Prevention of Significant Deterioration (PSD) Permit Modification Application. Outokumpu Stainless USA, LLC. June

Transcription

1

2 Prevention of Significant Deterioration (PSD) Permit Modification Application Outokumpu Stainless USA, LLC June

3 Outokumpu Stainless USA, LLC PSD Permit Modification Application June 2014 Project No Calvert, Alabama Ramesh Narasimhan Partner-in-Charge Deepu Dethan Project Manager Environmental Resources Management Southwest, Inc. 775 University Blvd. North, Suite 280 Mobile, Alabama (251) (Phone) (888) (Fax)

4 TABLE OF CONTENTS 1.0 INTRODUCTION BACKGROUND NITROGEN OXIDES - NO X SULFUR DIOXIDE - SO CONTROL TECHNOLOGY ANAYLSIS TOP-DOWN BACT ANALYSIS METHODOLOGY CONTROL TECHNOLOGY DETERMINATION FOR THE ELECTRIC ARC FURNACE BACT Determination for NO x Emissions from the Electric Arc Furnace BACT Baseline Step 1 Identify Potential Control Technologies Step 2 Eliminate Technically Infeasible Options Step 3 - Rank Remaining Technically Feasible Control Options Step 4 - Evaluating Remaining Control Technologies Step 5 Selection of BACT BACT Determination for SO 2 Emissions from the Electric Arc Furnace BACT Baseline Step 1 Identify Potential Control Technologies Step 2 Eliminate Technically Infeasible Options Step 3 - Rank Remaining Technically Feasible Control Options Step 4 - Evaluating Remaining Control Technologies Step 5 Selection of BACT FEDERAL AIR REGULATORY ANAYLSIS REQUESTED PERMIT CONDITION UPDATES 45 ii 2014\ \841rpt.doc

5 TABLE OF CONTENTS (CONT D) LIST OF APPENDICES A B C PROJECT EMISSION CALCULATIONS ADEM FORMS AIR QUALITY MODELING ANALYSIS REPORT LIST OF TABLES 3-1 RBLC Nitrogen Oxide Emission Limits from an Electric Arc Furnace 3-2 RBLC Sulfur Dioxide Emission Limits from an Electric Arc Furnace iii 2014\ \841rpt.doc

6 1.0 INTRODUCTION Outokumpu Stainless USA, LLC (Outokumpu) owns and operates a stainless steel mill located in Calvert, Alabama. The facility was previously owned and operated by ThyssenKrupp Stainless USA, LLC (TKL). TKL submitted Prevention of Significant Deterioration (PSD) permit applications for the stainless steel mill and obtained construction authorizations via PSD permits issued by the Alabama Department of Environmental Management (ADEM). Initial operation of certain sources at the facility commenced in June 2010 under Temporary Authorizations to Operate (TAOs) issued by ADEM. As per Alabama Administrative Code (AAC) (1), an initial Title V operating permit application was submitted within 12 months after the commencement of operations. Outokumpu acquired the facility in January of 2013, and filed the necessary transfer of ownership notifications. The Title V permit has not yet been issued for the facility. At the Outokumpu Mill, stainless steel is produced from scrap metal in a stainless steel melt shop that consists of an electric arc furnace (EAF), argon oxygen decarburization (AOD) converter, two ladle treatment stands (LTS), and a continuous caster (CC). Operations for the melt shop were initially permitted by ADEM on March 25, 2010 (Permit Number X001 under ThyssenKrupp Stainless Steel USA, LLC Ownership). On September 6, 2012, ADEM issued an amended air permit to reflect changes in particulate emissions from the melt shop (Permit Number X001). Most recently, on March 21, 2013 ADEM re-issued the melt shop permit to reflect the new facility name and ownership (Permit X001 under Outokumpu Stainless USA, LLC ownership). Condition 16 of Air Permit X001 establishes a nitrogen oxides (NO x ) emission limit of 0.35 pound per ton (lb/ton) and a sulfur dioxide (SO 2 ) emission limit of 0.15 lb/ton assessed at the outlet of the baghouse associated with the EAF (Emission Point ID: LO-1). This permit application is being submitted as an update to the original PSD permit application submitted in June 2007 to request more representative higher emissions limits for NO x and SO 2 for the EAF (LO-1) by modifying the currently effective PSD Permit Number X001. The following sections discuss the background and rationale for the request of increased emission limits. Environmental Resources Management Southwest, Inc \ \841rpt.doc

7 2.0 BACKGROUND The emission limits from Outokumpu s stainless steel melt shop are currently permitted under Permit Number X001 issued on March 21, Outokumpu is proposing to modify the NO x and SO 2 emissions limits established by this permit as follows: 1. Increase NO x Emission Limit from 0.35 lb/ton and lb/hr to 1.00 lb/ton and 126 lb/hr for EAF baghouse LO Increase SO 2 Emission Limit from 0.15 lb/ton and 18.9 lb/hr to lb/ton and lb/hr for EAF baghouse LO-1. Because existing PSD emission limits are being modified and due to the resulting increase in maximum potential emissions, the proposed modification will be subject to PSD review for NO x and SO 2. As part of PSD Review, a Best Available Control Technology (BACT) analysis must be conducted to evaluate potential controls options and appropriate emission limits for the stainless steel EAF. This application presents an updated BACT analysis for the EAF. Additionally, PSD review requires the applicant to perform an ambient air quality impact analysis to demonstrate compliance with the PSD increments and national ambient air quality standards (NAAQS). This application documents the required air dispersion modeling analysis that was performed for NO x and SO 2. This PSD air permit modification application includes the following: A detailed description of the proposed emission limit changes (Section 2); A complete BACT analysis and determination for the stainless steel EAF (Section 3); An air regulatory analysis of items relevant to the changes (Section 4); Updated emission calculations for the stainless steel EAF (Appendix A); ADEM permit application forms (Appendix B); and, Air quality modeling report (Appendix C). Please note, the only affected emission unit addressed in this application is the EAF (LO-1). This permit application does not propose any modifications to any other sources at the Outokumpu Mill. 2.1 NITROGEN OXIDES - NO X On May 28, 2013, Outokumpu received a quality assured emissions test report from Air Compliance Consultants Incorporated which provided the results for a stack test conducted at the outlet of Baghouse 1 installed on the EAF (Emission Point ID: LO-1). The stack test showed that the permitted NO x emission limit of 0.35 lb/ton (43.97 lb/hr) had been exceeded. On May 30, 2013, Outokumpu submitted a deviation report as required by Permit Condition 34 (within two Environmental Resources Management Southwest, Inc \ \841rpt.doc

8 working days of the stack test report). A copy of the stack test was included as an attachment in the deviation report. Upon filing the emissions deviation report, Outokumpu reviewed the stack testing data, evaluated the EAF process and maintenance procedures, and reexamined the BACT analysis presented in the original PSD permit application dated June 2007 submitted to ADEM by TKL to determine the underlying basis of the 0.35 lb/ton NO x emission limit for the EAF. Upon further review of the antecedent BACT analysis, Outokumpu concluded that the control technology determination and corresponding NO x emission limit for the EAF is not representative of the type of stainless steel production process at the Outokumpu Mill in Calvert. In the June 2007 PSD permit application, low NO x oxy-fuel burner technology was proposed as NO x BACT for the EAF which resulted in the NO x BACT limit of 0.35 lb/ton (43.97 lb/hr) in the current PSD permit. As per the description provided in the June 2007 PSD permit application, oxy-fuel burners reduce NO x emissions by replacing the ambient air used to support natural gas combustion with oxygen enriched air. Plant-grade oxygen contains 5-10% nitrogen (volume), compared to ambient air that contains 79% nitrogen (volume). By replacing ambient air with oxygen enriched air, the production of thermal NO x that occurs due to the reaction of nitrogen and oxygen present in the air at very high furnace temperatures is reduced because there will be much less available nitrogen to produce thermal NO x. The stainless steel EAF at the Outokumpu Mill utilizes oxygen lances, which provide small quantities of oxygen to the furnace and cause more uniform temperatures within the furnace, particularly near the slag door where the temperatures are typically lower. These oxygen lances cannot provide the large quantities of oxygen needed to replace ambient air in the EAF and therefore cannot cause similar levels of reduction in thermal NO x production as an oxy-fuel system would do. Outokumpu believes the selection of a BACT emission limit for the EAF based upon oxy-fuel technology is technically infeasible for this furnace. Furthermore, the EAF at the Outokumpu Mill does not use natural gas combustion as stated in the above quoted text from the 2007 PSD permit application. A detailed discussion on the technical infeasibility of oxy-fuel systems for a stainless steel EAF is provided in Section of this document. A representative and current-day NO x BACT analysis for the EAF is presented in Section of this document. 2.2 SULFUR DIOXIDE - SO 2 On May 16, 2013 and May 17, 2013, the Continuous Emissions Monitoring System (CEMS) associated with the EAF recorded SO 2 permit exceedances over a three hour averaging period. In response to the CEMS recordings, Outokumpu submitted a deviation report to ADEM on May 20, 2013; Outokumpu submitted the deviation report as required by Permit Condition 34 (within two working days of the exceedance). Environmental Resources Management Southwest, Inc \ \841rpt.doc

9 Upon filing of the emissions deviation report, Outokumpu reviewed the emissions data, evaluated the EAF process and maintenance procedures, and reexamined the BACT analysis presented in the original PSD permit application dated June 2007 submitted to ADEM to determine the underlying basis of the 0.15 lb/ton SO 2 emission limit for the EAF. In the original PSD permit application submitted by TKL, the SO 2 emission limit is based on data in the Reasonably Achievable Control Technology (RACT) / BACT/Lowest Achievable Emission Rate (LAER) Clearinghouse (RBLC) database without taking into consideration the key process differences between carbon steel and stainless steel production that impact the formation of SO 2. One of the main differences between these two production processes is that the sulfur content in the charge scrap for the stainless steel process is typically greater than that in the carbon steel process. Charge scrap for producing carbon steel consists of non-alloyed steel while charge scrap for stainless steel production consists of low or high alloyed metal. 1 Non-alloyed steel used as scrap for carbon steel production can have a sulfur content ranging from %. 2 Low-alloyed or high-alloyed steel, however, can have a sulfur content ranging from % according to the American Iron and Steel Institute. 3 Another fundamental difference between the carbon steel and stainless steel production processes is the formation of the foam slag layer. In the stainless steel production process, the chrome oxide content of the slag causes the slag to have a lower foam index as compared to a carbon steel slag. 4 This means that in a stainless steel production process, the slag layer is significantly less abundant than in the carbon steel production process. Most of the slag formation and removal in a stainless steel production process occurs in the AOD. As sulfur compounds are primarily removed as sulfides dissolved in the slag, the amount of SO 2 emitted from the process depends on the volume of slag generated; as slag production increases, SO 2 emissions decrease and vice versa. This is why a stainless steel EAF has higher SO 2 emissions compared to a carbon steel EAF, and the AOD in a stainless steel process has lower SO 2 emissions than the stainless steel EAF. Furthermore, the original PSD permit application provided no justification for the selection of the 0.15 lb/ton SO 2 BACT limit and appears to be largely arbitrary without evaluating the above discussed key factors that impact SO 2 emissions from a stainless steel EAF. A representative and current-day SO 2 BACT analysis for the EAF is presented in Section of this document. 1 The European Association Representing Metallurgical Slag Producers and Processors: Electric Arc Furnace Slag. Electronic Source: on ). 2 MSDS for Non-Alloyed Steel. Electronic Source: (accessed on ). 3 High Strength Low Alloyed Steel Document. Electronic Source: df/a764507a d23-b d31c0ba2 (accessed on ). 4 Metallurgical and Materials Transactions B August 2004, Volume 35, Issue 4, pp Environmental Resources Management Southwest, Inc \ \841rpt.doc

10 3.0 CONTROL TECHNOLOGY ANAYLSIS Outokumpu has researched the RBLC database, vendor data, and PSD permit applications from stainless steel production facilities in efforts to fully understand the fundamental cause of the SO 2 and NO x emission deviations and to amend the control technology determination for its stainless steel EAF to represent current-day BACT. The following sections present the BACT analysis for the EAF ameliorated by accounting for the unique aspects of Outokumpu s stainless steel production process. The Outokumpu Mill is located in Mobile County, Alabama which is designated as attainment or unclassifiable for all criteria pollutants. Alabama s PSD regulations are codified under ADEM Air Division - Air Pollution Control Program (Division 335-3) Air Permits Authorizing Construction in Clean Air Areas [Prevention of Significant Deterioration Permitting]. As the Outokumpu Mill is a major source of at least one regulated new source review (NSR) pollutant, the facility is subject to PSD regulations. Per (9)(c), the state of Alabama requires a BACT analysis to be conducted and applied for any facility modification that causes a significant net emissions increase of a regulated NSR pollutant. Because existing PSD emission limits are being modified and due to the resulting increase in maximum potential emissions, the proposed modification will be subject to BACT analysis for NO x and SO 2 emissions from the EAF. Per AAC (2)(1), ADEM defines BACT as follows: "Best Available Control Technology (BACT)" shall mean an emissions limitation (including a visible emission standard) based on the maximum degree of reduction for each regulated NSR pollutant which would be emitted from any proposed major stationary source or major modification which the Director, on a case-by-case basis, taking into account energy, environmental, and economic impacts and other costs, determines is achievable for such source or modification through application of production processes or available methods, systems and techniques, including fuel cleaning or treatment or innovative fuel combustion techniques for control of such pollutant. In no event shall application of BACT result in emissions of any pollutant which would exceed the emissions allowed by any applicable standard under 40 CFR 60 and 61. No BACT determination may be less stringent than an applicable New Source Performance Standard (NSPS), National Emissions Standards for Hazardous Air Pollutants (NESHAP), or State Implementation Plan (SIP) limit. The USEPA Assistant Administrator for Air and Radiation issued a memorandum on December 1, 1987, which implemented certain program initiatives designed to improve the effectiveness of NSR programs within the confines of existing regulations and SIPs. This memorandum developed the topdown methodology for determining BACT which remains in use today. The topdown process requires that available control technologies first be ranked in descending order of control effectiveness. The first step is to evaluate the most stringent or top alternative. This top alternative is represented as BACT unless it Environmental Resources Management Southwest, Inc \ \841rpt.doc

11 can be determined, and the permitting authority agrees, that technical considerations or energy, environmental, or economic impacts justify that the most stringent technology is not achievable for the specified source. If it is determined that the top alternative is not achievable, then the next most stringent alternative is considered, and so on, until a BACT control option is selected. 3.1 TOP-DOWN BACT ANALYSIS METHODOLOGY As discussed in the previous section, the USEPA has recommended a top-down approach in conducting a BACT analysis. This method evaluates progressively less stringent control technologies until an achievable level of control is reached; this method is based on the potential environmental, energy, and economic impacts. The five steps of a top-down BACT analysis are outlined below. Step 1 Identify Potential Control Technologies Identify all available control technologies with practical potential for application to the emission unit and regulated pollutant under evaluation. The list of potential technologies must be comprehensive. Step 2 - Eliminate Technically Infeasible Options A determination of technical infeasibility should be clearly documented and should demonstrate, based on physical, chemical, and engineering principles, that technical difficulties would preclude the successful implementation of the control option on the specific emissions source under review. Step 3 Rank Remaining Technically Feasible Control Options Remaining control technologies shall be ranked by effectiveness and a control hierarchy shall be tabulated. The following criteria shall be considered: Control effectiveness (percent pollutant removed); Expected emission rate (tons per year); Expected emissions reduction (tons per year); Energy impacts (Btu, kilowatt hour [kwh]); Environmental impacts (impacts to other media and/or the emission of toxic and hazardous air pollutants); and, Economic impacts (total cost effectiveness and incremental cost effectiveness, in annualized cost per ton of compound reduced). Step 4 Evaluate Remaining Control Technologies A case-by-case evaluation of each effective control shall be conducted taking into account consideration of energy, environmental, and economic impacts. If the top (most effective control) option is not selected as BACT, the next most effective control option shall then be evaluated. Environmental Resources Management Southwest, Inc \ \841rpt.doc

12 Step 5 Selection of BACT The most effective, practical remaining option shall be selected as BACT. In order to develop BACT determinations, information from the following databases and listings was considered to identify emission limits and control technologies that apply to the EAF located at the Outokumpu Mill: USEPA s RBLC database; Recent PSD permit applications for stainless steel production facilities. AP-42, Volume 1, Fifth Edition 5 ; Air pollution control technology vendors; Air Pollution Control Engineering, Second Edition (Noel De Nevers, 2000); The European Commission s Control of Nitrogen Oxide Emission at the Electric Arc Furnace Reference Document 6 ; and, European Commission Integrated Pollution Prevention Control (IPPC): Best Available Techniques Reference Document on the Production of Iron and Steel CONTROL TECHNOLOGY DETERMINATION FOR THE ELECTRIC ARC FURNACE Summary Results of Control Technology Analysis A detailed control technology analysis for the EAF is presented in the following sections of this application. A summary table of the proposed emission limits for NO x and SO 2 is presented below for quick reference. 5 AP 42, Fifth Edition: 6 Control of Nitrogen Oxide Emission at the EAF. Electronic Source: (accessed on May 12, 2014). 7 IPPC Manual. Electronic Source: (accessed on May 12, 2014). Environmental Resources Management Southwest, Inc \ \841rpt.doc

13 Pollutant NO x SO 2 Current BACT Emission Limit 0.35 lb/ton 0.15 lb/ton Proposed BACT Emission Limit 1.00 lb/ton (126 lb/hr) lb/ton (47.25 lb/hr) BACT Control Technology Direct evacuation control (with watercooled ductwork); Proper equipment design, proper operation, and good engineering practices Charge substitution (low sulfur coal with target sulfur content of 0.55% by weight); Chemical additive (lime); Proper equipment design, proper operation and maintenance, and good engineering practices Proposed Monitoring / Compliance Demonstration Stack Testing (Average from 3 test runs) SO 2 CEMS (3-hour average ) BACT Determination for NO x Emissions from the Electric Arc Furnace During the EAF process, there are periods when the electric arc operates in an oxidizing gas atmosphere, wherein the nitrogen in the air in combination with extremely high plasma temperatures leads to the formation of thermal NO x gas species in the arc. In addition, high temperature post combustion occurs with air in the furnace freeboard and primary off-gas system of the EAF, accounting for additional NO x formation. 8 The amount of NO x formed from each mechanism will vary depending on several factors including the number and length of the charges for each heat, the frequency of the charges, how well the furnace is insulated, and the use of oxygen lances. The following sections discuss the results of the detailed BACT analysis utilizing the sources list under Section 3.1 of this report BACT Baseline There are no specific regulatory requirements for NO x emissions from the EAF. Thus, baseline emissions are simply the uncontrolled emissions from the furnace. 8 Measurements and Simulation of NO x Formation in the Electric Arc Furnace by Thomas Echterhof, Jacqueline Gruber, Herbert Pfeifer. RWTH Aachen University, Dept. for Industrial Furnaces and Heat Engineering, Kopernikusstr. 10, D Aachen, Germany. Environmental Resources Management Southwest, Inc \ \841rpt.doc

14 Step 1 Identify Potential Control Technologies There are a number of technologies potentially applicable to conventional boilers and heaters that have demonstrated their ability to reduce both fuel nitrogen conversion to NO x and thermal formation of NO x from atmospheric nitrogen. While many of these are not applicable to the EAF at the Outokumpu Mill which relies on an electric arc rather than on fuel combustion based burners to provide the heat needed for melting steel scrap, they are nonetheless presented in this discussion in order to provide a thorough and comprehensive control technology analysis (Step 2 of the BACT analysis identifies when a specific control technology has not been effectively implemented on EAFs and is considered to be technically infeasible). Front-end NO x emissions are typically controlled by over-fire air, low excess air, and burners out of service, by controlling the amount of air allowed in the combustion source (also known as combustion controls). Exhaust controls may include Selective Catalytic Reduction (SCR), Selective Non-Catalytic Reduction (SNCR), or Non-Selective Catalytic Reduction (NSCR). Based on information obtained from the USEPA s RBLC database, recently submitted permit applications, and air pollution control guidance documents, a list of potential NO x controls for the EAF was developed. The potential control options followed by a brief description of each control alternative are outlined below: 1) Over-fire Air; 2) Low Excess Air; 3) Burners Out Of Service; 4) Flue Gas Recirculation; 5) Oxy-fuel Systems; 6) Direct Evacuation Control; 7) Selective Catalytic Reduction; 8) Selective Non-Catalytic Reduction; 9) Three-Way Catalysts (Non-Selective Catalytic Reduction); and, 10) Proper Equipment Design, Proper Operation, and Good Engineering Practices. Over-fire Air Over-fire air works by diverting a portion of the total combustion air away from the primary combustion zone. Over-fire air lowers the air-to-fuel ratio at the burners which in-turn lowers the nitrogen in the primary flame zone. This reduces the conversion of fuel nitrogen to NO x as well as reduces the formation of thermal NO x. 9 9 STEAG Energy Services LLC Overfire Air Systems. Electronic Source: (accessed on ). Environmental Resources Management Southwest, Inc \ \841rpt.doc

