can cost millions of dollars as changes to motors, cables, upstream power supply, drives, etc. may be needed. 

Commenter 0095-Steel Co stated the proposal included no analysis of impacts for closure of building openings. A review of the impacts on equipment heat stress and operational practices would be necessary and that the EPA had provided no justification for requiring melt shops to close all openings.

Response 2-4: The proposed rule 40 CFR part 60, subpart AAb does not require a closed roof nor a totally enclosed melt shop. The 0 percent opacity limit only necessitates that no particles be emitted from the shop and does not restrict air flow from exiting or entering the shop. Facilities would need to clean air of particles before allowing hot air to exit the shop opening. This is the basis for the addition of a partial roof canopy used in the proposed cost estimates.

Elevated wind conditions and changes in ambient pressures (not including tornados and hurricanes) are expected to cause only small changes in pressure within the melt shop, which can in turn be accommodated by small changes in the melt shop ventilation rate to ensure an air velocity vector caused by capture equipment that is always going into the melt shop. The revision to the proposal included in the final rule for charging allows facilities to meet the same 6 percent opacity standard currently being met during charging, when the roof may be open.

As similarly stated in the response to Comment #2-2, the shop 0 percent opacity limit in proposed 40 CFR part 60, subpart AAb did not include a specific requirement to plug holes in buildings nor does it require that buildings do not have doors or, if they have doors, that they can't use them. The 0 percent shop opacity limit in 40 CFR part 60, subpart AAb only requires that no particles be present in air that leaves the shop. Holes in walls and doors left open for long periods other than the moving in and out of equipment or open all the time should not be part of a facility's ventilation strategy or building management plan to achieve the proper ventilation for worker safety, equipment integrity and safety, and proper operation of capture and ventilation equipment. Particles exiting melt shops though holes in walls are a way to circumvent the shop limit. However, clean air exiting melt shops are not a compliance issue but could lead to reducing the structural integrity of the building. If a facility is concerned about worker breathing space air flow, there are industrial ventilation experts who are widely available to help facilities determine what devices can be used to ensure the correct amount of air for worker safety, equipment integrity safety, and proper operation of capture and ventilation equipment, all which also should minimize airborne particulate for health purposes.


Comment 2-5: The EPA does not have the authority to require existing plants to redesign and make structural changes to their melt shops without providing justification.

Commenter 0095-Steel Co said the EPA does not have the authority under CAA section 111(b) to require existing plants to redesign and make structural changes to their melt shops. The commenter said it was "arbitrary and capricious" for the EPA to depart from their prior policy without providing justification. 

Response 2-5: For the EPA's response to issues raised in this comment in regard to the proposed melt shop opacity standard, standard, see the EAF NSPS preamble to the final rule, section IV.B.2.


Comment 2-6: The EPA used a limited data set that was not indicative of continuous long-term performance and did not support a finding that 0 percent opacity was adequately demonstrated.

Commenters 0095-Steel Co. and 0097-Industry Trades stated the EPA's dataset for EAF steel mills is selective and not representative of the full scope of operations at these facilities. The commenters stated that the EPA purported to have based the proposed 40 CFR part 60, subpart AAb shop opacity limit on individual performance testing reports from a total of 13 of 31 EAF steel mills, which was less than half. Commenters noted that most facilities (16 out of 31) were unable to maintain 0 percent shop opacity throughout the duration of the performance tests. Thus, as the majority of facilities in the EPA's database did not maintain 0 percent opacity, for the short duration of performance testing plainly demonstrated that 0 percent shop opacity was not adequately demonstrated. Commenter 0093-Steel stated the EPA's proposal of 0 percent opacity from the melt shop was based on limited information from opacity tests conducted at 31 facilities, of which less than half achieved the 0 percent melt shop opacity requirement. 

Commenters 0095-Steel Co. and 0097-Industry Trades stated that the short-term observations conducted during a stack test were taken for a few hours under a specific set of conditions and were not representative of long-term compliance capability, and as such, could not account for routine operating variability and the full range of operating conditions that may affect opacity. Commenters stated that the subset of data the EPA relied upon did not include longer-term operating performance of the identified mills; yet NSPS, as defined by BSER, must account for what was achievable and adequately demonstrated by a wide variety of facilities operating under a wide variety of conditions, not simply show that the standard was achieved at a model plant for a short period of time. Commenters also noted that the data collected by the EPA generally showed that the more years of opacity data reviewed for a given facility, the higher the maximum melt shop opacity.
Response 2-6: For the EPA's response to issues raised in this comment in regard to the proposed melt shop opacity standard, standard, see the EAF NSPS preamble to the final rule, section IV.B.2.


Comment 2-7: The EPA's limited data set is not representative of performance during "charging and tapping"; 0 percent opacity should not apply to charging and tapping.

Commenters 0095-Steel Co. and 0097-Industry Trades stated the EPA's opacity data set did not adequately demonstrate that a 0 percent opacity limit could be consistently achievable across the full spectrum of expected operating conditions. Commenters said the vast majority of the opacity measurements in the data set were based on measurements taken during the melting and refining stage of production (as required under 40 CFR part 60, subpart AAa), and thus did not demonstrate that 0 percent opacity had been consistently achieved during charging or tapping, which was the established period with the greatest potential for uncaptured emissions to escape the melt shop. The commenters noted that most EAF steel mills were designed such that the primary emission controls (e.g., DEC) could not be engaged while the furnace roof was off during charging and tapping. 

Commenters 0095-Steel Co. and 0097-Industry Trades referenced previous rulemaking to corroborate their statements that EPA did not consider its own historical information. One commenter referred to background documents for earlier NSPS rulings stating that in those documents, the EPA concluded that facilities utilizing DEC were likely to have a visible plume during charging and tapping and could not meet 0 percent opacity on a continuous basis. One commenter referenced the 1983 rulemaking docket stating it included only seven hours of shop opacity data from some portion of the charging and tapping phase, and that such limited data from four decades ago was not representative of, or sufficient to, to characterize current melt shop operations. The commenters said these previous findings by the EPA contradict the current proposal that 0 percent opacity was achievable on a continuous basis. 

Commenter 0097-Industry Trades provided (CBI) summaries of long-term shop opacity data from the 13 facilities identified by the EPA as achieving the 0 percent standard, and noted that most of the opacity data was collected only during melting and refining, and not during charging and tapping. The commenter stated their summaries demonstrated that 0 percent melt shop opacity was not continuously achieved by the 13 mills cited as exemplars. The commenter noted in a reference that they would readily provide the confidential data to the EPA upon request.

Commenter 0098-Steel Co. stated that the current design at their facilities included DEC and a baghouse with a canopy, which under the proposed rule was considered the optimal design, yet it appeared the EPA did not include opacity data from their facilities in the limited data set. The commenter noted they fully complied with current limits in 40 CFR part 60, subparts AA, AAa, including opacity; but their facility data showed that compliance with a 0 percent opacity limit at all times per the proposed standard could not be met continuously due to the production process variability and the raw material inputs. The commenter stated it was possible for the melt shop to experience an opacity greater than 0 percent during charging and tapping when the DEC system was disengaged, and there were other sources of opacity from concurrent operations (e.g., vacuum tank degasser operations, the LMF, and the Caster). 

The commenter said the EPA in the proposed rule stated 0 percent opacity could be achieved utilizing a canopy over the furnace with an open roof monitor elsewhere. The commenter operated its facilities under such a configuration and did not meet 0 percent opacity on a continuous basis; thus, the EPA's data set was flawed and not representative of the steel manufacturing operation.

Response 2-7: For the EPA's response to issues raised in this comment in regard to the proposed melt shop opacity standard, standard, see the EAF NSPS preamble to the final rule, section IV.B.2.


Comment 2-8: The EPA relied on one mill for setting the limit based on BSER for shop opacity that is not representative of the industry.

Commenters 0095-Steel Co. and 0097-Industry Trades disagree with the EPA's inclusion of an EAF with a shaft furnace design as the single facility with permitted limits of 0 percent opacity and determined to be BACT, as the facility had a unique design and was not representative of the EAF steel industry. The commenters explained that a shaft furnace continually feeds scrap metal to the EAF via a closed conveyor, such that the primary emission control (e.g., DEC) was always engaged, including during charging, which was unlike the more common bucket-fed EAF steel furnace designs. Commenters also noted the shaft furnace operations had different emissions-profiles compared to traditional bucket-fed furnaces and also were limited to certain scrap types and production rates. Commenters stated the shaft design was outdated and uncommon; it was not the preferred design for recent construction or operation, as there were operational difficulties in the past. The commenters stated that just because a shaft furnace design could have lower opacity emissions did not make it a preferred technology for efficiently recycling scrap steel into steel products. Commenters 0095-Steel Co. stated the emissions profile of this facility and its shop opacity performance were not representative of the majority of steel mills, which are bucket-fed, and thus cannot be considered in setting the limit based on BSER for the sector. 

Commenters 0095-Steel Co. and 0097-Industry Trades stated if the shaft furnace facility was the basis by which the EPA presumes that 0 percent opacity can be met on a long-term basis, then the EPA must analyze the cost of replacing traditional EAFs with a shaft furnace configuration. Commenters said the EPA does not have the authority to change the design of facilities in this manner, and the cost for such a dramatic design change would be astronomical. Accordingly, the fact that a different type of furnace design may be able to meet 0 percent opacity cannot serve as a basis for determining that 0 percent opacity was adequately demonstrated for all EAF furnace types. 

Response 2-8: We acknowledge the commenter's contention that the NUCOR Marion, OH facility has a shaft furnace that is no longer in use within the EAF industry, and where the primary emission control (e.g., DEC) was always engaged, including during charging. However, the 0 percent melt shop opacity was a result of an order to enclose the melt shop to reduce/eliminate manganese emissions from the EAF shop within 150 days, not as a result of the shaft furnace. See the March 14, 2012, Director's Final Findings and Orders, Nucor Steel Marion, Inc., Ohio Environmental Protection Agency, section V. Orders, Melt Shop Enclosure:

      Respondent shall build the enclosure identified in PTI P0105283 to capture the particulate and Mn emissions from the Melt Shop housing emissions units P004, P005, P007, P012, P015, and P903, in accordance with the following_ schedule:
      a.	Within seven (7) days of the effective date of these Orders, Respondent shall perform a structural building analysis of the Melt Shop to determine load capacity of the existing trusses, columns, foundations, and soil loading;
      b.	Within ninety (90) days of the effective date of these Orders, Respondent shall:
            i.	design a canopy hood evacuation system for emissions unit P004, including structural members and ductwork with connection points;
            ii.	design a canopy hood evacuation system for emissions unit P903, including structural members and ductwork with connection points;
            iii.	develop equipment specifications for the purchase of ductwork, baghouse, fans, dampers, expansion joints, foundations, and stack; and
            iv.	solicit vendors and obtain quotes for the purchase of all equipment;
      c.	Within one hundred fifty (150) days of the effective date of these Orders, Respondent shall:
            i.	Finalize the process of equipment selection and purchase orders issuance; and
            ii.	complete the installation of all equipment, utilities, and accessories;
      d.	Within six hundred thirty (630) days of the effective date of these Orders, Respondent shall complete construction of the enclosure system for the Melt Shop and have the Melt Shop fully operational;

Further, the NUCOR Marion facility was not the only facility that was found via permits to be achieving 0% melt shop opacity. Finkl Steel in Chicago also has a 0 percent melt shop opacity limit (specific as "no visible emissions"), which was determined to be BACT for PSD as seen in their 2018 operating permit[10] and in their 2008 construction permit.[11] The construction permit describes how the EAF melt shop at Finkl also is totally enclosed and the furnace is not a shaft furnace design. In addition, as discussed above in Chapter 1, the Finkl Steel's 2018 operating permit included a limit of 0.19 lb/ton PM for melt shop baghouse emissions that is very similar to the PM limit based on BSER in this final rule. Additionally, Gerdau Macsteel in Monroe, MI, had a 2014 permit to install that included a limit on visible emissions from openings and vents in the upper half of the shop building to not exceed a 6-minute average of 0 percent opacity during operation of the electric arc furnace. The Gerdau Macsteel EAF also was not a shaft furnace. Lastly, the 19 EAF facilities in the EPA EAF dataset that achieved 0% opacity (rounded to one significant figure) also were not shaft design. See Table 5 and discussion in the preamble to the final rule as well as the Emissions Memoranda[6] for both proposed and final rules located in the docket for this rulemaking (Docket ID No. EPA-HQ-OAR-2002-0049).


Comment 2-9: The 0 percent opacity limit should not apply to hot metal charging facilities.

Commenter 0096-Steel Co. stated the EPA's proposed 0 percent melt shop opacity limit in 40 CFR part 60, subpart AAb was not achievable or adequately demonstrated for an EAF that employed hot metal charging and, [therefore], this type of EAF should be exempt from the opacity limit should it be finalized. The commenter noted the current industry standard approach for charging an EAF was to introduce scrap metal to the furnace, potentially along with fluxing agents, direct-reduced iron or other iron-bearing material; all of this material was solid and not molten. The commenter said in their process using hot metal charging, molten metal from a blast furnace was charged to the EAF in addition to scrap metal. The molten metal was transferred by a ladle and introduced to the EAF via a hot metal chute. The commenter stated that available historical data from 2005-2008 for a hot metal charging EAF demonstrated that a 0 percent opacity limit was not achievable nor adequately demonstrated. Additionally, the commenter stated that it would be inappropriate for the EPA to take a position that a source should not employ hot metal charging if doing so would prevent compliance with a proposed 0 percent shop opacity limit, citing New Source Review and NSPS. The commenter objected to the inclusion of a 0 percent shop opacity limit in proposed 40 CFR part 60, subpart AAb, but if the EPA included such a limit in the final rule, that limit must exclude EAFs that conduct hot metal charging and specify that the 6 percent shop opacity standard (as in existing 40 CFR part 60, subpart AAa) be applied to hot metal charging facilities. 

Commenter 0097-Industry Trades agreed with commenter 0096-Steel Co., above, that the proposed 0 percent shop opacity limit should not apply to an EAF that utilized "hot metal charging." The commenter clarified that they were aware of only one EAF steel facility operating in the U.S. that charged "hot metal" (i.e., molten iron from a blast furnace), and that the EPA had shop opacity data from this unique facility that included 77 nonzero 6-minute average opacity readings from 2005-2008.

Response 2-9: The EPA learned through follow up with Commenter 0097 (Industry Trades) that the EAF facility in question was no longer operating because the blast furnace (source of hot metal) was demolished a few years ago and that Commenter 0097 was asking for "another commenter." Commenter 0097 did not identify the other commenter or even provide certainty that another such blast furnace was being operated elsewhere. Therefore, we did not find it a basis to create a rule exemption based on a hypothetical future facility. Also, if this particular unit or another unit at another facility restarts in the future without modification, then the unit will not be subject to the new NSPS finalized in this action, because these standards only apply to facilities constructed, reconstructed, or modified after May 16, 2022]. We note that the commenter referred to historical data showing that 0 percent opacity was not achieved when the facility was active, but included no technical reason as to why hot molten iron would cause increased particulate emissions. It would seem that using molten iron to produce molten steel would not generate solid particle emissions because there is no phase transition expected with molten, i.e., liquid iron, or at least not more than an EAF charged with only dry solid scrap. In addition, the final EAF NSPS does not require 0 percent opacity during charging; instead, 6 percent opacity, as in current EAF NSPS subpart AAa, is the requirement for charging in new subpart AAb. Therefore, even if the EAF at the other facility mentioned by the commenter (re)starts with modification, it would be subject to the same level of opacity that it would have been required to meet, absent the new AAb subpart. 


Comment 2-10: The EPA must retain the current 6 percent melt shop opacity limit because the proposal ignored the practical operations of EAFs, including material handling.

Commenter 0093-Steel Co. stated the EPA should retain the current standard of 6 percent opacity without exceptions because the proposal ignored the practical operations of EAFs, including material handling. The commenter continued that melt shops contain various pieces of equipment and require the transportation of scrap into the shop, which may create situations where dust exits the melt shop. Differentiating the opacity resulting solely from any affected EAF would be impossible and lead to confusion and misrepresentation of emissions attributed only to the operations of any affected EAF or AOD vessel. 

Response 2-10: The EPA disagrees that the EAF NSPS rule should retain the current opacity standard during melting and refining. The opacity testing at 30 facilities used in the proposal to set the limit based on BSER was performed during typical or "normal operations" of EAF facilities as required by the NSPS. The CAA section 114 request in 2010 asked facilities to provide data representative of their normal operation. No facility that submitted data stated that the opacity test reports provided to the EPA were not performed during typical operation nor was there any indication in the test reports that the facility was not in normal operation. Therefore, we assume that the opacity data used in this rule represents normal operation of EAF. The results showed that 0 percent opacity was achieved by 15 of the 30 facilities, with very low opacity achieved by the other 15 facilities for an overall average of data from 30 facilities of 0.14 percent.

Additionally, it would be difficult to require facilities to purposely "handle their materials" during any shop testing performed to comply with a CAA section 114 information request because this might interfere with their normal operations and could be construed as manipulating the data. The opacity data acquired for this rulemaking was obtained from a CAA section 114 test request performed in 2010 to provide facility information to the EPA to evaluate mercury emission; therefore, opacity was not the focus of the CAA section 114 request. In order to comply with the timeline of the consent decree, it would not have been possible to send and receive, and analyze a CAA 114 request that was tailored to the goals of the NSPS. However, melt shop emissions, in general, are assumed to include all melt shop activities, which include material handling. See rule text below from 40 CFR part 60, subparts AA and AAa (and also in proposed 40 CFR part 60, subpart AAb):

      40 CFR 60.275a&b (h) Where emissions from any EAF(s) or AOD vessel(s) are combined with emissions from facilities not subject to the provisions of this subpart, determinations of compliance with 40 CFR 60.272b(a)(3) will only be based upon emissions originating from the affected facility(ies), except if the combined emissions are controlled by a common capture system and control device, in which case the owner or operator may use any of the following procedures during an opacity performance test and during shop opacity observations:
      
      (1) Base compliance on control of the combined emissions; or
      (2) Utilize a method acceptable to the Administrator that compensates for the emissions from the facilities not subject to the provisions of this subpart.


Comment 2-11: EPA did not properly evaluate the cost of compliance for a source as a result of modification.

Commenter 0093-Steel Co. commented the EPA had not properly evaluated the cost of compliance [with 0 percent opacity] for a source as a result of modification. 

Response 2-11: For the EPA's response to issues raised in this comment in regard to the proposed melt shop opacity standard, see the EAF NSPS preamble to the final rule, section IV.B.2.


Comment 2-12: The proposed rule should allow for alternative capture and opacity limits for EAF facilities unable to isolate EAF emissions from that of other processes.

Commenter 0091-State said measuring 0 percent opacity from a melt shop, which was unable to isolate [the EAF] from other processes, was impractical due to the presence of visible emissions from additional sources. The commenter added that the EPA should consider including provisions for sources to request alternative capture and opacity limits when specific factors make achieving the proposed limits infeasible. 
Response 2-12: We disagree with the commenter that alternative limits are needed. Moreover, there are no capture limits in the EAF NSPS rules. We also note that the final rule already has alternatives, consistent with those provided by the prior NSPS, to accommodate the situation of emissions from other processes present along with EAF/AOD emissions, and demonstrate compliance with 0 percent melt shop opacity solely due to EAF operations:
      40 CFR 60.275b
      (h) Where emissions from any EAF(s) or AOD vessel(s) are combined with emissions from facilities not subject to the provisions of this subpart, determinations of compliance with 40 CFR 60.272b(a)(3) will only be based upon emissions originating from the affected facility(ies), except if the combined emissions are controlled by a common capture system and control device, in which case the owner or operator may use any of the following procedures during an opacity performance test and during shop opacity observations:
      (1) Base compliance on control of the combined emissions; or
      (2) Utilize a method acceptable to the Administrator that compensates for the emissions from the facilities not subject to the provisions of this subpart.
      
Additionally, there is always the option to shut off the source of emissions from other processes for the opacity testing, which takes a total of six minutes. As a historical point, shutting down is still listed as an option in 40 CFR 60.275(g), option (2):

      (1) Base compliance on control of the combined emissions. 
      (2) Shut down operation of facilities not subject to the provisions of this subpart. 
      (3) Any combination of the criteria of paragraphs (g)(1) and (g)(2) of this section.


Comment 2-13: The EPA must exclude periods of malfunctions and upset conditions if a 0 percent opacity standard for melt shop emissions at all times is implemented.

Commenter 0093-Steel Co. stated if a 0 percent opacity standard for melt shop emissions at all times was implemented, the EPA must exclude periods of malfunctions and upset conditions. The commenter explained that malfunctions occurred during the melting and casting operations that required extraordinary measures for corrective action, such as a "breakout." In an extremely dangerous situation, breakouts occurred when molten steel escaped from one or more mold strands at the caster or during casting. The commenter stated after a breakout and subsequent corrective action, emissions were generated and may exit the melt shop, and those emissions should not be considered in determining compliance with the 0 percent opacity melt shop requirement. The commenter said the EPA's proposed approach lacked an understanding of the significant dangers, risks, and related emissions associated with "breakouts" and other malfunction events that occur during the steelmaking processes, and the EPA should reconsider the 0 percent melt shop opacity standard.
Response 2-13: For the EPA's response to issues raised in this comment in regard to the proposed melt shop opacity standard, standard, see the EAF NSPS preamble to the final rule, section IV.B.2.


Comment 2-14: Canopy hood control costs. 

Commenter 0097-Industry Trades stated the EPA did not examine advances in control technologies, process operations, design or efficiency improvements, or other systems of emission reduction, that are "adequately demonstrated". Rather, the EPA looked to a decades-old background information document (BID), concluded that "[c]anopy hoods are a common method of controlling fugitive EAF emissions," and assessed costs for: adding a partial roof canopy (segmented canopy hood, closed roof over furnace, open roof monitor elsewhere) to collect PM emissions that might otherwise escape through the shops to achieve complete control of melt shop fugitives. 

Commenter 0097-Industry Trades and 0095-Steel Co. stated the EPA did not analyze whether canopy hoods were used by the 19 facilities that recorded 0 percent opacity during performance testing or were absent from the nine facilities that recorded the highest opacity during performance tests. This information was available to the EPA in the docket for the 40 CFR part 63, subpart YYYYY NESHAP for EAFs  -  the same docket that supplied the majority of the performance test data the EPA used in this rule. The EPA's own review of the survey responses in the 40 CFR part 60, subpart YYYYY docket in June 2005 shows that EPA knows that canopy hoods were used to capture fugitive emissions from 32 of the 38 EAFs described in the CAA section 114 survey responses, and that the presence or absence of a partial roof canopy did not determine whether the facilities responding to the survey could achieve 0 percent opacity. Therefore, the EPA has no basis to now conclude for purposes of demonstrating achievability and cost effectiveness that the singular act of installing a partial roof canopy will "achieve complete control of melt shop fugitives." 

Commenter 0097-Industry Trades and 0095-Steel Co. stated the EPA's conclusion is also contradicted within the Agency's cost analysis. In order to estimate how much PM is emitted from a facility that emits 6 percent opacity, the EPA used the 1982 BID's estimate that EAFs emit an average of 29 lb/ton of uncontrolled PM emissions. The EPA then relied on the 1982 BID again to estimate that facilities emitting 6 percent opacity captured 90 percent of those emissions using a "segmented canopy hood, closed roof over furnace, open roof monitor elsewhere." This is the exact fugitive emission capture technology that the EPA's Cost Analysis presumes facilities with greater than 0 percent opacity can install to achieve 0 percent opacity. In other words, the EPA's Cost Analysis assumes that facilities with a "segmented canopy hood, closed roof over furnace, open roof monitor elsewhere" are emitting 6 percent opacity and if those facilities install a "segmented canopy hood, closed roof over furnace, open roof monitor elsewhere" they will achieve 0 percent opacity. Commenter 0097-Industry Trades and 0095-Steel Co. stated because it is arbitrary and unreasonable to assume that facilities will be able to achieve 0 percent opacity by doing nothing more than install the same systems that facilities already have installed without reaching 0 percent opacity, the EPA has failed to provide a cost estimate rationally related to reduction of opacity from 6 percent to 0 percent. 