15 Low Excess Air Low excess air is a combustion modification technique in which NO x formation is inhibited by reducing the excess air to less than normal ratios. 10 This control method reduces the local flame concentration of oxygen which decreases the formation of thermal and fuel NO x. The type of fuel fired, uniformity of the airto-fuel ratio, air and fuel control lags during load swings, and other combustion control features such as staging of fuel can affect this control process. 11 Burners Out Of Service This practice, if properly implemented, can result in off-stoichiometric combustion (OSC), which can reduce NO x formation by causing initial combustion in a fuel-rich zone and completing combustion in a lowertemperature, fuel-lean zone. One method of causing this OSC is to remove fuel from selective burners or rows of burners and allowing them to admit only air. This is referred to as Burners Out Of Service (BOOS). Combustion in the fuelrich burners occurs with low excess oxygen. Combustion at the air-only burners occurs at a reduced temperature. 12 Flue Gas Recirculation Flue Gas Recirculation (FGR) involves recirculating a small portion of the exhaust to the combustion air stream. Recirculation of combustion gas increases the concentration of inert gases into the primary combustion zone, which lowers the flame temperature and reduces formation of thermal NO x. Recirculation also reduces the oxygen concentration in the combustion zone slightly, again, reducing thermal NO x production. Oxy-fuel Systems Oxy-fuel systems reduce NO x emissions by replacing the ambient air with oxygen-enriched air via an injection system. As oxygen-enriched air contains significantly lower concentrations of nitrogen as compared to ambient air, oxyfuel systems can reduce NO x emissions by at least 85% Zeldovich, J. The Oxidation of Nitrogen in Combustion and Explosions. Acta. Physiochem. 21(4) U.S. Environmental Protection Agency. Control Techniques for Nitrogen Oxides Emissions from Stationary Sources- Revised Second Edition. EPA-450/ January Electronic Source: 001_stationary_sources.pdf (accessed on May 12, 2014) 12 Alternative Control Techniques Document - NOx Emissions from Iron and Steel Mills. Electronic Source: (accessed on ). 13 Industrial Heating : Regenerative Burners or Oxy-Fuel Burners for Your Furnace Upgrade. Electronic Source: (accessed on ). Environmental Resources Management Southwest, Inc \ \841rpt.doc

16 Oxy-fuel systems for EAFs are typically installed in conjunction with auxiliary natural gas fired burners and the combined system is often referred to as oxyfuel burners. By controlling the furnace draft, an oxy-fuel system facilitates a higher furnace temperature which in turn increases the radiant heat flux; a higher heat flux can reduce the melt time which can decrease energy consumption by 60%. Oxy-fuel systems also produce approximately 70% less flue gas than a traditional air based EAF which allows for longer flue gas residence times and less heat loss in the flue gas. 14 It should be noted that NO x emission reductions are predominantly yielded by the injection of oxygenenriched air, and the installation of natural gas fired burners is not the essential mechanism for achieving NO x emission reductions. Despite this, oxy-fuel systems are typically made available for EAFs in conjunction with natural gas fired burners and referred to as oxy-fuel burners. Direct Evacuation Control A Direct Evacuation Control (DEC) system helps maintain an optimum pressure within the EAF and then ducts the emissions to a control device such as a baghouse. Specifically, the DEC system manages the furnace draft in order to minimize the variation of pressure during the melt cycle. 15 The low pressure in the furnace facilitates off-stoichiometric combustion and can lower the formation of NO x. The selected design pressure maintained by a DEC system will vary depending on the specific process the DEC is controlling. If the DEC supplies a large negative pressure, increased air will be introduced into the EAF resulting in increased NO x formation (the high temperatures within the EAF facilitate a reaction between the nitrogen and oxygen in the entrained air). Conversely, large positive pressures lead to excessive opacity and increased CO formation. Thus, the pressure maintained by a DEC system must be well balanced in order to effectively reduce both NO x and CO emissions. 16 Selective Catalytic Reduction Selective Catalytic Reduction (SCR) is an off-gas treatment process that removes NO x from the exhaust gas stream by injecting ammonia (NH 3 ) into the exhaust gas upstream of a catalyst bed. On the catalyst surface, NH 3, NO x, and oxygen react to form diatomic nitrogen (N 2 ) and water. NO x reduction using SCR 14 Air Products: Advantages of Oxy-fuel Burner Systems for Aluminum Recycling Systems by Ludger Gluns and Siegfried Schemberg. Electronic Source: (accessed on ). 15 Illinois Environmental Protection Agency Bureau of Air, Permit Section Springfield, Illinois: Project Summary for a Construction Permit Application from A. Finkl & Sons Co. for a Specialty Steel and Forgings Facility in Chicago, Illinois. Electronic Source: (accessed on ). 16 Project Summary for a Construction Permit Application (EPA). Electronic Source: Environmental Resources Management Southwest, Inc \ \841rpt.doc

17 technology is only effective within a given temperature range. The optimum temperature range for a specific source will depend on the type of catalyst used and the composition of the flue gas. The required SCR temperature ranges from 480 F 800 F. If operated at the optimum temperature, SCR can achieve a NO x reduction efficiency of 70-90%. 17 Selective Non-Catalytic Reduction Selective Non-Catalytic Reduction (SNCR) controls NO x emissions by injecting ammonia or a urea solution into the flue gas stream, reducing NO x to molecular N 2 and water. Depending on the type of reactant, the optimum temperature range for operating a SNCR system is 1600 F 2300 F. 18 The capability of SNCR to reduce NO x emissions depends on the ability to uniformly mix the ammonia and flue gas as well as on the ability to maintain the appropriate temperature. With the utilization of SNCR technology, it is expected that NO x emissions will be reduced by %. 19 Three-Way Catalysts (Non-Selective Catalytic Reduction) Three-way catalysts are sometimes referred to as non-selective catalytic reduction, or NSCR. In NSCR systems, a catalyst reduces NO x in the flue gas stream by reacting with the organic compounds to form N 2 and water as shown by the equation below. Most commonly, the catalyst used in a NSCR system is composed of platinum or rhodium. 20 Nitrogen Oxides (NO x ) Nitrogen (N 2 ) + Water (H 2 O) NSCR systems have been proven applicable to rich burn engines that are capable of a simultaneous reduction of NO x, CO, and unburned hydrocarbons in a single catalyst due to the stoichiometric nature of the combustion process. In order to successfully remove NO x, the exhaust stream in these engines must contain very little oxygen; typical exhaust oxygen levels are less than 0.5% upstream of the 17 EPA Air Pollution Control Technology Fact Sheet for Selective Catalytic Reduction (SCR). Electronic Source: (accessed on ). 18 EPA Alternative Control Techniques Document - NOx Emissions from Iron and Steel, Mills Research Triangle Park, North Carolina 27711, September Electronic Source: (accessed on ). 19 Prevention of Significant Deterioration of Air Quality Preliminary Review, November 2010 for Osceola Steel Company Cook County, GA. pdf (accessed on ). 20 EPA: Summary of NOx Control Technologies and their Availability and Extent of Application, February Environmental Resources Management Southwest, Inc \ \841rpt.doc

18 catalyst. 21 While this technology has been demonstrated on rich burn engines, NSCR has not been proven as an effective emissions control on an EAF. Proper Equipment Design, Proper Operation, and Good Engineering Practices Implementing proper equipment design, proper operation, and good engineering practices on an EAF can maximize combustion efficiency, thus, reducing NO x emissions. These practices include, but are not limited to, the following: 22 Minimizing Air Filtration; Maintaining Furnace Draft During Melting and Refining Operations; Maintaining Combustion Equipment According to the Manufacturer s or Builder s Instructions; Continuously Monitoring and Adjusting the Fuel-to-Air Combustion Ratio of the Combustion Equipment per the Manufacturer s Specifications; Minimizing the Duration of the Roof being Open for Charging; Avoiding or Minimizing Operational Delays; Sealing the Furnace as Much as Possible; and, Operating at Optimal Pressure Step 2 Eliminate Technically Infeasible Options Over-fire Air The over-fire air control technique fires burners more fuel-rich as compared to normal firing. Additional air required to facilitate combustion is admitted into the combustion source through over-fire air ports. This technique has been proven on oil, coal, and natural gas-fired boilers, and can reduce NO x emissions by 24 59%. Based on USEPA literature, over-fire technology has only been proven on large utility boilers. The USEPA also has stated that this control has never been easily or effectively implemented on iron and steel process furnaces L. Sasadeusz, G. Arney. Operating Catalytic Emission Reduction Systems. Southern California Gas Company. Houston, Texas. January 30-31, Electronic Source: (accessed on ). 22 TCEQ: Current BACT Guidelines for Iron and Steel Industry. Electronic Source: w/bact/bact_ironsteel.pdf (accessed on ). 23 Letter and Attachments from Anderson, D.M., Bethlehem Steel Corporation to Jordan, B.C., USEPA/OAQPS. July 13, Response to Section 114 Letter on Iron and Steel Mills. Electronic Source: (accessed on ). Environmental Resources Management Southwest, Inc \ \841rpt.doc

19 Overall, the over-fire air technique is fundamentally inconsistent with the design criterion for an EAF which relies on an electric arc rather than on fuel combustion based burners to provide the heat needed for melting steel scrap. Thus, the use of over-fire air will be eliminated as a potential BACT for the EAF because of technical infeasibility. Low Excess Air The low excess air control technique reduces NO x formation by reducing the peak flame temperature and by reducing the amount of excess oxygen in the combustion source. By utilizing this control on a furnace, it is expected that NO x emissions will be reduced by only 13%. Based on USEPA literature outlined in Alternative Control Techniques Document: NO x Emissions from Iron and Steel Mills (September 1994), low excess air techniques have been installed exclusively on reheat furnaces in iron and steel manufacturing facilities. Low excess air technique is generally inconsistent with the design criterion for an EAF which relies on an electric arc rather than on fuel combustion based burners to provide the heat needed for melting steel scrap. The EPA document also states that low excess air techniques have the potential to increase CO emissions. 24 Based on the factors stated above, this control option will be eliminated as a potential BACT for the EAF because of technical infeasibility. Burners Out Of Service BOOS reduce emissions by taking a selected group of burners and making them inactive during the firing of a combustion source, thus reducing the furnace load. Overall, the BOOS process is fundamentally inconsistent with the design criterion for an EAF which relies on an electric arc rather than on fuel combustion based burners to provide the heat needed for melting steel scrap. As such, removing fuel from burners or a row of burners is not relevant to an EAF. Furthermore, based on USEPA literature outlined in Alternative Control Techniques Document: NO x Emissions from Iron and Steel Mills (September 1994), BOOS requires very accurate monitoring of the combustion process and flue gas or the efficiency and safety of the furnace will be compromised. 25 Based on the factors stated above, this control option will be eliminated as a potential BACT for the EAF because of technical infeasibility. Flue Gas Recirculation FGR recycles a portion of the cooled, outlet flue gas back into the primary combustion zone. By stimulating circulation, the concentration of oxygen available for combustion is decreased, and thermal NO x formation is lowered. The primary limitation of utilizing FGR on an EAF is the formation of cold spots which lowers the efficiency of the furnace. Furthermore, recirculating flue gas 24 USEPA Alternative Control Techniques Document: NOx Emissions from Iron and Steel Mills, Research Triangle Park, North Carolina 27711, September Electronic Source: (accessed on ). 25 USEPA Alternative Control Techniques Document - NOx Emissions from Iron and Steel, Mills Research Triangle Park, North Carolina 27711, September Environmental Resources Management Southwest, Inc \ \841rpt.doc

20 into the EAF has potential for contamination of steel being melted in the EAF by introducing undesirable particulate matter into the EAF. As additional natural gas fired burners would need to be installed on the EAF to account for the loss of even distribution of heat and due to potential risk of contamination of steel, FGR technology to reduce NO x emissions is not considered technically feasible. 26 Furthermore, based on USEPA literature outlined in Alternative Control Techniques Document: NO x Emissions from Iron and Steel Mills (September 1994), there is no evidence that FGR has been used to control NOx emissions from iron and steel process facilities. 23 Based on the factors stated above, this control option will be eliminated as a potential BACT for the EAF because of technical infeasibility. Oxy-fuel Systems Oxy-fuel systems reduce NO x by replacing the ambient air used to support natural gas combustion with oxygen enriched air. Thus, the use of oxy-fuel systems would increase the concentration of oxygen in the EAF at the Outokumpu Mill. In order to produce high-grade stainless steel, carbon must be oxidized from molten steel without also oxidizing large proportions of chromium. In order to control oxidation, maintaining the appropriate concentration of oxygen in the EAF is vital. The Outokumpu Mill uses a duplex process in which a high carbon melt is completed in the EAF followed by decarburization in the argon oxygen decarburization (AOD) vessel. In the AOD, a carefully designed oxygen-to-argon ratio is maintained in order to ensure that the selective oxidation of carbon is the dominant reaction and that the oxidation of chromium is minimized. 27 Operating the stainless steel EAF at the Outokumpu Mill in an oxygen enriched environment, as an oxy-fuel system would do, would cause the oxidation of chromium during the refining stage after the carbon has been oxidized. Ultimately, by injecting large amounts of oxygen into the EAF as an oxy-fuel system would do, the ability of the EAF to produce high-grade chrome stainless steel is compromised. The removal of carbon (decarburization) without detrimentally impacting the chromium is better accomplished in the duplex process via the use of an AOD following the EAF as done at the Outokumpu Mill. In order to confidently affirm that the use of oxy-fuel system is not technically feasible for a stainless steel EAF, the RBLC database, oxy-fuel vendors, and scholarly articles on oxygen enrichment technologies were researched. It should be noted that, Oxy-fuel systems for EAFs are typically installed in conjunction with auxiliary natural gas fired burners and the combined system is often referred to as oxy-fuel burners. NO x emission reductions in EAFs (other than 26 RACT for Nucor Steel Plymouth, Utah. Electronic Source: ocs/tsd/chapter5/5c/j-vulcraft%20tsd% pdf (accessed on ). 27 Chapter 12, Steelmaking and Refining Volume issued by The AISE Steel Foundation, 1998, Pittsburgh, PA. Environmental Resources Management Southwest, Inc \ \841rpt.doc

21 stainless steel EAFs) are predominantly yielded by the injection of oxygenenriched air, and the installation of natural gas fired burners is not an essential mechanism for achieving NO x emission reductions. Despite this, oxy-fuel systems are typically made available for EAFs in conjunction with natural gas fired burners and referred to as oxy-fuel burners. The RBLC database was researched in order to determine if any stainless steel facilities had installed oxy-fuel burners on an EAF. As shown in Table 3-1, the RBLC database returned several records for an EAF at a stainless facility; these records have been highlighted in light blue for easy recognition. Out of the stainless steel facilities, the Nucor Steel (Nucor) facility located in Indiana indicated the use of oxy-fuel burners for NO x control. Based on industry knowledge, Outokumpu assumes the Nucor facility does not utilize the oxy-fuel burners during stainless production in order to minimize chromium oxidation. Another stainless steel facility listed as having an EAF is the North American Stainless (NAS) steel mill located in Kentucky. The only NO x control technology utilized at this facility is a Direct Evacuation Control (DEC) system which has a NO x design emission factor of 1.32 lb/ton. Engineers from Stantec Global Technologies, Praxair, and the University of Toronto conducted plant-based research on the effectiveness of oxygen injection systems on the reduction of NO x in the steelmaking industry. The publication, NO x Emissions from EAF Steelmaking Assessment and Abatement, compared NO x emissions from an EAF before and after the installation of CoJet TM. 28 CoJet TM gas injection technology is a patented oxygen injection system typically installed to improve the efficiency of the furnace. Measurements were made before and after the installation of the CoJet TM system for comparison. As CoJet TM systems reduce NO x formation by increasing the level of oxygen and reducing the level of nitrogen available in the EAF, the study is relevant to the application of oxy-fuel system on the Calvert Mill EAF. The results of the study show that NO x levels were essentially unchanged in spite of doubling the oxygen consumption on an EAF. The study goes on to suggest that the best strategy to minimize NO x emissions is to control the furnace atmosphere through controlling the air ingress, monitoring the furnace off-gas, and controlling the fuel injection system (if installed) to maintain desired furnace conditions. 29 Based on the above, Outokumpu believes that operation of an oxy-fuel system (with or without natural gas fired burners) to provide an oxygen enriched environment in the EAF is not technically feasible as it will undermine Outokumpu s ability to produce high-grade chrome stainless steel. Therefore, 28 NO x Emissions from EAF Steelmaking Assessment and Abatement. Electronic Source: (accessed on May 12, 2014). 29 Goodfellow, Howard D., et al. "NOx Emissions from EAF Steelmaking Assessment and Abatement. EFC (2000): Print. Electronic Source: (accessed on ) Environmental Resources Management Southwest, Inc \ \841rpt.doc

22 oxy-fuel burner technology will be eliminated as a potential control of NO x emissions from the EAF at the Outokumpu Mill. Selective Catalytic Reduction NO x reduction using SCR technology is only effective under certain conditions. First, in order for SCR to reduce NO x emissions, the exhaust gas must be within the optimum temperature range from 480 F 800 F. 30 Second, the air flow, temperature, and NO x concentrations of the source must be relatively stable. Based on EPA guidance, the RBLC database, and current air permit applications for stainless steel facilities, SCR technology has been determined to be technically infeasible for the EAF at the Outokumpu Mill based upon the following: The required temperature for SCR ranges from 480 F 800 F. However, the average operating temperature for an EAF is outside the optimum SCR temperature range. o o Scenario One: Install SCR after the Furnace. Temperatures from an EAF during the steelmaking bath process can reach 3,000 F. 31 Thus, if a SCR were installed at the outlet of the furnace, temperatures would be too hot and above the required SCR range. Furthermore, at the outlet of the furnace high particulate matter emissions can poison the SCR catalyst. Metals in the form of particulates can damage the SCR catalyst. Particulates can form a film over the surface of the catalyst, preventing contact between the bed surface and the flue gas. This is known as catalyst poisoning. In order for a SCR to effectively control NO x emissions from the exhaust stream, that stream would need to be cooled from temperatures in excess of 3000 F to temperatures less than 800 F. A cooling system such as air quenching, cooling ducts, or spray coolers would have to be designed and installed to cool the exhaust gas stream to remain within the required SCR temperature range. 32 However, the potential for catalyst poisoning due to high particulate matter emissions will remain. Scenario Two: Install SCR after the Baghouse. If a SCR were installed after the baghouse which controls particulate matter emissions, the EAF exhaust gas would be too cool. Based on stack 30 USEPA Air Pollution Control Technology Fact Sheet for Selective Catalytic Reduction (SCR). Electronic Source: (accessed on ). 31 Temperature Homogenization in an Electric Arc Furnace Steelmaking Bath in an Electric Arc Furnace Steelmaking Bath. Electronic Source: (accessed on ). 32 Report on Best Available Techniques (BAT) in the Electric Steelmaking Industry. Electronic Source: 8.pdf (accessed on ). Environmental Resources Management Southwest, Inc \ \841rpt.doc