Response 2-14: For the EPA's response to issues raised in this comment in regard to the proposed melt shop opacity standard, standard, see the EAF NSPS preamble to the final rule, section IV.B.2.


Comment 2-15: Melt shop equipment needed to achieve 0 percent opacity. 

Commenter 0097-Industry Trades stated melt shop partitions of the size necessary to meaningfully contain EAF emissions within the melt shop are not feasible in many mills given other equipment and shop design, including cranes. In particular, sizable partition walls are not feasible at many EAF steel mills because they will interfere with overhead cranes that transport scrap metal to the furnace. Similarly, transfer ladles that are carried by crane to and from the furnace for tapping molten metal would be blocked by partition walls. 
 
Commenter 0097-Industry Trades said for existing facilities that may trigger an NSPS modification in the future, achieving 0 percent shop opacity would require extensive re-engineering that would be costly and introduce practical and worker safety concerns as well. For example, one [trade] association member stated that 0 percent shop opacity could only be achieved, if at all, with near total enclosure of the EAF and doubling the flow rate of the emission control system. 
 
Commenter 0097-Industry Trades stated that only very short (and therefore marginally effective) partition walls could be installed above the crane because of the lack of space between the crane and the roof. They also noted that such short partitions deteriorated quickly due to the heat and other elements. Thus, to increase the size and collection efficiency to meet a 0 percent opacity requirement, the facility would have to raise the roof of the structure at an undetermined cost (a cost that likely would trigger a "major modification"), and potentially enclose the entire monovent, which would likely create worker safety and heat stress issues. 
 
Commenter 0097-Industry Trades added, facilities would have to increase the number and volume of fans to the baghouse, as well as require new or additional fans in the shop and additional baghouses because the facility's current baghouses are operating at close to maximum capacity. Moreover, for servicing, cranes have to be moved to a different part of the melt shop due to the partitions being so close to the top of the cranes. To achieve compliance, existing facilities such as these also would have to enclose the large openings in the casting area to prevent winds from blowing through the shop or wall off the EAF operations. Neither option is feasible; melt shops are typically long buildings with EAF, LMS, and casting in the same structure. 

Response 2-15: For the EPA's response to issues raised in this comment in regard to the proposed melt shop opacity standard, standard, see the EAF NSPS preamble to the final rule, section IV.B.2.


Comment 2-16: The Ferroalloys NESHAP cost considerations. 

Commenter 0097-Industry Trades stated the docket for the Ferroalloys major source NESHAP should have apprised the EPA of the "irrationally [sic] of its control cost conclusions." In its economic analysis of the Ferroalloys NESHAP, the EPA conceded that discussion with vendors and ventilation experts and EPA research led to a conclusion that implementation of fugitive capture and control systems is complex and that system parameters are highly dependent on specific localized parameters (e.g., building volume, process locations, and airflow). And that any plans to implement a system for ventilation to control fugitive emissions should begin with a rigorous, systematic examination of the ventilation requirements throughout the building leading to design and implementation of an enhanced fugitive capture and control system (Harris Memo: Cost Impacts of Control Options Considered for the Ferroalloys Production NESHAP to Address Fugitive HAP Emissions, at 2).

Commenter 0097-Industry Trades said the Ferroalloys analysis is predicated on "[EPA's] expectations of the controls likely to be installed as a result of facility-specific ventilation analyses, the controls likely to be needed to address risk, and existing capture and control systems at the facilities." Thus, the EPA's Ferroalloys analysis is rooted in an understanding of context-specific control mechanisms that are unique to the facilities in which they must be installed. Said differently, the EPA did not intend for this cost-analysis of BSER [sic][MACT] proposed in the [NESHAP for the] ferroalloy industry to be universally applicable to all ferroalloy facilities, let alone those outside of the ferroalloy industry. 

Commenter 0097-Industry Trades stated the Ferroalloys analysis relies on a mere two ferroalloy facilities to conduct the bulk of its cost analyses, with the only supplemental information coming from the Agency itself. To rely on such limited input, with no support for the representativeness of the two facilities selected, is problematic as it can hardly be presumed to be representative of the larger industry. 

Commenter 0097-Industry Trades added having only two ferroalloy facilities present their cost analyses means the baseline understanding of costs adopts pre-existing measures these facilities employ. In other words, adopting only these two facilities cost analyses presupposes all the conditions present in these two facilities. The EPA itself recognizes the difficulty of retrofitting a one-size-fits-all solution onto a diverse milieu of facilities. 

Commenter 0097-Industry Trades stated not only is the assumption of similar conditions problematic but relying on only two ferroalloy facilities similarly means extensive gaps about specific sites exist in the data. The EPA acknowledges this and remedies it by filling in those gaps with both "general considerations" and "facility-specific considerations." While some of these considerations are reasonable, others are too tenuous and, therefore, of little use. 

Response 2-16: In the proposed 40 CFR part 60, subpart AAb, we used the mathematical approach presented in the ferroalloys production major source NESHAP (40 CFR part 63, subpart XXX), hereafter referred to as the "Ferroalloys NESHAP," to estimate canopy costs but tailored the model to steel-producing EAF by replacing the variables of air flow and labor rates to steel EAF and by scaling (upwards) all capital costs for steel-making EAF based on the ratio of air flow at steel-making EAF to air flow at EAF at ferroalloys production facilities. [See the Cost Memorandum prepared for the proposed[2] rule for details of this analysis] Note that there are only three Ferroalloy major sources in the U.S., therefore, using two facilities in the Ferroalloy cost analysis was a clear majority for the source category.

The commenter fails to identify the assumptions that are "too tenuous and, therefore, of little use" as opposed to the assumptions that are "reasonable"; therefore, we cannot address the commenters concerns with specificity. We also cannot do an individual cost analysis for individual EAF facilities to fill the alleged "extensive gaps about specific sites [that] exist in the data," as the commenter suggests, because the finalized NSPS standards apply to future facilities of unknown individual characteristics.. The approach and assumptions that were used to modify the cost estimating procedures taken from the Ferroalloys NESHAP for the proposal for the EAF NSPS are all reasonable and the best available to the EPA.

The EPA acknowledged the cost-effectiveness of local capture, such as canopy hoods, as compared to total enclosure in its response to comments on the proposed rule for the Ferroalloys NESHAP in 2015 ("Summary of Public Comments and Responses on Proposed Rule, (76 FR 72508, November 23, 2011) and Supplemental Proposal (79 FR 60238, October 6, 2014)" May 28, 2015. (Docket Item No. EPA-HQ-OAR-2010-0895-0302), on page (p.) 120, under section 4.6, "Fugitive Emissions Capture":

      ...full-building enclosure option with negative pressure may not be feasible and would have significant economic impacts on the facilities... [H]owever, we conclude[d] that the enhanced local capture option is a feasible and cost-effective approach to achieve significant reductions in fugitive HAP emissions and will achieve almost as much reductions as the full building enclosure option.... 

Further, in the same 2015 Ferroalloys NESHAP comment and response document, EPA explains how opacity would be used to tailor the capture system parameters to meet a rule's requirement for shop opacity: 

      ...the opacity standard is a main piece of the compliance strategy for the process fugitive emissions standards. A key aspect of the opacity limit is that it can be used by the source to establish a correlation between successful operation of capture system parameter values (capture system fan motor amperes and all capture system damper positions or volumetric flow rate through each separately ducted hood that comprises the capture system) and opacity levels, and then the source can continuously monitor the parameters to help ensure the opacity limits are achieved continuously and that the capture and control system is effectively capturing fugitives on a continuous basis. This combination yields an enforceable parameter limit that is designed to ensure continuing operation of the capture system at a level consistent with operation during the performance test.
      
Id. at 129, under section 4.8, "Opacity Standards."


Comment 2-17: The EPA's estimates of the PM associated with shop opacity are unsupported and improper. 

Commenter 0097-Industry Trades stated the EPA's quantification of PM emissions attributable to shop opacity and all resulting cost-effectiveness calculations rest on many assumptions for which the EPA provides no support. The EPA believes that the inverse relationship it presumes to exist between stack emissions (and therefore presumed capture efficiency) and shop opacity is borne out by the CAA section 114 performance test data for the 0 percent opacity and nonzero opacity facilities, but the opposite is true. A simple inspection of the data shows that 0 percent opacity is not correlated with higher stack emissions. 

Commenter 0097-Industry Trades stated when the PM stack emissions values for the 13, 0 percent opacity facilities in the EPA's database and the PM stack emissions values for those nine facilities that recorded opacity values above 3 percent are plotted in a bar graph, the highest stack emission values are the same for both 0 percent opacity and 3 percent opacity facilities, and the values for the 3 percent opacity facilities are completely subsumed within the values for the 0 percent opacity facilities, with just the two lowest values in the 3 percent opacity dataset outside of the range of the 0 percent opacity dataset. In view of this, the EPA's presumption that there are statistically significant differences in the medians of the two data sets is not supportable. It is more likely that the calculated differences reflect variability wholly unrelated to the correlation theory espoused by the EPA. The EPA uses their stack emission/opacity correlation as the basis for ascribing a PM emission reduction value to its proposed 0 percent opacity standard. 

Commenter 0097-Industry Trades and 0095-Steel Co. stated it is not possible to credibly correlate the measured opacity with the mass of filterable PM. The commenters noted melt shops contain numerous other equipment and activities (i.e., other than the EAF) which also can generate fugitive emissions that can be read as opacity. While it is correct that opacity is due to the presence of particles, quantifying the mass of particles that correspond to a given level of opacity is impossible because the observed opacity is a function of many variables include particle size and distribution, particle composition, ambient conditions, capabilities of the human (or non-human) observer, and others. The EPA falsely assumes a linear correlation between PM emissions and opacity, when in fact, at the low levels (i.e., under 6 percent) relevant here, opacity is often more a function of meteorology than emission level. Thus, even theoretically, there is no reason to presume a one-to-one correspondence between opacity and mass of PM. Much of the EPA's analysis supporting its proposed rule attempts to quantify the benefits (i.e., mass of PM reduced) as opacity is reduced from one level to a lower level  -  for example from 6 percent to 0 percent. 

Commenter 0097-Industry Trades stated the EPA estimated the mass of fugitive emissions from the melt shop relying on: the measured baghouse exhaust PM level, the assumed baghouse efficiency, and the assumed melt shop capture efficiency. The commenter notes that while the value of the assumed baghouse efficiency may have some basis (i.e., may be available from some tests where both inlet and outlet PM were measured), there is no basis to presume the value of the assumed melt shop capture efficiency. Even complex computational fluid dynamic (CFD) models cannot properly estimate melt shop capture efficiency. The mass of PM emitted from the EAF as fugitive PM from the melt shop cannot be correlated to any level of opacity. In total, the EPA's analysis of the emissions reductions that will result from reducing melt shop opacity from one level down to a lower level, or the cost-effectiveness that may accompany such reductions, are speculation made with unsupported assumptions.

Response 2-17: The EPA disagrees that it did not provide support in the proposal for the estimates of PM associated with shop opacity. The commenter is conflating the statements in the proposal about emissions with 0 percent shop opacity. The emission calculations used in the proposal for canopy hood cost estimates compare estimated uncontrolled shop emissions (which are not available in the EPA EAF data or elsewhere as actual measurements) for a theoretical facility and emission estimates for controlled emissions from the melt shop at 6 percent opacity for the same theoretical facility. In addition, the assumption that facilities with 0 percent opacity have higher baghouse emissions than if the same facility achieved 6 percent opacity, does not equate to comparing baghouse PM emissions at one facility with 0 percent shop opacity to baghouse PM emissions at another facility achieving a 6 percent opacity. As shown in Figure 1-5, the 13 facilities in the EPA EAF data that achieved 0.000 percent shop opacity have only slightly higher total facility baghouse PM emissions compared to all 30 EAF facilities in the EPA EAF data, with identical highest and lowest total facility baghouse PM values.

We mostly agree with the commenter that "while it is correct that opacity is due to the presence of particles, quantifying the exact mass of particles that correspond to a given level of opacity is impossible because the observed opacity is a function of many variables include particle size and distribution, particle composition, ambient conditions, capabilities of the human (or non-human) observer, and others." However, we do not agree with the commenter that the only effort needed to demonstrate 0 percent opacity at a facility that consistently demonstrates 6 percent opacity is to wait for the conditions of particle size and distribution, particle composition, ambient conditions, and capability of the observer to align in a fortuitous manner. 

For purposes of assessing cost estimates, simplifying assumptions and estimates for PM emissions need to be made in order to estimate the cost of control options. The need to make estimates also was evident in the Ferroalloys NESHAP cost estimates for fugitive EAF emissions, where controlled PM emission estimates were made based on assumptions, emissions factors, mass balance, and observations: 
      ...emissions reductions were estimated for each facility based on emissions data received in an information collection request sent to the industry along with application of emissions factors, and other data analyses and calculations. 
With more detail provided: 
      We have not identified any emissions factors based on the direct measurement of fugitive emissions from ferroalloys production furnaces [in order] to estimate these emissions. Therefore, we used a mass balance approach to estimate these emissions. 
      
      The furnace smelting fugitive emissions were estimated by calculating the uncontrolled PM emissions from the furnace and estimating the range of the uncontrolled PM emissions that are captured from the furnace. The difference between the uncontrolled pollutant emissions and the estimated captured pollutant emissions were assumed to be the furnace fugitive emissions. 
      The uncontrolled PM pollutant emissions were calculated using the furnace annual emissions for PM and the estimated control efficiency. The assumed fabric filter and scrubber control efficiencies ...are based on expected control efficiencies found on the EPA Clean Air Technology Center website .... [T]he assumed process fugitive capture efficiencies are based on visual observations of the fugitive capture during process operations.
      It was assumed that 92 percent of the furnace control device emissions are generated from the furnace and 8 percent of the furnace control device emissions are generated during tapping of the furnace. These assumptions are based on engineering judgment of the furnace operations and the tapping fugitive capture systems.
      
      
Comment 2-18: The EPA's cost estimates are not cost effective because they will result in only marginal reductions in PM emissions.

Commenter 0097-Industry Trades stated the EPA's proposal does not reasonably account for the costs associated with the emissions limits that the Agency suggests are "achievable through the application of the BSER." The EPA's own data and analyses demonstrate that the proposed 40 CFR part 60, subpart AAb requirements will result in marginal reductions in PM emissions that will be incredibly cost ineffective. The costs EPA ascribes to this proposal are conspicuously underestimated. The commenter further stated the EPA uses assumptions to weigh the costs and benefits of PM reductions associated with a change from 6 percent shop opacity to 0 percent shop opacity. But the EPA's data show that melt shops at facilities are not emitting 6 percent opacity; they are emitting an average of 0.14 percent opacity. This 0.14 percent average opacity is broadly consistent with the Associations' members experience and at the heart of our concerns with the proposed 0 percent shop opacity standard and its lack of necessity. Most facilities meet 0 percent opacity most of the time, but we have serious concerns about the technological feasibility (and cost) of a standard that requires all new, modified, and reconstructed facilities to achieve 0 percent opacity at all times. The difference between 0.14 percent [shop]opacity and 0 percent opacity is, therefore, exceptionally expansive (and expensive to achieve) from the standpoint of compliance and cost, but negligible from the perspective of environmental benefit in the form of reduced PM emissions. 

Commenter 0097-Industry Trades said the EPA estimates the changes from 6 percent [shop] opacity to 0 percent opacity will allow small facilities to reduce PM emissions by 56 tons per year (tpy) at a cost of $1,100 per ton, medium facilities reduce PM emissions by 730 tpy at a cost of $1,100 per ton, and large facilities reduce PM emissions by 4,000 tpy at a cost of $1,000 per ton. By evaluating the PM reduction potential from 0.14 percent instead of 6 percent, small facilities reduce PM emissions by 1.2 tpy at a cost of $50,000, medium facilities reduce PM emissions by 3.4 tpy at a cost of $235,294 per ton, and large facilities reduce PM emissions by 7.8 tpy at a cost of $512,821 per ton. 

Commenter 0095-Steel Co. added because the EPA has not appropriately calculated either the benefit or the cost, the Proposed Rule's 0 percent opacity limit violates the CAA section 111 requirement that a standard of performance "tak[e] into account the cost of achieving such reduction." (42 U.S.C. section 7411(a)(1))

Response 2-18: Analysis of BSER for the EAF NSPS required EPA to compare the NSPS rule being reviewed, i.e., 40 CFR part 60, subpart AAa, where 6 percent shop opacity is required, and the revised proposed BSER of a canopy hood in the melt shop that achieves 0 percent shop opacity. Therefore, 0.14 percent opacity was not used to represent EAF facilities in the calculations for the opacity limit that reflects BSER for melt shops and we disagree with the commenter that we should have used 0.14 percent. While the difference in emissions between 0.14 percent shop opacity may not be significant, the difference between 6 percent and 0 percent shop opacity is significant and, hence, the reason for revising the 40 CFR part 60, subpart AAa standard in the new NSPS subpart AAb. Because many EAF facilities are very close to the 0 percent shop limit during melting and refining (i.e., operation outside of periods of charging and tapping) even when the current rule requires 6 percent shop opacity, we expect that only small changes in hood and exhaust operation will be needed to reduce opacity from opacity levels such as 0.14 percent to 0 percent shop opacity and will not require an entirely new canopy hood, as in the proposal's cost estimates for the limit under BSER, which reflects a change in melt shop opacity from 6 percent to 0 percent. 

The EPA agrees with the commenter that "most facilities can meet 0 percent [shop] opacity most of the time." This underscores the likelihood that canopy hoods will not be needed in most cases to comply with the revised opacity limits in NSPS subpart AAb.


Comment 2-19: Monitoring opacity during the "period established to have the greatest potential for uncaptured emissions to escape the melt shop" is inappropriate and has not been demonstrated as achievable.

Commenter 0095-Steel Co. said the requirement to measure opacity at "the period established to have the greatest potential for uncaptured emissions to escape the melt shop" as opposed to measuring at a specific time (i.e., melting and refining, charging, or tapping) is vague and would create a "floating standard" that cannot be consistently met due to uncertainty of when the period of highest emissions may occur. The EPA did not explain what process should be used to establish the period of greatest potential emissions. The commenter said facilities may not be able to predict the highest period of potential opacity. The periods of time with potential for high in-shop opacity, such as charging, may not correspond to either the highest potential for uncaptured emissions from the melt shop or the highest visible emissions exiting from the melt shop. Emissions are influenced by ambient conditions like wind (airflow), pressure, temperature, and humidity. These ambient conditions often have greater impact on emissions than the EAF heat cycle. The commenter said it may be impossible to predict on a daily basis when the highest potential emissions or opacity from the melt shop will occur.

For plants with multiple EAFs within the same building, the commenters said it could be impossible to predict the time period during which the highest opacity will occur because the EAFs operate independently from each other and may be at different stages of the production cycle at any given time. The commenters said this provision was especially problematic if the highest opacity occurs during charging because the duration of charging is short relative to other periods of the cycle and not predictable enough to be able to measure that period every day for the several consecutive minutes required by Method 9. The commenter said it was not feasible to identify any particular time period that consistently and predictably has the "greatest potential for uncaptured emissions to escape the melt shop" where multiple EAFs are operated.

Commenter 0095-Steel Co. said the EPA had not demonstrated that monitoring opacity during the "period established to have the greatest potential for uncaptured emissions to escape the melt shop" was "achievable" and "adequately demonstrated" as required for standards published under section 111(a)(1) of the CAA. To be achievable, the commenter asserted that "a uniform standard must be capable of being met under most adverse conditions which can reasonably be expected to recur and which are not or cannot be taken into account in determining the `costs' of compliance." (National Lime Ass'n v. EPA, 627 F.2d 416, 431, n. 46 (D.C. Cir. 1980). The EPA must also account for routine operating variability associated with performance of the system on which the performance the standard is based. (National Lime Ass'n v. EPA, 627 F.2d at 431-433; and 79 FR at 39245) The EPA must show how a standard is achievable under the range of relevant conditions that may affect the emissions anywhere in the country and cannot simply show that the standard was achieved at a model plant for a short period of time. (National Lime Ass'n v. EPA, 627 F.2d at 433; Electric Arc Furnaces and Argon-Oxygen Decarburization Vessels in Steel Industry -Background Information for Proposed Revisions to Standards, Preliminary , at 2-6, 2-7, June 1982). 

In addition, commenter 0095-Steel Co. said the proposed provision increased the likelihood of missed opacity observations since facilities would have to schedule their opacity observations to coincide with anticipated production schedules. Operational issues occur that could result in a shutdown of an EAF for the rest of the compliance period. If the visible emissions observation had not already been taken, it would be impossible for the observation to be performed during the "greatest potential for uncaptured emission to escape the melt shop" because the highest opacity event would have already occurred. The commenter recommended the EPA define a specific period of the heat cycle for shop opacity measurements to be taken rather than requiring facilities to measure opacity at the "greatest potential" time period. The commenter urged the EPA to address how any opacity requirement will be met at melt shops with multiple EAFs within a single melt shop building.

Response 2-19: As shown in the opacity data in the EPA/EAF data set, which included data from 31 facilities and numerous test reports from more than a 5-year span, the 0 percent melt shop opacity standard during melting and refining is achievable in a range of conditions anywhere in the country and is capable of being met under most adverse conditions. For the full EPA response to the issues raised in this comment in regard to melt shop opacity, see the EAF NSPS preamble to the final rule, section IV.E.2, IV.B.2 and B.3, for the EPA responses to comments on this subject.



3.0	REVISIONS TO 40 CFR PART 60, SUBPARTS AA AND AAa

Comment 3-1: The current compliance demonstration requirements using the fan amperage and damper position monitoring in 40 CFR part 60, subparts AA and AAa are the best methods for assuring compliance with the melt shop NSPS standards. 

Commenters 0096 and 0093-Steel Co. asserted that the existing fan amperage and damper position monitoring in combination with opacity observations are the best methods of assuring compliance with the NSPS standards for the melt shop. The commenters opposed the proposed new monitoring requirements, including: 
 Installation of bag leak detection monitoring systems on all baghouses, including multi-stack baghouses; 
 Monitoring and operational restriction for furnace static pressure monitoring based on 15-minute averages on all EAFs; 
 Monitoring and operational restriction for volumetric flow rate or static pressure at each separately-ducted hood, based on 15-minute averages on all EAFs; 
 Removal of the option for monitoring and operational restriction for fan amps; 
 Adding inspections and maintenance requirements for holes or other openings in the melt shop building; and 
 Mandate for shop opacity observations to be made during charging and tapping or during the period of the heat cycle that generates the greatest uncaptured emissions. 
      
The commenters considered these new monitoring requirements to be unnecessary, expensive, and, some cases, impractical. The commenters said the existing monitoring requirements in 40 CFR part 60, subparts AA and AAa are adequate for demonstrating compliance with the standards. The commenters stated that the existing fan amperage and damper position monitoring have worked efficiently and effectively for many years and the proposed new monitoring would be less effective and would impose "extreme technical and engineering complications" on EAF plants.