23 testing done on the emission point associated with the EAF (LO- 1), the exhaust gas temperature of the stack is 114 F, which is below the optimal SCR temperature range. In order for a SCR system to effectively operate after the baghouse, the gas stream would have to be re-heated. In order for SCR to effectively reduce NO x, the air flow, temperature, and concentration of NO x in the EAF must be relatively stable. The exhaust gas stream of the EAF at Outokumpu s Mill, however, is variable. As the EAF undergoes different tap-to-tap cycles, air flow, temperature, and NO x concentrations are dependent on which type of cycle the EAF is undergoing. For example, the conditions of the initial charging cycle can be exceedingly different than those that occur during the melting or refining cycles. Based on the EPA factsheet for SCR, the technology is only proven in the United States on boilers, heaters, and gas turbines. 33 The RBLC database was researched to identify potential controls for the EAF. No records of a SCR installed to control NO x on an EAF were found. Selective Non-Catalytic Reduction In order for SNCR to effectively remove NO x from the exhaust gas stream, several conditions must be in place. First, very stable gas conditions are required. Second, the combustion source must operate at a temperature range of 1600 F 2300 F. 34 Third, the capability of SNCR to reduce NO x emissions depends on the ability to uniformly mix the ammonia and flue gas. At ideal conditions, it is anticipated that NO x emissions will be reduced by %. 35 Based on USEPA guidance, the RBLC database, and current air permit applications for stainless steel facilities, SNCR technology has been determined to be technically infeasible for the EAF at the Outokumpu Mill based upon the following: The required temperature range for SNCR ranges from 1600 F 2300 F. o Scenario One: Install SNCR after the Furnace. Temperatures from an EAF during the steelmaking bath process can reach 33 USEPA Air Pollution Control Technology Fact Sheet for Selective Catalytic Reduction (SCR). Electronic Source: (accessed on ). 34 USEPA Alternative Control Techniques Document - NOx Emissions from Iron and Steel, Mills Research Triangle Park, North Carolina 27711, September Prevention of Significant Deterioration of Air Quality Preliminary Review, November 2010 for Osceola Steel Company Cook County, GA. Electronic Source: pdf (accessed on ). Environmental Resources Management Southwest, Inc \ \841rpt.doc

24 o 3,000 F. 36 Thus, if a SNCR were installed at the outlet of the furnace, temperatures would be too hot. USEPA has documented that at temperatures above the optimal SNCR temperature range, additional NO x is formed. In order for a SNCR to effectively remove NO x from the exhaust stream, that stream would need to be cooled from temperatures in excess of 3000 F to temperatures less than 2300 F. In order for the SNCR system to work, a cooling system such as air quenching, cooling ducts, or spray coolers would have to be designed and installed to cool the exhaust gas stream to remain within the required SNCR temperature range. 37 Scenario Two: Install SNCR after the Furnace Baghouse. If a SNCR were installed after the baghouse which controls particulate matter emissions, the EAF exhaust gas stream would be too cool. Based on stack testing done on the emission point associated with the EAF (LO-1), the exhaust gas temperature of the stack is 114 F, which is below the optimal SNCR temperature range. In order to effectively operate the SNCR system after the baghouse, the gas stream would have to be re-heated. In order for SNCR to effectively reduce NO x, the concentration of NO x in the exhaust gas must be relatively stable. The exhaust gas stream of the EAF at Outokumpu s Mill, however, is variable as previously discussed under the SCR section. The variance in the exhaust gas streams makes it difficult to ensure the SNCR reagent is properly distributed. NO x reduction is dependent upon uniform mixing of NH 3 and the flue gas. Thus, the variability of the EAF exhaust stream jeopardizes the ability of SNCR to reduce NO x emissions. Based on the USEPA factsheet for SNCR, the technology is only proven in the United States on boilers, thermal incinerators, municipal and hazardous solid waste combustion units, cement kilns, process heaters, and glass furnaces not electric arc furnaces. 38 The RBLC database was researched to identify potential controls for the EAF. No records of a SNCR installed to control NO x on an EAF were found. As a result, SNCR is eliminated as technically infeasible. 36 Temperature Homogenization in an Electric Arc Furnace Steelmaking Bath in an Electric Arc Furnace Steelmaking Bath. Electronic Source: (accessed on ). 37 Report on Best Available Techniques (BAT) in the Electric Steelmaking Industry. Electronic Source: 8.pdf (accessed on ). 38 USEPA Air Pollution Control Technology Fact Sheet for Selective Non-Catalytic Reduction (SNCR). Electronic Source: (accessed on ). Environmental Resources Management Southwest, Inc \ \841rpt.doc

25 Three-Way Catalysts (Non-Selective Catalytic Reduction) Based on research conducted during this BACT analysis, it was concluded that an NSCR system is technically infeasible due to a lack of established precedence for the use of the control technology on similar applications outside of experimentation. Specifically, NSCR technologies have been demonstrated almost exclusively on rich burn engines. 39 Technologies that have been permitted as BACT were obtained from the USEPA s RBLC database; an NSCR was not a listed NO x control for any of the records found within the database. Additionally, all NSCR systems are equipped with a Lambda Sensor, which is similar to the oxygen sensor in an automobile. This sensor is responsible for ensuring the stoichiometric balance of nitrogen and carbon that is vital to the emission reduction process. This sensor requires cumbersome, frequent tuning and maintenance. If the sensor is not working properly, the NSCR becomes very ineffective at reducing emissions. 40 Due the above factors and lack of proven NO x control on an EAF, this technology is considered to be technically infeasible Step 3 - Rank Remaining Technically Feasible Control Options The remaining control technologies are ranked based upon records identified in the RBLC database; these records are presented in Table 3-1 for reference. The data presented in Table 3-1 includes EAFs operating at both carbon and stainless steel production facilities. Facilities producing stainless steel have been highlighted in a light blue color for ease of identifying. While control technologies and corresponding emission limits for carbon steel EAFs are presented in Table 3-1, emission factors for carbon steel EAFs are not representative of a stainless steel EAF. Similarly, the control options appropriate for a carbon steel EAF are not necessarily appropriate for a stainless steel EAF. For example, as discussed in Section 2.1, operating a stainless steel EAF in an oxygen enriched environment, as an oxy-fuel system would do, would cause the oxidation of chromium during the refining stage after the carbon has been oxidized; injecting large amounts of oxygen into a stainless EAF compromises the ability of that EAF to produce high-grade chrome stainless steel. This demonstrates that while an oxy-fuel burner system would work for a carbon steel EAF, the same technology is not feasible for a stainless steel EAF. Based upon records found within the RBLC database, the use of a DEC system to maintain the optimal design pressure within the EAF has been shown to effectively reduce NO x emissions. Also, proper equipment design, operation, and good engineering practices are generally feasible to control NO x. Refer to Table 3-1 for a full set of RBLC records. 39 MIRATECH: Oxidation Catalyst NSCR. Electronic Source: (accessed on ). 40 L. Sasadeusz, G. Arney. Operating Catalytic Emission Reduction Systems. Southern California Gas Company. Houston, Texas. January 30-31, Electronic Source: (accessed on ). Environmental Resources Management Southwest, Inc \ \841rpt.doc

26 Table RACT/BACT/LAER/Clearinghouse Database Records for Nitrogen Oxides (NO x ) Emissions and Control Technologies for an Electric Arc Furnace (1) ERM ADDED ERM ADDED Stainless Steel Facility? RBLC ID Facility Name Facility State Permit Number Permit Issuance Date Process Name Control Method Emission Limit (lb/ton) No PA-0251 ELLWOOD NATIONAL STEEL PA 62-32B 8/18/2006 ELECTRIC ARC FURNACE NONE PROVIDED 0.14 No CO-0054 CF & I STEEL L.P. DBA ROCKY MOUNTAIN STEEL MILLS CO 02PB0492 6/21/2004 ELECTRIC ARC FURNACE (EAF) GOOD COMBUSTION PRACTICES No CO-0061 CF&I STEEL L.P. DBA ROCKY MOUNTAIN STEEL MILLS CO 02PB0492 6/21/2004 EAF #5 OPERATING PRACTICES 0.15 No IA-0087 GERDAU AMERISTEEL WILTON IA /29/2007 ELECTRIC ARC FURNACE OXY-FUEL BURNERS AND DIRECT EVACUATION CONTROL (DEC) 0.19 No OH-0339 HARRISON STEEL PLANT OH P /29/2010 Electric Arc Furnace (2) NONE PROVIDED 0.20 No OH-0342 FAIRCREST STEEL OH P /29/2010 Electric Arc Furnace NONE PROVIDED 0.20 No MI-0404 Gerdau Macsteel, Inc. MI /4/2013 EUEAF, EULMF, AND TWO EUVTD REAL TIME PROCESS OPTIMIZATION (COMBUSTION CONTROLS) AND USE OF OXY-FUEL BURNERS 0.20 No PA-0214 J & L SPECIALTY STEEL, INC. PA PA A 4/2/2003 ELECTRIC ARC FURNACE, (2) LOW NOX BURNERS 0.20 No NY-0094 NUCOR AUBURN STEEL NY / /22/2004 EAF NONE PROVIDED 0.25 No NY-0099 NUCOR STEEL AUBURN INC NY /5/2006 ELECTRIC ARC FURNACE (STEEL RECOVERY) NONE PROVIDED 0.27 No MN-0070 MINNESOTA STEEL INDUSTRIES, LLC MN /7/2007 ELECTRIC ARC FURNACE/MELT SHOP NONE PROVIDED 0.30 No OK-0128 MID AMERICAN STEEL ROLLING MILL OK C(M-1) PSD 9/8/2008 Electric Arc Furnaces NONE PROVIDED 0.30 No OH-0315 NEW STEEL INTERNATIONAL, INC., HAVERHILL OH /6/2008 ELECTRIC ARC FURNACE (2) LOW NOX OXY-FUEL BURNERS 0.31

27 Table RACT/BACT/LAER/Clearinghouse Database Records for Nitrogen Oxides (NO x ) Emissions and Control Technologies for an Electric Arc Furnace (1) ERM ADDED ERM ADDED Stainless Steel Facility? RBLC ID Facility Name Facility State Permit Number Permit Issuance Date Process Name Control Method Emission Limit (lb/ton) No OH-0276 CHARTER STEEL OH /10/2004 ELECTRIC ARC FURNACE DIRECT EVACUATION CONTROL AND LOW NOX OXY-FUEL BURNERS 0.33 No DE-0019 CITISTEEL USA, INC. DE AQM-003/ /7/2004 ELECTRIC ARC FURNACE None 0.34 No AR-0077 BLUEWATER PROJECT AR 2062-AOP-R0 7/22/2004 ELECTRIC ARC FURNACE (EAF) LOW NOX BURNERS 0.35 No AL-0202 CORUS TUSCALOOSA AL X005,X008 6/3/2003 ELECTRIC ARC FURNACE NONE PROVIDED 0.35 No AL-0218 NUCOR STEEL TUSCALOOSA, INC. AL /6/2006 ELECTRIC ARC FURNACE NONE PROVIDED 0.35 No GA-0142 OSCEOLA STEEL CO. GA P /29/2010 Electric Arc Furnace Low NOx Burners with FGR Technology and Good Combustion/Operating practices Yes (2) IN-0108 NUCOR STEEL IN /21/2003 EAF, AOD VESSELS, DESULFURIZATION, & OTHER PROCESS NATURAL GAS FIRED OXY FUEL BURNERS, COMPLIANCE METHOD: NOX CONTINUOUS EMISSIONS MONITOR Yes (3) PA-0284 AK Steel PA /30/2013 ELECTRIC ARC FURNACE 5 None Specified 0.35 Yes (4) AL-0230 ThyssenKrupp AL X001 THRU X026 8/17/2007 ELECTRIC ARC FURNACE LOW NOX OXYFUEL BURNERS 0.35 No NC-0112 Nucor NC 08680T09 11/23/2004 ELECTRIC ARC FURNACE, LADLE METALLURGY STATION, AND CONTINUOUS SLAB CASTER None Specified 0.36 No AR-0096 NUCOR YAMATO STEEL AR 883-AOP-R8 1/31/2008 ELECTRIC ARC FURNACE LOW NOX BURNERS 0.38 No OH-0316 V & M STAR OH P /23/2008 ELECTRIC ARC FURNACE NONE PROVIDED 0.40 No SC-0127 NUCOR STEEL CORPORATION (DARLINGTON PLANT) SC DE 12/29/2006 ELECTRIC ARC FURNACE OXY-FUEL BURNERS FOR SCRAP METAL PREHEATING, A SCRAP METAL MANAGEMENT PLAN, AND LIMITING THE TOTAL ANNUAL PRODUCTION OF RESULFURIZED STEEL TO 105,200 BILLET TONS. 0.41

28 Table RACT/BACT/LAER/Clearinghouse Database Records for Nitrogen Oxides (NO x ) Emissions and Control Technologies for an Electric Arc Furnace (1) ERM ADDED ERM ADDED Stainless Steel Facility? RBLC ID Facility Name Facility State Permit Number Permit Issuance Date Process Name Control Method Emission Limit (lb/ton) Yes (3) PA-0284 AK Steel PA /30/2013 ELECTRIC ARC FURNACE 2 None Specified 0.42 No AL-0231 NUCOR AL /12/2007 TWO (2) ELECTRIC ARC FURNACES AND THREE (3) LADLE METALLURGY FURNACES WITH TWO (2) MELTSHOP BAGHOUSES None Specified 0.42 No OH-0341 NUCOR STEEL MARION, INC. OH P /23/ electric arc furnace, including charging, melting, tapping, slag skimming; continuous casting operations, including torch cutting; 4 natural gas filed ladle preheaters; and 2 natural gas fired tundish preheaters. None Specified 0.43 No *OH-0350 REPUBLIC STEEL OH P /18/2012 Electric Arc Furnace NONE PROVIDED 0.50 No MI-0376 MACSTEEL DIVISION MI G 12/8/ ELECTRIC ARC FURNACES NONE PROVIDED 0.53 No OH-0292 WHEELING PITTSBURGH STEEL CORPO OH /6/2005 ELECTRIC ARC FURNACE NONE PROVIDED 0.54 No OH-0285 NORTH STAR BHP STEEL, LTD OH /20/2005 ELECTRIC ARC FURNACE WITH TWO LADLE MELT FURNACES DIRECT EVACUATION CONTROL SYSTEM WITH AIR GAP, AND COOLED POST COMBUSTION CHAMBER WITH BURNERS 0.57 No AR-0078 NUCOR STEEL, ARKANSAS AR 1139-AOP-R5 6/9/2003 EAF NATURAL GAS FIRED OXY-FUEL BURNERS 0.58 No TX-0399 NUCOR JEWETT PLANT TX PSD /5/2003 EAF, LMF, CASTER MELTSHOP NONE PROVIDED 0.90 Yes KY-0094 NORTH AMERICAN STAINLESS KY V /1/2003 EAF THE CONTROL EQUIPMENT IS A DIRECT EVACUATION CONTROL 1.32 Yes OH-0331 AK Steel OH /11/2010 Melt Shop, 2 EAFs, 1 LMF, 2 Ladle Pre-heaters NONE PROVIDED 1.43 Notes: (1) Records from the RBLC were obtained on July 30, The following categories in the RBLC were searched: General - Electric Arc Furnace; Ferroalloy Production; Steel Production EAF. (2) The Nucor Steel facility located in Montgomery County, Indiana is permitted to produce both stainless steel and carbon steel. (3) The AK Steel Butler Mill produces both low carbon and stainless steel and is, therefore, not a purely stainless steel mill. The emission limits for EAF#2 and #5 listed in the RBLC were in units of lb/hr. The emisison limits in units of lb/ton were obtained from AK Butler's Permit (Permit Number issued on (4) ThyssenKrupp previously owned the facility now operated and owned by Outokumpu. The NOx limit listed in the table above is the current BACT/PSD limit discussed within this permit modification application.

29 The remaining technologies have been ranked based on emission limits obtained from the USEPA s RBLC database. Refer to Table 3-1 for emission limits and corresponding controls. 1. Direct Evacuation Control; and, 2. Proper Equipment Design, Proper Operation, and Good Engineering Practices Step 4 - Evaluating Remaining Control Technologies To evaluate the effectiveness of the remaining control technologies, emission factors from the RBLC database along with permitted NO x emission factors from stainless steel facilities were compared. North American Stainless - Kentucky As shown in Table 3-1, the NAS facility located in Kentucky uses a DEC system to control NO x and has a permitted NO x emission factor of 1.32 lb/ton. NAS s permit (Permit Number V ) was reviewed to evaluate the NO x control system in place at the Kentucky mill. Emissions at NAS are controlled by a DEC consisting of ductwork which draws the emissions from the furnace during melting to a baghouse. Good engineering practice is also an established NO x control for the NAS EAF. Please note, in May of 2005 the NAS facility submitted a permit modification application that proposed a NO x emission factor on the EAF of 1.00 lb/ton. It could not be determined if a modified permit was issued. Nucor Steel Indiana Also shown in Table 3-1, the Nucor Steel facility located in Indiana uses oxy-fuel burners and a NO x CEMS to monitor and control NO x emissions from the facility s EAF. The control methods and emission rate from this facility, however, will not be considered as BACT for the EAF at the Outokumpu Mill because the Nucor mill primarily produces carbon steel as noted in Table 3-1. Carpenter Tech Pennsylvania Another stainless steel facility researched was the Carpenter Tech mill located in Reading, Pennsylvania. The Title V Air Permit (Permit Number ) indicated that this facility operates a total of six stainless steel EAFs. Five of these furnaces have a permitted NO x emission limit of 2.16 lb/ton. The sixth EAF has a NO x emission limit of 3.15 lb/ton. All six furnaces are controlled by Environmental Resources Management Southwest, Inc \ \841rpt.doc

30 good combustion practices coupled with the exclusive combustion of clean natural gas. 41 AK Steel Ohio The control technologies in place at the AK Steel Mill located in Mansfield, Ohio, were also reviewed as this stainless steel facility has a process similar to that of the Outokumpu Mill. The Mansfield Mill is permitted with an emissions cap on the entire melt shop (Permit Number P ). The NO x emission factor associated with the EAF in the melt shop is 1.43 lb/ton and is controlled by good combustion practice only. AK Steel Pennsylvania AK Steel also owns a steel mill in Butler, Pennsylvania, which produces stainless steel and flat rolled electrical steel. In order to evaluate the equipment and control technologies in place at the Butler mill, the Pennsylvania Department of Environmental Quality was contacted in order to obtain a copy of the facility s air permit. Based on Permit Number issued on April 30, 2013, the Butler facility operates two EAFs. The first EAF has a NO x emission limit of 0.42 lb/ton, and the second has a NO x limit of 0.35 lb/ton. The state also provided a copy of the facility s control technology review which was part of a plan approval application. This document indicates that BACT for NO x emissions from EAF #5 is oxy-fuel burners with oxygen lance. BACT for Outokumpu s stainless steel EAF will not be based on the AK Steel Butler EAFs based on the following: The electrical steel produced at Butler is a high-permeability, grainoriented, electrical steel used for power and distribution transformers. 42 Per ASTM A876, electrical steel is a type of steel that has high permeability and its primary application is for use in transformer cores operating at moderate to high magnetic flux densities at commercial power frequencies. ASTM A876 also states that this type of steel is a carbon steel. 43 This grade of steel produced by the Butler facility is different from the stainless steel produced at the Outokumpu Mill. 41 In order to calculate NOx emission factors for the EAFs at the Carpenter Tech mill, the following methodology was used: Five EAFs at the Carpenter facility are permitted together; the five EAFs must not exceed 54 lb/hr of NOx. To calculate an emission factor, the capacities of all five furnaces were summed ( = 25 tph total) then 54 lb/hr was divided by the total capacity. (Example Calc: 54 lb/hr / 25 tons/hr = 2.16 lb/ton). The same methodology was used for the sixth furnace. 42 AK Steel Website: (accessed on ) 43 ASTM: (accessed on ) Environmental Resources Management Southwest, Inc \ \841rpt.doc

31 Permit Number issued on April 30, 2013 requires that when stack testing is conducted on the EAFs, at least one run must be performed when carbon steel is being produced (there is no requirement to run stainless steel during stack test). 44 As discussed previously in this application, the production of carbon steel produces a foamy slag which helps decrease NO x emissions. As such, the stack test results from running carbon steel at the Butler facility would not be representative of stack testing results from Outokumpu s stainless steel EAF. As stated previously, the Butler facility s control technology review document which was part of a plan approval application indicated that at least one EAF (#5) has a BACT determination of oxy-fuel burners with oxygen lance. As stated previously, the Butler facility produces stainless steel in addition to flat rolled electrical carbon steel. As previously discussed, operating a stainless steel EAF in an oxygen enriched environment, as an oxy-fuel system would do, would cause the oxidation of chromium during the refining stage after the carbon has been oxidized. This would compromise the ability of the EAF to produce high-grade chrome stainless steel. As such, it is expected that the oxy-fuel burners with oxygen lance is not used by AK Steel during stainless steel production to avoid oxidizing the chromium in the steel. State of Texas Current BACT Guidelines On March 04, 2013, the Texas Commission on Environmental Quality (TCEQ) published current BACT guidelines for the iron and steel industry. This document proposes a NO x emission limit of 0.43 lb/ton for an EAF. However, the TCEQ guidelines do not differentiate between carbon and stainless steel processes. Outokumpu believes that the suggested BACT NO x limit provided by TCEQ is not representative of a stainless steel EAF. The TCEQ BACT guidelines list the following suggested NO x controls for an EAF based on proper equipment design, proper operation, and good engineering practices: 45 Minimizing air filtration; Maintaining the furnace draft during melting and refining operations; and, Following good combustion practices. For reference, a summary table listing each NO x emission limit implemented on an EAF at a stainless mill is presented below. 44 Permit Number , page 81 of 98, condition IIb. 45 TCEQ: Current BACT Guidelines for the Iron and Steel Industry. Electronic Source: ct/bact_ironsteel.pdf (accessed on ). Environmental Resources Management Southwest, Inc \ \841rpt.doc

32 Facility Location Emission Limit BACT Kentucky 1.32 lb/ton (or 1.00 lb/ton per last permit application) North American Stainless Direct evacuation control (with water-cooled ductwork); Good engineering practices. Carpenter Tech Pennsylvania 2.16 lb/ton (for 5 furnaces); 3.15 lb/ton (for 1 furnace) Good combustion practices; use of clean natural gas fuel in burners. AK Steel Ohio 1.43 lb/ton Good combustion practices. Nucor Steel (1) Indiana 0.35 lb/ton Not representative of Outokumpu AK Steel (2) Pennsylvania 0.35 lb/ton (for one furnace); 0.42 lb/ton (for a second furnace) Not representative of Outokumpu (see above discussion) NO x CEMS; Oxy-fuel Burners; Combustion of Natural Gas. Oxy-fuel Burners; DEC. (3) (1) Based on industry knowledge, the Nucor Steel facility located in Montgomery County, Indiana does not use oxy-fuel burner system when stainless steel batches are run. (2) The AK Steel Butler Mill produces both carbon and stainless steel and is, therefore, not a purely stainless steel mill. (3) While the RBLC database listed no NO x controls for the AK Steel Butler EAF, correspondence from the state of Pennsylvania s Department of Environmental Quality provided a copy of the PSD permit application from the Butler Mill which listed oxy-fuel burners with oxygen lances as BACT for EAF #5. Additionally, the facility s air permit lists a DEC as a control for EAF#5. Based on research conducted for this BACT analysis, none of the control technologies identified in Step 3 will be eliminated as they each are viable and proven controls. Additionally, based on the available data located during this study, it is believed that proper equipment design, proper operation, and good engineering practices and a DEC system work effectively in unison to control NOx emissions Step 5 Selection of BACT The BACT for the Outokumpu facility s EAF will be primarily based on the NO x emission limit for NAS s natural gas-fired EAF at the Ghent, Kentucky, Stainless Steel Mill. The BACT of 1.00 lb/ton (126 lb/hr) is proposed for the EAF at Environmental Resources Management Southwest, Inc \ \841rpt.doc

33 Outokumpu s Mill EAF (LO-1). To achieve this limit the furnace will be equipped with the following control technologies: The EAF will be equipped with a DEC, which will draw out emissions from the furnace during the melting phase and route the emissions to a baghouse (LO-1). The ductwork of the DEC system will be water-cooled in order to decrease the temperature of the furnace exhaust gas before the exhaust stream enters the baghouse (LO-1). The Outokumpu facility will establish and implement proper equipment design, proper operation, and good engineering practices for the EAF. Good engineering practices include, but are not limited to the following: o o o o o o o o Minimizing Air Filtration; Managing Furnace Draft During Melting and Refining Operations; Maintaining Proper Operating Temperature in the Primary Combustion Zone; Maintaining Sufficient Residence Time; Maintaining Equipment According to the Manufacturer s Instructions; Sealing the Furnace as much as Possible, which will Include Maximizing the Time that the Slag Door is Closed; Operating at Optimal Design Pressure; and Continuously Monitoring the Static Pressure of the DEC System BACT Determination for SO 2 Emissions from the Electric Arc Furnace The formation of SO 2 from an EAF is primarily associated with the combustion of sulfur compounds contained in the scrap and any other materials charged into the furnace. The quantity of SO 2 emissions generated by operating an EAF depends on the sulfur content of the scrap material, the sulfur content of the oil on the surface of the scrap metal and sulfur content of any substituted charge materials. In order to identify potential SO 2 controls and determine the BACT for Outokumpu s EAF, the following resources were researched: 1. The USEPA s RBLC database; 2. European Commission Integrated Pollution Prevention Control (IPPC): Best Available Techniques Reference Document on the Production of Iron and Steel; and, 3. Recent PSD permit applications for stainless steel production facilities. The following sections will discuss the emission limits and control technologies identified utilizing the sources above. Environmental Resources Management Southwest, Inc \ \841rpt.doc

34 BACT Baseline There are no specific regulatory requirements for SO 2 emissions from the EAF. Thus, baseline emissions are simply the uncontrolled emissions from the furnace Step 1 Identify Potential Control Technologies The EAF has the potential to generate SO 2 during the melting phase of production. The amount of SO 2 generated will be dependent upon the sulfur content of all raw materials charged in the furnace. Based on information obtained from the USEPA s RBLC database, recently submitted permit applications, and air pollution control guidance documents, a list of potential SO 2 controls for the EAF was developed. The potential control options followed by a brief description of each control alternative are outlined below: 1) Charge Substitution; 2) Flue Gas Desulfurization; 3) Sorbent Injection; 4) Chemical Additives; 5) Scrap Management Plan; and, 6) Proper Equipment Design, Proper Operation and Maintenance, and Good Engineering Practices. Charge Substitution Different materials such as pre-reduced pellets, direct reduced iron (DRI), hot briquetted iron (HBI), coal, pig iron, hot metal, and iron carbide are often charged into an EAF along with steel scrap for various reasons such as providing energy savings, increased productivity, improved slag foaming, increased carbon content in the charge, etc. SO 2 emitted from an EAF can be reduced by using lower sulfur alternatives to such charged materials. Because the quantity of SO 2 emitted from an EAF is directly correlated to the sulfur content of the materials charged in the EAF, by reducing the sulfur content of such charged materials in the furnace, the overall SO 2 emissions can be reduced. Similarly, the quantities charged into the EAF of low sulfur charge substitutes can be increased, if feasible, to offset high sulfur steel scrap in an effort to reduce the overall sulfur content of the furnace charge. Alternative charge materials such as DRI and HBI have sulfur contents as low as 0.015%. 46 Coals have sulfur contents ranging from 0.5-5% depending on the type of coal. 47 Use of coal with sulfur content towards 46 Potentialities of alternative charge materials for the electric arc furnace. Electronic Source: (accessed on ). 47 Special Variation of Organic Coal in Sulfur. C. A. Wert, B. H. Tseng. K. C. Hsieh, M. Buckentin and Y. P. Ge Department of Materials Science and Engineering and Materials Research Laboratory University of Illinois. Electronic Source: (paper accessed on ). Environmental Resources Management Southwest, Inc \ \841rpt.doc

35 the lower end of this range is a good method to reducing SO 2 emissions from an EAF. Flue Gas Desulfurization The term flue gas desulfurization (FGD) traditionally refers to scrubbers that remove SO 2 emissions from large electric utility boilers. FGD technologies are primarily used for industrial coal-fired boilers. 48 FGD systems may be broadly classified as wet or dry systems. Wet FGD systems are characterized by low flue gas outlet temperatures, saturated flue gas conditions, and a wet sludge reaction product which is dewatered before reuse or disposal. 49 Wet FGD can be further classified as non-regenerable or regenerable. Non-regenerable FGD systems are the most common type, but result in a waste product that requires proper disposal. Regenerable FGD converts the waste byproduct into a marketable product, such as sulfur or sulfuric acid. Dry FGD systems are characterized by outlet flue gas temperatures about 20 to 50 F above the saturation point, or about 150 C to 180 C. Additionally, dry FGD systems require less water than wet FGD systems. Most FGD systems employ two separate phases, a fly ash removal process and a SO 2 removal system. 50 Numerous operating variables affect the SO 2 removal rate of the scrubber. These include the following: liquid to gas ratio, ph, gas velocity, residence time, gas distribution, scrubber design, turndown, moisture content, sulfur content, ash content, and chlorine content. 51 The overall SO 2 emissions reduction in wet FGD systems ranges from % reduction. The overall SO 2 emissions reduction in dry FGD systems ranges from 70 95% removal. 41 Sorbent Injection Sorbent injection involves injecting a dry sorbent into the upper part of a furnace in order to facilitate a reaction between the SO 2 in the flue gas and the injected dry sorbent. Typically, limestone (CaCO 3 ) or hydrated lime (Ca(OH) 2 ) is used as 48 Reference Source: Boiler Emission Guide by Cleaver Brooks. Electronic Source: sights/boiler%20emissions%20guide.pdf (accessed on ). 49 We Energies - Air Pollution Control Construction Permit Application for a 50 MW Biomass Fuels-Fired Cogeneration Facility, RTP Environmental Associates, Inc. Appendix B: Control Technology Review for the Natural Gas-Fired Boilers. March USEPA Document, Electronic Source: 07c0e03a85256b6c006caf64/$FILE/si412c_lesson9.pdf (accessed on ). 51 Ponderetal 1979 and Leivo, Electronic Source: 07c0e03a85256b6c006caf64/$FILE/si412c_lesson9.pdf (accessed on ). Environmental Resources Management Southwest, Inc \ \841rpt.doc

36 sorbent. As the furnace heats to temperatures above 1382 F, SO 2 in the flue gas reacts with oxygen and calcium to form calcium sulfate (CaSO 4 ). To effectively reduce SO 2 formation, the sorbent must be evenly distributed throughout the flue gas stream; uniform distribution of the sorbent is possible only between 1382 F 2282 F. If temperatures exceed 2282 F, the surface of the sorbent is sintered, and the chemical structure of the sorbent is compromised. At temperatures below 1382 F, the reaction of SO 2, oxygen, and calcium cannot proceed, and SO 2 formation cannot be adequately reduced. SO 2 removal efficiencies for sorbent injection range from 50 95% depending on the type of sorbent used, the residence time, and the molar ratio of the sorbent particle size. 52 Chemical Additives Quicklime, both high calcium and dolomitic can be used as a flux in purifying steel produced via an EAF process. Lime is effective at removing sulfur, silica, and manganese. 53 Adding lime to the EAF charge facilitates a reaction between the additive and sulfur containing compounds in the scrap metal. As lime is composed of 90% calcium oxide (CaO), the addition of lime to the EAF charge reduces the formation of SO 2 as shown in the following chemical balance: CaO + SO 2 CaSO 3 As shown in the chemical reaction above, the utilization of lime as an additive reduces SO 2 emissions by promoting the formation of calcium sulfite (CaSO 3 ). 54 Scrap Management Plan A scrap management plan aims to eliminate undesirable compounds from scrap metal used in the steel production process to maintain the desired high quality of stainless steel produced. Facilities can only control the scrap content by visual inspections, scrap metal testing, or by selecting reputable scrap dealers offering a guarantee on the metal composition. Currently, it is not common practice for scrap vendors to provide the sulfur content of scrap metal; at best, vendors may provide a maximum allowable sulfur concentration for scrap sold IEA Clean Coal Center: Sorbent injection systems for SO 2 control Electronic Source: (accessed on ). 53 National Lime Association: Iron and Steel, Electric Arc Furnaces Benefits of Lime Flux. Electronic Source: (accessed on ). 54 United States Office of Patents - Method for reducing the SO 2 emission from a plant for manufacturing cement clinker and such plant. Electronic Source: (accessed on ). 55 CMC Commercial Metals guarantees that the sulfur content of their scrap is not to exceed 0.05%. Electronic Source: (accessed on ). Environmental Resources Management Southwest, Inc \ \841rpt.doc

37 Typical scrap used for the production of stainless steel consists of 200 series stainless steel scrap solids, 300 series stainless clips and solids, stainless steel turnings, and 400 series grade stainless steel solids. 56 Proper Equipment Design, Proper Operation, and Good Engineering Practices The formation of SO 2 generated from an EAF is primarily caused by the sulfur content of the scrap metal used as furnace charge. However, emissions can be minimally controlled by decreasing the surface tension of the slag and by increasing the viscosity of the slag. By doing so, the slag s ability to foam will be optimized. 57 SO 2 emissions are dependent upon the foam index of the slag when slag properly foams, sulfur containing compounds are oxidized and SO 2 is minimized Step 2 Eliminate Technically Infeasible Options Flue Gas Desulfurization FGD systems use an alkaline reagent to facilitate a reaction such that SO 2 emissions are reduced by the production of solid sulfur compounds. According to the USEPA, wet and dry FGD units have been successfully installed on industrial boilers, waste incinerators, cement and lime kilns, metal smelters, glass furnaces, and sulfuric acid manufacturing equipment. These systems have not, however, been proven in the United States on an EAF. For wet FGD systems, the inlet gas temperature into the wet scrubber must be between 300 F 700 F. 58 For dry FGD systems, the temperature of the flue gas exiting the absorber must be 20 F 30 F above the adiabatic saturation temperature. While it is very difficult to estimate the adiabatic saturation temperature, the EPA suggests that the optimal temperature for SO 2 removal for dry systems ranges from 300 F 350 F. 59 Based on the EAF stack temperature measured during stack testing along with the air modeling conducted during this BACT study, the average temperature of the exhaust gas is 114 F. 56 A. Javaid and E. Essadiqi. Final Report on Scrap Management, Sorting, and Classification of Steel. December Government of Canada. Electronic Source: (accessed on ) 57 Foamy Slag Fundamentals and Their Practical Application to Electric Furnace Steelmaking. Electronic Source: their%20practical%20application%20to%20electric%20furnace%20steelmaking.pdf 58 FETC, Electric Utility Engineer s FGD Manual, Volume 1: FGD Process Design. Department of Energy, Federal Energy Technology Center. Morgantown, WV, EPA Guidance Document on FGD. Electronic Source: (accessed on ). Environmental Resources Management Southwest, Inc \ \841rpt.doc

38 In order for a FGD system to effectively control SO 2 emissions after the baghouse, the gas stream would have to be re-heated. In order for a FGD to effectively control SO 2 emissions closer to the the outlet of the EAF, the exhaust stream would need to be cooled from temperatures in excess of 3000 F to temperatures within the FGD optimal range. A cooling system such as air quenching, cooling ducts, or spray coolers would have to be designed and installed to cool the exhaust gas stream to remain within the required FGD temperature range. Due to the above factors and because FGD systems have not been proven in the United States on an EAF, this technology is considered technically infeasible. Sorbent Injection Sorbent injection involves injecting CaCO 3 or Ca(OH) 2 into the upper part of a furnace in order to facilitate a reaction between the SO 2 in the flue gas and the injected sorbent. In order to reduce SO 2, the uniform distribution of the sorbent must be guaranteed. It is important to note that homogeneous dispersion of the sorbent is only possible within the temperature range of 1382 F 2282 F. If temperatures exceed 2282 F, the surface of the sorbent is sintered, and the chemical structure of sorbent is compromised. 60 Thus, if temperatures in the EAF rise above 2282 F, SO 2 is not removed. According to the American Iron and Steel Institute (AISI), the production of molten steel occurs at temperatures between 2900 F F. 61 This indicates that in order to produce stainless steel an EAF must exceed the maximum temperature where sorbent injection technology is effective. Furthermore, mathematical equations can be used to approximate thermodynamic equilibrium of the chromium to carbon content of the steel production batch. Simkovich and McCoy developed such an equation for stainless steel production; this equation assumes a temperature range of 3100 F 3300 F. 62 This again indicates that temperatures within the EAF can greatly exceed the maximum allowable temperature of 2282 F for the use of sorbent injection. As the temperature of the EAF at the Outokumpu Mill can exceed the optimum temperature range needed for sorbent injection, this control technology is considered technically infeasible. Scrap Management Plan A scrap management plan may be used by a steel mill as a guide to producing the desired quality of product while attempting to accept the widest range of raw 60 IEA Clean Coal Center: Sorbent injection systems for SO 2 control. Electronic Source: (accessed on ). 61 American Iron and Steel Institute CR Electronic Source: on ). 62 Key to Metals: Production of Steel Part One. Electronic Source: (accessed on ). Environmental Resources Management Southwest, Inc \ \841rpt.doc

39 materials available. As a whole, a scrap management plan is based on maximizing the furnace charge within the bounds of the operability of the unit. Typically, environmental and safety limits included in a scrap management plan restrict the materials that are regulated under 40 CFR 63 National Emission Standards For Hazardous Air Pollutants For Source Categories (Subpart EEEEE - National Emission Standards For Hazardous Air Pollutants For Iron And Steel Foundries and Subpart FFFFF - National Emission Standards For Hazardous Air Pollutants For Integrated Iron And Steel Manufacturing Facilities). However, Outokumpu is not subject to either 40 CFR 63 Subpart EEEEE or Subpart FFFFF. As such, there is no regulatory driver for Outokumpu to develop and maintain a scrap management plan. The aforementioned federal regulations set forth regulatory guidelines on the composition of scrap material used for steel production; the regulated pollutants include chlorinated plastics, polychlorinated biphenyls, free organic liquids, organic metals, lead, and mercury. There is currently no regulatory limit restricting the sulfur contained in scrap material. As there is no regulatory limit on the sulfur content of scrap, several scrap management plans were reviewed to gauge the industry standard on scrap composition. In conclusion of the study, it was determined that a restriction on sulfur bearing scrap material is not evidenced by any of the plans reviewed. Furthermore, scrap materials used to produce high alloy steel are selected based on market availability and the belief that the type of scrap metal chosen will produce the desired high quality stainless steel. A restriction on the variety of scrap a mill can accept as charge can limit the ability of the mill to ensure production, utilization and steel quality of the plant. Lastly, the practicality of ensuring a maximum allowable sulfur content for all scrap metal charged in the EAF was considered. At best, scrap vendors can only provide an estimated range of the sulfur content of the scrap based on the types of recycled material within the scrap. 63 Thus, in order for Outokumpu to validate the actual amount of sulfur being processed in the EAF, facility personnel would have to test each scrap batch that entered the facility. As this practice is not practical due to market demanded production timelines, the use of a scrap management plan to reduce the sulfur content of scrap metal is considered practically infeasible Step 3 - Rank Remaining Technically Feasible Control Options The remaining control technologies are ranked based upon records identified in the RBLC database; these records are presented in Table 3-2 for reference. The data presented in Table 3-2 includes EAFs operating at both carbon and stainless steel production facilities. Facilities producing stainless steel have been highlighted in a light blue color for ease of identifying. While control technologies and corresponding emission limits for carbon steel EAFs are 63 CMC Scrap Management Plan Electronic Source: Environmental Resources Management Southwest, Inc \ \841rpt.doc

40 presented in Table 3-2, SO 2 emission factors for carbon steel EAFs are not representative of a stainless steel EAF. Similarly, the SO 2 control options appropriate for a carbon steel EAF are not necessarily appropriate for a stainless steel EAF. As discussed previously in Section 2.1, SO 2 emissions from a stainless steel EAF are typically higher than from a carbon steel EAF for the following reasons: The sulfur content in the charge scrap for the stainless steel process is typically greater than that in the carbon steel process. Charge scrap for producing carbon steel consists of non-alloyed steel while charge scrap for stainless steel production consists of low or high alloyed metal. 64 Non-alloyed steel used as scrap for carbon steel production can have a sulfur content ranging from %. 65 Low-alloyed or high-alloyed steel, however, can have a sulfur content ranging from % according to the American Iron and Steel Institute. 66 In the stainless steel production process, the chrome oxide content of the slag can cause the slag to have a lower foam index as compared to a carbon steel slag. 67 This means that in a stainless steel production process, the slag layer is significantly less abundant than in the carbon steel production process. Most of the slag formation and removal in a stainless steel production process occurs in the AOD. As sulfur compounds are primarily removed as sulfides dissolved in the slag, the amount of SO 2 emitted from the process depends on the volume of slag generated; as slag production increases, SO 2 emissions decrease and vice versa. This is why a stainless steel EAF has higher SO 2 emissions compared to a carbon steel EAF. 64 The European Association Representing Metallurgical Slag Producers and Processors: Electric Arc Furnace Slag. Electronic Source: on ). 65 MSDS for Non-Alloyed Steel. Electronic Source: (accessed on ). 66 High Strength Low Alloyed Steel Document. Electronic Source: df/a764507a d23-b d31c0ba2 (accessed on ). 67 Metallurgical and Materials Transactions B August 2004, Volume 35, Issue 4, pp Environmental Resources Management Southwest, Inc \ \841rpt.doc