Similarly, commenter 0097-Industry Trades urged the EPA to keep the current requirement in 40 CFR part 60, subparts AA, AAa for monitoring fan amperage in place, because they said this parameter directly correlates to the air flow to the control device, via the fan curve, unique to each site. 

Response 3-1: For the EPA's response to issues raised in this comment in regard to the proposed BLDS requirement, see the EAF NSPS preamble to the final rule, section IV.E.2, BLDS Response. For the EPA's response to issues raised in this comment in regard to damper position monitoring, see the EAF NSPS preamble to the final rule, section IV.E.2, Damper Position Response. For the EPA's response to issues raised in this comment in regard to fan amperage monitoring: See the EAF NSPS preamble to the final rule, section IV.E.2, Fan Amperage Response. For the EPA's response to issues raised in this comment in regard to furnace static pressure monitoring based on 15-minute averages: See the EAF NSPS preamble to the final rule, section IV.E.2, Furnace Static Pressure Response. For the EPA's response to issues raised in this comment in regard to melt shop inspections: See the EAF NSPS preamble to the final rule, section IV.E.2, Melt Shop Inspection Response. For the EPA's response to issues raised in this comment in regard to melt shop opacity, see the EAF NSPS preamble to the final rule, section IV.E.2, Melt Shop Opacity Response/ For the EPA's response to issues raised in this comment in regard to volumetric flow and static pressure monitoring, see the EAF NSPS preamble to the final rule, section IV.E.2, Volumetric Flow and Static Pressure Response.


Comment 3-2: The CAA limits the EPA's NSPS revision authority to only new sources.

Commenter 0096-Steel Co stated there is no denying that the proposed monitoring and associated revisions to 40 CFR part 60, subparts AA and AAa are materially substantive and are not error corrections, clarifications, or clerical adjustments. These changes far exceed the EPA's legal authority to revise a NSPS applicable to existing sources. The commenter continued, the EPA's legal authority under the CAA is very limited as it relates to revisions to existing NSPSs. The 40 CFR part 60, subparts AA and AAa [proposed] revisions (BLDS, furnace static pressure monitoring, volumetric flow monitoring, etc.) all constitute an "emission limitation" as defined by the CAA and, therefore, constitute a "standard of performance." The EPA has no authority under the CAA to make any such revisions to a "standard of performance" unless those revisions are expressly applicable to "new sources." Commenter 0097-Industry Trades said the proposed changes to the melt shop monitoring requirements were arbitrary and violate the basic premise of NSPS that revisions apply only to facilities that qualify as new, modified, or reconstructed after proposal of the NSPS requirements. 

Commenter 0095-Steel Co. also said the EPA does not have the authority to add new monitoring requirements for charging and tapping operations. The existing shop opacity monitoring in 40 CFR part 60, subpart AAa verifies efficiency of the DEC during normal operations. By expanding monitoring requirements to cover tapping and charging (a time period the furnace roof is rolled back, and the DEC control is not engaged), the EPA was creating new monitoring requirements designed to monitor a standard that was not included in the original rule. The proposed monitoring during charging and tapping cannot evaluate DEC capture efficiency, as the shop opacity observations were originally designed to do. Hence, the addition of the new monitoring, the commenter said, represents an unlawful revision to the existing NSPS standard.

Commenter 0097-Industry Trades was concerned the EPA was adding entirely new installation, monitoring, and maintenance requirement for charging and tapping furnace modes, including requirements for operators to install, calibrate, and maintain monitoring devices that continuously record the capture system damper position(s) and either the volumetric flow rate through each separately ducted hood or the rolling 15-minute average static pressure at each separately ducted hood. The commenter said these requirements are unnecessary and that they ignore the 1999 rulemaking that provided alternative monitoring methods. The commenter also argued the EPA failed to provide a reasonable explanation for these changes, had not explained why the additional monitoring is needed, had not explained the EPA's change in position from prior EAF steel NSPS rulemakings, and had neglected to account for any costs associated with the monitoring requirements. 

Response 3-2: For the EPA's response to issues raised in this comment in regard to melt shop opacity, see the EAF NSPS preamble to the final rule, section IV.E.2, Melt Shop Opacity Response and General Monitoring Response. For the EPA's response to issues raised in this comment in regard to melt shop opacity, see the EAF NSPS preamble to the final rule, section IV.E.2, General Monitoring Response.


Comment 3-3: Proposed changes to 40 CFR part 60, subparts AA and AAa are substantial and not "corrections and clarifications."

Commenter 0096-Steel Co. stated the EPA proposes substantial revisions to 40 CFR part 60, subparts AA and AAa for existing EAF. In the preamble, the EPA identifies these substantial revisions as being proposed to "clarify and refine the rule requirements by adding, removing, or revising ambiguous or outdated definitions, compliance, measurement, monitoring, and reporting requirements to "clarify the rule and enhance compliance and enforcement." The proposed revisions to 40 CFR part 60, subparts AA and AAa are anything but clarifications, refinements, or minor changes. The commenter further stated the EPA has inherent authority to correct mistakes or make other clerical revisions or clarifications to existing regulations. (e.g., Citizens Against Pellissippi Parkway Extension, Inc. v. Mineta, 375 F.3d 412, 416 (6th Cir. 2004) (citations omitted); Dun & Bradstreet Corp. Found. v. U.S. Postal Serv., 946 F.2d 189, 193 (2nd Cir. 1991) (citations omitted)). However, that the proposed monitoring and associated revisions to 40 CFR part 60, subparts AA and AAa are materially substantive and are not error corrections, clarifications, or clerical adjustments. 

Commenters 0096 and 0095-Steel Co. stated that CAA section 111 states that "standards of performance or revisions thereof shall become effective upon promulgation." (42 U.S.C. section 7411(b)(1)(B)) Therefore, quite plainly, if 40 CFR part 60, subparts AA and AAa are finalized as proposed, the vast array of [proposed] new monitoring and operational restrictions will become effective on the date of final promulgation. However, it will obviously be impossible to comply on the date of final promulgation. As proposed, BLDS, pressure monitors and flow monitors would all need to be procured, installed, calibrated, and tested. This of course cannot happen immediately but would take a considerable amount of time (likely years). As the EPA has no statutory authority under CAA section 111 to extend that compliance date, the only conclusion is that the EPA has no legal authority to impose such substantive revisions in an existing NSPS.

Commenter 0097-Industry Trades said that steel companies would need time to comply with the shop opacity limits for charging and tapping, parametric monitoring requirements, and BLDS requirements. There must be sufficient time to allow the required equipment to be designed and/or engineered, procured, installed, tested, and connected to data systems. There must be sufficient time after installation for the new monitoring equipment to collect data in preparation for the compliance stack test. Several years will be needed to complete these tasks. The commenter recommended the EPA provide existing facilities complying with 40 CFR part 60, subparts AA or AAa at least three years to comply with the new NSPS requirements. The commenter also suggested the EPA include a provision allowing facilities to request additional time if needed. 

Commenters 0097-Industry Trades and 0095-Steel Co. stated unless the EPA's CAA section 111 authority to impose new regulatory requirements is limited to "new sources" that commence construction, modification, or reconstruction after the date the proposed standards are published, the CAA section 111(b)(1)(b) requirement that "[s]tandards of performance or revisions thereof...become effective upon promulgation" (42 U.S.C. section 7411(b)(1)(B)) would be altogether unworkable. Substantial new regulatory requirements can perhaps feasibly take effect upon promulgation when the affected facilities are in the midst of construction, modification, or reconstruction. Such requirements are decidedly infeasible when the affected facility is not undergoing these types of changes or any changes at all. For instance, the commenter questioned how might an existing source that currently does not operate a BLDS timely comply with a requirement to install, calibrate, and operate a BLDS that is "effective upon promulgation." 

Commenters 0097-Industry Trades stated in the case of the EPA's proposed new requirement to conduct VE monitoring during "charging and tapping, or during the period established to have the greatest potential for uncaptured emissions to escape the melt shop," (Proposed revised 40 CFR 60.273(d) and 60.273a(d) would be altered to require shop opacity observations to be taken "during charging or tapping, or during the period established to have the greatest potential for uncaptured emissions to escape the melt shop," rather than "when the furnace is operating in the meltdown and refining period.") as opposed to the period of "melting and refining," EPA's proposed change effectively represents an entirely new emissions standard. Courts recognized that changes in monitoring requirements can effectively change the stringency of a standard. (Clean Air Implementation Project v. EPA, 150 F.3d 1200, 2014 (D.C. Cir. 2000))

Commenters 0097-Industry Trades stated while some aspects of the EPA's proposal could be considered as clarifying existing requirements, none of the proposed changes to the existing monitoring requirements in 40 CFR part 60, subparts AA and AAa can be plausibly construed as clarifications. These are completely new obligations that impose entirely new requirements and conditions on sources' ability to operate. As the D.C. Circuit has acknowledged, "[c]onverting a periodic standard into a continuous one makes the standard more rigorous because ... continuous monitoring will capture all the fluctuations and variability inherent in emissions and thus increase each source's number of `violations.'" (Clean Air Implementation Project v. EPA, 150 F.3d 1200, 2014 (D.C. Cir. 2000)) 

Commenters 0097-Industry Trades and 0095-Steel Co. stated in its 1983 revisions to the NSPS, the EPA used its authority to establish a new 40 CFR part 60, subpart to ensure that new regulatory requirements and obligations were appropriately constrained to those new and modified facilities, the construction, modification, or reconstruction of which commenced after publication of the proposed revision. (49 FR 43838 (Oct. 31, 1984); the 1999 Direct Final Rule revising the NSPS made changes to both 40 CFR part 60, subparts AA and AAa to add alternative requirements for the monitoring of EAF capture systems in response to recommendations made by the Common Sense Initiative (CSI) subcommittee on iron and steel. The CSI was established by the Administrator to bring together affected stakeholders to find cleaner, cheaper, and smarter environmental management solutions. These editorial changes did not affect the applicability or requirements of the rule. 64 FR 10105 (Mar. 2, 1999). Similarly, the 2005 revisions to the NSPS were adopted in response to a petition from the Steel Associations and did not add or increase the stringency of requirements applicable to facilities subject to 40 CFR part 60, subparts AA or AAa. 70 FR 8523 (Feb. 22, 2005).) 

Commenters 0097-Industry Trades stated this [proposed] standard requires agencies to "examine the relevant data and articulate a satisfactory explanation for its action including a `rational connection between the facts found and the choice made.'" That said, if a "new policy rests upon factual findings that contradict those which underlay [an agency's] prior policy," the agency "must" provide "a more detailed justification" for its action. (F.C.C. v. Fox Television Stations, Inc., 556 U.S. at 515) The same is true if the agency's "prior policy has engendered serious reliance interests that must be taken into account." (F.C.C. v. Fox Television Stations, Inc., 556 U.S. at 515) In such cases, in order to offer "a satisfactory explanation" for its action, as part of providing "a rational connection between the facts found and the choice made," ( Motor Vehicle Mfrs. Ass'n of U.S., Inc. v. State Farm Mut. Auto. Ins. Co., 463 U.S. at 43 (1983) (internal quotations omitted)) the agency must give "a reasoned explanation ... for disregarding facts and circumstances that underlay or were engendered by the prior policy." (F.C.C. v. Fox Television Stations, Inc., 556 U.S. at 515) Here, for the proposed NSPS revisions, the EPA provides no explanation at all for its decision to impose new regulatory requirements and obligations on existing facilities through proposed amendments to the existing 40 CFR part 60, subparts (AA and AAa). 

Commenters 0097-Industry Trades stated the EPA fails to articulate the problem(s) that its proposed changes purport to solve. Nor does the EPA's record reflect that the Agency meaningfully accounted for or even considered how the 40 CFR part 60, subparts AA and AAa monitoring requirements engendered serious reliance interests among sources that already installed, calibrated, and currently operate emissions monitoring technology required by the existing NSPS standards; incorporated those monitoring requirements into their Title V permit; engineered and operate their affected facility and emissions control technology to ensure compliance as measured by those monitoring requirements; and reasonably expected that these compliance monitoring provisions would not change unless the sources modify or reconstruct their EAF or AOD. 

Commenter 0096-Steel Co. stated, the EPA presumably has authority to add optional monitoring requirements to a NSPS, as it has done previously with 40 CFR part 60, subpart AAa. (64 FR 10105 (March 2, 1999); 70 FR 8523 (February 22, 2005)). Doing so arguably does not change the performance standard, since an existing source subject to the standard can continue complying with existing terms by simply "opting out" of new, optional terms. There would be no need to "retrofit" an existing source to meet such optional standards, thus avoiding a conflict with Congress' stated intent for the NSPS program. The changes that the EPA proposes to 40 CFR part 60, subparts AA and AAa, however, are not optional changes. Rather, they are all mandatory revisions that significantly alter a source's compliance obligations. Because the EPA's proposed revisions to AA and AAa would apply to existing sources, those revisions are unlawful and arbitrary. Thus, the EPA must remove all such changes prior to finalizing the proposed rule.

Commenter 0097-Industry Trades and 0095-Steel Co. stated the proposed rule misleadingly characterizes these fundamental revisions to the existing NSPS standards (40 CFR part 60, subparts AA and AAa) as "minor" and "editorial and clarifying changes." This is not an accurate description of the proposed revisions, which are, in fact, dramatic deviations from the current standards and would require substantial additional modifications to the melt shop facility and furnace operations. Given this mischaracterization of the proposed changes, we question the validity of the Paperwork Reduction Act (PRA) analysis for the proposed rule and urge the Office of Management and Budget (OMB) to review our forthcoming comments in detail to ensure an accurate understanding of the proposed rule requirements and their cost impacts. 

Response 3-3: The changes that are being included after consideration of comments are in the nature of possible additional recordkeeping requirements that don't impose significant additional burden because they are timed with already existing recordkeeping requirements in these subparts. For the EPA's full response to the issues raised in this comment, see the EAF NSPS preamble to the final rule, section IV.E.2, Melt Shop Opacity Response, and General Monitoring Response. 


Comment 3-4: The EPA's revisions to 40 CFR part 60, subparts AA and AAa constitute prohibited retroactive rulemaking.

Commenters 0097-Industry Trades and Commenters 0096 and 0095-Steel Co. stated the monitoring and operational revisions (i.e., revisions to 40 CFR part 60, subpart AAa that delete the portion of 40 CFR section 60.273a(d) that exempts facilities from the 40 CFR section 60.274a(f) requirement; remove the option to monitor control system amperes in lieu of flow rate and remove the option to measure flowrate at the baghouse rather than through each separately ducted hood; proposed requirement for static pressure monitoring devices to be installed in the duct immediately downstream from the hood exit; changing the frequency of monitoring from once-per-shift to continuous; and add ASTM D7520-16 to 40 CFR, part 60, subparts AA and AAa) that the EPA proposes to 40 CFR part 60, subparts AA and AAa constitute retroactive rulemaking that is inconsistent with the CAA, multiple court decisions interpreting the Act, and the EPA's long-standing implementation of CAA section 111, generally and with respect to the EAF NSPS, specifically. The Administrative Procedures Act (APA) prohibits retroactive rulemaking, which occurs when a new rule "attaches new legal consequences to events completed before [the new rule's] enactment." (National Min. Ass'n v. Department of Labor, 292 F.3d 849, 860 (D.C. Cir 2002)) Courts have held that the CAA gives no indication that Congress provided the EPA with the authority to implement rules retroactively. (Sierra Club v. Whitman, 285 F.3d 63, 68 (D.C. Cir. 2002)) The EPA should rescind those proposed NSPS Revisions that would retroactively impose new requirements on sources that have not been constructed, modified, or reconstructed on or after May 16, 2022. 

Commenters 0097-Industry Trades stated a provision operates retroactively when it "impair[s] rights a party possessed when he acted, increase[s] a party's liability for past conduct or impose[s] new duties with respect to transactions already completed." (Landgraf v. USI Film Prods., 511 U.S. 244, 280 (1994)) 

Commenters 0097-Industry Trades and 0095-Steel Co. stated agencies are permitted to change policy positions and adopt new regulatory interpretations, but they cannot do so in the manner the EPA is proposing. (F.C.C. v. Fox Television Stations, Inc., 556 U.S. 502 (2009)) New and changed policy positions are subject to the same judicial review standards (F.C.C. v. Fox Television Stations, Inc., 556 U.S. 502 (2009) at 515 ("The [APA] makes no distinction ... between initial agency action and subsequent agency action undoing or revising that action.")) under which "a reviewing court shall ... hold unlawful and set aside agency action, findings, and conclusions found to be ... arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law." (5 U.S.C. section 706(2)(A)) 

Commenters 0095-Steel Co. and 0097-Industry Trades stated the EPA cannot rely on other rulemakings or court decisions (such as Title V or 40 CFR section 112 (e.g., MACT standards)) for determining whether increased monitoring constitute new substantive control requirements.

Response 3-4: We are not finalizing any changes to 40 CFR part 60, subparts AA and AAa that require the purchase of equipment and other significant costs. The changes that are being included after consideration of comments are in the nature of possible additional recordkeeping requirements that don't impose significant additional burden because they are timed with already existing recordkeeping requirements in these subparts. Further, as indicated in the preamble, facilities already complying with 40 CFR part 60, subparts AA and AAa will not be held responsible for past conduct that may contravene the finalized updates to 40 CFR part 60, subparts AA and AAa, and thus, these revisions will not have a retroactive effect. Rather, the compliance date for all of the updates to 40 CFR part 60, subparts AA and AAa is 180 days from the effective date of the final rule amendments for facilities complying with 40 CFR part 60, subparts AA and AAa. For the EPA's full response to the issues raised in this comment, see the EAF NSPS preamble to the final rule, section IV.E.2, General Monitoring Response. 


Comment 3-5: The EPA cannot change monitoring requirements for 40 CFR part 60, subparts AA and AAa for charging and tapping operations and the decision is inconsistent with prior rulemakings.

Commenters 0095-Steel Co. and 0097-Industry Trades opposed the proposed changes to the monitoring requirements for the melt shop because they would increase the stringency of the existing opacity standard and associated monitoring requirements. 

Commenter 0095-Steel Co. said that charging and tapping typically have higher opacity than the melting and refining phase. For 40 CFR part 60, subpart AA, the EPA initially set different opacity limits depending on the operations, with 0 percent opacity standard during most operations, 20 percent opacity during charging, and 40 percent opacity during tapping. (40 FR 43850 and 43853, September 23, 1975.) In 1999, the shop opacity was decreased to a uniform 6 percent opacity standard in 40 CFR part 60, subparts AA and AAa to ensure that compliance could be demonstrated by opacity measurements taken only during the melting and refining period. (64 FR 10105 and 10107, March 2, 1999) Since the shop opacity is based on an arithmetic average of 24 Method 9 observations and Method 9 monitoring is also based on an average of measurements, limiting observations to only the highest periods of potential opacity skews the average measurement higher than if it were to account for the average opacity throughout all heat cycles. Commenter 0097-Industry Trades said the EPA was shifting the point of compliance from the time period when the primary emission control is operational (during melting and refining when the DEC is in place) for a long duration during the heat cycle to a much shorter time period where fugitive emissions may be greater.

Commenter 0097-Industry Trades said that some facilities may not be able to comply with the proposed 40 CFR part 60, subpart AAa melt shop opacity standard and may have to re-engineer how they control emissions within the melt shop. At most EAF steel mills, the primary emission control system (e.g., DEC) is operational only during the melting and refining phase. The DEC is disconnected during charging and tapping when the furnace roof is swung open to allow the introduction of scrap metal and other feedstocks during charging or to allow pouring of molten metal from the EAF during tapping. By measuring shop opacity during melting and refining, the EPA adopted a shop opacity standard for 40 CFR part 60, subpart AAa facilities for the period when the primary emissions control system was operational. Fugitive emissions that escape from the EAF during melting and refining (as well as charging and tapping) are controlled by secondary pollution equipment, such as canopy hoods and scavenger ducts.

Commenter 0095-Steel Co. said the EPA cannot justify monitoring solely during periods with higher potential for opacity based on the EPA's prior assertion in the 1983 40 CFR part 60, subpart AAa rulemaking that the 6 percent opacity standard can be met during charging and tapping based on "full heat cycle" data, because there is no evidence in the record that the test data relied on over 40 years ago actually included complete charging and tapping events and even if they did, an average across solely higher opacity periods would be higher than an average across the "full heat cycle" If measurements account only for the highest periods, the commenter asserted the standard effectively becomes more stringent.

Similarly, commenter 0097-Industry Trades noted that in the 1982 40 CFR part 60, subpart AAa rulemaking the EPA said emissions during charging are more difficult to control than during melting and refining. The EPA has also acknowledged in previous rulemakings that the level of potential fugitive emissions may be greater during charging and tapping because the DEC is not in use and for most mills the DEC cannot be used during charging and tapping events. Requiring fugitive emission observations when the DEC is not engaged or when emissions otherwise "have the greatest potential" to be uncaptured is at odds with the original purpose of viewing the shop opacity observation requirement as a means to ensure proper operation of the primary control system. The commenter said the EPA appeared be using melt shop opacity monitoring as a means of enforce a fugitive emissions limit and as a surrogate for making the melt shop a total enclosure. The proposed new monitoring requirements are not an indicator of the proper operation of the primary control system but are instead a fundamental change to the purpose, substance, and stringency of the existing shop opacity standard. 
Response 3-5: For the EPA's response to the issues raised in this comment in regard to melt shop opacity, see the EAF NSPS preamble to the final rule, section IV.E.2, Melt Shop Opacity Response.


Comment 3-6: Monitoring opacity during charging and tapping is inappropriate.

Commenter 0095-Steel Co. disagreed with measuring opacity during the charging and tapping of the EAF. The commenter noted that for many EAFs, charging may be the time period with the greatest visible emissions inside the melt shop due to the rapid thermal shift involved in charging and because the DEC may not be operating with furnaces designed for open top charging. The commenter said that charging period does not necessarily correspond to the highest visible emissions exiting from the shop. Averages taken during charging may be skewed higher than an average taken across all heat cycles. The charging period only occurs one time per cycle, varies in duration, and can often be as short as 30 seconds. The short duration of any opacity associated with charging is not representative of average EAF opacity emissions or environmental performance.

Similarly, commenter 0091-State was concerned monitoring opacity during charging and tapping would not demonstrate proper operation of the EAF or result in emissions reductions because charging and tapping events occur for only a few minutes at a time.

Response 3-6: For the EPA's response to some of the issues raised in this comment in regard to melt shop opacity, see the EAF NSPS preamble to the final rule, section IV.E.2, Melt Shop Opacity Response. In addition, the EPA acknowledges that the data obtained through the CAA section 114 requests only consisted of data collected during melting and refining, which is the time period required to test opacity in the current EAF NSPS rules. Therefore, we are allowing in the final rule for 40 FCR part 60, subpart AAb the current rule limit of 6 percent opacity to apply during charging and tapping, and retaining the proposed 0 percent melt shop opacity during melting and refining. We estimate that this change will decrease the emision reductions in the proposed rule by approximately 15 percent. Details of the changes to emissions and reductions for the final rule are discussed in the memorandum titled "Particulate Matter Emissions from Electric Arc Furnace Facilities," hereafter called the "Emissions Memorandum," and memorandum "Cost and Other Analyses to Determine BSER for PM Emissions and Opacity from EAF Facilities," hereafter referred to as the "Cost Memorandum." Both memoranda have been updated for the final rule and are located in the docket for this rule (Docket No. EPA-HQ-OAR-2002-0049). Note that we also are revising the method used to measure opacity during charging due to its short time period. See Comment #2-7 above for more details.


Comment 3-7: The proposed volumetric flow and static pressure monitoring requirements are unworkable.

Commenters 0096 and 0093-Steel Co. and 0097-Industry Trades opposed the proposed requirement to measure volumetric flow or static pressure at each ducted hood to obtain 15-minute averages. Commenter 0096-Steel Co. said that this monitoring would be unworkable due to the complex structure of many melt shops. As an example, the commenter noted that the Mansfield Works melt shop contains two EAFs, an Argon-Oxygen Decarburization (AOD), and a LMF, and is controlled by two baghouses. The two EAFs and AOD are all subject to 40 CFR part 60, subpart AAa requirements, while the ladle metallurgy furnace is not. The commenter said it would be impracticable to establish a minimum volumetric flow rate or static pressure at a facility with these complex set of overlapping operations. The commenter added that the melt shop operations are variable, and a 15-minute average would not adequately capture the variability. The commenter added that technical complications with flow monitors used in areas of extreme heat or in dusty conditions would increase the need for re-calibrations and likely lead to periodic failures of the monitors. Some shops (e.g., Butler Works) have long runs of ductwork on the roof of the melt shop that pose access and safety difficulties for installing and maintaining flow monitors. Depending on the technical specifications for where flow or pressure monitors must be installed, redesigns and retrofits of hoods and ductwork could be necessary.

Commenters 0096, 0093, and 0098-Steel Co. said they are not aware of any melt shops that currently effectively measure furnace static pressure based on 15-minute averages, as the harsh environment in the furnace makes pressure monitoring difficult. Commenter 0096-Steel Co. noted that their experience monitoring pressure in furnaces showed it was difficult to generate useful data on furnace operation and capture efficiency. This commenter suggested measurement of volumetric flow or static pressure at each separately ducted hood or furnace pressure would not generate useful data that will enable an understanding of how well the capture system is operating. 

Commenters 0093 and 0098-Steel Co. said that maintaining flow meters in the highly abrasive and corrosive EAF ducts is difficult and that equipment fouling and damage is a problem. Commenter 0098-Steel Co. said that in their experience the monitoring equipment was "virtually impossible to operate and maintain" because the monitoring equipment does not survive. Commenter 0093-Steel Co. said that few EAFs are equipped with volumetric flow rate equipment due to the high particulate loadings and corresponding fouling. Where the volumetric flowrate monitors are present, the flowrate monitors are typically post-baghouse to take advantage of the cleaner air stream or a single volumetric flowrate monitoring system is installed in a common duct. Requiring a volumetric flowrate monitoring device in each separately ducted hood would be difficult to maintain due to the highly abrasive exhaust prior to the control device. 

Commenter 0097-Industry Trades stated that the 1999 rulemaking allowed daily visible emissions monitoring as an alternative to installation of a furnace static pressure monitoring device. In that rulemaking, the EPA recognized that furnace static pressure monitors presented practical problems with maintenance and installation. The commenter said the EPA recognized that daily visible emissions observations were as good or better an indicator of control device capture efficiency. The free space above an EAF is subject to severe conditions of high temperature and dust. Several owners and operators have had problems with frequent plugging of the pressure monitoring sensor. Due to the location of the sensor, maintenance and repair can be both difficult and dangerous. The EPA does not address the difficulties with installing, operating, and maintaining monitors on a continuous basis in the harsh furnace environment. Commenter 0097-Industry Trades said that any requirement for furnace pressure monitoring would have to provide for the regular downtime that would be needed to maintain the monitors. In addition, the electronics that would be used to transmit the pressure data on a continuous basis to a data recorder must be able to survive the same harsh environment. These issues are the primary reason most EAFs do not use furnace static pressure monitors to comply with the existing requirements.

Commenter 0097-Industry Trades stated that the significant costs of installing and maintaining furnace static pressure monitors were not considered in the proposal.

All commenters supported the existing approach that uses a combination of damper positions and fan motor amperes because they said it provides an adequate assurance of compliance. 

Response 3-7: For the EPA's response to this comment, see the EAF NSPS preamble to the final rule, section IV.E.2, Furnace Static Pressure Response and General Monitoring Response. 


Comment 3-8: The EPA should not remove options to monitor either control system fan motor amperes or volumetric flow rate at the control device inlet (baghouse) or to replace with requirement to monitor volumetric flow rate or static pressure at each separately ducted hood.

Commenter 0095-Steel Co. added that using amperage monitoring to monitor flow rate is more efficient because the amperage measurements are proportionally related and monitoring the fans and motors is required anyway to evaluate evacuation system performance, operational reliability, power management, and preventative maintenance and replacement projections. The commenter said that 40 CFR part 60, subpart AAa has always allowed monitoring of amperage and the alternative options to monitor flowrate through each duct or flowrate at the control device inlet were added as alternative methods in 1999 to increase compliance flexibility. 

Response 3-8: For the EPA's response to this comment, see the EAF NSPS preamble to the final rule, section IV.E.2, General Monitoring Response.


Comment 3-9: Recommended changes to the proposed volumetric flow monitoring provisions.

Commenter 0097-Industry Trades opposed the proposed requirements for direct flow measurements, but recommended changes be made if the proposed provisions are incorporated into the final rulemaking. First, the EPA should allow placement of flow meters either in the clean side of the baghouse (for those baghouses that have stacks) or in the inlet side of the baghouse (for baghouses without stacks). Second, an 8-hour (or once-per-shift) averaging of the flow should be allowed to compare against set points to determine deviations. Flow rates to the baghouse can vary considerably based on heat-to-heat variations due to the dynamic nature of EAF operations. The EPA's proposal to assess flow on a 15-minute basis, the commenter said, will not capture the full extent of variability of the flows to the baghouse. 

Response 3-9: For the EPA's response to this comment, see the EAF NSPS preamble to the final rule, section IV.E.2, General Monitoring Response.


Comment 3-10: The EPA conducted no cost analysis for its proposed changes to 40 CFR part 60, subparts AA and AAa. 

Commenters 0097-Industry Trades and 0095-Steel Co. stated the EPA's Economic Impact Analysis considered "costs and emissions impacts" only associated with the proposed 40 CFR part 60, subpart AAb. The EPA assumes its changes to 40 CFR part 60, subparts AA and AAa will have no impact on emissions and will result in no new costs. 

Commenters 0097-Industry Trades continued as such, while "the requirement that an agency provide a reasoned explanation for its action would ordinarily demand that it display awareness that it is changing position," (F.C.C. v. Fox Television Stations, Inc., 556 U.S. 502, 515) the Agency's proposed NSPS Revisions reflect no acknowledgement that the EPA is proposing a fundamental change that is both substantive and costly.

Commenter 0097-Industry Trades stated what the EPA labels as "editorial and clarifying changes," (87 FR at 29721 and 29726) are, in reality, substantive new requirements with real costs and burdens that are centrally relevant to a reasoned consideration of whether to impose regulatory burdens on facilities under 40 CFR part 60, subparts AA and AAa. (Michigan v. EPA, 135 S.Ct. 2699) The EPA failed to engage in reasoned decision-making by wholly ignoring the costs of these proposed new requirements. 

Commenter 0097-Industry Trades stated the following are some of the proposed new 40 CFR part 60, subparts AA and AAa requirements that the EPA impermissibly excluded from its cost and economic impact analyses: 

 Conduct daily observations of "Shop Opacity" "during charging or tapping, or during the period established to have the greatest potential for uncaptured emissions to escape the melt shop;" (Revised proposed 40 CFR 60.273(d), 60.273a(d), and 60.273b(d)) 
 Monitor and record on a continuous basis the rolling 15-minute average furnace static pressure; (Proposed 40 CFR 60.274(b)(1)(i), 60.274a(b)(1)(i), and 60.274b(b)(1)(i)). 
 Install, calibrate, and maintain a monitoring device that continuously records damper position(s); and, (Proposed 40 CFR 60.274(b)(1)(ii), 60.274a(b)(1)(ii), and 60.274b(b)(1)(ii)).
 Install, calibrate, and maintain a monitoring device that continuously records on a rolling 15-minute average basis either the volumetric flow rate through each separately ducted hood or the static pressure at each separately ducted hood, (Proposed 40 CFR 60.274(b)(1)(iii), 60.274a(b)(1)(iii), and 60.274b(b)(1)(iii).)
 Install and utilize BLDS on all baghouses, rather than only for single stack fabric filters. (Proposed 40 CFR 60.273(c ) and (e), and 60.273a (c) and (e)).

Commenter 0095-Steel Co. stated promulgating such revisions without assessment of cost would also be unlawful for the independent reason that CAA section 111 requires the EPA to "tak[e] into account the cost of achieving such reduction and any non-air quality health and environmental impact and energy requirements" when issuing any standard under CAA section 111. (42 U.S.C. section 7411(a)(1)) And although the EPA does have some discretion when considering costs, "the law does require the EPA to `cogently explain why it has exercised its discretion in a given manner'." (Nat'l Parks Conserv'tion Ass'n v. EPA, 788 F.3d 1134, 1142-43 (9th Cir. 2015) (internal citations omitted) (in regional haze context, striking down a BART determination where the EPA provided no supporting rationale for why one cost level was acceptable, but another was not).) Furthermore, although the EPA is entitled to make assumptions in its cost analyses, it has a "duty to examine [and justify] key assumptions as part of its affirmative `burden of promulgating and explaining a nonarbitrary, non-capricious rule... '" (Appalachian Power Co. v. EPA, 135 F.3d 791, 818 (D.C. Cir. 1998).) and the EPA may not "promulgate rules on the basis of inadequate data, or on data that, to a critical degree, is known only to the agency." (Portland Cement Ass'n v. Ruckelshaus, 486 F.2d 375, 391-93 (D.C. Cir. 1973)) The EPA's failure to consider any costs associated with its proposed revisions to 40 CFR part 60, subparts AA and AAa runs afoul of these principles.

Commenter 0095-Steel Co. added that the EPA failed to evaluate costs associated with enclosing melt shops and that it was unlawful for the EPA to incorporate the building inspection and maintenance requirements into 40 CFR part 60, subparts AA, AAa, or proposed AAb without completing a full analysis of the costs.

Response 3-10: For the EPA's response to this comment, see the EAF NSPS preamble to the final rule, section IV.E.2, General Monitoring Response.


Comment 3-11: The EPA did not provide a rational explanation for changing the monitoring frequency for fan amperage and damper positions for 40 CFR part 60, subparts AA and AAa from once-per-shift to continuous. 

Commenters 0095-Steel Co. and 0097-Industry Trades opposed changing the frequency of monitoring from once-per-shift to continuous. Commenter 0097-Industry Trades said that the EPA has failed to provide a rational explanation for changing the monitoring frequency for fan amperage and damper positions. The commenter added the EPA had failed to explain how continuously recording damper positions provides assurance for meeting an ambient air limit. They said the proposed provision imposes more data reporting requirements on the industry for no additional emissions reduction or monitoring benefit. 

Response 3-11: For the EPA's response to issues raised in this comment in regard to damper position monitoring, see the EAF NSPS preamble to the final rule, section IV.E.2, Damper Position Response. In addition, the purpose of the CAA section 111 standards is not to meet ambient air limits but to establish emission limitations that will apply to new sources or sources that are modified or reconstructed.


Comment 3-12: The EPA should define the term "material impact" (40 CFR 60.274(e), 60.274a(d)) in terms of opacity limits and only require repairs when holes result in noncompliance with the opacity standards.

Commenter 0095-Steel Co. recommended the EPA better define a "material impact" on the capture system because they said the phrase was too vague. Any airflow changes, they said, may theoretically impact capture efficiency to some extent, but fluctuations that do not affect the compliance of the facility with the substantive emission and opacity standards should not be prohibited. The EPA should define material impacts in terms of opacity limits by revising 40 CFR part 60, subparts AA, AAa, and AAb to only require repairs to openings that lead to noncompliance with the opacity standards.

Response 3-12: See the EAF NSPS preamble to the final rule, section IV.E.2, Melt Shop Inspection Response. For the EPA's response to issues raised in this comment in regard to melt shop opacity, see the EAF NSPS preamble to the final rule, section IV.B.2.


Comment 3-13: The proposed BLDS requirements are unworkable for multiple-outlet baghouses that allow alternative monitoring.

Commenter 0096-Steel Co. disagreed with the proposed new requirements for BLDS. The commenter said the EPA failed to provide adequate explanation for the proposed changes, which the commenter said made it difficult to provide comments. The Mansfield Works and Steelton plant currently comply with 40 CFR part 60 subpart AAa using the shop opacity observations and monitoring fan amperages and damper positions. The Butler Works demonstrates compliance using an alternate monitoring plan under which the plant performs additional shop opacity observations and monitors fan amperage but not damper positions. None of these mills employ BLDS. The commenter noted the Butler Works, Mansfield Works and Steelton plant all have multiple-outlet baghouses. The number of bag leak detection devices needed to monitor the baghouse outlets under the proposed rule would be 20, 28 and 34 devices, respectively, for the three plants. Due to false alarms from BLDS, industry commonly use backup bag leak detection devices. Hence, the number of BLDS would be 40, 56 and 68 devices, respectively, for each of the three plants. The commenter added that operating so many BLDS would be burdensome and add operational complications that the EPA did not recognize in the proposal.

Response 3-13: For the EPA's response to this comment, see the EAF NSPS preamble to the final rule, section IV.E.2, BLDS Response. In addition, if facilities have an alternative monitoring plan, revisions to the corresponding monitoring section(s) of 40 CFR part 60 subparts AA, AAa, and AAb will not affect these plans.


Comment 3-14: Requirement to use BLDS on all baghouses reverses prior Agency decisions.

Commenter 0097-Industry Trades opposed the proposed new requirement to install BLDS on all baghouses. When adopted in 1975, 40 CFR part 60 subpart AA originally required baghouse opacity to be monitored using a COMS. However, the 40 CFR part 60 subpart AAa amendments eliminated the requirement to use a COMS for emissions from modular, multiple-stack, negative-pressure or positive-pressure fabric filters provided the facility conducts daily Method 9 observations. The EPA determined that the use of visible emission standards was "technically sound and provides the most practical and inexpensive means to ensure that affected facilities are properly maintained and operated." (49 FR 43838 and 43840, October 31, 1984). The commenter also noted that the EPA acknowledged in the 1984 rulemaking that the capital and operating costs to install multiple COMS was unreasonable for modular, multi-stack baghouses. Amendments to the EAF steel NSPS in 2005 allowed mills to use a BLDS on single stack baghouses as an alternative to using COMS. (70 FR 8523, February 22, 2005). Facilities using BLDS were required to have site-specific monitoring plans, adopt corrective action plans, and conduct daily Method 9 observations when the EAF was in operation. 

The commenter added that the EPA's proposal requiring facilities to install and operate BLDS ignores the reasons why the BLDS requirement was limited to single-stack fabric filter baghouses. BLDS have never previously been required for modular, multi-stack baghouses. BLDS are feasible for a baghouse that has one stack and requires only one BLDS. However, some EAF baghouses do not have a stack at all but have exhaust monitors from each compartment in the roof of the baghouse. It is not uncommon for baghouses to have dozens of compartments; therefore, facilities may need dozens of BLDS to monitor each individual compartment. This would substantially increase installation and operating costs. In 2005, the EPA determined that BLDS in conjunction with visible emission observations were appropriate for single-stack baghouses as an alternative to COMS. (70 FR 8523, February 22, 2005). The EPA did not propose BLDS on modular, multi-stack baghouses in 2005 because these facilities were exempt from the COMS requirement under the 1984 rulemaking due to the impracticality and unreasonableness of requiring a COMS on such are large number of stacks. Requiring multiple BLDS for all baghouses. The commenter said it was inconsistent with prior determinations to require multiple BLDS for all baghouses. The proposed change is particularly unreasonable because the EPA provides no explanation for it.

Response 3-14: For the EPA's response to the BLDS issue in this comment, see the EAF NSPS preamble to the final rule, section IV.E.2, BLDS Response.


4.0	MONITORING AND COMPLIANCE


4.1 Monitoring Requirements in Proposed 40 CFR Part 60, Subpart AAb

Comment 4-1: The current compliance demonstration requirements using the fan amperage and damper position monitoring [in 40 CFR part 60, subparts AA and AAa] are the best methods for assuring compliance with the melt shop NSPS standards, and therefore oppose the proposed new monitoring requirements for [40 CFR part 60, subpart] AAb.

Commenters 0093 and 0096-Steel Co. asserted that the existing fan amperage and damper position monitoring in combination with opacity observations are the best methods of assuring compliance with the NSPS standards for the melt shop. 

The commenters opposed the proposed new monitoring requirements for [40 CFR part 60, subpart] AAb: 
 Installation of bag leak detection monitoring systems on all baghouses, including multi-stack baghouses; 
 Monitoring and operational restriction for furnace static pressure monitoring based on 15-minute averages on all EAFs; 
 Monitoring and operational restriction for volumetric flow rate or static pressure at each separately-ducted hood, based on 15-minute averages on all EAFs; 
 Removal of the option for monitoring and operational restriction for fan amps; 
 Adding inspections and maintenance requirements for holes or other openings in the melt shop building; and 
 Mandate for shop opacity observations to be made during charging and tapping or during the period of the heat cycle that generates the greatest uncaptured emissions. 
      
The commenters considered these new monitoring requirements to be unnecessary, expensive, and, some cases, impractical. The commenters said the existing monitoring requirements (in 40 CFR part 60, subparts AA and AAa) are adequate for demonstrating compliance with the standards. The commenters stated that the existing fan amperage and damper position monitoring have worked efficiently and effectively for many years and the proposed new monitoring would be less effective and would impose "extreme technical and engineering complications" on EAF plants.

Similarly, commenter 0097-Industry Trades urged the EPA to keep the current requirement for monitoring fan amperage in place, because they said this parameter directly correlates to the air flow to the control device, via the fan curve, unique to each site. 

Response 4-1: For the EPA's response to the issues raised in this comment in regard to on BLDS, see the EAF NSPS preamble to the final rule, section IV.E.2, BLDS Response. For the EPA's response to issues raised in this comment in regard to damper position monitoring, see the EAF NSPS preamble to the final rule, section IV.E.2, Damper Position Response. For the EPA's response to issues raised in this comment in regard to fan amperage monitoring: See the EAF NSPS preamble to the final rule, section IV.E.2, Fan Amperage Response. For the EPA's response to issues raised in this comment in regard to furnace static pressure monitoring based on 15-minute averages: See the EAF NSPS preamble to the final rule, section IV.E.2, Furnace Static Pressure Response. For the EPA's response to issues raised in this comment in regard to melt shop inspections for 40 CFR part 60, subpart AAb: See the EAF NSPS preamble to the final rule, section IV.E.2, Melt Shop Inspection Response, as well as EPA's response in section IV.E.2 to the proposed requirements in 40 CFR 60.274b(d) for subpart AAb in regard to allowing operators discretion in an inspection as to what issues "materially impact" the capture system performance. For the EPA's response to issues raised in this comment in regard to melt shop opacity: See the EAF NSPS preamble to the final rule, section IV.E.2, Melt Shop Opacity Response. For the EPA's response to the comment concerning whether the opacity standards in 40 CFR part 60, subpart AAb should apply at all times: See the EAF NSPS preamble to the final rule, section IV.D.2, in regard to EPA's response for 40 CFR part 60, subpart AAb. For the EPA's response to issues raised in this comment in regard to volumetric flow and static pressure monitoring, see the EAF NSPS preamble to the final rule, section IV.E.2, Volumetric Flow and Static Pressure Response. See also EPA's response to issues raised in this comment in the EAF NSPS preamble to the final rule, section IV.E.2, General Monitoring Response.


Comment 4-2: Requiring both static pressure monitoring and shop opacity observations for the melt shop is unnecessary. Facilities should be given the flexibility to monitor either static pressure or opacity.

Comment 0095-Steel Co. opposed the [portion of the new EAF NSPS rule, 40 CFR part 60, subpart AAb] that removed the previous exemption for facilities to install a furnace static pressure monitor [as described in 40 CFR 60.273a(d)] provided the facility conducts regular shop opacity monitoring. The commenter argued that removing the flexibility to select a compliance method is inconsistent with the EPA's prior statements made in the 1999 revision of 40 CFR part 60, subpart AAa in which the EPA acknowledged that furnace static monitors created operational and safety issues in some EAFs and that static pressure monitors are less reliable than shop opacity observations. (64 FR 10105 and 10106, March 2, 1999). The commenter said they agreed with the EPA's 1999 statements and added that in their experience most facilities use opacity monitoring because they have not overcome the technical problems associated with furnace static pressure monitors. The commenter said they considered opacity monitoring to be a more direct means of monitoring capture effectiveness of the DEC and that the EPA has not provided any rationale or support to justify reversing the Agency's prior decision by requiring facilities to install and operate a furnace static monitor.

Similarly, commenter 0093-Steel Co. noted that proposed 40 CFR 60.274b(f) and (g) begin with the phrase "Except as provided in 60.273b(d)". The commenter said that proposed 40 CFR 60.273b(d) requires shop opacity monitoring for all shops and does not say that a furnace static pressure device is not required if shop opacity is conducted. The commenter also said that their experience with furnace static pressure monitoring has shown them to be highly unreliable and subject to frequent fouling. The commenter recommended that proposed 40 CFR 60.274b(f) and (g) be deleted or made optional.

Response 4-2: For the EPA's response to issues raised in this comment in regard to furnace static pressure monitoring based on 15-minute averages: See the EAF NSPS preamble to the final rule, section IV.E.2, Furnace Static Pressure Response, as well as revised 40 CFR 60.273b(d) in the final rule for 40 CFR subpart AAb. 

In addition, we note that in 1999 40 CFR part 60, subpart AA was amended to be consistent with 40 CFR part 60, subpart AAa to allow the pressure monitoring sensor to be moved to a location that would potentially result in fewer maintenance issues due to plugging, as follows: 
      
      64 FR 10106: Concerns were raised to EPA (and the Common Sense Initiative) regarding the use of a pressure monitoring system in the free space above an EAF when it is equipped with a direct shell evacuation system. The free space above an EAF is subject to severe conditions of high temperature and dust. Several owners and operators have had problems with frequent plugging of the pressure monitoring sensor. Due to the location of the sensor, maintenance and repair can be both difficult and dangerous. Industry representatives sought a more practical means of monitoring. 
      
      64 FR 10107: The EPA amended 40 CFR part 60, subpart AA in 1999 to allow owners or operators to locate the furnace static pressure monitor in the DEC duct prior to the introduction of ambient air. This change is consistent with the requirements in 40 CFR part 60, subpart AAa, and provided greater flexibility in locating the pressure sensors where plugging may not be as serious of a problem. 

Therefore, we believe the amendments in1999 that allowed relocation of the furnace static pressure monitor provided a solution to the plugging issue that made furnace static pressure monitoring less reliable, as the commenter asserts. 

The EPA proposed requiring continuous monitoring of furnace static pressure because it would provide information about capture at the EAF on a more frequent basis and because many facilities already use this equipment. However, because using this monitoring method/compliance option would involve the purchase of equipment not currently available or installed at all subject facilities, we agree with the commenters who stated that these changes should not be made as corrections nor included in 40 CFR part 60, subparts AA, AAa, and AAb. Therefore, the requirement for continuous furnace static pressure monitoring is not required in the final rules for 40 CFR part 60, subparts AA, AAa, and AAb because of potential significant cost implications. 

In addition, it is stated in the final rule at 40 CFR 60.273b(d) that furnace static pressure monitoring is not required if opacity is performed. Lastly, it is stated in the final rule at 40 CFR 60.274b(f) that facilities using a DEC system can measure opacity instead of using a furnace static pressure monitor.


Comment 4-3: The EPA should clarify how the proposed new monitoring requirements improve compliance demonstration.