41 Table RACT/BACT/LAER/Clearinghouse Database Records for Sulfur Dioxide (SO 2 ) Emissions and Control Technologies for an Electric Arc Furnace (1) Stainless Steel Facility? RBLCID FACILITY NAME FACILITY STATE PERMIT NUMBER PERMIT ISSUANCE DATE PROCESS NAME POLLUTANT CONTROL METHOD DESCRIPTION EMISSION LIMIT (3) (lb/ton) No OH-0302 REPUBLIC ENGINEERED PRODUCTS, INC. OH /30/2005 ELECTRIC ARC FURNACE Sulfur Dioxide (SO2) NONE PROVIDED 0.07 No OH-0315 NEW STEEL INTERNATIONAL, INC., HAVERHILL OH /6/2008 ELECTRIC ARC FURNACE (2) Sulfur Dioxide (SO2) NONE PROVIDED 0.10 Yes AL-0230 THYSSENKRUPP STEEL AND STAINLESS USA, LLC AL X001 THRU X026 8/17/2007 TPH ELECTRIC ARC FURNACE WITH DEC & ELEPHANT HOUSE VENTED TO BAGHOUSE 3 (LA1) (MULTIPLE EMISSION POINTS) Sulfur Dioxide (SO2) NONE PROVIDED 0.15 Yes AL-0230 THYSSENKRUPP STEEL AND STAINLESS USA, LLC AL X001 THRU X026 8/17/2007 TPH ELECTRIC ARC FURNACE WITH DEC & ELEPHANT HOUSE VENTED TO BAGHOUSE 3 (LA1) (MULTIPLE EMISSION POINTS) Sulfur Dioxide (SO2) NONE PROVIDED 0.15 No MN-0070 MINNESOTA STEEL INDUSTRIES, LLC MN /7/2007 ELECTRIC ARC FURNACE/MELT SHOP Sulfur Dioxide (SO2) NONE PROVIDED 0.15 No OH-0246 THE TIMKEN COMPANY/FAIRCREST PLANT OH PTI /20/2003 EAF Sulfur Dioxide (SO2) NONE PROVIDED 0.15 No GA-0142 OSCEOLA STEEL CO. GA P /29/2010 Electric Arc Furnace Sulfur Dioxide (SO2) Fuel Selection (firing natural gas exclusively) Charge Material Selection (Use of low-sulfur containing feed material No MI-0404 Gerdau Macsteel, Inc. MI /4/2013 EUEAF, EULMF, AND TWO EUVTD Sulfur Dioxide (SO2) NONE PROVIDED 0.2 No AR-0077 BLUEWATER PROJECT AR 2062-AOP-R0 7/22/2004 ELECTRIC ARC FURNACE (EAF) Sulfur Dioxide (SO2) LOW SULFUR COKE AND SCRAP MANAGEMENT 0.20 No AR-0096 NUCOR YAMATO STEEL AR 883-AOP-R8 1/31/2008 ELECTRIC ARC FURNACE Sulfur Dioxide (SO2) LOW SULFUR COKE AND SCRAP MANAGEMENT 0.20 No OH-0276 CHARTER STEEL OH /10/2004 ELECTRIC ARC FURNACE Sulfur Dioxide (SO2) PRODUCTION LIMITS. SEE NOTE 0.20 No AR-0077 BLUEWATER PROJECT AR 2062-AOP-R0 7/22/2004 ELECTRIC ARC FURNACE (EAF) Sulfur Dioxide (SO2) LOW SULFUR COKE AND SCRAP MANAGEMENT 0.20 No AR-0086 NUCOR-YAMATO STEEL COMPANY, BLYTHEVILLE MILL AR 883-AOP-R4 6/11/2004 EAF #1 BAGHOUSE, SN-01 Sulfur Dioxide (SO2) LOW SULFUR COKE USAGE 0.20 No NY-0094 NUCOR AUBURN STEEL NY / /22/2004 EAF Sulfur Dioxide (SO2) NONE PROVIDED 0.23 No CO-0054 CF & I STEEL L.P. DBA ROCKY MOUNTAIN STEEL MILLS CO 02PB0492 6/21/2004 ELECTRIC ARC FURNACE (EAF) Sulfur Dioxide (SO2) ALTERNATIVE RAW MATERIALS + PROCESS CONTROLS 0.25 No OH-0294 NUCOR STEEL MARION, INC. OH /18/2005 ELECTRIC ARC FURNACE (STACK EMISSIONS) Sulfur Dioxide (SO2) NONE PROVIDED 0.25 No OH-0316 V & M STAR OH P /23/2008 ELECTRIC ARC FURNACE Sulfur Dioxide (SO2) NONE PROVIDED 0.25 No CO-0054 CF & I STEEL L.P. DBA ROCKY MOUNTAIN STEEL MILLS CO 02PB0492 6/21/2004 ELECTRIC ARC FURNACE (EAF) Sulfur Dioxide (SO2) ALTERNATIVE RAW MATERIALS + PROCESS CONTROLS 0.25 No CO-0061 CF&I STEEL L.P. DBA ROCKY MOUNTAIN STEEL MILLS CO 02PB0492 6/21/2004 EAF #5 Sulfur Dioxide (SO2) Yes IN-0108 NUCOR STEEL (2) IN /21/2003 EAF, AOD VESSELS, DESULFURIZATION, & OTHER PROCESS Sulfur Dioxide (SO2) ALTERNATIVE RAW MATERIALS AND PROCESS CONTROLS SCRAP MANAGEMENT PLAN. COMPLIANCE METHOD: SO2 CEM No OH-0285 NORTH STAR BHP STEEL, LTD OH /20/2005 ELECTRIC ARC FURNACE WITH TWO LADLE MELT FURNACES Sulfur Dioxide (SO2) PROCESS OPERATIONS AND SCRAP MANAGEMENT PLAN 0.25 No OH-0328 V & M STAR OH P /10/2009 ELECTRIC ARC FURNACE Sulfur Dioxide (SO2) NONE PROVIDED 0.27 No OH-0292 WHEELING PITTSBURGH STEEL CORPORATION OH /6/2005 ELECTRIC ARC FURNACE Sulfur Dioxide (SO2) NONE PROVIDED 0.30 No OK-0128 MID AMERICAN STEEL ROLLING MILL OK C(M-1) PSD 9/8/2008 Electric Arc Furnaces Sulfur Dioxide (SO2) cleaned scrap 0.30 No OH-0292 WHEELING PITTSBURGH STEEL CORPORATION OH /6/2005 ELECTRIC ARC FURNACE Sulfur Dioxide (SO2) NONE PROVIDED 0.30 No AR-0090 NUCOR STEEL, ARKANSAS AR 1139-AOP-R6 4/3/2006 EAF'S LMF'S Sulfur Dioxide (SO2) NONE PROVIDED 0.30 No IN-0140 NUCOR IN /8/2010 ELECTRIC ARC FURNACES 1 & 2, TWO CONTINUOUS CASTERS, DESULFURIZATION STATION, LADLE DRYER, LADLE PREHEATER, ONE ARGON OXYGEN DECARURIZATION VESSEL, ONE LMF EU-13C, 2 LMFS (EU13A & EU13B) Sulfur Dioxide (SO2) NONE PROVIDED 0.33 No TX-0417 NUCOR CORP. TX PSD-TX /15/2003 ELECTRIC ARC FURNACE Sulfur Dioxide (SO2) LOW SULFUR FUEL 0.35 No TX-0463 JEWETT PLANT STEEL MILL TX P1029 1/15/2003 EAF BAGHOUSE STACK Sulfur Dioxide (SO2) NUCOR WILL USE SWEET NATURAL GAS AS THE SUPPLEMENTAL FUEL SOURCE FOR THE EAF AND WILL EMIT NOT MORE THAN 0.35 LB SO2/TON STEEL 0.35 No NC-0112 NUCOR NC 08680T09 11/23/2004 ELECTRIC ARC FURNACE, LADLE METALLURGY STATION, AND CONTINUOUS SLAB CASTER Sulfur Dioxide (SO2) SCRAP MANAGEMENT 0.35 No AR-0078 NUCOR STEEL, ARKANSAS AR 1139-AOP-R5 6/9/2003 EAF Sulfur Dioxide (SO2) SCRAP MANAGEMENT 0.36 No *OH-0350 REPUBLIC STEEL OH P /18/2012 Electric Arc Furnace Sulfur Dioxide (SO2) NONE PROVIDED 0.39 No OH-0339 HARRISON STEEL PLANT OH P /29/2010 Electric Arc Furnace (2) Sulfur Dioxide (SO2) NONE PROVIDED 0.44 No AL-0218 NUCOR STEEL TUSCALOOSA, INC. AL /6/2006 ELECTRIC ARC FURNACE Sulfur Dioxide (SO2) UTILIZATION OF A SCRAP MANAGEMENT PROGRAM 0.46

42 Stainless Steel Facility? RBLCID FACILITY NAME FACILITY STATE PERMIT NUMBER PERMIT ISSUANCE DATE PROCESS NAME POLLUTANT CONTROL METHOD DESCRIPTION EMISSION LIMIT (3) (lb/ton) No OH-0341 NUCOR STEEL MARION, INC. OH P /23/ electric arc furnace, including charging, melting, tapping, slag skimming; continuous casting operations, including torch cutting; 4 natural gas filed ladle preheaters; and 2 natural gas fired tundish preheaters. Sulfur Dioxide (SO2) NONE PROVIDED 0.50 No OH-0342 FAIRCREST STEEL OH P /29/2010 Electric Arc Furnace Sulfur Dioxide (SO2) NONE PROVIDED 0.52 No PA-0251 ELLWOOD NATIONAL STEEL PA 62-32B 8/18/2006 ELECTRIC ARC FURNACE Sulfur Dioxide (SO2) NONE PROVIDED 0.55 No IL-0104 NUCOR STEEL KANKAKEE IL /12/2007 ELECTRIC ARC FURNACE Sulfur Dioxide (SO2) IMPLEMENTATION OF SCRAP MANAGEMENT PROGRAM No AL-0202 CORUS TUSCALOOSA AL X005,X008 6/3/2003 ELECTRIC ARC FURNACE Sulfur Dioxide (SO2) NONE PROVIDED 0.62 No AL-0231 NUCOR DECATUR LLC AL /12/2007 TWO (2) ELECTRIC ARC FURNACES AND THREE (3) LADLE METALLURGY FURNACES WITH TWO (2) MELTSHOP BAGHOUSES Sulfur Dioxide (SO2) NONE PROVIDED 0.62 No SC-0127 NUCOR STEEL CORPORATION (DARLINGTON PLANT) SC DE 12/29/2006 ELECTRIC ARC FURNACE Sulfur Dioxide (SO2) SCRAP METAL MANAGEMENT PROGRAM, AND FOR RESULFURIZED STEEL PRODUCTION, LIMITING THE TOTAL ANNUAL PRODUCTION OF RESULFURIZED STEEL 0.67 TO 105,200 BILLET TONS IN ADDITION TO SCRAP METAL MANAGEMENT. No MI-0376 MACSTEEL DIVISION MI G 12/8/ ELECTRIC ARC FURNACES Sulfur Dioxide (SO2) NONE PROVIDED 1.00 No TX-0399 NUCOR JEWETT PLANT TX PSD /5/2003 EAF, LMF, CASTER MELTSHOP Sulfur Dioxide (SO2) NONE PROVIDED 1.76 Notes: (1) Records from the RBLC were obtained on July 30, 2013 and re-verified on April 21, The following categories in the RBLC were searched: General - Electric Arc Furnace; Ferroalloy Production; Steel Production EAF. (2) The Nucor Steel facility located in Montgomery County, Indiana is permitted to produce both stainless steel and carbon steel. (3) Where applicable the emission limit (lb/hr) was divided by the throughput (ton/hr) in order to obtain the emission factor (lb/ton).

43 As shown in Table 3-2, records from the RBLC database were sorted in ascending order of emission limit. The first listed control is charge substitution which entails using low sulfur feed materials. This control technology is proposed on a natural gas-fired EAF at the yet to be built Osceola Steel Mill in Cook County, Georgia. Based on the RBLC database, charge substitution is the most effective way to reduce SO 2 emissions from an EAF. As such, this control technology is ranked first. Please note, while a record for the Osceola Steel Mill is recorded in the RBLC, the plant has not been built. Thus, the control technologies presented in the RBLC for the Osceola EAF are not yet demonstrated. The remaining technologies have been ranked based on process knowledge and engineering judgment. 1. Charge Substitution; 2. Chemical Additives; and, 3. Proper Equipment Design, Proper Operation and Maintenance, and Good Engineering Practices Step 4 - Evaluating Remaining Control Technologies To evaluate the effectiveness of the remaining control technologies, control technology guidance documents, the RBLC database, and permitted SO 2 emission limits from stainless steel facilities were compared. Universal Stainless Pennsylvania The Universal Stainless steel mill located in Bridgeville, Pennsylvania, operates one EAF. Based on the Title V Operating Permit (Permit Number 0027) issued on December 20, 2005, the SO 2 controls in place on the EAF are as follows: Chemical Addition of Lime to the Furnace Charge; and, Good Engineering and Air Pollution Control Practices. With these controls in place, the permitted SO 2 emission limit for the EAF is 1.62 lb/hr with a maximum design rate of 23.1 tons of steel/hr. 68 Please note, the SO 2 permit limit at Universal Stainless is a Title V Air Permit limit, not a BACT/PSD limit. Nucor Steel Indiana According to the facility s PSD Permit (Permit Number ) issued on November 21, 2003, Nucor s melt shop operates two EAFs with the following SO 2 controls: 68 In order to determine an emission factor for the Universal Stainless EAF, the hourly limit is divided by the maximum design rate to calculate a factor in lb/ton (1.62/23.1 = 0.07). The permitted limits are as follows: 1.62 lb/hr and 6.14 tons/yr. Environmental Resources Management Southwest, Inc \ \841rpt.doc

44 Chemical Addition of Lime; Charge Substitution of DRI, HBI, and Pig Iron; and, Good Engineering Practices. With the listed controls in place, the total SO 2 emissions from the Nucor melt shop EAF Baghouses (1 and 2) shall not exceed 0.25 lb/ton and 125 lb/hr, based on a 3-hour block average. As the Nucor mill is primarily a carbon steel mill, the Nucor SO 2 emission limit is not adequately representative value for the EAF at the Outokumpu Mill. State of Texas Current BACT Guidelines On March 04, 2013, the TCEQ published current BACT guidelines for the iron and steel industry. This document proposes an SO 2 emission limit of 0.24 lb/ton for an EAF. However, the TCEQ guidelines do not differentiate between carbon and stainless steel processes. As such, Outokumpu believes that the suggested BACT SO 2 limit provided by TCEQ is not representative of a stainless steel EAF. The TCEQ BACT guidelines do not list any suggested SO 2 controls for an EAF. 69 European Commission IPPC Manual The European Commission s IPPC Best Available Techniques for the Production of Iron and Steel was consulted in order to evaluate internationally accepted control systems and emission factors for SO 2. Based on this document, SO 2 emission factors for an EAF typically range from grams/ton ( lb/ton) based on actual industry data. 70 These values presented in the IPPC are provided for informational purposes and are not regulatory enforced emission limits. For an EAF, there is no proposed SO 2 control system recommended as a Best Available Technique from the IPPC Manual. Conclusion Based on research conducted for this BACT analysis, it is believed that the utilization of charge substitution and chemical additives coupled with proper equipment design, proper operation and maintenance, and good engineering practices will work most effectively in unison. Thus, no further technologies will be evaluated, and the remaining technologies will be implemented together. 69 TCEQ: Current BACT Guidelines for the Iron and Steel Industry. Electronic Source: ct/bact_ironsteel.pdf (accessed on ). 70 European Commission IPPC Manual December 2001, Table 9.1 page 281. Electronic Source: (accessed December 2, 2013). Environmental Resources Management Southwest, Inc \ \841rpt.doc

45 Step 5 Selection of BACT The BACT for the Outokumpu facility s EAF will be based on the viable control technologies identified from the RBLC database, from permitted stainless steel facilities, and from process engineering knowledge. The EAF (LO-1) at the Outokumpu Mill will be equipped with the following control technologies: The EAF will utilize charge substitution by using only low-sulfur coal as charge. While coals can have sulfur contents ranging from 0.5-5%, the coal charged into the EAF at the Outokumpu Mill will meet the lower end of this limit by having a target sulfur content of 0.55%. Lime will be added to the EAF to facilitate a reaction between the calcium oxide in the lime and sulfur containing compounds in the scrap metal and other charged materials. The utilization of this chemical additive will reduce SO 2 emissions by promoting the formation of calcium sulfite. Proper equipment design, proper operation and maintenance, and good engineering practices will be implemented on the EAF and the associated control systems. With the aforementioned controls in place, a BACT emission limit of lb/hr calculated based on a 3-hour average is proposed for the EAF (LO-1) at Outokumpu s Mill in Calvert. The proposed hourly limit of lb/hr is calculated based on a lb/ton SO 2 emission limit. 71 This limit is the best SO 2 emission rate that can be consistently achieved and was determined by Outokumpu s engineers based on evaluation and application of above controls and a thorough review of actual historical CEMS data. 71 To calculate the proposed hourly emission rate, a SO 2 emission factor of lb/ton based on the BACT analysis was multiplied by the Outokumpu EAF production rate of 126 tons per hour. See emission calculations in Appendix A for more details. Environmental Resources Management Southwest, Inc \ \841rpt.doc

46 4.0 FEDERAL AIR REGULATORY ANAYLSIS As part of Outokumpu s reconciliation of its EAF permit limits for NO x and SO 2, a regulatory analysis was performed to determine what, if any, changes are required to ensure compliance. The following sections discuss the potentially applicable federal air regulations for Outokumpu s EAF. New Source Performance Standards New Source Performance Standards (NSPS) require new, modified, or reconstructed sources to control emissions to the level achievable by the best demonstrated technology as specified in the applicable provisions. An applicability analysis of potentially applicable NSPS subparts is presented below. Subpart A General Provisions Sources subject to source-specific NSPS are also subject to the general provisions of NSPS Subpart A. In general, NSPS Subpart A may require facilities subject to a source-specific NSPS to be subject to the following: Initial construction/reconstruction notifications; Initial startup notifications; Performance tests; Performance test date initial notifications; General monitoring requirements; General recordkeeping requirements; and, Semiannual monitoring system and/or excess emissions reports. 40 CFR 60 Subpart AAa - Standards of Performance for Steel Plants: Electric Arc Furnaces and Argon-Oxygen Decarburization Vessels Constructed After August 17, 1983 The provisions of this subpart apply to EAFs, AODs, and dust handling systems in steel plants producing carbon, alloy, or specialty steels. The rule applies to any such unit installed after August 17, 1983 and sets forth the following provisions: The volume of particulate matter emitted from an EAF or AOD is limited to grain per dry standard cubic foot (gr/dscf). The opacity from the EAF and AOD control device is limited to 3%. The opacity from the melt shop is limited to 6%. The dust handling system cannot exceed 10% opacity. The rule requires installation of a continuous opacity monitoring system (COMS) on each baghouse controlling an EAF or AOD. Alternatively, a COMS is not required for any single stack fabric filter or any modular, Environmental Resources Management Southwest, Inc \ \841rpt.doc

47 multi-stack fabric filter if opacity observations are made by a certified Method 9 observer and if the owner installs and operates a bag leak detection system. If the alternative method is selected, visible emissions observations must be conducted at least once per day for at least three 6- minute periods when the EAF is operating in the melting and refining period. o Permit Condition 30 of Permit Number X001 specifies that a COMS must be installed on Baghouse LO-1, and the COMS must be installed, maintained, and operated in accordance with the requirements outlined in 40 CFR 60 Appendix B. The rule also requires monitoring the static pressure monitor in the duct of the direct evacuation canopy (DEC) system, fan amps, and damper setting at least once per shift. Alternatively, a continuous monitoring device that monitors the air flow-rate being processed by the baghouse may be used. o A static pressure monitoring device is not required on the DEC system if a certified observer conducts observations at least once per day when the furnace is operating in the meltdown and refining period. The initial PSD permit application presented standards for particulate matter and opacity control as part of the facility s BACT analysis. Each of the control measures equals or supersedes the standards presented in this rule. Outokumpu s emissions test performed in May 2013 demonstrated compliance with this rule. Permit Condition 27 of Permit Number X001 requires that the monitoring standards specified in 40 CFR a(a), (b), and (d) and 40 CFR a(a) through (h) must be adhered to at all times. National Emission Standards for Hazardous Air Pollutants The National Emission Standards for Hazardous Air Pollutants (NESHAP) are codified under 40 CFR 63. NESHAP are emission standards for Hazardous Air Pollutants (HAP) that are generally applicable to major sources of HAPs, but also apply to certain area sources of HAPs. A HAP major source is defined as having potential emissions in excess of 10 tons per year (tpy) for any individual HAP and/or 25 tpy for total HAPs. NESHAP apply to specific pollutant sources, to sources in specifically regulated industrial source categories, or to facilities not regulated as a specific industrial source type on a case-by-case basis. Outokumpu is a major source for HAPs. An applicability analysis of potentially applicable NESHAP subparts is presented below. Subpart A General Provisions All affected sources are subject to the general provisions of NESHAP Subpart A unless specifically excluded by the source-specific NESHAP. NESHAP Environmental Resources Management Southwest, Inc \ \841rpt.doc