Commenter 0091-State said it was unclear how the additional monitoring requirements in the proposed [40 CFR part 60, subpart AAb] rule will improve data or accuracy in demonstrating compliance with applicable requirements. The proposed NSPS 40 CFR part 60, subpart AAb requires monitoring of parameters that are not required to be monitored under the existing NSPS 40 CFR part 60, subparts AA and AAa standards. The commenter recommended the EPA explain the benefits of new monitoring techniques and additional monitoring parameters. The monitoring requirements should provide enough data to accurately demonstrate compliance with applicable requirements. The commenter added that the EPA must consider the cost to air agencies and facilities and associated benefits to compliance before requiring additional monitoring. Additional monitoring with no clear benefit, is burdensome for both regulated facilities and permitting authorities due to additional equipment, maintenance, and operator costs. 

Response 4-3: See the responses to Comment #4-1 above for same responses to this comment. In addition, we have removed the proposed requirements that would significantly impact EAF facility operating costs in the situations where existing sources are making modifications or being reconstructed in a manner such that compliance with 40 CFR part 60, subpart AAb is required. See EPA's response to issues raised in this comment in the EAF NSPS preamble to the final rule, section IV.E.2, General Monitoring Response.


Comment 4-4: The EPA should allow more flexibility by letting facilities decide where static pressure monitoring devices are installed. Installation "in the duct immediately downstream of the hood exit" may not be the best location. 

Commenter 0095-Steel Co. said they opposed the requirement directing facilities to install the static pressure monitoring device immediately downstream of the hood exit. The commenter stated the EPA had not demonstrated the feasibility of installing the static pressure monitoring devices in the duct immediately downstream of the hood exit and had also not demonstrated this as the best location. The commenters recommended that the rule allow facilities more flexibility in selecting the best location for the static pressure monitoring devices and added that the EPA had not provided any reason why performance, maintenance and health and safety considerations should not be considered when selecting the location. 

The commenter added that the EPA provides flexibility for the furnace static pressure monitor, which may be installed "in any appropriate location in the EAF or DEC duct prior to the introduction of ambient air such that reproducible results will be obtained." The EPA has not explained why the same flexibility cannot be provided to installation in other ducts [such as at hood exit] and that it would be arbitrary to not provide similar flexibility in [this] installation location. The commenter said when a facility installs monitoring equipment it is important that its installation location be sited for consistent performance, efficient maintenance, or subjecting workers responsible for maintenance to more significant health and safety risks. The EPA should accommodate such considerations of maintenance convenience and feasibility so long as they do not compromise the accuracy of the monitor. The EPA has not identified a basis in the record for ignoring such performance, maintenance or health and safety considerations. The commenter recommended proposed 40 CFR part 60, subpart AAb be changed to allow facilities to consider feasibility and maintenance when deciding where to install the monitoring devices. The commenter recommended the EPA revise the proposed text to read "downstream as close to the hood exit as feasible, taking into account performance, maintenance and health and safety considerations." 

Response 4-4: We agree with the commenter that the proposed rule as written would allow a facility to monitor furnace static pressure in the duct after it combines with other ducts, which would be inappropriate. Therefore, we modified the language for the 40 CFR part 60, subparts AA, AAa, and AAb final rules (40 CFR 60.274(f), 60.274a(f), 60.274b(f)) to allow monitoring, as follows:

      ...in any appropriate location in the EAF or DEC duct prior to combining with other ducts and prior to the introduction of ambient air at a location that has no flow disturbance due to the junctions.


Comment 4-5: Monitoring opacity during charging and tapping is inappropriate.

Commenter 0095-Steel Co. disagreed with measuring opacity during the charging and tapping of the EAF. The commenter noted that for many EAFs charging may be the time period with the greatest visible emissions inside the melt shop due to the rapid thermal shift involved in charging and because the DEC may not be operating with furnaces designed for open top charging. The commenter said that charging period does not necessarily correspond to the highest visible emissions exiting from the shop. Averages taken during charging may be skewed higher than an average taken across all heat cycles. The charging period only occurs one time per cycle, varies in duration, and can often be as short as 30 seconds. The short duration of any opacity associated with charging is not representative of average EAF opacity emissions or environmental performance.

Similarly, commenter 0091-State was concerned monitoring opacity during charging and tapping would not demonstrate proper operation of the EAF or result in emissions reductions because charging and tapping events occur for only a few minutes at a time.

Response 4-5: For the EPA's response to some of the issues raised in this comment in regard to melt shop opacity, see the EAF NSPS preamble to the final rule, section IV.E.2, Melt Shop Opacity Response as well as other responses in section IV.E.2. For the response to the comment concerning whether the opacity standards, as they are written in 40 CFR part 60, subparts AA, AAa, and AAb should apply at all times: See the EAF NSPS preamble section IV.D.2.


Comment 4-6: Monitoring opacity during the "period established to have the greatest potential for uncaptured emissions to escape the melt shop" is inappropriate and has not been demonstrated as achievable.

Commenter 0095-Steel Co. said the requirement to measure opacity at "the period established to have the greatest potential for uncaptured emissions to escape the melt shop" as opposed to measuring at a specific time (i.e., melting and refining, charging, or tapping) is vague and would create a "floating standard" that cannot be consistently met due to uncertainty of when the period of highest emissions may occur. The EPA did not explain what process should be used to establish the period of greatest potential emissions. The commenter said facilities may not be able to predict the highest period of potential opacity. The periods of time with potential for high in-shop opacity, such as charging, may not correspond to either the highest potential for uncaptured emissions from the melt shop or the highest visible emissions exiting from the melt shop. Emissions are influenced by ambient conditions like wind (airflow), pressure, temperature, and humidity. These ambient conditions often have greater impact on emissions than the EAF heat cycle. The commenter said it may be impossible to predict on a daily basis when the highest potential emissions or opacity from the melt shop will occur.

For plants with multiple EAFs within the same building, the commenters said it could be impossible to predict the time period during which the highest opacity will occur because the EAFs operate independently from each other and may be at different stages of the production cycle at any given time. The commenters said this provision was especially problematic if the highest opacity occurs during charging because the duration of charging is short relative to other periods of the cycle and not predictable enough to be able to measure that period every day for the several consecutive minutes required by Method 9. The commenter said it was not feasible to identify any particular time period that consistently and predictably has the "greatest potential for uncaptured emissions to escape the melt shop" where multiple EAFs are operated.

Commenter 0095-Steel Co. said the EPA had not demonstrated that monitoring opacity during the "period established to have the greatest potential for uncaptured emissions to escape the melt shop" was "achievable" and "adequately demonstrated" as required for standards published under section 111(a)(1) of the CAA. To be achievable, the commenter asserted that "a uniform standard must be capable of being met under most adverse conditions which can reasonably be expected to recur and which are not or cannot be taken into account in determining the `costs' of compliance." (National Lime Ass'n v. EPA, 627 F.2d 416, 431, n. 46 (D.C. Cir. 1980). The EPA must also account for routine operating variability associated with performance of the system on which the performance the standard is based. (National Lime Ass'n v. EPA, 627 F.2d at 431-433; and 79 FR at 39245) The EPA must show how a standard is achievable under the range of relevant conditions that may affect the emissions anywhere in the country and cannot simply show that the standard was achieved at a model plant for a short period of time. (National Lime Ass'n v. EPA, 627 F.2d at 433; Electric Arc Furnaces and Argon-Oxygen Decarburization Vessels in Steel Industry -Background Information for Proposed Revisions to Standards, Preliminary , at 2-6, 2-7, June 1982). The commenter said that all compliance testing and opacity emission reporting relied on by the EPA for setting the new opacity limits in 40 CFR part 60, subpart AAb are based on observations taken during the melting and refining phase as required under NSPS 40 CFR part 60, subpart AAa, and therefore, cannot be used as the basis for determining that a 0 percent opacity limit is consistently and predictably achievable during "the period established to have the greatest potential for uncaptured emissions to escape the melt shop." 

In addition, commenter 0095-Steel Co. said the proposed provision increased the likelihood of missed opacity observations since facilities would have to schedule their opacity observations to coincide with anticipated production schedules. Operational issues occur that could result in a shutdown of an EAF for the rest of the compliance period. If the visible emissions observation had not already been taken, it would be impossible for the observation to be performed during the "greatest potential for uncaptured emission to escape the melt shop" because the highest opacity event would have already occurred. The commenter recommended the EPA define a specific period of the heat cycle for shop opacity measurements to be taken rather than requiring facilities to measure opacity at the "greatest potential" time period. The commenter urged the EPA to address how any opacity requirement will be met at melt shops with multiple EAFs within a single melt shop building.

Response 4-6: For the EPA's response to issues raised in this comment in regard to melt shop opacity, see the EAF NSPS preamble to the final rule, section IV.E.2, Melt Shop Opacity Response.


Comment 4-7: The proposed volumetric flow and static pressure monitoring requirements are unworkable.

Commenters 0096 and 0093-Steel Co. and 0097-Industry Trades opposed the proposed requirement to measure volumetric flow or static pressure at each ducted hood to obtain 15-minute averages. 

Commenter 0096-Steel Co. said that this monitoring would be unworkable due to the complex structure of many melt shops. As an example, the commenter noted that the Mansfield Works melt shop contains two EAFs, an argon-oxygen decarburization (AOD), and a ladle metallurgy furnace (LMF), and is controlled by two baghouses. The two EAFs and AOD are all subject to 40 CFR part 60, subpart AAa requirements, while the LMF is not. The commenter said it would be impracticable to establish a minimum volumetric flow rate or static pressure at a facility with these complex set of overlapping operations. The commenter added that the melt shop operations are variable, and a 15-minute average would not adequately capture the variability. The commenter added that technical complications with flow monitors used in areas of extreme heat or in dusty conditions would increase the need for re-calibrations and likely lead to periodic failures of the monitors. Some shops (e.g., Butler Works) have long runs of ductwork on the roof of the melt shop that pose access and safety difficulties for installing and maintaining flow monitors. Depending on the technical specifications for where flow or pressure monitors must be installed, redesigns and retrofits of hoods and ductwork could be necessary.

Commenters 0093, 0096, and 0098-Steel Co. said they are not aware of any melt shops that currently effectively measure furnace static pressure based on 15-minute averages, as the harsh environment in the furnace makes pressure monitoring difficult. Commenter 0096-Steel Co. noted that their experience monitoring pressure in furnaces showed it was difficult to generate useful data on furnace operation and capture efficiency. This commenter suggested measurement of volumetric flow or static pressure at each separately ducted hood or furnace pressure would not generate useful data that will enable an understanding of how well the capture system is operating. 
Commenters 0093 and 0098-Steel Co. said that maintaining flow meters in the highly abrasive and corrosive EAF ducts is difficult and that equipment fouling and damage is a problem. Commenter 0098-Steel Co. said that in their experience the monitoring equipment was "virtually impossible to operate and maintain" because the monitoring equipment does not survive. Commenter 0093-Steel Co. said that few EAFs are equipped with volumetric flow rate equipment due to the high particulate loadings and corresponding fouling. Where the volumetric flowrate monitors are present, the flowrate monitors are typically post-baghouse to take advantage of the cleaner air stream or a single volumetric flowrate monitoring system is installed in a common duct. Requiring a volumetric flowrate monitoring device in each separately ducted hood would be difficult to maintain due to the highly abrasive exhaust prior to the control device. 

Commenter 0097-Industry Trades stated that the 1999 rulemaking allowed daily visible emissions monitoring as an alternative to installation of a furnace static pressure monitoring device. In that rulemaking, the EPA recognized that furnace static pressure monitors presented practical problems with maintenance and installation. The commenter said the EPA recognized that daily visible emissions observations were as good or better an indicator of control device capture efficiency. The free space above an EAF is subject to severe conditions of high temperature and dust. Several owners and operators have had problems with frequent plugging of the pressure monitoring sensor. Due to the location of the sensor, maintenance and repair can be both difficult and dangerous. The EPA does not address the difficulties with installing, operating, and maintaining monitors on a continuous basis in the harsh furnace environment. Commenter 0097-Industry Trades said that any requirement for furnace pressure monitoring would have to provide for the regular downtime that would be needed to maintain the monitors. In addition, the electronics that would be used to transmit the pressure data on a continuous basis to a data recorder must be able to survive the same harsh environment. These issues are the primary reason most EAFs do not use furnace static pressure monitors to comply with the existing requirements.

Commenter 0097-Industry Trades stated that the significant costs of installing and maintaining furnace static pressure monitors were not considered in the proposal.

All commenters supported the existing approach that uses a combination of damper positions and fan motor amperes because they said it provides an adequate assurance of compliance. 

Response 4-7: For the EPA's response to issues raised in this comment in regard to furnace static pressure monitoring based on 15-minute averages: See the EAF NSPS preamble to the final rule, section IV.E.2, Furnace Static Pressure Response.


Comment 4-8: The EPA should not remove options to monitor either control system fan motor amperes or volumetric flow rate at the control device inlet (baghouse) or to replace with requirement to monitor volumetric flow rate or static pressure at each separately ducted hood.

Commenter 0095-Steel Co. added that using amperage monitoring to monitor flow rate is more efficient because the amperage measurements are proportionally related and monitoring the fans and motors is required anyway to evaluate evacuation system performance, operational reliability, power management, and preventative maintenance and replacement projections. The commenter said that 40 CFR part 60, subpart AAa allowed monitoring of amperage; the alternative options to monitor flowrate through each duct or flowrate at the control device inlet were added as alternative methods in 1999 to increase compliance flexibility. 

Commenter 0093-Steel Co. recommended the phrase "As required, ..." be added to the beginning of 40 CFR 60.274b(h)(7) because the volumetric airflow through each separately ducted hood should be an optional requirement. 

Response 4-8: For the EPA's response to issues raised in this comment in regard to furnace static pressure monitoring based on 15-minute averages: See the EAF NSPS preamble to the final rule, section IV.E.2, Furnace Static Pressure Response. For the EPA's response to issues raised in this comment in regard to volumetric flow and static pressure monitoring, see the EAF NSPS preamble to the final rule, section IV.E.2, Volumetric Flow and Static Pressure Response.


Comment 4-9: Recommended changes to the proposed volumetric flow monitoring provisions.

Commenter 0097-Industry Trades opposed the proposed requirements for direct flow measurements, but recommended changes be made if the proposed provisions are incorporated into the final rulemaking. First, the EPA should allow placement of flow meters either in the clean side of the baghouse (for those baghouses that have stacks) or in the inlet side of the baghouse (for baghouses without stacks). Second, an 8-hour (or once-per-shift) averaging of the flow should be allowed to compare against set points to determine deviations. Flow rates to the baghouse can vary considerably based on heat-to-heat variations due to the dynamic nature of EAF operations. The EPA's proposal to assess flow on a 15-minute basis, the commenter said, will not capture the full extent of variability of the flows to the baghouse. 

Response 4-9: Placement of flow meters on the inlet or outlet of baghouse only provides a total flow from the system and doesn't provide specific information on the capture at each emission point. Duct positions change throughout the melt cycle; an 8-hour average would not provide useful information to understand whether capture during normal operations was consistent with capture during initial compliance demonstration. Therefore, the EPA disagrees with the commenter and is not including in the final rules the changes suggested by the commenter.


Comment 4-10: Melt shop opacity observations in 40 CFR 60.274b(b) are unnecessary and duplicative of 40 CFR 60.273b(d).

Commenter 0093-Steel Co. said that daily shop opacity readings are already required at 40 CFR 60.273b(d) [Emission Monitoring]. 

      60.273b(d) Shop opacity observations shall be conducted at least once per day during charging or tapping, or during the period established to have the greatest potential for uncaptured emissions to escape the melt shop. Shop opacity shall be determined as the arithmetic average of 24 consecutive 15-second opacity observations of emissions from the shop taken in accordance with EPA Method 9, or, as an alternative, according to ASTM D7520-16 (incorporated by reference, see 40 CFR 60.17), with the caveats described under Shop opacity in 40 CFR 60.271, or as the total duration of visible emissions measured according to EPA Method 22 over a 6 minute period, modified to require the recording of the aggregate duration of visible emissions at 15 second intervals. Shop opacity shall be recorded for any point(s) where visible emissions are observed. Where it is possible to determine that a number of visible emission points relate to only one incident of visible emissions, only one observation of shop opacity will be required. In this case, the shop opacity observations must be made for the point of highest opacity that directly relates to the cause (or location) of visible emissions observed during a single incident.

They said it was unclear why the EPA was requiring a second set of daily shop opacity readings under proposed 40 CFR 60.274b(b) [Monitoring of Operations]. 

      60.274b(b) Except as provided under paragraph (e) of this section, the owner or operator subject to the provisions of this subpart shall conduct observations of shop opacity at least once per day when the furnace is operating in the period of the heat cycle that generates the greatest uncaptured emissions.... 

The provisions in proposed 40 CFR 60.273b(d) [Emission Monitoring] are clearer. The commenter recommended the shop opacity language in 40 CFR 60.274b(b) should be deleted.

Response 4-10: The EPA agrees with the commenter that the daily opacity reading requirement from the proposed rule could be confused with adding additional requirements. While the two sections of the rule identified by the commenter have a different focus, both address issues with opacity. The EPA also agrees with the commenter that emission monitoring explains the details of opacity measurements. Therefore, we have deleted reference to opacity in 40 CFR 60.274b(b) in an effort to make the final rule clear.


Comment 4-11: Recommended changes to 40 CFR 60.274b(h).

Commenter 0093-Steel Co. recommended the following changes to 40 CFR 60.274b(h):

 40 CFR 60.274b(h)(5) should be deleted because it was redundant to 40 CFR 60.274b(h)(2):
      60.274b(h) During any performance test required under 40 CFR 60.8 or 40 CFR 60.272b(d), and for any report thereof required by 40 CFR 60.276b(f) of this subpart, or to determine compliance with 40 CFR 60.272b(a)(3) of this subpart, the owner or operator shall monitor the following information for all heats covered by the test:
         ........
         (h)(2) Heat times, including start and stop times, and a log of process operation, including periods of no operation during testing and, if a furnace static pressure monitoring device is operated pursuant to 40 CFR 60.274b(f), the pressure inside an EAF when DEC systems are used;
         ........
         (h)(5) Furnace static pressure;
         
 "As necessary," should be added to the beginning of 40 CFR 60.274b(h)(4) because the EAF test run is 4 hours, but the opacity demonstration under 40 CFR 60.11 is only three hours. This section implies that 12 hours (3 runs x 4 hours) of Method 9 or ASTM D7520-16 data are required.

      (h)(4) Continuous opacity monitor or EPA Method 9 data, or, as an alternative to EPA Method 9, according to ASTM D7520-16 (incorporated by reference, see 40 CFR 60.17), with the caveats described under Shop opacity in 40 CFR 60.271.
      
Response 4-11: We agree, in part, with the commenter and have removed the reference in 40 CFR 60.274b(h)(5) to furnace static pressure, but have retained the reference in 40 CFR 60.274b(b)(h)(2) but with the qualifying language: "if furnace static pressure monitoring device is operating pursuant to 40 CFR 60.274b(f)....".

We disagree with the commenter's interpretation that in the phrase "continuous opacity monitor or EPA Method 9 data" the word "continuous" modifies both opacity monitoring and Method 9 performance. In this case, the phrase "continuous opacity monitor" refers to an actual instrument, called a continuous opacity monitor (COM). To clarify, we have amended the regulatory text at 40 CFR 60.274b(h)(4) to indicate that this is referring to the instrument and not a description of the data collection. The readings done according to EPA Method 9 should be done to coincide with PM testing, but according to the time frames described in EPA Method 9 and in 40 CFR 60.271b under Shop Opacity.


4.2 Bag Leak Detection System (BLDS) 

Comment 4-12: Facilities should be allowed 24 hours to respond to BLDS alarms and to complete the response as soon as practical in 40 CFR part 60, subparts AA, AAa, and AAb. 

Commenter 0093-Steel Co. and 0097-Industry Trades disagreed with the proposed 40 CFR part 60, subpart AAb provisions that would require facilities to determine the cause of all BLDS alarms within one hour and alleviate the cause of the alarm within three hours by taking the necessary response action. The commenters recommended the EPA adopt a 24-hour timeframe to initiate corrective action and to require that response actions be completed as soon as practicable. This approach would recognize the practical realities in identifying and responding to BLDS alarms. The commenters added that this approach is the same as that used in the Integrated Iron and Steel NESHAP, and is consistent with 40 CFR part 63, subparts X (NESHAP From Secondary Lead Smelting), DDD (NESHAP for Mineral Wool Production), EEE (NESHAP from Hazardous Waste Combustors), MMM (NESHAP for Pesticide Active Ingredient Production), RRR (NESHAP for Secondary Aluminum Production), and TTT (NESHAP for Primary Lead Smelting). The commenters said there was no justification for the proposal to be different from other existing rules. They added that the proposed three-hours was arbitrary and ignored the numerous scenarios in which it can take longer than three hours to identify and fix the cause of an alarm. Allowing facilities 24 hours to identify the cause and requiring facilities to alleviate the cause of the alarm "as soon as practicable" is more practical, particularly where many baghouse compartments must be inspected to determine the cause of an alarm.

Commenter 0097-Industry Trades noted that there are situations in which more than three hours is needed to respond to a BLDS alarm and address its cause. Because many mills calibrate their BLDS to be very sensitive, the likelihood that a BLDS alarm will be falsely triggered is increased. The commenter included the following examples of situations in which false alarms can occur: 

 Weather  -  BLDS alarms will occasionally trigger during a heavy downpour or when there are significant changes in temperature or humidity. 
 Bag Cleaning Cycle  -  BLDS alarm may trigger at the end of the baghouse cleaning cycle due to the temporary absence of dust in the bags. 
 New Bag Start  - BLDS alarms can be triggered following a replacement of some or all of the bags in the baghouse. 
 Systems Checks/Testing  -  Some facilities may run systems checks on their BLDS that cause the system to alarm. For example, a facility may check the sensitivity of a BLDS by introducing a handful of flour into a port upstream from the probe. Facilities also evaluate and optimize their BLDS performance through drift checks, response tests, calibration exercises, and other quality assurance procedures. Some of these procedures require the alarm to be triggered in order to test performance, but in other instances the BLDS alarm may be inadvertently triggered during testing. 
 Electrical Malfunctions  -  As BLDS detection is based on contact electrification, alarms can be triggered due to electrical surges impacting the sensors, processing electronics, or the connections between the sensor and processing electronics. These surges can either be environmental (lightning) or from variations/malfunctions in the BLDS system, its software, or its power source. Additionally, the abrasive environment in the baghouse duct can deteriorate the BLDS probe, probe housing, and housing insulation, which can cause an increase in malfunctions. BLDS alarms may be triggered during temporary power lapses or brief connectivity issues between the sensor and the processing electrics, or between the processing electronics and the system output/alarm. The BLDS can experience brief mechanical or software glitches/errors, including with respect to the sensor's signal amplification or with the configuration of the processing electronics. 
 Repair/Maintenance  -  Some baghouse repair and maintenance activities may be conducted while the baghouse is in operation. In some of these cases, proper inspection and repair requires the baghouse to be operating in order to observe and repair malfunctions/maintenance issues. Often these activities are coordinated with a baghouse operator observing the BLDS readout in real time in order to identify the cause of an earlier alarm or to proactively identify maintenance or performance issues. Baghouse repair and maintenance activities sometimes must be conducted when the baghouse is operating because the repair/maintenance is urgently needed, and it is infeasible to quickly shut down the baghouse. These activities will cause BLDS alarms to trigger. Work on baghouse compartments and conveyances can introduce particulates into the system or dislodge caked or accumulated dust which triggers alarms. Sounding of BLDS alarms also can be caused by maintenance and repair activities conducted when the baghouse is not operating. These activities can introduce foreign material or dislodge accumulations of material from ducts, conveyances, access panels, joints, and other components of the system upstream from the probe. Then when the baghouse is restarted, the newly introduced or dislodged material can cause the BLDS alarm to be triggered. 

Commenter 0097-Industry Trades pointed out that it is possible for a baghouse to operate within its emission and opacity limits even if the cause of a BLDS alarm is not identified and corrective measures taken. For example, if a broken bag in a compartment causes an alarm, the compartment can be isolated and shut down without affecting the rest of the baghouse.

Commenter 0097-Industry Trades noted that determining the cause of the alarm often requires operators to undertake a multi-step troubleshooting process of elimination requiring multiple rounds of physical inspections and diagnostic efforts. This process of elimination often requires more than three hours to complete. The process can be very time-consuming, particularly when the BLDS alarm lasts only a short time. Identifying the cause of a brief BLDS alarm, the commenter said, can be very difficult and sometimes proves impossible. Some baghouses in the EAF industry have 25 or more compartments housing 5,000 or more individual bags. Some mills do not have BLDS with detection capability in each separate compartment because the baghouse design does not allow for such monitoring (e.g., multiple compartments sharing common exit plenum). In these instances, mills must continue running and sequentially isolate compartments to determine which compartment may have caused the BLDS alarm. Facilities must then physically examine the compartment(s), which may contain 150 or more individual bags. If a bag has a significant rupture or has been dislodged, the cause of the alarm will likely be readily apparent. However, some alarms can be triggered by extremely small holes in bags and in these cases, finding the leak by physical inspection can take a long time. 

Commenter 0097-Industry Trades said that EAF mills can have difficulty responding to multiple, intermittent alarms of short duration. The commenter noted that EAF facilities record the alarm as resolved where investigation shows no evidence of a bag leak. While the facility may be able to respond to each separate alarm in under three hours, the commenter said they are aware of one instance in which an enforcement authority determined the company was in violation of the three-hour response requirement because the total time the facility spent responding to each of the separate intermittent alarms exceeded three hours. The commenter said the enforcement authority misinterpreted the three-hour response requirement. This and provided this example was provided to show how a three-hour response requirement presents a compliance risk even when individual responses are completed within the three-hour window.

Commenters 0093-Steel Co. and 0097-Industry Trades recommended the following additions (shown in bold and underline) be made to 40 CFR 60.273b(f)(6) and (7) in all three rules in regard to the leak monitors to clarify false alarm situations.

      (6) Establishing to the extent acceptable by the delegated authority that the alarm was a false alarm and not caused by a bag leak or other malfunction that could reasonably result in excess particulate emissions, in which case alarms due to the [leak] monitor malfunctioning are not subject to the 3-hour response action requirement so long as the [leak] monitor malfunction is timely corrected;

      (7) Shutting down the process producing the particulate emissions; provided, however, that process unit shut down is not required if an operator reasonably believes repetitive alarms are the result of a [leak] monitor malfunction so long as the [leak] monitor malfunction is timely repaired. 

Commenter 0093-Steel Co. also recommended 40 CFR 60.276b(h)(3) be revised to specify that malfunction of the [leak] monitor connected to the alarm can be a basis for not resolving an issue within three hours [if change to 24 hours response is not made]. 

Response 4-12: For the EPA's response to this comment, see the EAF NSPS preamble to the final rule, section IV.E.2.


Comment 4-13: The proposed BLDS requirements in proposed 40 CFR part 60, subpart AAb are unworkable for multiple-outlet baghouses that allow alternative monitoring. 

Commenter 0096-Steel Co. disagreed with the proposed new requirements for BLDS. The commenter said the EPA failed to provide adequate explanation for the proposed changes, which the commenter said made it difficult to provide comments. The Mansfield Works and Steelton plant currently comply with 40 CFR part 60, subpart AAa using the shop opacity observations and monitoring fan amperages and damper positions. The Butler Works demonstrates compliance using an alternate monitoring plan under which the plant performs additional shop opacity observations and monitors fan amperage but not damper positions. None of these mills employ BLDS. The commenter noted the Butler Works, Mansfield Works and Steelton plant all have multiple-outlet baghouses. The number of bag leak detection devices needed to monitor the baghouse outlets under the proposed rule would be 20, 28 and 34 devices, respectively, for the three plants. Due to false alarms from BLDS, industry commonly use backup bag leak detection devices. Hence, the number of BLDS would be 40, 56 and 68 devices, respectively, for each of the three plants. The commenter added that operating so many BLDS would be burdensome and add operational complications that the EPA did not recognize in the proposal.

Response 4-13: For the EPA's response to this comment, see the EAF NSPS preamble to the final rule, section IV.E.2. BLDS Response. In addition, if facilities have an alternative monitoring plan, revisions to the corresponding monitoring section(s) of 40 CFR part 60, subparts AA, AAa, and AAb will not affect these plans.


Comment 4-14: Requirement in proposed 40 CFR part 60, subpart AAb to use BLDS on all baghouses reverses prior Agency decisions. 

Commenter 0097-Industry Trades opposed the proposed new requirement to install BLDS on all baghouses. When adopted in 1975, 40 CFR part 60, subpart AA originally required baghouse opacity to be monitored using a COMS. However, the 40 CFR part 60, subpart AAa amendments eliminated the requirement to use a COMS for emissions from modular, multiple-stack, negative-pressure or positive-pressure fabric filters provided the facility conducts daily Method 9 observations. The EPA determined that the use of visible emission standards was "technically sound and provides the most practical and inexpensive means to ensure that affected facilities are properly maintained and operated." (49 FR 43838 and 43840, October 31, 1984). The commenter also noted that the EPA acknowledged in the 1984 rulemaking that the capital and operating costs to install multiple COMS was unreasonable for modular, multi-stack baghouses. Amendments to the EAF steel NSPS in 2005 allowed mills to use a BLDS on single stack baghouses as an alternative to using COMS. (70 FR 8523, February 22, 2005). Facilities using BLDS were required to have site-specific monitoring plans, adopt corrective action plans, and conduct daily Method 9 observations when the EAF was in operation. 

The commenter added that the EPA's proposal requiring [new] facilities to install and operate BLDS ignores the reasons why the BLDS requirement was limited to single-stack fabric filter baghouses. BLDS have never previously been required for modular, multi-stack baghouses. BLDS are feasible for a baghouse that has one stack and requires only one BLDS. However, some EAF baghouses do not have a stack at all but have exhaust monitors from each compartment in the roof of the baghouse. It is not uncommon for baghouses to have dozens of compartments; therefore, facilities may need dozens of BLDS to monitor each individual compartment. This would substantially increase installation and operating costs. In 2005, the EPA determined that BLDS in conjunction with visible emission observations were appropriate for single-stack baghouses as an alternative to COMS. (70 FR 8523, February 22, 2005). The EPA did not propose BLDS on modular, multi-stack baghouses in 2005 because these facilities were exempt from the COMS requirement under the 1984 rulemaking due to the impracticality and unreasonableness of requiring a COMS on such are large number of stacks. Requiring multiple BLDS for all baghouses. The commenter said it was inconsistent with prior determinations to require multiple BLDS for all baghouses. The proposed change is particularly unreasonable because the EPA provides no explanation for it.

Response 4-14: For the EPA's response to this comment in regard to BLDS, see the EAF NSPS preamble to the final rule, section IV.E.2, BLDS Response.


4.3 Building Inspection Requirement in Proposed Subparts 40 CFR Part 60, Subparts AA, AAa, and AAb

Comment 4-15: The proposed building inspection requirement is vague, unnecessary and unworkable.

Commenters 0093 and 0095-Steel Co. and 0097-Industry Trades opposed the proposed building inspection requirements in 40 CFR [60.274(e), 60.274a(d), and] 60.274b(d). The commenters said the requirements are vague, too broad, impractical, unnecessary (there is already a 6 percent shop opacity limit) and exacerbate concerns about the impact of high temperatures on workers and equipment. 

Commenter 0097-Industry Trades said they agreed with the EPA that maintaining the melt shop building is important, but they said they opposed the proposed building inspection requirement because it imposes a de facto "total building enclosure" and would prohibit any hole or opening that allows dust to escape from any part of the building housing the melt shop. The commenter expressed concern that the proposed requirements may result in inspectors wrongfully alleging violations. The commenter added that companies would be forced to demonstrate that emissions of dust from small holes in the building do not materially compromise the dust capture system. They may also have to prove the emissions are not from EAF or AOD processes. They said the requirement was too subjective, open to abuse by the EPA and state enforcement inspectors, and could create "...excessive busy work for the facility." They added that the EPA and state inspectors may not be familiar enough with EAF steel mill operations to understand what is and is not material under the proposed standard. 

Commenters 0093 and 0095-Steel Co. and 0097-Industry Trades said the proposed building inspection requirements are impractical as the EAF melt shops have numerous openings for ventilation, for moving large pieces of equipment and materials in and out of the building, for utility ducting, cooling air, piping, and doors. Commenters also noted that melt shop buildings can be very large (often tens of thousands of square feet and potentially a quarter of a mile long or more) and house many different operations in addition to the EAF and AOD. Given the size of the building, repairs are occasionally needed to patch holes or seams in the structure but a hole in the roof or wall should not automatically be considered a problem. Commenters noted that the melt shop building helps contain emissions for capture by the canopy hood systems and the buildings are not under positive pressure. In the roof area near the EAF, the local pressure is generally negative due to the draw of the canopy. The DEC system creates negative pressure that keeps melting and refining emissions from drifting away from the EAF. A hole or tear in the building structure, the commenter said, does not mean emissions are being released.

Response 4-15: For the EPA's response to issues raised in this comment in regard to melt shop inspections for 40 CFR part 60, subparts AA, AAa and AAb: See the EAF NSPS preamble to the final rule, section IV.E.2, Melt Shop Inspection Response. 


Comment 4-16: The EPA should clarify the proposed building inspection requirements in 40 CFR part 60, subpart AAb.

Commenter 0097-Industry Trades said if the EPA decides to incorporate the proposed melt shop maintenance requirement in the final rule, they recommend that the EPA define what qualifies as a "material" or "significant" hole or opening in the melt shop. The definition should include evidence that the opening was associated with fugitive dust emissions from the EAF and/or AOD in excess of permitted shop opacity limits.

Commenter 0093-Steel Co. recommended the provision be restricted to openings above the EAF deck and be limited to "holes or openings where EAF emissions are observed to escape."

Response 4-16: For the EPA's response to issues raised in this comment in regard to melt shop inspections for 40 CFR part 60, subpart AAb: See the EAF NSPS preamble to the final rule, section IV.E.2, Melt Shop Inspection Response, as well as EPA's response in the EAF NSPS preamble section IV.E.2 to the proposed requirements in 40 CFR 60.274b(d) for subpart AAb in regard to allowing operators discretion in an inspection as to what issues "materially impact" the capture system performance. 
Comment 4-17: The EPA should define the term "material impact" (40 CFR 60.274e, 60.274a(d), and 60.274b(d)) in terms of opacity limits and only require repairs when holes result in noncompliance with the opacity standards.

Commenter 0095-Steel Co. recommended the EPA better define a "material impact" on the capture system because they said the phrase was too vague. Any airflow changes, they said, may theoretically impact capture efficiency to some extent, but fluctuations that do not affect the compliance of the facility with the substantive emission and opacity standards should not be prohibited. The EPA should define material impacts in terms of opacity limits by revising 40 CFR part 60, subparts AA, AAa, and AAb to only require repairs to openings that lead to noncompliance with the opacity standards.

Response 4-17: For the EPA's response to issues raised in this comment in regard to melt shop inspections for 40 CFR part 60, subpart AAb: See the EAF NSPS preamble to the final rule, section IV.E.2, Melt Shop Inspection Response, as well as EPA's response in section IV.E.2 to the proposed requirements in 40 CFR 60.274b(d) for subpart AAb in regard to allowing operators discretion in an inspection as to what issues "materially impact" the capture system performance. 


4.4 Calculation for Zero Percent Opacity Under Method 9
Comment 4-18: The EPA should clarify the calculation for determining compliance with the opacity limits when using Method 9 for 40 CFR part 60, subpart AAb.
Commenters 0097-Industry Trades and 0095-Steel Co. asked the EPA to clarify in the rule how facilities should determine compliance with the opacity limits. The commenters noted that compliance with the shop opacity limits will be determined based on the arithmetic average of 24 consecutive 15-second opacity observations over a six-minute period. The commenters said that it is their understanding that the proposed zero opacity standard does not require all 24 15-second Method 9 observation periods to be zero percent and that some of the 24 readings may exceed zero percent provided the arithmetic average rounds down to zero. Similarly, in calculating compliance with the existing 6 percent shop opacity standard, some readings can exceed 6 percent provided the arithmetic average rounded down is below 6 percent. The commenters asked the EPA to confirm their interpretation of the rule is correct. 

The commenters noted that this approach is consistent with prior NSPS rulemakings, including 40 CFR part 60, subpart KK (Lead-Acid Battery Manufacturing) and 40 CFR part 60, subpart NN (Phosphate Rock Plants). However, the commenters said the EPA specified in 40 CFR part 60, subparts KK and NN that compliance with the opacity standard is determined by taking the average opacity over a 6-minute period, according to EPA Test Method 9, and rounding the average to the nearest whole percentage. (45 FR 2790 and 2794, January 14, 1980; 47 FR 16564, 16566, 16582 and 16586, Apr. 16, 1982). The commenters recommended the EPA add the same explanation provided in these earlier NSPS in the final shop opacity limit.

Response 4-18: For the EPA's response to issues raised in this comment: See the EAF NSPS preamble to the final rule, section IV.B.2.


4.5 ASTM D7520-16 Digital Camera Opacity Technique
Comment 4-19: The EPA should clarify whether ASTM D7520-16 Digital Camera Opacity Technique (DCOT) can be used by third parties to determine a facility's compliance with the proposed opacity standard.
Commenter 0095-Steel Co. asked the EPA to clarify whether ASTM D7520-16 could be used by a third party (e.g., agency staff) to challenge a facility's Method 9 compliance demonstrations. The commenter was concerned that allowing third parties to use ASTM D7520-16 to determine compliance would make it possible for the EPA or third parties to challenge a facility's determination of compliance made using Method 9. If the EPA allows a third party to use ASTM D7520-16 to determine compliance, then facilities would have to make opacity measurements under both Method 9 and ASTM D7520-16. If ASTM D7520-16 is included as an alternative compliance method, the commenter recommended the EPA clarify that compliance will be demonstrated only by the method chosen by the facility.

Response 4-19: For the purposes of this rule, the method ASTM D7520-16 Digital Camera Opacity Technique (DCOT) or EPA ALT-082 can be used by any facility or any state, local, or delegated authority as an alternative to EPA Method 9. Alternative methods provided in EPA rules are always allowed to be used for compliance purposes unless otherwise specified. 


Comment 4-20: Method ASTM D7520-16 (DCOT) should not be included in 40 CFR part 60, subparts AA, AAa, and AAb.

Commenters 0095-Steel Co. and 0097-Industry Trades opposed the inclusion of method ASTM D7520-16. They said there are serious problems with ASTM D7520-16 (DCOT) and recommended ASTM D7520-16 should not be incorporated into 40 CFR part 60, subparts AA, AAa, and AAb. The reasons provided by the commenters are:
 The EPA has not demonstrated the method is accurate and reliable for measuring fugitive emissions from EAF sources. Use of DCOT to evaluate performance of sources against a zero percent opacity standard (as proposed for 40 CFR part 60, subpart AAa) or 6 percent opacity standard (as proposed for 40 CFR part 60, subpart AAb) is not adequately demonstrated. The commenters said the EPA had not evaluated the potential for false-positive readings that are likely to occur due to the complexity of measuring shop opacity at EAF sources. 
 DCOT is not currently used by any EAF steel producing facility subject to the NSPS standards.
 The DCOT method does not rely on the same observation limits of the human eye and has accuracy and reliability issues. The human eye cannot differentiate between zero percent opacity and near zero opacity and for this reason Method 9 observations are rounded to the closest 5 percent opacity reading. Hence, the Method 9 readings relied on by the EPA showing zero percent during certain compliance demonstrations may reflect observations averaged down from 2.5 percent or higher due to the detection limit of the human eye. A zero percent opacity measured by DCOT has not been demonstrated by the EPA.
 DCOT has not been proven accurate in enough settings, including those conditions that routinely exist in the EAF steel industry. Fugitive emissions at EAF steel mills may be emitted from one or more broad roof vents that are typically total over 300 feet in length. Fugitive emissions may also come from other non-point sources. The low-velocity fugitive plume can vary significantly over time and are often more diffuse than those emitted from stacks. The plume color will also vary. Opacity measurements must be taken in variable ambient wind conditions, cloudy days, and other conditions that do not provide high contrast between the fugitive emission plume and background. A high contrast between the plume and background is the ideal condition for measuring opacity and the characteristics of the fugitive emissions from EAF steel mills make opacity measurement difficult. The DCOT method is not well suited to achieve accurate opacity readings in non-ideal conditions, particularly for measuring very low contrast near the proposed low opacity limits.
 DCOT has a higher error rate than Method 9, especially below 10 percent opacity. The error rate skews high, which would create a more stringent standard while decreasing predictability and accuracy.
 DCOT measurements are susceptible to errors because the observations are taken from a distance and depending on the location and angle the image is taken from, there may be errors due to other nearby plumes. For example, the plume measured may be from a location other than a melt shop. The fugitive emissions must be observed over emissions points, such as roof monitors, that may include additional sources that may generate plumes of particulate or water vapor of differing colors. An experienced and knowledgeable Method 9 observer can adjust to these situations, while a DCOT system cannot.
 DCOT is not suitable for large sources with fugitive emissions for the same reason COMS measurements are not appropriate. Commenter 0097-Industry Trades cited court case (American Iron and Steel Institute et. al vs. EPA (72 FR 53769 (Sept. 20, 2007)) in which the EPA acknowledged COMS can be affected by calibration drift, misaligned mirrors, and dust on the windows or mirrors that can cause positive or negative measurement errors. They added that these distortions and other issues are amplified the further the distance of the camera from the source.
 ASTM D7520-16 has not gone through the standard process for approving test methods. Incorporating the method into 40 CFR part 60, subparts AA, AAa and AAb in this rulemaking, Commenter 0095-Steel Co. said, circumvents the usual review of the validity, accuracy, precision, consistency, and applicability of the test method. The commenter said it was arbitrary and unreasonable to incorporate ASTM D7520-16 without first completing the review the EPA requires for all other test methods.
 DCOT is available from only one vendor. Commenter 0095-Steel Co. was concerned the DCOT vendor was pushing for broader use of method ASTM D7520-16, which they said raised conflict of interest and monopoly concerns. The commenter referenced a petition submitted by Eramet Marietta Inc. for a ferroalloys production rulemaking in 2017 (82 FR 5401), which they said raised concerns the sole-source vendor for certified ASTM D7520-16 technology was "attempting to extract monopolistic pricing and consulting fees, and refusing to warranty that the method would satisfy the applicable testing requirements at issue in that rulemaking". The commenter also noted the Salt River Project Agricultural Improvement and Power District  -  Navajo Generating Station, NSR Appeal No. 16-01 (E.A.B. Aug. 30, 2016), which the commenter said was an example of a sole source provider of ASTM D7520-16 technology filing an appeal to attempt to force a facility to use its ASTM D7520-16 technology to verify compliance instead of Method 9.
In support of their comments, the commenters referenced two studies. The first is titled "Evaluation of the Use of Digital Still Cameras and Human Observers to determine Ambient Plume Opacity during Smoke School Evaluations" prepared by Dr. Mark J. Rood and Associates for Paul Balserak, American Iron and Steel Institute (August 25, 2016). The commenters said that Rood had found DCOT measurements "... are biased to be higher than the reference transmissometer opacity values, since all AOD values are positive at any opacity range." and "... are consistently larger than the corresponding values for Average Opacity Error (AOE) values from human observers for all six sets of results." Dr. Rood's study used the same camera, software and field tests used by the ASTM workgroup and involved 6 tests comprising two plume colors (white and black) varied at opacity ranges 0-100 percent, 0-20 percent and 0-10 percent)).

The second study is titled "Fugitive Emissions Opacity Determination Using the Digital Opacity Compliance System (DOCS)" by Michael J. McFarland, Arthur C. Olivas , Sally G. Atkins , Robert L. Kennedy and Kalpesh Patel (published in the Journal of the Air & Waste Management Association, 57:11, 1317-1325, 2007), which they said found that "... on average, field results indicated that the DOCS technology consistently yielded opacity values that were greater than those reported by Method 9-certified human opacity observers.", that "... local wind shear would often redirect the fog plume away from traveling in a direction perpendicular to the line of sight of the DOCS cameras and Method 9-certified opacity observers." and that "... placement of dedicated targets to observe fugitive emissions was impossible because of the inability to anticipate wind shear direction and magnitude." The commenter noted that the McFarland study found DCOTs recorded higher opacity measurements compared to Method 9 at the mid-offset distances of 90 and 150 ft. They added that McFarland's study had shown that meteorological conditions made DCOT measurements invalid in 38 of the 100 opacity measurements they made due to "... the adverse effects of localized wind shear on particle dispersion and transport." 

Although the EPA has included DCOT in other NESHAPs (e.g., ASTM D7520 - 13 in 40 CFR part 63, subpart XXX for Ferroalloy major sources and ASTM D7520-16 in 40 CFR part 63, subpart FFFFF for Integrated Iron and Steel Manufacturing), the commenters said that in those rulemakings the EPA had not addressed "... fundamental limitations of the DCOT methodology" and had "... misconstrued the findings of the McFarland Report both with respect to the accuracy of DCOT compared to Method 9 and with regard to the implications of establishing an appropriate observation point." Commenter 0097-Industry Trades disagreed with the EPA's statements (1) that a mounted camera can be moved as easily as a human observer to adjust changing wind direction, sunlight, contrast and other factors because a camera must be mounted and calibrated; (2) that Method 9 and DCOT give comparable measurements; and (3) that DCOT measurements were not accurate at the 300 ft distances common in measurements for EAF melt shop roofline emissions. Commenter 0097-Industry Trades added that sources subject to 40 CFR part 63, subpart XXX are not currently using DCOT.

Commenter 0095-Steel Co. noted the opacity limits for those rules are higher (between 8 and 20 percent opacity) than those proposed for EAF sources.

Commenter 0097-Industry Trades said that Method 9 is the preferable compliance measurement method for measuring fugitive emission opacity under the "non-ideal conditions" at EAF sources. They added that Method 9 is more flexible because it allows human observers to move locations, adjust to wind conditions, attend to the presence/absence and angle of sun light, etc. to achieve the best opacity reading.