48 Subpart A requires initial notification, performance testing, recordkeeping, and monitoring, and mandates general control device requirements for all other subparts, as applicable. As Outokumpu is subject to another 40 CFR 63 subpart, the provisions of Subpart A are also applicable. 40 CFR 63 Subpart EEEEE - National Emission Standards for Hazardous Air Pollutants for Iron and Steel Foundries The provisions of 40 CFR 63 Subpart EEEEE NESHAP for Iron and Steel Foundries regulate existing and new iron and steel foundries. As per 40 CFR , the rule applies to owners and operators of iron and steel foundries that are (or are part of) a major source of HAP emissions. Iron and steel foundry is defined in 40 CFR as follows: Iron and steel foundry means a facility or portion of a facility that melts scrap, ingot, and/or other forms of iron and/or steel and pours the resulting molten metal into molds to produce final or near final shape products for introduction into commerce. Research and development facilities and operations that only produce noncommercial castings are not included in this definition. Mold is defined in 40 CFR as follows: Mold or core making line means the collection of equipment that is used to mix an aggregate of sand and binder chemicals, form the aggregate into final shape, and harden the formed aggregate. This definition does not include a line for making green sand molds or cores. The Outokumpu Mill does not perform these types of molding activities. Instead of a mold, Outokumpu utilizes a continuous caster to produce slabs of steel that will later be processed in either a hot or cold rolling mill. Therefore, this regulation is not applicable. 40 CFR 63 Subpart FFFFF - National Emission Standards for Hazardous Air Pollutants for Integrated Iron and Steel Manufacturing Facilities The provisions of 40 CFR 63, Subpart FFFFF National Emission Standards for Hazardous Air Pollutants for Integrated Iron and Steel Manufacturing Facilities apply to integrated iron and steel manufacturing facilities. As per 40 CFR , this rule applies to owners and operators of integrated iron and steel manufacturing facilities that are (or are part of) a major source of HAP emissions. Integrated iron and steel manufacturing facility is defined in 40 CFR as follows: Integrated iron and steel manufacturing facility means an establishment engaged in the production of steel from iron ore. Outokumpu does not produce steel from iron ore; the raw material used at the Outokumpu Mill consists of scrap metal. The reconciliation of emissions associated with the EAF in the melt shop does not affect the applicability of this regulation. Therefore, this regulation is not applicable. Environmental Resources Management Southwest, Inc \ \841rpt.doc

49 40 CFR 63 Subpart YYYYY - National Emission Standards for Hazardous Air Pollutants for Area Sources: Electric Arc Furnace Steelmaking Facilities The provisions of 40 CFR 63 Subpart YYYYY applies to EAF steelmaking facilities that are an area source of HAPs. As the Outokumpu Mill is classified as a major source of HAPs, per 40 CFR (a) the EAF at Outokumpu s Mill does not qualify as an affected source under Subpart YYYYY. Therefore, this regulation is not applicable. Note, as Subpart YYYYY is not applicable to the Outokumpu EAF, there is no regulatory requirement for the Outokumpu Mill to implement a scrap management plan. Environmental Resources Management Southwest, Inc \ \841rpt.doc

50 5.0 REQUESTED PERMIT CONDITION UPDATES Outokumpu requests the following amendments to the conditions in the current PSD permit (Permit Number X001) issued on March 21, As discussed in detail in Sections 2.2 and 3.2 of this application, Outokumpu is proposing to increase the SO 2 emission limit for baghouse LO-1 associated with the stainless steel EAF. Outokumpu is proposing an emission limit of lb/hr based on a 3-hour average measured by CEMS. The lb/hr limit is calculated based on the SO 2 emission factor of lb/ton determined from the BACT analysis enclosed within this application. Outokumpu is proposing that the enforceable emission limit for LO-1 be the pound per hour limit. This is because the actual production rate of steel in the EAF varies during the short-term tap-to-tap cycles and can make the calculation of short-term emissions on a lb/ton basis infeasible (e.g. periods when steel production rate is zero). As such, Outokumpu requests that the SO 2 emission limit in units of pound per ton be removed from the permit condition. 2. As discussed in detail in Sections 2.1 and 3.2 of this application, Outokumpu is proposing to increase the NO x emission limit for emissions point LO-1 associated with the stainless steel EAF. Outokumpu is proposing an emission limit of 126 lb/hr based on an average of three stack test runs. The 126 lb/hr limit is calculated based on the NO x emission factor of 1.00 lb/ton determined from the BACT analysis enclosed within this application. Consistent with the above request for SO 2, Outokumpu is proposing that the enforceable NO x emission limit for LO-1 be the pound per hour limit. As such, Outokumpu requests that the NO x emission limit in units of pound per ton be removed from the permit condition. 3. Outokumpu requests that Permit Condition 20 be removed. Permit Condition 20 states, There shall be no visible emissions allowed from the roof or any openings of the meltshop enclosing the electric arc furnace and AOD vessel. Per 40 CFR a(a)(3), the opacity standard for gases emitted from a melt shop (housing EAFs and AODs) is limited to 6% opacity. As such, Outokumpu believes there is no regulatory basis for Permit Condition 20 and requests that it be removed. 4. Outokumpu requests that Permit Condition 33 be removed. Permit Condition 33 states, The permittee shall inspect incoming scrap as outlined in the scrap management plan. The plan may be revised upon approval by the Department. Records documenting all scrap inspections shall be kept in a form suitable for review for a period of at least five years from the date of inspection. As discussed in Section of this application, a scrap management plan is deemed as an infeasible method to reduce emissions from melt shop operations. Additionally, as discussed in Section 4.0 of this application, there is no regulatory driver for Outokumpu to implement and maintain a scrap management plan (Outokumpu is not subject to NESHAP Subpart EEEEE, FFFFF, or YYYYY). Environmental Resources Management Southwest, Inc \ \841rpt.doc

51 Project Emission Calculations Appendix A Project No Outokumpu Stainless USA, LLC Calvert, Alabama Environmental Resources Management-Southwest, Inc. 775 University Blvd. North, Suite 280 Mobile, Alabama (251)

52 Outokumpu Stainless Steel Mill Potential Emissions Calculations for Electric Arc Furnace Calvert, Alabama Source: LO-1 Description: Electric Arc Furnace 1 Inputs: Description Value Units Steel Production (1) 1,100,000 tons/yr Maximum Hourly Steel Production (2) 126 tons/hr Annual Operating Time 8,760 hr/yr Potential Emissions Summary: Maximum Hourly Average Annual Average Hourly Emissions Pollutant Emissions Emissions lbs AVG /hr lbs MAX /hr tpy SO NO X Emission Factors: Pollutant (3) Value Units NO X 1.00 lb/ton Notes: SO lb/ton (1) The steel production is based on Permit Condition 21 of Permit Number X001 issued on March 21, (2) The maximum hourly steel production rate represented in 2007 PSD Permit Application and validated by Outokumpu in November (3) Emissions factors from the updated BACT analysis presented in this PSD permit modification application. Page 1 of \ \841rpt\Appendix A

53 ADEM Forms Appendix B Project No Outokumpu Stainless USA, LLC Calvert, Alabama Environmental Resources Management-Southwest, Inc. 775 University Blvd. North, Suite 280 Mobile, Alabama (251)

54 ALABAMA DEPARTMENT OF ENVIRONMENTAL MANAGEMENT (AIR DIVISION) Do not Write in This Space Facility Number - CONSTRUCTION/OPERATING PERMIT APPLICATION FACILITY IDENTIFICATION FORM 1. Name of Facility, Firm, or Institution: Outokumpu Stainless USA, LLC Facility Physical Location Address Street & Number: 1 ThyssenKrupp Drive City: Calvert County: Mobile Zip: Facility Mailing Address (If different from above) Address or PO Box: 1 ThyssenKrupp Drive, P.O. Box City: Calvert State: Alabama Zip: Owner's Business Mailing Address 2. Owner: Outokumpu Stainless USA, LLC Street & Number: 1 ThyssenKrupp Drive, P.O. Box City: Calvert State: Alabama Zip: Telephone: Responsible Official's Business Mailing Address 3. Responsible Official: Michael Wallis Title: Senior Vice President Street & Number: 1 ThyssenKrupp Drive, P.O. Box City: Calvert State: Alabama Zip: Telephone Number: Address: mick.wallis@outokumpu.com Plant Contact Information 4. Plant Contact: Ronald Denton Title: Manager, Environmental Services Telephone Number: (251) Address: ronald.denton@outokumpu.com 5. UTM Coordinates: 405,462 m E-W 3,446,821 m N-S ADEM Form 103

55 6. Permit application is made for: Existing source (initial application) Modification New source (to be constructed) Change of ownership Change of location Other (specify) Existing source (permit renewal) If application is being made to construct or modify, please provide the name and address of installer or contractor Date construction/modification to begin Telephone 7. Permit application is being made to obtain the following type permit: Air permit Major source operating permit Synthetic minor source operating permit General permit to be completed 8. Indicate the number of each of the following forms attached and made a part of this application: (if a form does not apply to your operation indicates "N/A" in the space opposite the form). Multiple forms may be used as required. 0 ADEM INDIRECT HEATING EQUIPMENT 1 ADEM MANUFACTURING OR PROCESSING OPERATION 0 ADEM REFUSE HANDLING, DISPOSAL, AND INCINERATION 0 ADEM STATIONARY INTERNAL COMBUSTION ENGINES 0 ADEM LOADING, STORAGE & DISPENSING LIQUID & GASEOUS ORGANIC COMPOUNDS 0 ADEM VOLATILE ORGANIC COMPOUND SURFACE COATING EMISSION SOURCES 1 ADEM AIR POLLUTION CONTROL DEVICE 0 ADEM SOLVENT METAL CLEANING 0 ADEM CONTINUOUS EMISSION MONITORS 1 ADEM COMPLIANCE SCHEDULE 9. General nature of business: (describe and list appropriate standard industrial classification (SIC) and North American Industry Classification System (NAICS) ( code(s)) 1 : SIC = 3312 Steel Works, Blast Furnaces and Rolling Mills SIC = 3316 Cold Rolled Steel Sheet, Strip, Bars SIC = 3471 Electroplating, Plating, Polishing, Anodizing, and Coloring NAICS = Stainless Steel 1 SIC and NAICS codes were obtained from Outokumpu s Change of Ownership Permit Application Submitted to ADEM on January 14, ADEM Form 103

56 10. For those making application for a synthetic minor or major source operating permit, please summarize each pollutant emitted and the emission rate for the pollutant. Indicate those pollutants for which the facility is major. Regulated pollutant Potential Emissions * (1) (tons/year) Major source? yes/no Nitrogen Oxides (NOX) 1,636.8 Yes Sulfur Oxides (SO2) Yes * Potential emissions are either the maximum allowed by the regulations or by permit, or, if there is no regulatory limit, it is the emissions that occur from continuous operation at maximum capacity. (1) Facility-wide emissions are shown exclusively for NO x and SO 2 as these are the only pollutants impacted by the proposed modifications outlined in this air permit application. 11. For those applying for a major source operating permit, indicate the compliance status by program for each emission unit or source and the method used to determine compliance. Also cite the specific applicable requirement. Emission unit or source: Meltshop Baghouse Associated with the EAF (LO-1) Emission Point No. Method used to Pollutant 4 Standard 5 Program 1 determine compliance Compliance Status IN 2 OUT 3 LO-1 NOx lb/hr (126 lb/hr proposed emission limit) PSD BACT Limit Test Method 7E: Average of 3 Test Runs X 18.9 lb/hr LO-1 SO2 (47.25 lb/hr proposed emission limit) PSD BACT Limit CEMS: 3 Hour Average X (Intermittent) 1 PSD, non-attainment NSR, NSPS, NESHAP (40 CFR Part 61), NESHAP (40 CFR Part 63), accidental release (112(r)),SIP regulation, Title IV, Enhanced Monitoring, Title VI, Other (specify) 2 Attach compliance plan 3 Attach compliance schedule (ADEM Form-437) 4 Fugitive emissions must be included as separate entries 5 Only NOx and SO2 emissions were affected by this application and are represented above. ADEM Form 103

57 12. List all insignificant activities and the basis for listing them as such (i.e., less than the insignificant activity thresholds or on the list of insignificant activities). Attach any documentation needed, such as calculations. No unit subject to an NSPS, NESHAP or MACT standard can be listed as insignificant 1. Insignificant Activity Basis 1 Facility wide insignificant activities as represented in Title V permit application submitted to ADEM on June 27, 2007 were not affected by the requested modifications presented in this application. As such, this table is left blank. ADEM Form 103 Page 4 of 5

58

59 PERMIT APPLICATION FOR MANUFACTURING OR PROCESSING OPERATION - - Do not write in this space 1. Name of firm or organization: Outokumpu Stainless USA, LLC 2. Briefly describe the operation of this unit or process in your facility: (separate forms are to be submitted for each type of process or for multiple units of one process type. If the unit or process receives input material from, or provides input material to, another operation, please indicate the relationship between the operations.) An application should be completed for each alternative operating scenario. The production of stainless steel starts with the processing of stainless steel scrap and various alloys in an electric arc furnace (EAF). Numerous additives are also included in the melt such as lime to remove impurities. To produce stainless steel, elemental carbon and sulfur must be removed from the molten steel. This is accomplished in the Argon-Oxygen-Decarburization (AOD) vessel. Adjustment of the final composition of specific grades of stainless steels is accomplished by adding chromium, nickel, and other alloying agents in ladle metallurgical stations (LMS). Once the melt has attained the specified composition at the ladles (which have a preheater), the molten steel is poured into a tundish (with preheaters) which feeds a continuous caster. The caster produces a continuous bar. Water is sprayed on the bar to solidify the steel. A torch then cuts the bar into individual slabs that are taken to a slab grinder before proceeding to a hot strip. Throughout the melting and refining process in the EAF and AOD, slag is produced and recycled in the slag yard. Large pieces of slag must be cut for transport. The refractory linings of the EAF, AOD, LMS, and tundish is recycled for reuse. 3. Type of unit or process (e.g., calcining kiln, cupola furnace): Melt Shop EAF LO-1 Make: Siemens Model: Rated process capacity (manufacturer's or designer's guaranteed maximum) in gallons/hour: Manufactured date: Jan 2009 Proposed installation date: N/A Original installation date (if existing): Jan 2009 Reconstruction or Modification date ( if applicable): N/A 4. Normal operating schedule: Hours per day: 24 Days per week: 7 Weeks per year: 52 Peak production season (if any): ADEM Form 105 Page 1 of 4

60 5. Materials (feed input) used in unit or process (include solid fuel materials used, if any): Process weight Maximum Quantity Material average (lb/hr) tons/year Raw Materials 126 tons/hour 1,100, Total heat input capacity of process heating equipment (exclude fuel used by indirect heating equipment previously described on Form ADEM-104): MMBtu/hr Fuel Heat Content Units Max. % Sulfur Max. % Ash Grade No. [fuel oil only] Supplier [used oil only] Coal Btu/lb Fuel Oil Btu/gal Natural Gas Btu/ft 3 L. P. Gas Btu/ft 3 Wood Btu/lb Other (specify) 7. Products of process or unit: Products Quantity/year Units of production Stainless Slabs 1,100,000 Tons per year 8. For each regulated pollutant, describe any limitations on source operation which affects emissions or any work practice standard (attach additional page if necessary): (1) Per 40 CFR 60 Subpart AAa, the opacity from the EAF and AOD control device is limited to 3%. (2) Per 40 CFR 60 Subpart AAa, the opacity from the melt shop is limited to 6%. (3) Per 40 CFR 60 Subpart AAa, the dust handling system cannot exceed 10% opacity. (4) 40 CFR 60 Appendix B specifies that a COMS must be installed on Baghouse LO-1 (also permit condition 30 of permit number X001). ADEM Form 105 Page 2 of 4

61 9. Is there any emission control equipment on this emission source? Yes No (Where a control device exists, Form ADEM-110 must be completed and attached). 10. Air contaminant emission points: (each point of emission should be listed separately and numbered so that it can be located on the attached flow diagram): Emission Point Height Above Grade (Ft) Stack Base Elevation (Ft) Diameter (Ft) Gas Exit Velocity (Ft/Sec) Volume of Gas Discharged (ACFM) Exit Temperature (ºF) LO , * std temperature is 68ºF 11. Air contaminants emitted: basis of estimate (material balance, stack test, emission factor, etc.) must be clearly indicated on calculations appended to this form. Fugitive emissions must be included and calculations must be appended (1). Potential Emissions Regulatory Emission Limit Emission Basis of (units of standard) Point Pollutants (lb/hr) (Tons/yr) Calculation LO-1 SO BACT lb/hr LO-1 NO x BACT lb/hr (1) Emissions are based on the updated BACT analysis presented in this permit application. Emissions for NO x and SO 2 are exclusively shown as these are the only pollutants affected by the proposed modifications presented in this application. 12. Using a flow diagram: (1) Illustrate input of raw materials, (2) Label production processes, process fuel combustion, process equipment and air pollution control equipment, (3) Illustrate locations of air contaminant release so that emission points under item 10 can be identified. (Attach extra pages as needed) ADEM Form 105 Page 3 of 4

62 Process flow diagram 13. Is this unit or process in compliance with all applicable air pollution rules and regulations? yes no (if "no", a compliance schedule, Form ADEM-437 must be completed and attached.) 14. Does the input material or product from this process or unit contain finely divided materials which could become airborne? yes no 15. If "yes, is this material stored in piles or in some other facility as to make possible the creation of fugitive dust problems? yes no List storage piles or other facility (if any): Type of material Particle size (diameter or screen size) Pile size or facility (average tons) Methods utilized to control fugitive emissions (wetted, covered, etc.) Name of person preparing application: Sarah Backes Signature: Date: June 18, 2014 ADEM Form 105 Page 4 of 4

63 PERMIT APPLICATION FOR AIR POLLUTION CONTROL DEVICE 1. Name of firm or organization Outokumpu Stainless USA, LLC - - Do not write in this space 2. Type of pollution control device: (if more than one, check each; however, separate forms are to be submitted for each specific device.) Settling chamber Afterburner Cyclone Absorber Condenser Wet scrubber (kind): Stage 1 - Vapor balance (type) Other (describe): Electrostatic precipitator Baghouse Multiclone Adsorber 3. Control device manufacturer's information: Wet Suppression Name of manufacturer Siemens Model no. 4. Emission source to which device is installed or is to be installed: EAF and Raw Material Handling and Storage - LO (LO1) 5. Emission parameters: Pollutant #1 Pollutant #2 Pollutants removed PM Mass emission rate (#/hr) Uncontrolled 9,750 Designed 9.75 Manufacturer's guaranteed Mass emission rate (units of the Standard) Required by regulation Manufacturer's guaranteed Removal efficiency (%) gr/dscf gr/dscf gr/dscf Designed Manufacturer's guaranteed The designed removal efficiency is based off the EPA PM Calculator s value which was used to estimate emission calculations. Refer to emission calculations submitted in Exhibit A of the permit application submitted June 27, 2007 and Air Permit No X001 for further details. ADEM Form 110 Page 1 of 3

64 6. Gas conditions: Inlet Intermediate Locations Outlet Volume (SDCFM, 68ºf, 29.92" hg) 866, ,951 (ACFM, existing conditions) 883, ,000 Temperature (ºF) Velocity (ft/sec) Percent moisture - - Pressure drop (inches H 2 0) Stack dimensions: Height above grade 200 (feet) Inside diameter at exit (feet) Base Elevation 48.9 (feet) 8. Draw a flow diagram which includes gas exit from process, each control device, location of by-pass, fan or blower, each emission point, exits for collected pollutants, and location of sampling ports. 9. Enclosed are: Blueprints Manufacturer's literature Emissions test of existing installation Other Particle size distribution report Size-efficiency curves Fan curves 10. If the pollution control device is of unusual design, please provide a sketch of the device. 11. List below the important operating parameters for the device. (For example: air/cloth ratio and fabric type, weight, and weave for baghouse; throat velocity and water use rate for a venturi scrubber; etc.) Filter media and pressure drop to be specified 12. By-pass (if any) is to be used when: N/A ADEM Form 110 Page 2 of 3