Response 4-20: We disagree with the commenter on the points mentioned above or find the comments not relevant to the use of the DCOT method as an alternative method for EAF. Our rationale is below: 

 The DCOT method (EPA Alt Method 082) is an approved alternative to EPA Method 9 with the caveats described in the proposed EAF rule in 40 CFR 60.271b under Shop Opacity. In addition, the EPA determined from an evaluation of potential voluntary consensus standards equivalent to the standards included in the proposed rule, as required by the National Technology Transfer and Advancement Act (15 U.S.C. §3701 et seq. (1996)), that ASTM Method D7520-16 (equivalent to EPA Alt Method 082) is an approved alternative for EPA Method 9, as documented within the memorandum, Voluntary Consensus Standard Result for Standards of Performance for Steel Plants: Electric Arc Furnaces and Argon-Oxygen Carburization Vessels Constructed After August 17, 1983, located in the docket for this rule at EPA-HQ-OAR-2002-0049-0066).
 The DCOT method has been approved by the EPA (EPA Method Alt-082) as well as the American Society for Testing and Materials (ASTM). Both organizations have strict protocols for approving methods and seek outside comments before they finalize a method. See https://www.epa.gov/sites/default/files/2020-08/documents/alt082.pdf and https://www.astm.org/about/overview/fact-sheet.html:
 The DCOT method is not required to be used, but is available as an option for facilities that choose to do so. 
 It is not a criterium for approval of a method that the method needs to be in use by a facility in the source category. The DCOT method was approved for the iron and steel industry in 2020 (85 FR 42074) when no steel facility had been using the method.
 It is unclear as to what calibration requirements the commenter is referring to as there are no calibration requirements for the camera in the DCOT method. Additionally, neither EPA Method Alt-082 nor ASTM Method D7520-16 require the use of a mounted camera, but a hand-held or tripod-mounted camera may be used (see Section 8.2 of ASTM D7520-16), allowing for the same flexibilities as described for a manual observer. 
 The location requirements for ASTM D6720-16 are identical to that of EPA Method 9, and in cases where the location may be problematic, the facility retains the ability to use EPA Method 9, as ASTM D6720-16 is provided as an option and is not required. 
 The EPA disagrees with the commenter that COMS are an appropriate comparison to the DCOT. The digital camera requires no mirrors that can be misaligned, and there is no calibration of the camera, so calibration drift is not an issue, and, since the camera itself is not within the source as is a COMS, it can be easily wiped off in the case where dust may have accumulated, which should be less of an issue since it is not directly in the gas stream. 
 In the case of distance from the measurement location, proper siting of the camera is a key component of the DCOT methodology, including appropriate distance. 
 With respect to the effect of localized meteorology, as referenced in McFarland, it appears the camera was not properly sited, and a Method 9 observer would have the same issues if placed in the same location. 
 The hand-held camera option of ASTM D6720-16 allows for the same flexibility in location as the EPA Method 9 observer, and the EPA agrees that a fixed location camera would not have the same flexibility of an EPA Method 9 observer, but the option ASTM D6720-16 for a hand-held camera mitigates that issue. 
 The current availability of a single vendor may limit the overall adoption of the methodology within the EAF industry, but this does not inhibit the EPA in offering it as a compliance method option. For facilities that do not wish to use ASTM D6720-16, the use of EPA Method 9 is still available. 


Comment 4-21: Incorporating Method ASTM D7520-16 (DCOT) into 40 CFR part 60, subparts AA and AAa is not allowed under CAA section 111.

Commenter 0095-Steel Co. said the EPA cannot add ASTM D7520-16 to 40 CFR, part 60, subparts AA and AAa because a CAA section 111 rulemaking cannot be used to retroactively make a rule more stringent. Allowing facilities to use ASTM D7520-16 would add a way for the EPA and third parties to challenge a facility's determination of compliance under Method 9. The commenter added that courts frequently recognize that different titles of the CAA were designed for different purposes, such that the EPA cannot assert the same authority under a different regulation, where Congress has placed different constraints on the EPA's authority, citing Portland Cement Ass'n v. E.P.A., 665 F.3d 177, 183 (D.C. Cir. 2011) (differentiating between Section 111 (NSPS) and Section 112 (NESHAP) regulations, noting that "NESHAP alone governs the regulation of existing sources, and NSPS alone governs the regulation of modified sources," that the two provisions call for different standards of EPA review for setting the emission standards, and treat considerations of cost and discretion differently); Cf. Utility Air Regulatory Group v. Environmental Protection Agency, 573 U.S. 302 (2014) (permitting regulation of GHGs under some parts of Title 1 PSD regulations, but not others, and not allowing under Title V at all).

Response 4-21: Because the DCOT method is only being added as an alternative method to the current regulatory requirements in 40 CFR part 60, subparts AA and AAa (and thus is not mandatory), this addition to those subparts does not make the standards more stringent, either retroactively or prospectively. It merely adds a flexibility for facilities to demonstrate compliance with those standards. The commenter does not otherwise explain how the addition of this voluntary alternative to EPA Method 9 undermines a facilities decision to instead use EPA Method 9, as currently provided in the current EAF NSPS rules.


Comment 4-22: In references to ASTM D7520-16 (DCOT), the EPA should use "points of highest opacity" instead of "sites of highest opacity".

Commenter 0095-Steel Co. recommended the EPA replace "sites of highest opacity" with "points of highest opacity" in the rule language describing method ASTM D7520-16. The commenter said this change would make the rule more consistent.

Response 4-22: We agree with the commenter that the change is appropriate. This language has been changed in the final rule.


4.6 EPA Method 22 (Visible Emissions)
Comment 4-23: Method 22 is inappropriate for measuring compliance with the zero percent opacity limit.
Commenter 0097-Industry Trades and 0095-Steel Co. disagreed with using Method 22 for determining compliance with the shop opacity limit. Commenter 0097-Industry Trades noted that Method 22 does not require certification of the observer and is a qualitative test that does not quantify the level of opacity. Both commenters said that Method 22 is less rigorous than Method 9 because it determines only whether opacity is present. Method 22 was not needed, the commenters said, because the rule allows facilities to use the more reliable and well-established Method 9 to determine compliance with shop opacity limits. 

Commenter 0097-Industry Trades said Method 22 is prone to error when used to determine low levels of emissions or where emissions occur for brief periods of time. Method 22 is also difficult to apply where other sources of emissions are located nearby or where there are wind-borne fugitive emissions from off site. Because Method 22 does not require certification, compliance with the shop opacity limit would be determined by untrained staff. The commenter said that allowing facilities to use Method 22 would lead to less precise and potentially incorrect visible emissions observations and could potentially result in confusion in determining compliance with the shop opacity standard.

Commenter 0095-Steel Co. said the EPA failed to demonstrate that a zero percent opacity can be demonstrated using Method 22. The commenter noted that Method 9 was used in previous measurements showing zero opacity and that the human eye cannot differentiate between zero percent opacity and near zero opacity. The Method 9 readings relied on by the EPA showing zero percent during certain compliance demonstrations may reflect observations averaged down from 2.5 percent or higher due to the detection limit of the human eye. Method 22 averages across time periods should not be used to verify 0 percent opacity because Method 22 is qualitative, is significantly error prone when used to determine low levels of emissions that occur for brief periods of time, especially when opacity may be caused by other emissions sources. Because Method 22 does not require certification of the observer, the commenter said that untrained observers would be allowed to determine compliance with opacity regulations.
Response 4-23: The rule allows Method 22 to be used in cases when zero opacity is expected, to be performed as the total duration of visible emissions measured according to EPA Method 22 over a six-minute period, modified to require the recording of the aggregate duration of visible emissions at 15 second intervals. EPA Method 22 is an approved method specifically for determining zero visible emissions and has training requirements specified in the rule, as follows: 
      Section 2.3 At a minimum, the observer must be trained and knowledgeable regarding the effects of background contrast, ambient lighting, observer position relative to lighting, wind, and the presence of uncombined water (condensing water vapor) on the visibility of emissions. This training is to be obtained from written materials found in References 1 and 2 or from the lecture portion of the Method 9 certification course.
In addition, EPA Method 22 is specifically designed to measure cases where there are low levels of emissions because zero visible emissions are the target parameter in the method and it is being allowed as an option to provide a less complicated, less expensive, and less demanding method where no visible emissions are expected. Also, EPA Method 22 is not required to be used. Both EPA Method 9 and 22 measure visible emissions. In the case of EPA Method 22, only the presence or absence of visible emissions is being determined; in EPA Method 9, the degree to which the visible emissions obscures the background is being determined. Lastly, if opacity from other sources are expected, EPA Method 9 also will indicate visible emissions. All three EAF NSPS rules include a procedure to account for emissions from other sources (40 CFR 60.275(g), 60.275a(h), and 60.275b (h)) and is not a reflection of which test method is being used. 

5.0	ECONOMIC IMPACTS 


Comment 5-1: COVID-19 impacts on costs. 

Commenter 0097-Industry Trades stated the EPA readjusted the Ferroalloys NESHAP's costs from 2012 dollars to 2020 dollars, this recalibration is already outdated. The COVID-19 pandemic has disrupted supply chains in unforeseen fashion with the building and construction industries facing unique shortages, skyrocketing the prices of material. This will only be exacerbated by the fact that many canopy hoods and other ventilation mitigation equipment must, by the EPA's own admission, be custom-designed for facility specific use. Moreover, the COVID-19 pandemic resulted, and continues to exacerbate, a major disruption in America's labor force, with America's construction industry taking a disproportionately massive hit. 

In October 2021, 402,000 construction positions were unfilled, the second-highest level recorded since data collection began in December 2000.[24] In fact, experts anticipate that the construction labor shortage is set to intensify over the next 6 months, stalling countless construction projects across the country. Consequently, scaling the ferroalloy estimates from 2012 USD to 2020 USD is inadequate, as it wholly neglects the unprecedented shift to higher material costs from supply chain disruptions, as well as higher installation costs from a colossal labor shortage and a likely long-term upward readjustment of labor rates. 

Commenter 0097-Industry Trades stated though EPA did anticipate "unforeseen changes in industry and economic shocks" (Economic Impact Analysis) in its Economic Impact Analysis, the paradigm shifts left in the wake of COVID-19 were unprecedented and continue into uncharted territory. In other words, past trends have and will be broken and, accordingly, are of limited utility for future extrapolation.

Response 5-1: As noted in the Economic Impact Analysis (located in the docket for this rule, EPA-HQ-OAR-2002-0049), there is uncertainty associated with compliance costs estimates due to, among other reasons, unforeseen changes in the regulated industry and economic shocks. While we lack data to project the specific impact of economic shocks, such as those described in the comment, on steel production activities at a granular level, in finalizing the proposed rule we updated the compliance cost estimates to incorporate the best available economic information, including information on prevailing interest rates and inflation. 


Comment 5-2: Demand for labor impacts on costs, economics.

Commenter 0097-Industry Trades stated in comparing costs against capital expenditures and revenue, the EPA introduces another interesting way labor costs might rise. In its Economic Impact Analysis, the EPA concedes that "demand for labor employed in steel production activities and associated industries... might experience adjustments as there may be increases in compliance-related labor requirements such as labor associated with the manufacture, installation, and operation of pollution control as well as changes in employment due to quantity effects in directly regulated sectors and sectors that consume EAF-steel. For this proposal, however, the Agency does not have the data and analyses available to quantify these potential labor impacts." (See the Economic Impact Analysis located in the docket for this rule, EPA-HQ-OAR-2002-0049.) 

Response 5-2: As noted in the Economic Impact Analysis for this rulemaking, demand for labor in steel production activities and associated industries might experience adjustments due to increases in compliance-related labor requirements as well as changes in employment due to changes in production from EAF steel manufacturers and sectors that consume EAF-produced steel. While we do not have the data and analysis available to quantify these potential labor impacts, the proposed rule is unlikely to cause large shifts in steel production or prices since the expected compliance costs associated with it comprise only a small portion of total industry revenues.


Comment 5-3: Inflation impacts on costs.

Commenter 0097-Industry Trades stated that, though the adjustment of 2012 USD to 2020 USD in the cost-analyses considered inflation adjustments, inflation rates from 2020 through 2022 are more than those of 2012 through 2020 combined. Indeed, inflation rates in total from 2012 through 2020 are 13.1 percent, while inflation rates from 2020 through 2022 alone are 17.5 percent. Annual rates of inflation are calculated at https://www.usinflationcalculator.com/inflation/current-inflation-rates by using 12-month selections of the Consumer Price Index, which is published monthly by the Labor Department's Bureau of Labor Statistics. Alarmingly, many experts anticipate inflation getting worse, noting "a real possibility that inflation rates will not come down to [policymaker's preferred targets] for many years." This unprecedented spike in inflation grossly undermines the Agency's adjustment of the ferroalloy cost analyses from 2012 to steel-making cost analyses of 2020 using trend-based extrapolation. 

Response 5-3: The EPA recognizes that inflation has been higher than normal from 2020 to the present. While we do not have the available data to quantify the relative impact of inflation on the steel industry versus other industries, in finalizing the rule we updated the estimated compliance costs to incorporate inflation through 2022. 

6.0	MISCELLANEOUS

6.1 Definitions

Comment 6-1: Revision to definition of direct-shell evacuation control system (DEC system).

Commenter 0093-Steel Co. stated the proposed definition of the direct-shell evacuation control system (DEC system) referenced in 40 CFR 60.271b, 60.271a, and 60.271(n), should be amended and applied in 40 CFR part 60, subparts AA, AAa, and AAb, where appropriate. The commenter explained that due to gas generation rates, a DEC system may not maintain "a negative pressure within the EAF shell during melting and refining" at all times. Rather, the system would reduce the pressure and capture the majority of emissions during a charging period, and the remainder of the emissions would be captured in the canopy hood and routed to the control device. The commenter proposes the definition be amended to state, "Direct-shell evacuation control system (DEC system) means a system that creates and maintains a negative or reduced pressure within the electric arc furnace shell during melting and refining, and transports emissions to the control device" to ensure compliance with the requirement as a practicable matter.

Response 6-1: We agree with the commenter that the definition can be made clearer. We have amended the definition in all three subparts (40 CFR part 60, subparts AA, AAa, AAb).


Comment 6-2: Revision to definition of charging period.

Commenter 0097-Industry Trades requested an amendment to the proposed definition of the charging period, which was defined as "the time period when iron and steel scrap or other materials are added into the top of an EAF until the melting and refining period commences." The commenter explained the proposed definition was longer than the true charging period for most bucket-charged EAFs, which, as reflected in the definition of "charge," was the time during "the addition of iron and steel scrap or other materials into the shell of an EAF or the addition of molten steel or other materials into the top of an AOD vessel." The commenter stated the true "charging period" concluded when the furnace roof was closed, at which point no further scrap metal or other materials could be introduced (i.e., charged).

Response 6-2: The EPA disagrees with the commenter that changes to the proposed definition of charging period are necessary. If the definition of charging period was changed as suggested, the definition of "melting and refining period" would also need to change so there is no gap in time between "charging period" and "melting and refining period." If we changed charging period definition as suggested, then the "melting period" would have to be defined as beginning before actual melting starts. Therefore, we are finalizing the proposed definition of charging period to all EAF rules (40 CFR part 60, subparts AA, AAa, AAb) as follows: "Charging period means the time period when iron and steel scrap or other materials are added into the top of an EAF until the melting and refining period commences."


Comment 6-3: Remove or append the definition of modification.

Commenter 0097-Industry Trades requested that the EPA either remove the proposed definition of "modification" from the applicability sections or add a reference to 40 CFR 60.14. The commenter said the EPA restated from 40 CFR 60.2 only the definition of "modification" in this section, even though the terms "construction" and "affected facility" as used in this section were also defined terms in the General Provisions of Part 60. The commenter noted the term "modification" was also substantively explained in 40 CFR 60.14, and that although restating just the definition of "modification" and not also citing 40 CFR 60.14 can in no way eliminate the applicability of 40 CFR 60.14 as it relates to EAFs, this proposed edit raises a potential contradictory circumstance.

Response 6-3: We added the definition of modification in the rule because its sometimes forgotten or is unclear, whereas "new", "constructed," and "affected facility" seem to be well known. We will add a reference to 40 CFR 60.14 in the final rule's definition of modification.


6.2 Startup, Shutdown, and Malfunction 

Comment 6-4: Removal of exemptions for startup, shutdown, and malfunction in CAA section 112 rules should not be extended to CAA section 111.

Commenters 0095-Steel Co. and 0097-Industry Trades stated the EPA cannot extend interpretations to remove startup, shutdown, and malfunction provisions applied to the NESHAPs in CAA section 112 to NSPS in CAA section 111 of the CAA without explanation or justification to address the differences between the two programs. Commenters noted that unlike the NESHAPs under which emissions reduction limits must be maximized and continuous, emissions limits under NSPS need only be continuous if continuous compliance irrespective of SSM was based on the application of the BSER and was achievable and adequately demonstrated. 

Response 6-4: As discussed in more detail in the preamble (section III.A.4 and IV.D), the EPA has determined that the reasoning in the court's decision in Sierra Club v. EPA, 551 F.3d 1019 (D.C. Cir. 2008), which vacated the SSM exemption in CAA section 112, applies equally to CAA section 111. Therefore, we disagree with the commenter on the applicability of this decision to CAA section 111. While the EPA recognizes the differences between the NESHAP and NSPS programs, the court in Sierra Club held that under section 302(k) of the CAA, emissions standards or limitations must be continuous in nature, and the definition of emission or standard in CAA section 302(k) and the requirement for continuous standards applies to both NESHAP and NSPS. 


Comment 6-5: The EPA must provide work practice standards if they remove startup, shutdown, and malfunction exemptions.

Commenters 0095-Steel Co. and 0097-Industry Trades stated the EPA must provide work practice standards if they remove startup, shutdown, and malfunction exemptions. First, the commenters stated, subjecting startup, shutdown, and malfunction periods to the same limit as those during normal operations was not adequately demonstrated as required per CAA section 111(a)(1) and was not provided in the docket prior to promulgation as per CAA section 307(d). Commenters stated that the dataset of stack tests from 33 facilities did not include adequate testing to demonstrate that startup, shutdown, and malfunction periods consistently met the limits proposed in 40 CFR part 60, subpart AAb. Commenters noted that stack tests were not conducted during startup, shutdown, and malfunction periods, and as such could not provide a basis for concluding that emissions during shutdown and startup could comply with the proposed limits. Commenters stated if the EPA cannot show that compliance with a numerical limit was adequately demonstrated during periods of startup, shutdown, and malfunction, and provide that data in the record, then the EPA did not have the legal authority under CAA section 111 to subject those emissions to such a standard.

Second, the commenters stated, the EPA should adopt work practice standards for periods of startup, shutdown, and malfunction. The commenters noted the EPA has historically promulgated work practice standards requiring a source to follow the manufacturer's recommended procedures for minimizing periods of startup and shutdown due to physical limitations and the short duration of startup and shutdown periods, which made it technologically infeasible to conduct the requisite testing for numeric emissions limits. Commenters said the EPA's departure from that approach here was not explained or justified with data from the record, and thus the EPA should implement work practice standards for startup and shutdown as approved by the D.C. Circuit. 

The commenters also stated that malfunction events outside of the control of the operator must be recognized and accounted for as part of the rulemaking process to avoid unnecessary enforcement activity arising out of these events. Additionally, commenters said it was a violation of constitutional due process to mandate liability for malfunctions regardless of cause and rely solely on enforcement discretion to determine in which cases to bring enforcement actions. Commenters stated that specifically recognizing malfunction events in a work practice standard or guidance would help avoid prolonged and expensive debates over what was "reasonably preventable," and provide facilities with greater assurance of how to comply during such events. The EPA has the discretion to adopt a work practice standard under CAA section 111(h) to address periods of malfunction and should do so.

Additionally, one commenter (0097-Industry Trades) stated an example work practice standard under CAA section 111(h) could include the following: 

 Notification within a prescribed period of time to the appropriate regulatory after a malfunction event; 
 Documentation that any excess emissions were caused by a sudden, short, infrequent, and unavoidable failure of air pollution control and monitoring equipment, process equipment, or a process to operate in a normal or usual manner; 
 Explanation that the malfunction event was not part of a recurring pattern indicative of inadequate design, operation, or maintenance; 
 Repair of the equipment as expeditiously as possible when the applicable emission limitations are exceeded; 
 Showing that steps were taken to minimize the frequency, amount, and duration of the excess emissions to the extent practicable; 
 Showing that all reasonable steps were taken to minimize the impact of the excess emissions on ambient air quality, the environment, and human health. 

The commenter noted that the EPA also should specify malfunction events that would qualify for a work practice standard or the EPA enforcement discretion, which should include: 
 Utility outages/power interruptions,
 Safety requirements,
 Unpredictable material failures, 
 Severe weather, and
 Other events beyond the control of the operator.

Response 6-5: For the EPA's response to issues raised in this comment in regard to SSM: See the EAF NSPS preamble to the final rule, section IV.D.2. Regarding commenters' assertion that it would be a violation of constitutional due process to mandate liability for malfunctions regardless of cause and rely solely on enforcement discretion to determine in which cases to bring enforcement actions, we disagree. 

As defined in 40 CFR 60.2, a malfunction means any sudden, infrequent, and not reasonably preventable failure of air pollution control equipment, process equipment, or a process to operate in a normal or unusual manner. Failures that area caused in part by poor maintenance or careless operation are not malfunctions. As described at proposal (87 FR 29710, 29721), the EPA interprets CAA section 111 as not requiring emissions that occur during periods of malfunction to be factored into development of CAA section 111 standards. While the EPA accounts for variability in setting emissions standards, the EPA is not required to treat a malfunction in the same manner as the type of variation in performance that occurs during routine operations of a source. 

This EPA rulemaking establishing emissions limits that operate on a continuous basis does not negate due process as outlined in 40 CFR 60.11(d) in determining whether acceptable operating and maintenance procedures were being used at the time of a malfunction. At all times, including periods of SSM, owners and operators shall, to the extent practicable, maintain and operate any affected facility including associated air pollution control equipment in a manner consistent with good air pollution control practice for minimizing emissions. Determination of whether acceptable operating and maintenance procedures are being used, including at the time of a malfunction, will be based on information available to the Administrator which may include, but is not limited to, monitoring results, opacity observations, review of operating and maintenance procedures, and inspection of the source. 


6.3 Response to the EPA's Solicitation for Information on EAF NSPS Facility List for 2022

Comment 6-6: EAF NSPS Facility List.

Commenter 0093-Steel Co. responded to the EPA's request for information regarding the 88 EAF facilities thought to be subject to the NSPS and have provided an update to the Excel[TM] file document named ``EAF NSPS Facility List 2022'', as requested.

Response 6-6: The EPA agrees with the commenter that corrections to the record were needed regarding the EPA's list of EAF facilities. We have included the revised facility list in our demographic and economic analyses for the final rules. 


6.4 Environmental Justice

Comment 6-7: The EPA's analysis of EO 12898 for environmental justice was insufficient.

Commenter 0094-Enviro asserted that the EPA's analysis of EO 12898 for environmental justice was insufficient because the EPA failed to identify the impacts from the continued operation of existing EAF on communities of color and low-income. The commenter said the EPA focused only on whether the proposed regulatory changes created new disproportionate impacts or exacerbated existing disproportionate impacts, without also addressing whether the proposed rule would allow existing impacts to continue. 

The commenter provided two detailed case studies as examples of environmental justice communities in Alabama, whose health and environment are allegedly being adversely affected by existing EAF, and emphasized that the EPA must conduct a comprehensive analysis to identify all such communities and address their impacts. The commenter provided maps from the EPA Environmental Justice (EJ) Screen Tool (that uses ambient data) for communities around two EAF facilities in Alabama and stated the communities around these facilities are overburdened by air pollution, based on ambient air measurements.
Response 6-7: In this action, the EPA is promulgating improved and more stringent standards that will apply nationwide to new, modified, and reconstructed sources and address the potential impacts thereof.. Furthermore, we are promulgating a requirement that opacity must be maintained at 0 percent during melting and refining operations at affected facilities. 

As described in the proposal, the EPA followed the statutory requirements in reviewing and revising the PM standard of performance for new, modified, and reconstructed EAF and AOD sources. The EPA considered the required factors for BSER and standard of performance as described in preamble sections for the final rule at IV. A, B, C, D, and E. 