65 13. Disposal of collected air pollutants: Solid waste Solid waste Liquid waste Liquid waste Volume Composition Is waste hazardous? Method of disposal Final destination Unknown Metal Dust No Off Site Land Fill If collected air pollutants are recycled, describe: Name of person preparing application Sarah Backes Signature Date June 18, 2014 ADEM Form 110 Page 3 of 3

66 PERMIT APPLICATION FOR COMPLIANCE SCHEDULE - - Do not write in this space 1. Name of firm or organization: Outokumpu Stainless USA, LLC 2. Compliance schedule for: Electric Arc Furnace (LO-1) 3. Compliance schedule (include schedule of remedial measures leading to compliance) and schedule for submittal of progress reports (must be at least once every six months): In response to NO x and SO 2 exceedances on the baghouse associated with Outokumpu s EAF (LO-1), Outokumpu discussed a reconciliatory permitting strategy with ADEM during a meeting on June 7, Following the meeting, Outokumpu developed this permit modification application including a full updated BACT analysis and air quality modeling for the EAF. In this permit application, Outokumpu has proposed revised NO x and SO 2 BACT limits for the EAF. It is expected that Outokumpu will achieve compliance with the standards upon the expected date of the modified PSD permit issuance; Outokumpu anticipates this date to be in October In the meantime, Outokumpu operates the EAF in an efficient manner to minimize NO x and SO 2 emissions. Refer to the permit application for details on methods used by Outokumpu to minimize NO x and SO 2 emissions. Outokumpo will submit progress reports at least once every six months. This application is intended to serve as the first progress report. The next progress report is therefore due by November 27, Describe method(s) to be used to determine compliance: Stack Testing for NO x emissions SO 2 CEMS for SO 2 emissions monitoring 5. Date by which item will be in complete compliance with all applicable air pollution control rules and regulations: October 2014 month/day/year Name of person preparing schedule: Sarah Backes Signature: Date: June 18, 2014 ADEM Form 437 1/06 m1 Page 1 of 1

67 Air Quality Modeling Analysis Report Appendix C Project No Outokumpu Stainless USA, LLC Calvert, Alabama Environmental Resources Management-Southwest, Inc. 775 University Blvd. North, Suite 280 Mobile, Alabama (251)

68 Air Dispersion Modeling Assessment For a PSD Permit Modification at Outokumpu Stainless USA, LLC June 2014

69 Outokumpu Stainless USA, LLC Air Dispersion Modeling Assessment for a PSD Permit Modification June 2014 Calvert, Alabama Ramesh Narasimhan Partner-In-Charge Deepu Dethan Project Manager Environmental Resources Management Southwest Inc. 775 N. University Blvd., Suite 280 Mobile, Alabama T: F:

70 TABLE OF CONTENTS 1.0 INTRODUCTION PROJECT OVERVIEW PSD AIR QUALITY MODELING ANALYSES AIR QUALITY DISPERSION MODEL RECEPTORS AND TERRAIN BUILDING DOWNWASH REPRESENTATIVE METEOROLOGICAL DATA OUTOKUMPO SOURCE INVENTORY SIGNIFICANT IMPACT ANALYSIS SO 2 PRELIMINARY IMPACT ANALYSIS NO 2 PRELIMINARY IMPACT ANALYSIS NO to NO 2 Conversion Significance Analysis Results CONCLUSIONS OF PRELIMINARY IMPACT ANALYSIS FULL IMPACT ANALYSIS SIGNIFICANT IMPACT AREA DEVELOPMENT REPRESENTATIVE BACKGROUND CONCENTRATIONS OFF-SITE INVENTORY HOUR NO 2 NAAQS MODELING RESULTS CLASS I ASSESSMENT NON-APPLICABILITY ADDITIONAL AIR QUALITY IMPACT ANALYSIS 18 LIST OF TABLES 8.1 ADDITIONAL GROWTH ANALYSIS SOILS, VEGETATION, AND WILDLIFE ANALYSIS IMPACT ON VISIBILITY PSD Applicability Determination Summary 2-2 Summary of Short Term Emission Rate Changes 3-1 Ambient Air Quality Standards 4-1 Downwash Structure Parameters 4-2 SIL Analysis On-Site Source Parameters 4-3 NAAQS Analysis On-Site Source Parameters 5-1 Summary of H1H Annual SO 2 Impacts Environmental Resources Management Southwest, Inc. ii 2014\ \841rpt.doc

71 5-2 Summary of H1H 24-Hour SO 2 Impacts 5-3 Summary of H1H 3-Hour SO 2 Impacts 5-4 Summary of H1H 1-Hour SO 2 Impacts 5-5 Summary of Site Specific ISR 5-6 Summary of H1H Annual NO 2 Impacts 5-7 Summary of H1H 1-Hour NO 2 Impacts 6-1 NO 2 1-Hour NAAQS Modeling Results Summary 8-1 Predicted Air Quality Impacts Compared to NO 2 Vegetation Impact Thresholds APPENDIX A: FIGURES A-1 Site Aerial Photograph A-2 Receptor Grid for Ambient Air Quality Impact Analyses A-3 Site Plot Plan A-4 1-Hour NO 2 Significant Impact Area APPENDIX B: ELECTRONIC MODELING FILES APPENDIX C: FEDERAL LAND MANAGER CORRESPONDENCE Environmental Resources Management Southwest, Inc. iii 2014\ \841rpt.doc

72 1.0 INTRODUCTION Outokumpu Stainless USA, LLC (Outokumpu) owns and operates a stainless steel mill in Calvert, Alabama. The facility was previously owned and operated by ThyssenKrupp Stainless USA, LLC (TKL). TKL submitted Prevention of Significant Deterioration (PSD) permit applications for the stainless steel mill and obtained construction authorizations via PSD permits issued by the Alabama Department of Environmental Management (ADEM). The first operation of certain sources at the facility commenced in June 2010 under Temporary Authorizations to Operate (TAOs) issued by ADEM. As per Alabama Administrative Code (AAC) (1), an initial Title V operating permit application was submitted within 12 months after the commencement of operations; the Title V permit has not yet been issued. Outokumpu acquired the facility in January 2013, and procured the aforementioned permits formerly associated with TKL. Operations for the melt shop were initially permitted by ADEM on March 25, 2010 (Permit Number X001 under ThyssenKrupp Stainless Steel USA, LLC Ownership). On September 6, 2012, ADEM issued an amended air permit to reflect changes in particulate emissions from the melt shop (Permit Number X001). Most recently, on March 21, 2013 ADEM re-issued the melt shop permit to reflect the new facility name and ownership (Permit X001 under Outokumpu Stainless USA, LLC ownership). Condition sixteen of Air Permit X001 establishes a Nitrogen Oxides (NO X) emission limit of 0.35 pound per ton (lb/ton) and a Sulfur Dioxide (SO 2) emission limit of 0.15 lb/ton for the baghouse associated with the Electric Arc Furnace (EAF)(Emission Point ID: LO-1). Outokumpu is proposing to modify the NO X and SO 2 emissions limits established by this permit as follows: 1. Increase NO X Emission Limit from lb/hr to 126 lb/hr. 2. Increase SO 2 Emission Limit from lb/hr to lb/hr. Because existing PSD emission limits are being modified and due to the resulting increase in maximum potential emissions, the proposed modification will be subject to PSD review for NO X and SO 2. PSD review requires the applicant to perform ambient air quality impact analyses to demonstrate compliance with the PSD increments and national ambient air quality standards (NAAQS). This report documents the required air dispersion modeling analysis that was performed for NO X and SO 2. Environmental Resources Management Southwest, Inc \ \841rpt.doc

73 2.0 PROJECT OVERVIEW The Outokumpu facility is located in Mobile County, which is currently designated by EPA as attainment or unclassifiable relative to applicable NAAQS. An area map of the existing facility and surrounding area is included as Figure A-1 of this report. As described in Section 1.0, Outokumpu is proposing to modify the NO X and SO 2 Best Available Control Technology (BACT) emission limits on the EAF (LO-1). The proposed emission increases exceed the PSD significant emission rates (SER) and therefore trigger PSD review for both NO X and SO 2. Based on maximum potential emissions, the proposed modification will be subject to PSD review for NO X and SO 2. The annualized rates below are based on the existing permitted capacity of 1,100,000 tons per year of steel throughput through the EAF. Table 2-1 contains a summary of the PSD Applicability Determination. TABLE 2-1: PSD Applicability Determination Summary EMISSION RATE (tpy) SO 2 NO X Existing Potential Emissions Proposed Potential Emissions Emission Increase a PSD SER Subject to PSD Review Yes Yes a. As per discussions with ADEM, the increase in potential emissions of the EAF was used for this evaluation. The PSD modeling analyses were performed in a manner that generally conforms to the following EPA and the ADEM modeling regulations and guidance documents. Guideline on Air Quality Models 40 CFR Part 51, Appendix W, Revised November 9, AERMOD Implementation Guide, Revised March 19, PSD Air Quality Analysis AERMOD Modeling Guidelines, ADEM, Air Division, Planning Branch, Meteorological Section, Revised May, The remainder of this report describes the modeling techniques and data resources Outokumpu used to meet PSD air quality analysis requirements and demonstrate compliance with NO 2 and SO 2 air quality standards. Table 2-2 summarizes the maximum hourly emission rates based on the existing permitted capacity of 1,100,000 tons per year of steel throughput to the EAF, 8,760 hours of annual operation and 126 tons per hour of steel throughput to the EAF. Outokumpu is not proposing any change in throughput on either a short term or annual basis. Environmental Resources Management Southwest, Inc \ \841rpt.doc

74 TABLE 2-2: Summary of Short Term Emission Rate Changes Electric Arc Furnace (EAF) EMISSION RATE Model ID: L01 SO 2 NO X Existing Permit Limit (lb/hr) Proposed Permit Limit (lb/hr) Increase (lb/hr) a Increase (g/s) a a. As per discussions with ADEM, the increase in potential emissions of the EAF was used for this evaluation. Environmental Resources Management Southwest, Inc \ \841rpt.doc

75 3.0 PSD AIR QUALITY MODELING ANALYSES Air quality impact analyses to support the proposed modifications to the permit limits was performed to demonstrate compliance with the 1-hour and annual NO 2, 1-hour SO 2, and the 3-hour SO 2 NAAQS as well as 24-hour PSD increment standard for SO 2 and the annual PSD increment standards for SO 2 and NO 2. Table 3-1 summarizes the NO 2 and SO 2 ambient air quality standards. TABLE 3-1: Ambient Air Quality Standards Pollutant Averaging Period SIL (μg/m 3 ) SMC (μg/m 3 ) NAAQS (μg/m 3 ) Class II Increment (μg/m 3 ) NO 2 Annual Hour Annual 1 -- Revoked Hour 5 13 Revoked 91 SO 2 3-Hour Hour The Significant Impact Level (SIL) analysis was performed by modeling the increase in maximum hourly SO 2 and NO 2 emission rates from proposed permit modifications to compute the maximum ambient impact among all receptors and meteorological data years for comparison to the SIL. Under the ADEM and EPA guidance, demonstrating ambient impacts less than the SIL is sufficient to conclude that the proposed project would not cause or contribute to an exceedance of the applicable NAAQS for NO 2 and SO 2. Environmental Resources Management Southwest, Inc \ \841rpt.doc

76 4.0 AIR QUALITY DISPERSION MODEL All modeling was performed using the regulatory version (12345) of EPA s American Meteorological Society/Environmental Protection Agency Regulatory Model (AERMOD). AERMOD is a steady-state plume dispersion model that is recommended by EPA s Guideline on Air Quality Models. 1 All SO 2 modeling analyses were performed using regulatory default options. The NO 2 modeling analyses were performed using the Tier 3 non-default Plume Volume Molar Ratio Method (PVMRM) option. 4.1 RECEPTORS AND TERRAIN For the preliminary impact analysis, a receptor grid was developed that extends to approximately 10 kilometers (km) from the center of the facility in each direction. The following receptor spacing was used: meter (m) spacing along the mill fence line; m spacing from fence line to 5,000 m; m spacing from 5,000 m to 7,000 m; and m spacing from 7,000 m to 10,000 m. Terrain heights for each receptor point were derived from the latest National Elevation Data (NED) obtained from the U.S. Geological Survey. Elevations for buildings and structures were based on information provided by the mill. Figure A-2 in Appendix A illustrates the location and extent of receptors used in this analysis. The land use around the facility is determined to be rural. 4.2 BUILDING DOWNWASH AERMOD allows for simulation of multiple sources (and source types) simultaneously, while making the correct accounting for building downwash and building cavity effects. Emission sources at the Outokumpu facility were evaluated in terms of their proximity to nearby buildings/structures to determine if stack discharges might become caught in the turbulent wakes of these buildings/structures utilizing EPA s Building Profile Input Program PRIME (BPIP-PRIME, version 04274). The direction specific building dimensions are calculated using BPIP-PRIME. The heights of buildings evaluated for downwash are shown in Table Revisions to the Guideline on Air Quality Models: Adoption of a Preferred General Purpose (Flat and Complex Terrain) Dispersion Model and Other Revisions, 40 CFR Part 51, November 9, Environmental Resources Management Southwest, Inc \ \841rpt.doc

77 TABLE 4-1: Downwash Structure Parameters Building Model ID Number of Tiers Tier Number Base Elevation (m) Building/Tier Height (m) AODEAF CAPL CAPL LCB CRM CRM HAPL LFB LFB LFB LMS LHRM BLDG BLDG BLDG Figure A-3 in Appendix A illustrates the mill emission sources in relation to building structures considered in the downwash analysis. Environmental Resources Management Southwest, Inc \ \841rpt.doc

78 4.3 REPRESENTATIVE METEOROLOGICAL DATA The meteorological database used for the dispersion model consists of 5 years ( ) of surface observations from the Mobile Regional Airport National Weather Service (NWS) station and upper air data from the Slidell, Louisiana NWS station. The pre-processed, hourly NWS data were processed using AERMET by the ADEM Air Division and provided to Outokumpu on April 22, 2014 as a single set of meteorological data to be used in this air dispersion modeling assessment. 4.4 OUTOKUMPO SOURCE INVENTORY Table 4-2a and Table 4-3 summarize the on-site emission sources included in the analysis and relevant stack parameters (e.g., location, emission rate, stack height, stack diameter, exhaust temperature, and velocity) to simulate each emission point in the dispersion model. All sources at the facility are point sources with vertical exhaust. Source coordinates are represented in Universal Transverse Mercator (UTM) coordinates in Zone 16 referenced North American Datum 1983 (1983). TABLE 4-2: SIL Analysis On-Site Source Parameters Source ID Easting (X) Northing (Y) Base Elevation Stack Height Temp. Exit Velocity Stack Diameter NO 2 SO 2 (m) (m) (m) (m) (K) (m/s) (m) (g/s) (g/s) LO1 406,683 3,447, TABLE 4-3: NAAQS Analysis On-Site Source Parameters Source ID Easting (X) Northing (Y) Base Elevation Stack Height Temp. Exit Velocity Stack Diameter NO 2 (m) (m) (m) (m) (K) (m/s) (m) (g/s) LO1N 406,683 3,447, LA1N 406,393 3,447, LO2N 406,674 3,447, LA2N 406,383 3,447, LO26N 405,908 3,447, LO27N 405,902 3,447, LA21N 406,316 3,447, LA24N 406,184 3,447, LA25N 406,153 3,447, LA26N 406,116 3,447, LA27N 406,077 3,447, LA29N 406,018 3,446, LO41A 406,051 3,446, LO41B 406,043 3,446, LO42N 406,550 3,447, Environmental Resources Management Southwest, Inc \ \841rpt.doc

79 Source ID Easting (X) Northing (Y) Base Elevation Stack Height Temp. Exit Velocity Stack Diameter NO 2 (m) (m) (m) (m) (K) (m/s) (m) (g/s) LO43N 405,833 3,447, LA43N 405,934 3,447, LO47N 406,074 3,447, LA47N 406,172 3,447, LO53N 405,808 3,447, LA53N 405,762 3,447, LO57N 405,682 3,447, LA57N 405,645 3,447, LA63N 405,996 3,447, LA64N 405,948 3,446, LO72N 405,921 3,447, LA72N 405,800 3,447, Environmental Resources Management Southwest, Inc \ \841rpt.doc

80 5.0 SIGNIFICANT IMPACT ANALYSIS This section describes the preliminary impact analyses conducted for SO 2 and NO 2. The procedure is as follows. The highest first high (H1H) impacts due to emission sources associated with the proposed project are estimated through modeling and compared with the SILs for all applicable pollutants. For the NO 2 1-hour and SO 2 1-hour NAAQS, the modeled H1H is averaged over five years of meteorology for the purpose of comparison to the SIL. For all other NAAQS and PSD increment averaging periods included in this demonstration, the H1H concentration modeled for all receptors for each meteorological data year simulated is computed and compared to the SIL. If the H1H impacts from the project source equal or exceed the SIL for a given pollutant, the significant impact area (SIA) for that pollutant and averaging period is calculated. The SIA for each pollutant and averaging period is determined by calculating the maximum distance at which impacts are greater than the SIL. Exceedances of the SIL triggers NAAQS and PSD increment analyses within the SIA. The NAAQS and PSD Increment analyses are carried out by modeling all receptors at which Outokumpu contributes significant impacts (i.e., impacts greater than the respective SILs). For NAAQS analyses, impacts due to off-site sources as well as ambient air monitored background concentrations are added to modeled concentrations for comparison with the NAAQS. Detailed discussions of each analysis are included in the following sections on a pollutant by pollutant basis. 5.1 SO 2 PRELIMINARY IMPACT ANALYSIS The SO 2 preliminary impact analysis evaluates the project increase of SO 2 emissions from the modified EAF as identified in Section 2.0 of this report, Table 2-2. The H1H concentration is estimated through modeling and compared to the PSD SIL for the 1-hour, 3-hour, 24-hour, and annual averaging periods. Annual SO 2 SIL Analysis: Table 5-1 shows the result of annual average SO 2 SIL analysis. TABLE 5-1: Summary of H1H Annual SO 2 Impacts CY SO 2 H1H Impact (μg/m 3 ) SIL (μg/m 3 ) SIL Exceeded (Yes/No) SMC (μg/m 3 ) Representative Air Quality Data Must Be Obtained (Yes/No) No N/A N/A No N/A N/A No N/A N/A No N/A N/A No N/A N/A Environmental Resources Management Southwest, Inc \ \841rpt.doc

81 As demonstrated in Table 5-1, the maximum H1H modeled impact of SO 2 over the five year period is less than the SIL. Each year over the five year period was modeled separately. Therefore, a full impact analysis is not required and compliance with PSD increment is demonstrated. There is not currently a SMC or NAAQS set for the SO 2 annual averaging period. 24-Hour SO 2 SIL Analysis: Table 5-2 shows the result of 24-hour average SO 2 SIL analysis. TABLE 5-2: Summary of H1H 24-Hour SO 2 Impacts CY SO 2 H1H Impact (μg/m 3 ) SIL (μg/m 3 ) SIL Exceeded (Yes/No) SMC (μg/m 3 ) Representative Air Quality Data Must Be Obtained (Yes/No) No 13 No No 13 No No 13 No No 13 No No 13 No As demonstrated in Table 5-2, the maximum H1H modeled impact of SO 2 over the five year period is less than the SIL. Each year over the five year period was modeled separately. Therefore, a full impact analysis is not required and compliance with PSD increment is demonstrated and no monitoring is required. There is not currently a NAAQS set for the SO 2 24-hour averaging period. 3-Hour SO 2 SIL Analysis: Table 5-3 shows the result of 3-hour average SO 2 SIL analysis. TABLE 5-3: Summary of H1H 3-Hour SO 2 Impacts CY SO 2 H1H Impact (μg/m 3 ) SIL (μg/m 3 ) SIL Exceeded (Yes/No) SMC (μg/m 3 ) Representative Air Quality Data Must Be Obtained (Yes/No) No N/A N/A No N/A N/A No N/A N/A No N/A N/A No N/A N/A As demonstrated in Table 5-3, the maximum H1H modeled impact of SO 2 over the five year period is less than the SIL. Each year over the five year period was modeled separately. Therefore, a full impact analysis is not required and compliance with NAAQS and PSD increment is demonstrated. There is not currently a SMC set for the SO 2 3-hour averaging period. Environmental Resources Management Southwest, Inc \ \841rpt.doc

82 1-Hour SO 2 SIL Analysis: The SO 2 1-hour averaging period concentration represents the H1H impact averaged over five years. Table 5-4 shows the results of 1-hour SO 2 SIL analysis. TABLE 5-4: Summary of H1H 1-Hour SO 2 Impacts SO 2 H1H Impact (μg/m 3 ) SIL (μg/m 3 ) SIA Analysis Required (Yes/No) SMC (μg/m 3 ) Representative Air Quality Data Must Be Obtained (Yes/No) No N/A N/A As demonstrated in the Table 5-4, the SO 2 H1H, 1-hour modeled impact averaged over the five-year period is less than the SIL. Therefore, a full impact analysis is not required and compliance with NAAQS is demonstrated. There is not currently a SMC or PSD Increment set for the SO 2 1-hour averaging period. 5.2 NO 2 PRELIMINARY IMPACT ANALYSIS The NO 2 preliminary impact analysis evaluates the project increases of NO 2 emissions from the EAF as identified in Section 2.0 of this report, Table 2-2. The H1H concentration is estimated through modeling and compared to the PSD SIL for both the 1-hour and annual averaging periods NO to NO 2 Conversion The NO 2 NAAQS compliance demonstration follows the Tier 3 screening approach recommended by USEPA to obtain 1-hour estimates of NO 2 from point sources. The Tier 3 screening approach is applied in the preliminary determination of the significant impact area and for all required full impact analyses. The Tier 3 screening approaches, Ozone-Limiting Method (OLM) and Plume Volume Molar Ratio Method (PVMRM) included as non-default options in the AERMOD dispersion model, are currently considered to be detailed screening methods under Tier 3. Outokumpu used the PVMRM method as the refined approach to account for the NO to NO 2 conversion. The use of PVMRM options in AERMOD requires the specification of an in-stack ratio (ISR) of NO 2/NO X for each source. Outokumpu used stack testing results completed for onsite sources, inclusive of the EAF, to determine the proper ISR for sources where such data are available. For sources without more appropriate sourcespecific information, Outokumpu used the default ISR of 0.5 as dictated in the EPA memorandum titled Additional Clarification Regarding Application of Appendix W Modeling Guidance for the 1-hour NO 2,National Ambient Air Quality Standard (March 2011). The following table demonstrates the modeled site-specific ISR in both the Outokumpu preliminary and full impact analyses. Environmental Resources Management Southwest, Inc \ \841rpt.doc

83 TABLE 5-5: Summary of Site Specific ISR SOURCE EAF and Raw Material Handling and Storage LO (Baghouse 1) Model ID NO 2/NO X IN-STACK RATIO LO1N 0.15 Boiler No. 1 LO26N 0.15 Boiler No. 2 LO27N 0.15 Passive Annealing (1) LO41A 0.10 Passive Annealing (2) LO41B 0.10 Hot Annealing and Pickling Line - LO (HAPL1) - Annealing Furnace Hot Annealing and Pickling Line - LA (HAPL2) - Annealing Furnace Cold Annealing and Pickling Line - LO (CAPL1) - Annealing Furnace Cold Annealing and Pickling Line - LA (CAPL2) - Annealing Furnace LO43N 0.10 LA43N 0.10 LO53N 0.10 LA53N 0.10 The facility has conducted stack testing on a number of on-site sources that was used to calculate specific NO 2/NO X ratios for the purpose of this modeling exercise. For the EAF, the stack testing data shows an in-stack ratio of There was also testing conducted on the Annealing Furnace (LO43) on the Hot Annealing and Pickling Line (HAPL). There is a second HAPL annealing furnace (LA43), two Annealing Furnaces (LO53, LA53) on the Cold Annealing and Pickling Line (CAPL), and two Passive Annealing Furnaces (LO41A, LO41B) with the same operating characteristics as the HAPL annealing furnace and thus the stack tested NO 2/NO X ratio of 0.10 was used. Testing conducted on the two natural gas boilers (LO26, LO27) showed an NO 2/NO X ratio of This ratio is also consistent with NO 2/NO X ratios from similarly sized natural gas boilers as found in the latest version of US EPA s NO 2/NO X ISR Database 2. When using the PVMRM for estimating the 1-hour NO 2 concentrations, ambient ozone concentrations from five ozone monitoring stations during the period modeled in the PSD analysis were used to represent hourly background ozone levels. Complete hourly data from the Pensacola, FL station (AQS # ) and the Ellyson Industrial Park station (AQS # ) served as the baseline data, with incomplete hourly ozone data from three Alabama monitors; Chickasaw site (AQS # ), Fairhope site (AQS # ), and Mobile - Bay Road site (AQS # ) being used to fill data gaps. The Alabama ozone data is considered incomplete because the available data for 2 Environmental Resources Management Southwest, Inc \ \841rpt.doc

84 these monitors covers only the summer months. The baseline data was evaluated for completeness, with gaps in the data being filled by the available Alabama data based on proximity to the facility. These hourly ozone data were used within the PVMRM option of AERMOD to simulate the atmospheric chemistry of ozone reacting with nitric oxide (NO) emitted from the stack to form NO 2. The model disperses the initial NO X emissions (which are mostly NO) each hour of the day over each year of meteorological data (e.g., 8,760 hours in a 365-day year). AERMOD allows replacement of any missing data in the ozone data file using a single value or the import of an ozone data file where the missing hourly ozone concentrations have already been filled. To more realistically predict downwind NO 2 concentrations based on the ozone diurnal pattern, missing hours of ambient ozone concentration were filled using the hourly background ozone data according to the aforementioned procedures Significance Analysis Results Annual NO 2 SIL Analysis: Table 5-6 shows the result of annual average NO 2 SIL analysis. TABLE 5-6: Summary of H1H Annual NO 2 Impacts CY NO 2 H1H Impact using PVMRM (μg/m 3 ) SIL (μg/m 3 ) SIL Exceeded (Yes/No) SMC (μg/m 3 ) Representative Air Quality Data Must Be Obtained (Yes/No) No 14 No No 14 No No 14 No No 14 No No 14 No As demonstrated in Table 5-6, the maximum H1H modeled impact of NO 2 over the five year period is less than the SIL and SMC. Each year over the five year period was modeled separately. Therefore, a full impact analysis for annual averaging period for NO 2 is not required and compliance with NAAQS and PSD increment is demonstrated. 1-Hour NO 2 SIL Analysis: The NO 2 1-hour averaging period concentration represents the H1H impact averaged over five years. The modeled impacts are then compared to EPA s interim SIL of 4 ppb to determine whether further analysis is required. This value equates to a concentration of 7.5 µg/m 3 at 25 C and 760 mmhg. Table 5-7 shows the results of 1-hour NO 2 SIL analysis. Environmental Resources Management Southwest, Inc \ \841rpt.doc

85 TABLE 5-7: Summary of H1H 1-Hour NO 2 Impacts NO 2 H1H Impact using PVMRM (μg/m 3 ) SIL (μg/m 3 ) SIA Analysis Required (Yes/No) SMC (μg/m 3 ) Representative Air Quality Data Must Be Obtained (Yes/No) Yes N/A N/A As demonstrated in the Table 5-7, the NO 2 H1H, 1-hour modeled impact averaged over the five-year period is greater than the SIL. Therefore, a full impact analysis is required to demonstrate compliance with the NAAQS for the 1-hour averaging period for NO 2. There is currently no SMC or PSD increment set for the NO 2 1-hour averaging period. 5.3 CONCLUSIONS OF PRELIMINARY IMPACT ANALYSIS The preliminary impact analysis demonstrated that a full impact analysis was required for the 1-hour NO 2 averaging period only. The results of this analysis are presented in the next section. Environmental Resources Management Southwest, Inc \ \841rpt.doc

86 6.0 FULL IMPACT ANALYSIS This section describes the methodology and results of the full impact analysis for the 1-hour NO 2 averaging period. 6.1 SIGNIFICANT IMPACT AREA DEVELOPMENT The significant impact area is defined as the area in which predicted concentrations, due to the proposed modification, exceed specified significant impact levels on a pollutant-specific basis. Only the 1-hour NO 2 averaging period resulted in modeled significant impacts. Therefore, the significant impact area analysis was limited to the 1-hour averaging period for NO 2. For the NO 2 1-hour preliminary model, the conversion of NO to NO 2 was evaluated for significant impact area determination using the Tier 3 calculation methodology as per USEPA s aforementioned March 1, 2011 memorandum. Figure A-4 in Appendix A provides a visual representation of the NO 2 1-hour significant impact receptors, with the most distant impacted receptor being located 4 km from the facility centroid. It is important to note that, as shown in the figure, the furthest receptor was not on the outer boundary of the receptor set; as such, extension of the receptor set was not required. The preliminary model results used to calculate the significant impact areas for each pollutant are provided on the accompanying modeling CD in Appendix B of this report. 6.2 REPRESENTATIVE BACKGROUND CONCENTRATIONS Representative background concentrations were added to the maximum predicted concentrations due to major emission sources for comparison with the NAAQS. The background concentration of 24 µg/m 3 for 1-hour NO 2 was provided by ADEM and used in the NAAQS modeling. 6.3 OFF-SITE INVENTORY An inventory of off-site NO 2 sources was requested and received from ADEM. The maximum predicted concentration from the dispersion modeling analysis (onsite sources + off-site inventory) is added to the representative background concentration and compared with NAAQS HOUR NO 2 NAAQS MODELING RESULTS A NAAQS modeling analysis using five years of meteorological data is performed for the NO 2 1-hour averaging period. The NAAQS analysis is carried out by modeling the Outokumpu source parameters and emission rates; modeling an off-site source inventory as provided by ADEM; and adding the background concentration to modeled concentrations for comparison to the Environmental Resources Management Southwest, Inc \ \841rpt.doc

87 NAAQS. The NO 2 1-hour design value is represented as the highest 8 th high (H8H) modeled impact over the concatenated five-year period. Table 6-1 summarizes the results of the NAAQS assessment. TABLE 6-1: NO 2 1-Hour NAAQS Modeling Results Summary H8H Model Predicted Concentration with PVMRM (µg/m 3 ) Background Value (µg/m 3 ) Total Concentration (µg/m 3 ) NAAQS (µg/m 3 ) Less than NAAQS (Yes/No) Yes The result demonstrates that the Outokumpu facility is in compliance with the applicable NO 2 1-hour NAAQS at the NO 2 emission limit proposed in this PSD permit modification application. As such, the facility will not cause or contribute to any violations of the NAAQS. There is no PSD Increment set for the 1-hour NO 2 averaging period. Therefore, no further review is required. Environmental Resources Management Southwest, Inc \ \841rpt.doc

88 7.0 CLASS I ASSESSMENT NON-APPLICABILITY The nearest PSD Class I area is the Breton Wildlife Refuge, located approximately 130 km south-southwest of the mill. Because the Class I area is greater than 100 km from Outokumpu, ADEM does not require Outokumpu to address the modification s impact on PSD Increment at the Class I Area. The Federal Land Managers use the following approach to determine whether a PSD project should provide detailed dispersion modeling impact analyses for air quality related values (AQRVs): [SO 2 + NO X + PM 10 emissions (tpy)]/ distance (km) > 10 Using the emissions increases associated with the modification, we evaluated the above equation and the corresponding results are (124 tpy of SO tpy of NO X + 0 tpy of PM 10)/130 km = Because this factor is well below the screening level of 10, AQRV analyses for the Class I area are not required. The Federal Land Manager (U.S. Fish and Wildlife Service) of the Class I area was contacted and provided with details of the proposed modification. The FLM response confirming an AQRV modeling analysis is not required is included in Appendix C. Environmental Resources Management Southwest, Inc \ \841rpt.doc

89 8.0 ADDITIONAL AIR QUALITY IMPACT ANALYSIS A qualitative assessment of the impacts on general growth, soil, and vegetation associated with the proposed modification was performed. The additional impact analysis has been conducted to evaluate the following: Analysis of additional growth associated with the proposed modification, Analysis of potential detrimental effects to soils, and Analysis of potential detrimental effects to vegetation with economic or recreational value. 8.1 ADDITIONAL GROWTH ANALYSIS The proposed modification to increase EAF BACT emission limits will not involve an expansion of plant activities. The proposed project will not result in increased residential or commercial growth in the area. Outokumpu will continue to follow the current rate of hiring from the existing workforce in the local area. Therefore, there will be no increase in emissions due to growth and conducting additional modeling is not required. 8.2 SOILS, VEGETATION, AND WILDLIFE ANALYSIS Both direct and indirect (formed in the atmosphere) air emissions could have impacts on soils, vegetation, and wildlife surrounding a new development or modified existing source. Nitrate formed due to chemical transformation of NO X from the project could potentially deposit in the soil surrounding the facility. As the impacts from the proposed modification will be less than all NAAQS, which are intended to protect human health and are more stringent than standards intended to protect soil or vegetation, the project is not expected to have a significant impact on the surrounding soil. Modeled impacts of SO 2 and annual NO 2 are less than the SIL. In addition, the project is not expected to emit any toxic pollutants in significant quantities which could potentially impair surrounding vegetation nor is the project is expected to impact wildlife in the surrounding area. In summary, the project is not expected to result in significant impacts on soil, vegetation or wildlife in the area surrounding the facility as clearly evidenced in the results shown in Tables 8-1. Environmental Resources Management Southwest, Inc \ \841rpt.doc

90 TABLE 8-1: Predicted Air Quality Impacts Compared to NO 2 Vegetation Impact Thresholds NO 2 Averaging Period Predicted Impact (µg/m 3 ) Threshold for Impact to Vegetation (µg/m 3 ) Applicability 1-hour ,000 1 Leaf Injury to Plants Protects all vegetation Annual Metabolic and growth impacts to plants 1 Diagnosing Injury Caused by Air Pollution, EPA , Prepared by Applied Science Associates, Inc. under contract to the Air Pollution Training Institute, Research Triangle Park, North Carolina Secondary National Ambient Air Quality Standard (ug/m 3 ) which is a limit set to avoid damage to vegetation resulting in economic losses in commercial crops, aesthetic damage to cultivated trees, shrubs, and other ornamentals, and reductions in productivity, species richness, and diversity in natural ecosystems to protect public welfare (Section 109 of the Clean Air Act). These thresholds are the most stringent of those found in the literature survey. 3 Air Quality Criteria for Oxides of Nitrogen, EPA/600/8-91/049aF-cF.3v, Office of Health and Environment Assessment, Environmental Criteria and Assessment Office, U.S. Environmental Protection Agency, Research Triangle Park, NC IMPACT ON VISIBILITY Any facility emitting significant amounts of TSP/PM 10 and/or NO X has a potential adverse impact on visibility through atmospheric discoloration or reduction of visual range due to increased haze. The Clean Air Act Amendments of 1977 require evaluation of visibility impairment in the vicinity of PSD Class I areas due to emissions from new/modified air pollution sources. Since the Q/D ratio for this project is less than 10, no visibility analysis was conducted as part of this analysis. A visibility analysis was not required using the VISCREEN model as there are no regional airports or scenic vistas located within the significant impact area of the proposed modification. The closest identified Class II area is Meaher State Park, which is greater than 40 km away and well outside the 1-hour NO 2 SIA. As the increase in emissions associated with the operation of the facility will result in only a very small increase in air quality impacts in a small area surrounding the facility, it is unlikely that the operation of the facility will cause any impairment to visibility at any location. Effects on Class II Visibility A review of land uses within the NO 2 1-hour SIAs verified that there are no airports, state or federal parks, or other land uses that would be impeded due to emissions from the modification. The facility will comply with all applicable federal and state visible emissions regulations. As stated in the permit application, Outokumpu will comply with all opacity limitations. Therefore, no degradation to visibility within the impact area is expected from this project. Environmental Resources Management Southwest, Inc \ \841rpt.doc

91 Impacts on Nonattainment Areas The modification will take place in Mobile County, which is classified as being in attainment for all pollutants. From the results of the analysis, it is clearly demonstrated that the project will not significantly impact any non-attainment area. Environmental Resources Management Southwest, Inc \ \841rpt.doc

92 Appendix A Figures Project No Outokumpu Stainless Steel, USA Calvert, Alabama Environmental Resources Management Southwest, Inc. 775 N. University Blvd., Suite 280 Mobile, Alabama (251)

93 o Miles Source: Esri, DigitalGlobe, GeoEye, i-cubed, USDA, USGS, AEX, Getmapping, Aerogrid, IGN, IGP, swisstopo, and the GIS User Community DESIGN: DATE: O. LEACH 5/19/2014 Environmental Resources Management DRAWN: SCALE: O. LEACH AS SHOWN CHKD.: REVISION: J. SMITH 0.0 FIGURE A-1 SITE AERIAL PHOTOGRAPH OUTOKUMPU STAINLESS USA, LLC CALVERT, ALABAMA Legend Property Boundary FILE: S:\holdtank\Leach\GIS\Outokumpu Modeling\Outokumpu_sources_ mxd

94

95 o LA1N LA2N LA53N LO53N LA57N LO57N LO26N LO27N LO47N LA47N LA21N LO1N LO2N LO72N LO42N LA72N LO43N LA43N LA25N LA24N LA26N LA63N LA27N LA64N LA29N LO41A LO41B Feet Source: Esri, DigitalGlobe, GeoEye, i-cubed, USDA, USGS, AEX, Getmapping, Aerogrid, IGN, IGP, swisstopo, and the GIS User Community DESIGN: DATE: O. LEACH 5/27/2014 Environmental Resources Management DRAWN: SCALE: O. LEACH AS SHOWN CHKD.: REVISION: J. SMITH FIGURE A-3 PLOT PLAN OUTOKUMPU STAINLESS USA, LLC Legend Property Boundary 0.1 CALVERT, ALABAMA Buildings!> Sources FILE: S:\holdtank\Leach\GIS\Outokumpu Modeling\Outokumpu_sources_ mxd

96 o 0 1,000 2,000 Feet Source: Esri, DigitalGlobe, GeoEye, i-cubed, USDA, USGS, AEX, Getmapping, Aerogrid, IGN, IGP, swisstopo, and the GIS User Community 5/27/2014 O. LEACH AS SHOWN S:\holdtank\Leach\GIS\Outokumpu Modeling\Outokumpu_sources_ mxd FILE: DATE: DESIGN: DRAWN: SCALE: CHKD.: REVISION: Environmental Resources Management CALVERT, ALABAMA OUTOKUMPU STAINLESS USA, LLC 1-HOUR NO 2 SIGNIFICANT IMPACT AREA Legend J. SMITH Property Boundary FIGURE A-4 O. LEACH 0.0 Receptors of Significant Impact

97 Appendix B Electronic Modeling Files Project No Outokumpu Stainless Steel, USA Calvert, Alabama Environmental Resources Management Southwest, Inc. 775 N. University Blvd., Suite 280 Mobile, Alabama (251)

98 Appendix C Federal Land Manager Correspondence Project No Outokumpu Stainless Steel, USA Calvert, Alabama Environmental Resources Management Southwest, Inc. 775 N. University Blvd., Suite 280 Mobile, Alabama (251)

99 Brad Arnold From: Sent: To: Cc: Subject: Webster, Jill Tuesday, December 10, :10 PM Brad Arnold Denton, Ronald Ramesh Narasimhan; Nicole Sullivan; Deepu Dethan Outokumpu Stainless USA, LLC Calvert Facility Air Quality Assessment of Breton National Wildlife Refuge Class I Area Mr. Arnold, Thank you for sending the information regarding Outokumpu Stainless USA, located in Calvert, Alabama. Assuming that the emissions provided in your letter dated December 3, 2013 are based on maximum 24 hour potential, and based on the distance to the Breton Wilderness, the Fish and Wildlife Service anticipates that modeling would not show any significant additional impacts to air quality related values (AQRV) at the Class I area. Therefore, we are not requesting that any AQRV analysis be included in the PSD permit application. However, we disagree that no Class I increment analysis is required for projects beyond 100km of the nearest Class I boundary. We highly encourage you to contact the EPA Region IV office regarding increment for the Breton Wilderness. Should the emissions or nature of this project change, please contact me directly so that we might re-evaluate the revised proposed project. Thank you for keeping us informed and involving the Fish and Wildlife Service in the project review. -- Jill Webster, Environmental Scientist US Fish and Wildlife Service National Wildlife Refuge System Branch of Air Quality 7333 W. Jefferson Ave., Suite 375 Lakewood, CO (303) fax: (303)

100

101

102

103

104

105

106

107

108

109

110

111

112

113

114

115

116

117

118

119

120

121

122

123

124

125

126

127

128

129

130

131

132

133

134

135

136

137

138

139

140

141

142

143 Prevention of Significant Deterioration (PSD) Permit Modification Application Addendum Outokumpu Stainless USA, LLC January The business of sustainability