At proposal, the EPA presented the results of the demographic proximity analysis conducted for the proposal. In that analysis, we assessed the population living in areas around existing EAF facilities because EPA did not know the locations of future new, modified, or reconstructed facilities that are affected by this rule. 
The demographics analysis for the proposed and final rules showed that the national average U.S. population demographics for whites is 60 percent of the U.S. population, with the average population within 5 km of the 88 EAF facilities being 63 percent white. For minorities, the national average demographics is 40 percent of the U.S. population, with the average population within 5 km of the 88 EAF facilities being 37 percent. Of the 88 facilities, 46 facilities (55 percent) are in areas with greater than 70 percent white population and less than 30 percent minority population.

In summary, this final rule provides benefits of PM reduction to all affected populations, including EJ communities, where new, modified, or reconstructed facilities may be located. The final rule will also require that the total EAF facility PM emissions from control devices at EAF must be at or lower than the standard specified in the NSPS subpart AAb.


Comment 6-8: The EPA should require fenceline monitoring at EAFs neighboring environmental justice communities.

Commenter 0094-Enviro stated the EPA should require additional fenceline monitoring at EAFs neighboring environmental justice communities per Executive Order 12898. The commenter referenced EPA Administrator Regan's January 2022 announcement that the EPA would hold companies more accountable for their actions in overburdened communities with increased monitoring and oversight of polluting facilities. The commenter said the EPA should require more monitoring at the fenceline of EAFs, particularly for PM and SO2.

Response 6-8: We disagree that the EPA should establish fenceline monitoring as part of this review. As described in the proposal, the EPA followed the statutory requirements in reviewing and revising the PM standard of performance for new, modified, and reconstructed EAF and AOD sources. The EPA considered the required factors for BSER and standard of performance as described in preamble sections for the final rule at IV. A, B, C, D, and E. The statute does not require fenceline monitoring.

In reviewing NSPS subpart AAa, as described in the proposal, we reviewed information from the EPA's RACT/BACT/LAER Clearinghouse, facility operating permits, and other regulatory development efforts. From this review, the EPA did not identify any facilities in this source category conducting fenceline monitoring. We have no knowledge that any facilities in the category are conducting fenceline monitoring. Fenceline monitoring is only effective for ground level or close-to-the-ground emission release points, which is not the case for EAF. Therefore, fenceline monitoring was not considered in our
We note that the EPA has only required fenceline monitoring for three categories to date, all of which address one pollutant -- benzene -- that is not a regulated pollutant under CAA section 111.. 


6.5 Integrated 15-minute Block Pressure Averages in 40 CFR Part 60, Subparts AA and AAa

Comment 6-9: The EPA should reduce the integrated 15-minute pressure averages in 40 CFR part 60, subparts AA and AAa to 1-minute pressure averages. 

Commenter 0088-Citizen stated proposed 40 CFR 60.264(f) and 60.264a(f) would require 15-minute integrated block averages of EAF pressure for EAFs controlled by use of a DEC system. No data have been presented to indicate that the 15-minute time interval is relevant to overall particulate emissions for an individual cycle. Because of the cyclic operation of EAFs, it is possible that the bulk of the particulate emissions occur during a much shorter time interval and that the 15-minute averaging period would obscure the pressure fluctuations indicative of high emission rate. In the absence of data supporting the 15-minute block average, the EPA should require more conservative 1-minute pressure averages. The same instrumentation used to measure pressure and calculate 15-minute averages is capable of calculation 1-minute averages.

Response 6-9: This is not the correct citation. The EAF NSPS starts at 40 CFR 60.270. It's likely the commenter meant 60.274(f) and 60.274a(f). The "cyclic" operation that the commenter states actually is a batch process, where a batch lasts 4 to 6 hours and not minutes. Therefore a 15-minute average is not out of line. However, in response to other comments regarding the proposed averaging periods of 15 minutes for furnace static pressure described in Chapters 3 and 4, the EPA has modified the averaging period language in the final rule for 40 CFR part 60, subparts AA, AAa, and AAb (40 CFR 60.274(f), 60.274a(f), and 60.274b(f)) to be "no greater than 15 minutes." This modification allows greater flexibility for establishing monitoring setpoints to capture the variability that may occur during short periods that are less than 15 minutes.

Because we did not propose 1-minute averages, and the commenter provides no data to support the assertion that they are needed, we are not requiring shorter timeframes in the final rules. Other timed measurements in the NSPS include the following and refer to 15-minute or longer periods:

 Melting and Refining: Monitor and record on a continuous basis the rolling 15-minute average furnace static pressure (if a DEC system is in use, and a furnace static pressure gauge is installed);
 Melting and Refining: Maintain a monitoring device that continuously records on a rolling 15-minute average basis either the volumetric flow rate through each separately ducted hood or the static pressure at each separately ducted hood;
 Charging and Tapping: "...install, calibrate, and maintain a monitoring device that continuously records either the volumetric flow rate through each separately ducted hood or the rolling 15-minute average static pressure at each separately ducted hood.
 Visible emission observations shall be conducted at least once per day of the control device for at least three 6-minute periods [total equals 18 minutes];
 Visible emission observations shall be conducted at least once per day of the control device for at least three 6-minute periods when the furnace is operating in the melting and refining period [total equals 18 minutes].
 In addition, PM compliance tests, consisting or three runs of Method 5, typically occur over three hours or more of source operation and sampling for PM. 

In summary, nothing in the rule record supports shorter time intervals than 15 minutes for required assessments of EAF operation.


6.6 Criteria for Exempting Capture System Operations Ducted to a Control Device 

Comment 6-10: The EPA should include exemption criteria for capture system operations ducted to a control device. 

Commenter 0088-Citizen stated proposed 40 CFR 60.264(h) [in 40 CFR part 60, subpart AA] exempts operators of EAFs with capture systems designed and operated such that all emissions are captured and ducted to a control device from the requirements of 40 CFR 60.264. No criteria are provided for an operator to demonstrate that the system is being operated in this manner. In the absence of such criteria, permitting authorities will have no way to confirm that the capture systems are designed and operated in this manner, and operators will have no way to demonstrate that they are eligible for this exemption. If the EPA believes that Method 204 (criteria for and verification of a permanent or temporary total enclosure) is appropriate, this should be referenced. If not, the EPA should propose a procedure for establishing eligibility for this exemption.

Response 6-10: The commenter likely has used an incorrect citation. The EAF NSPS rules start with 40 CFR 60.270. It is likely the commenter meant 40 CFR 60.274(h) in 40 CFR part 60, subpart AA. The EPA did not address, and therefore did not reopen, this aspect of the rule (in 40 CFR 60.274 (h)) in the 2022 proposal, where if a capture system that is designed and operated such that all emissions are captured and ducted to a control device, the owner or operator is not subject to the requirements of the section. This comment is therefore outside the scope of this rulemaking. Moreover, the commenter does not provide any information as to why a change is needed for 40 CFR part 60, subpart AA. On the contrary, shop opacity would be elevated if it were the case that the emissions were not sufficiently captured and ducted and an exceedance of the shop opacity standard would be noted. Therefore, no additional monitoring requirements would be warranted for 40 CFR part 60, subpart AA. Note that this aspect of the rule was not repeated in 40 CFR part 60, subpart AAa and AAb. We believe there is only one operating EAF that is subject to AA (and a few other EAF facilities that are not subject to any NSPS). 


6.7 Procedural--Comment Period Extension

Comment 6-11: The EPA should extend the comment period from 7/15/22 to a later date.

Commenter 0087-Enviro stated the undersigned groups respectfully request that the EPA extend the public comment period for the recently published proposed revisions to CAA standards of performance for EAFs and AOD vessels, EPA-HQ-OAR-2002-0049, 87 FR 29710 (May 16, 2022). The EPA set a deadline of July 15, 2022 for submission of public comment. The commenter stated the undersigned groups, who collectively represent hundreds of community members who live in communities neighboring EAF facilities, respectfully request a 30-day extension of the comment period for the proposed rule. Based on our review of the EPA's regulatory docket, this proposed rule appears to be the product of years of communications between the EPA and the regulated industry, dating back to the previous Administration, but our groups were not informed of the scope of this rulemaking until it was published in the FR a few weeks ago. In order to offer informed comments here, our groups need the additional time to communicate with our members, many of whom live in environmental justice communities and have direct experience with pollution from EAF facilities. In addition, the proposed rule involves highly technical and complex subject matter, requiring expertise in fields such as engineering to fully evaluate, and our groups may need to consult with one or more technical experts to most effectively evaluate the proposed rule and offer own comments on proposed regulatory changes. 

Commenter 0097-Industry Trades (on behalf of the Steel Manufacturers Association (SMA), the American Iron and Steel Institute (AISI), and the Specialty Steel Industry of North America (SSINA)) and Commenter 0095-Steel Co. stated we respectfully request that the EPA extend by at least 60 days the comment period for the Agency's proposed new and revised NSPS for EAF steel manufacturing (NSPS Revisions). The Commenters stated the EPA's current deadline for comments on the proposed NSPS Revisions is July 15, 2022. Given the overlap of this comment period with the comment period for the proposed "Good Neighbor Rule" (Federal Implementation Plan Addressing Regional Ozone Transport for the 2015 Primary Ozone National Ambient Air Quality Standard), which runs through June 21, 2022, and includes first-ever proposed coverage of the steel industry, the additional time is warranted and necessary. An extension of the comment period also is justified by the fact that the proposed NSPS Revisions are more complex than anticipated, notably due to proposed changes that would fundamentally rewrite the long-existing compliance standards (40 CFR part 60, subparts AA and AAa) for facilities that are neither new nor modified and have been complying with the existing NSPS standards for years, if not decades. Accordingly, we request that EPA accept comments until at least September 13, 2022. The Commenters also stated the overlapping comment periods for these two major rulemakings (Good Neighbor and the NSPS Revisions) is a dramatic strain on the resources of the U.S. EAF steel manufacturing industry. Both rules involve highly complex proposed emission limits and other amendments to the primary air emission limitations that apply broadly across the industry. Preparing meaningful comments on both rule proposals requires the time and resources of the same set of company and industry experts, consultants, and legal counsel. To respond to the Good Neighbor Rule proposal alone, during its very tight 75-day comment period, has occupied the industry's full time and attention over the last several weeks. It is therefore imperative that the EPA grant the additional requested time to enable meaningful review and comment on the proposed NSPS Revisions. 

Commenter 0097-Industry Trades and 0095-Steel Co. stated we appreciate that the EPA has been cooperative and open to industry feedback during development of the proposed NSPS Revisions. While several proposed changes were anticipated, such as a tighter PM standard and more stringent requirements for facilities "new" or "modified" after the date of proposal (May 16, 2022), the proposed NSPS Revisions go much farther than anticipated or, as detailed in our other comments, appear to exceed the EPA's authority with regard to revisions to the existing NSPS standards in 40 CFR part 60, subparts AA and AAa. Commenter 0097-Industry Trades and 0095-Steel Co. stated for the foregoing reasons, we respectfully request that the EPA extend the comment period on the proposed NSPS Revisions until at least September 13, 2022.

Response 6-11: The EPA extended comment period on the EAF NSPS proposal until August 15, 2022, which provided a total comment period lasting 90 days from publication of the proposal on May 16, 2022 (87 FR 29710) to the final comment due date. A pre-publication version of the preamble was posted online at https://www.epa.gov/stationary-sources-air-pollution/electric-arc-furnaces-eafs-and-argon-oxygen-decarburization on April 29, 2022. The original publication of the proposal provided 60 days for comments beginning with publication of the proposed rule in the Federal Register. The EPA then extended the comment period on July 15, 2022 (87 FR 49796), providing an additional 30 days to review the proposal, for a total of 90 days from publication in the Federal Register to the close of the comment period, and a total 105 days between posting of the proposal online on April 29, 2022, and the final comment due date of August 15, 2022. In light of this extensive comment period, the EPA did not find it necessary or appropriate to grant a longer extension of the comment period. 


6.8 Adoption of Other Commenter's Comments

Comment 6-12: Acceptance of other comments.

Commenter 0093-Steel Co. stated they support and adopt comments submitted by AISI and SMA. Issues and concerns raised through our comments, including comments submitted by our associations, can serve as an objection pursuant to section 307(d)(7)(B) of the CAA. (33 42 U.S.C. section 7607)

Response 6-12: The EPA has responded to the comments submitted by AISA and SMA elsewhere in this document and in the preamble. The requirements of CAA section 307(d)(7)(B) only apply to proceedings that may post-date this final action, and any questions of whether such requirements are met would be determined in those separate proceedings. 


6.9 Regulation of Other NSPS Pollutants, Major Sources of HAP, and Greenhouse Gases

Comment 6-13: The EPA should list EAFs as a category of major sources of HAP under CAA section 112 and promulgate HAP emissions standards. 

Commenter 0094-Enviro stated according to the data in the record, many EAFs appear to be major sources of HAP, and yet the EPA has never listed EAFs as a major source category under CAA section 112(c)(1) and promulgated national emissions standards for the HAP that major source EAFs emit. The EPA's failure to list EAFs as a major source category  -  notwithstanding its listing of area source EAFs  -  is long overdue and violates the EPA's non-discretionary duties under the CAA. Many EAFs are located in close proximity to low-income communities and communities of color, making the need for these facilities to be covered by standards limiting the emissions of dangerous air toxics all the more important.

Response 6-13: To clarify, the proposed rule on which the commenter has submitted comments is an NSPS, which by law (CAA section 111) does not differentiate between major and other sources. The comment about CAA section 112 is outside of the scope of this action. 


Comment 6-14: The EPA must also engage in a new CAA section 111(d) rulemaking to address greenhouse gas emissions from the entire steelmaking industry, including EAFs. 

Commenter 0094-Enviro stated the EPA must initiate a new rulemaking under CAA section 111(d) to address greenhouse gas emissions from the entire steelmaking industry, including but not limited to EAFs. Iron and steel production was responsible for over 62 million tons CO2e emitted in 2020, (EPA, Greenhouse Gas Reporting Program (GHGRP): GHGRP Metals, https://www.epa.gov/ghgreporting/ghgrpmetals (last visited Aug. 8, 2022).) which was 9.1 percent of the total US emissions that year. (EPA, Sources of Greenhouse Gas Emissions, https://www.epa.gov/ghgemissions/sources-greenhouse-gasemissions (last visited Aug. 8, 2022).) Even after the U.S. Supreme Court's recent decision in West Virginia v. EPA, there can be no doubt that the U.S. steel industry is a major contributor to climate endangering greenhouse gas emissions, and that the EPA has a responsibility to address emissions from this industry under CAA section 111(d). The most efficient and effective way to regulate would be to establish facility-specific, inside-the-fenceline limits (consistent with West Virginia v. EPA, 577 U. S. 1126 (2016)) that would regulate the industry comprehensively along the same timeframe, so that some portions of the industry are not placed on an uneven regulatory playing field with other portions.
Commenter 0094-Enviro continued EAFs are likely to benefit from greenhouse gas (GHG) regulations, as they produce substantially less CO2e per ton than conventional steel facilities. For example, a recent study by CRU Group for the Steel Manufacturers Association found that steel produced by EAFs had approximately 75 percent lower carbon intensity than traditional blast furnace steelmakers. (Steel Mfrs. Ass'n, Press Release, Independent Study Validates that Steelmaking by Electric Arc Furnace Manufacturers in U.S. Produces 75 percent Lower Carbon Emissions (July 25, 2022). https://steelnet.org/independentstudy-validates-that-steelmaking-by-electric-arc-furnace-manufacturers-in-u-s-produces-75-lower-carbon-emissions/ (last visited Aug. 8, 2022). The carbon benefits of EAF steelmaking will only increase with greater reliance on renewable energy and green hydrogen-powered direct reduction ironmaking  -  both of which the EPA should consider requiring in a CAA section 111(d) rulemaking. However, the high likelihood that steelmaking will shift away from traditional blast furnace production to EAFs as the industry decarbonizes only makes it more important that the EPA use this current NSPS rulemaking to require good air pollution control practices that specifically address SO2, NOx, CO, and other pollutants from both existing and new EAFs. Because these facilities are the future of the industry, it is imperative that the EPA act now to address their health and environmental impacts. 

Response 6-14: As described in the proposal and the final rule preamble, the EPA followed the statutory requirements in reviewing and revising the PM standard of performance for new, modified, and reconstructed EAF and AOD sources under 111(b). In this action, EPA's obligation is limited to reviewing and revising the PM limits within 40 CFR part 60, subpart AAa, and we are thus not obligated to consider other pollutants, such as GHG, SO2, NOx, or CO at this time. Due to the deadline imposed upon the EPA by the consent decree in Our Children's Earth Foundation v. Wheeler, No. 18-cv-04765 (N.D. Cal. entered May 10, 2019), there was insufficient time to collect the data necessary to address pollutants other than the currently regulated PM in this review of the EAF NSPS. 


Comment 6-15: 40 CFR part 60, subparts AA and AAa should include limits for sulfur dioxide (SO2), nitrogen oxides (NOx), carbon monoxide (CO), and other pollutants.

Commenter 0094-Enviro stated the EPA's NSPS regulations for existing EAFs (40 CFR part 60, subparts AA and AAa) specifically limit only particulate matter (PM) and opacity. With respect to other pollutants such as SO2, NOx, and CO, the EPA has never determined what level of pollution reduction would be achievable using good air pollution control practices. The EPA through this rulemaking should supplement 40 CFR part 60, subparts AA and AAa to set limits for SO2, NOx, CO, and other pollutants, as well as to add limits at least as stringent for new sources to the proposed 40 CFR part 60, subpart AAb. The commenter continued it is possible to establish low emission rates for EAFs based solely on good work practices that are implemented by the best-performing sources, without expensive add-on controls. For example, one way to better control SO2 emissions from EAFs would be to require that they ensure the best quality scrap is used and produce monthly reports on the quality of their scrap to their permitting agencies. Such requirements would be similar to pollution prevention practices already required of EAFs [in 40 CFR part 63, subpart YYYYY] that are area sources of hazardous air pollutants (HAP) to control chlorinated plastics, lead, free organic liquids, and mercury in the scrap they use (40 CFR 63.10685). Adding sulfurous compounds to the list of materials that EAFs are required to address through a pollution prevention plan should not impose a significant new regulatory burden on the industry and would result in significant public health benefits to communities neighboring EAF facilities. (It is also worth noting that EAFs that rely on direct reduction ironmaking (DRI) instead of scrap may be able to eliminate emissions of SO2, mercury, and other pollutants.) 

Commenter 0094-Enviro continued substantial reductions of NOx and CO also can be achieved through good operating practices. The good operating practices used at best performing facilities such as CF&I Steel in Pueblo, Colorado, are consistent with more efficient steel making, such as minimizing introduction of air into the furnace and reducing lost energy due to production delays. The EPA's RACT/BACT/LAER Clearinghouse reflects that CF&I Steel achieves average emissions rates of 0.15 lb NOx per ton of steel produced using good combustion practices and process controls and 2 lbs CO. (https://cfpub.epa.gov/rblc/index.cfm?action=PermitDetail.PollutantInfo&Facility_ID=25898&Process_ID=103430 &Pollutant_ID=149&Per_Control_Equipment_Id=137234; 
https://cfpub.epa.gov/rblc/index.cfm?action=PermitDetail.PollutantInfo&Facility_ID=25898&Process_ID=103430 &Pollutant_ID=48&Per_Control_Equipment_Id=137236) 

Response 6-15: As described in the proposal and the final rule preamble, the EPA followed the statutory requirements in reviewing and revising the PM standard of performance for new, modified, and reconstructed EAF and AOD sources. The points raised by the commenter about regulating other NSPS pollutants from EAF may be worth exploring at some point in the future. Because the regulation of other pollutants was not proposed, the EPA is prevented from introducing any new standards into the final EAF NSPS rules.. Also, the time allotment enforced by law onto the EPA in typical court orders for regulations prevents the EPA from considering changes to current NSPS. Even considering other pollutants into a new NSPS is difficult under time-constrained rulemakings because a CAA section 114 request is typically needed to find a good selection of advancements in pollution control because not all state agencies report advancements in control levels to the RACT/BACT/LAER Clearinghouse.


Comment 6-16: The EPA should revise 40 CFR part 60, subparts AA and AAa to clarify that facilities' general duty to minimize emissions requires EAFs to use good air pollution control practices to minimize SO2 and other pollutants. 

Commenter 0094-Enviro stated for decades, the EPA's NSPS regulations have included a "general duty" requirement that polluting facilities "shall, to the extent practicable, maintain and operate any affected facility including associated air pollution control equipment in a manner consistent with good air pollution control practice for minimizing emissions". The commenter further stated that existing EAF facilities can impose significant public health and environmental burdens on their neighboring communities, especially when good air pollution control practices are not followed. The EPA should use this rulemaking to make a specific determination of what good air pollution control practices require at both existing and new EAFs.

Response 6-16: The general provisions to NSPS (section 111) already has a "general duty" requirement that is stated to be based on monitoring results, opacity observations, review of operating and maintenance procedures, and inspection of the source. See below from the general provisions to NSPS (underline added).

      Section 111 General Provisions. 40 CFR 60.11 Compliance with standards and maintenance requirements. (d) At all times, including periods of startup, shutdown, and malfunction, owners and operators shall, to the extent practicable, maintain and operate any affected facility including associated air pollution control equipment in a manner consistent with good air pollution control practice for minimizing emissions. Determination of whether acceptable operating and maintenance procedures are being used will be based on information available to the Administrator which may include, but is not limited to, monitoring results, opacity observations, review of operating and maintenance procedures, and inspection of the source.
      
The EAF NSPS address control device operation and emissions, with emission and operating limits and specific and detailed procedures required to ensure proper operation of capture and control systems. 
6.10 Reissuing of Rule

Comment 6-17: The EPA must reissue the rule for additional comment if substantive changes in approach are made in the final rule.

Commenter 0095-Steel Co. stated the rulemaking procedures at section 307(d) of the CAA specifically require that a proposed rulemaking must "include a summary of -- (A) the factual data on which the proposed rule is based; (B) the methodology used in obtaining the data and in analyzing the data; and (C) the major legal interpretations and policy considerations underlying the proposed rule" and "All data, information, and documents referred to in this paragraph on which the proposed rule relies shall be included in the docket on the date of publication of the proposed rule." (42 U.S.C. section 7607(d)(3)) Furthermore, any final "promulgated rule may not be based (in part or whole) on any information or data which has not been placed in the docket as of the date of such promulgation." (42 U.S.C. section 7607(d)(6)(C)) Relatedly, the EPA has "an initial burden of promulgating and explaining a non-arbitrary, non-capricious rule" including an obligation to "explain how the standard proposed is achievable under the range of relevant conditions which may affect the emissions to be regulated." (National Lime Ass'n v. EPA, 627 F.2d 416, 433 (D.C. Cir. 1980) and Nat. Res. Def. Council v. EPA, 755 F.3d 1010, 1023 (D.C. Cir. 2014)) 

The commenter continued, if the EPA makes any revisions to the proposed rule that are not supported by the data in the docket, including but not limited to changing opacity standards to something other than the limits specified in the current text of 40 CFR part 60, subparts AA and AAa or the proposed text of 40 CFR part 60, subpart AAb, the EPA should issue a new proposed rule providing the opportunity for public comment on the basis for any such newly proposed changes. Furthermore, it would be arbitrary for the EPA to reverse any of the determinations it has made in this proposed rule, such as by determining that it has authority to issue any revision to 40 CFR part 60, subpart AAa other than "minor" and "editorial and clarifying changes" (87 FR at 29721 and 29726) without first re-issuing such changes in the form of a new proposed rule for additional public comment.

Response 6-17: The EAF final rule finalizes requirements that are consistent with or a logical outgrowth of the proposed rule and public comments receive thereon. Therefore, the EPA did not find it necessary to issue a new proposed rule as to any of the final elements of this action. 
