Appendix E Calculation Methods Appendix

Appendix E
Calculation Methods Appendix $-$ Part II.A and II.B

II. Compiling a Facility-Level Emissions Inventory and Allocating to Subprocesses and Unit Processes

This section sets out the data and calculation steps the Commission used to build a facility-level emissions inventory, and then the steps it took to allocate those emissions to different subprocesses and unit processes in preparation for computing a product-level emissions inventory. These calculations represent the steps described in stage 1 and the first part of stage 2 in chapter 3 (see “Stage 1: Compiling a Facility-Level Emissions Inventory” and “Allocation of Facility-Level Emissions to Unit Processes”). As described in chapter 3, the facility-level inventory is comprised of scope 1 process and fuel combustion emissions, scope 2 emissions associated with purchased energy, and scope 3 emissions associated with the material inputs into production of covered steel and aluminum products within the system boundary the Commission has established. The calculations the Commission uses to generate these emissions inventory data from activity data and emissions factors and the calculations that reporters to the U.S. Environmental Protection Agency’s (EPA’s) Greenhouse Gas Reporting Program (GHGRP) use to calculate their own emissions data are both described.485F[482]

II.A. Process Emissions for Steel

This subsection provides more detail on how the Commission calculated scope 1 process emissions from the production of iron and steel. The Commission’s scope 1 process emissions calculation for steel production applies to multiple subprocesses, including lime and dolime production, metallurgical coke production, iron sintering, and steelmaking, for the purposes of product-level allocation of emissions.

Scope 1 process emissions at steelmaking facilities can occur from the operation of a basic oxygen furnace (BOF), electric arc furnace (EAF), non-recovery coke oven battery, sinter plant, decarburization vessel, or direct reduction furnace (see chapter 2, “Steel Production,” for more information on these processes). In this report, process emissions also include those from flaring of blast furnace gas and coke oven gas.486F[483] There are other sources of process emissions from iron and steel facilities as well, including those that produce fugitive emissions in steelmaking facilities.487F[484]

To generate process emissions estimates for facilities producing covered steel products, the Commission relied on two data sources: (1) GHGRP data, which contains data for almost all steel producers with process emissions, and (2) data from the Commission’s questionnaire for the few facilities that did not report to the GHGRP in 2022 but nonetheless had process emissions from steelmaking.

II.A.1. Scope 1 Process Emissions Reported Under the GHGRP

Facility-level reports under EPA’s GHGRP are the primary source of steel process emissions data for the Commission’s investigation. The GHGRP subpart Q requires iron and steel-producing facilities with annual emissions of over 25,000 mt of CO₂e to report their emissions to EPA under this program.488F[485] The EPA estimated that this threshold captures 100 percent of BOF producers and the vast majority of EAF producers in the United States.489F[486]

Of the steelmaking facilities (i.e., those facilities using an on-site BOF or EAF to make semifinished steel) that responded to the Commission’s questionnaire, 95.6 percent of steelmaking facilities also reported to the GHGRP.

II.A.1.a GHGRP Iron and Steel Production Calculations

The GHGRP regulation provides steelmaking facilities a variety of options to calculate their process emissions under subpart Q. These calculation options are broadly grouped into four techniques: default emissions factors, site-specific emissions factor, mass-balance equations approach, and continuous emissions monitoring systems (CEMS).

· Default emissions factors:490F[487] Default emissions factors are provided by the GHGRP and are based on the average emissions that occur per unit of consumption of raw material or per unit of output. In the case of steel, a default emissions factor is used to determine fugitive CO₂ emissions from one process: coke pushing.491F[488]

· Site-specific emissions factors: Under the GHGRP, steel facilities can calculate GHG emissions from a particular process by using a site-specific emissions factor that they have calculated specific to that process. They can determine an emissions factor from a performance test that measures CO₂ emissions from all exhaust stacks for the process, and also measure either the feed rate of materials into the process or the production rate during the test in metric tons per hour. Under this approach, the site-specific emissions factor is multiplied by annual feedstock use or production to determine annual CO₂ emissions from the process.492F[489]

· Mass-balance equations: Mass-balance equations generally measure (1) the carbon entering a process through inputs or feedstocks (the product of the carbon content of inputs and the quantity of those inputs used in the process); and (2) the carbon exiting the same process through products and by-products. These equations then subtract the carbon outputs from the carbon inputs and assume the carbon difference is either directly released or oxidized and then released as CO₂.493F[490] Separate mass-balance equations are provided within the GHGRP regulations for various processes used to produce iron and steel (described in 40 C.F.R. Part 98 Subpart Q).494F[491]

· Continuous emissions monitoring system (CEMS): A CEMS is a device that continually collects information on the quantity of a gas being emitted, including GHGs. CEMS devices will often collect GHG emissions information covering both process and fuel combustion emissions.495F[492]

Facilities report the specific methods they use to calculate subpart Q emissions associated with different production processes throughout their facilities. Facilities may report one or multiple quantities of emissions for each production process depending on the complexity of their operations. The Commission assigned these emissions to specific subprocesses used in this investigation ( $S 1 P_{s u b p r o c}$ ) based on the methodology the facility used to calculate those emissions under subpart Q. Table E.2 shows the specific methods used by facilities to report process emissions and the subprocesses to which the Commission assigned those emissions.

Table E.2 Steel process emissions data reported under the Greenhouse Gas Reporting Program (GHGRP) and associated subprocess

GHGRP basis for calculating or measuring GHG emissions	Regulatory citation for calculation or measurement	Assigned subprocess
Carbon mass-balance method for a decarburization vessel	40 C.F.R. § 98.173(b)(1)(vi)	Steelmaking
Carbon mass-balance method for an electric arc furnace (EAF)	40 C.F.R. § 98.173(b)(1)(v)	Steelmaking
Carbon mass-balance method for basic oxygen process furnaces	40 C.F.R. § 98.173(b)(1)(ii)	Steelmaking
Carbon mass-balance method for nonrecovery coke ovens	40 C.F.R. § 98.173(b)(1)(iii)	Metallurgical coke production
Carbon mass-balance method for direct reduction furnace	40 C.F.R. § 98.173(b)(1)(vii)	Rotary hearth furnace
Carbon mass-balance method for sinter processes	40 C.F.R. § 98.173(b)(1)(iv)	Iron sinter production
Default emissions factor for coke pushing emissions	40 C.F.R. § 98.173(c)	Metallurgical coke production
Continuous emissions monitoring system (CEMS), where a facility report includes reference to an EAF associated with the reported emissions	40 C.F.R. § 98.173(d)	Steelmaking
CEMS, where facility report includes reference to iron sinter processes associated with the reported emissions	40 C.F.R. § 98.173(d)	Iron sinter production
Site-specific factor, where facility report includes reference to EAF processes associated with the reported emissions	40 C.F.R. § 98.173(b)(2)	Steelmaking
Site-specific factor, where the facility report includes reference to a decarburization vessel associated with the reported emissions	40 C.F.R. § 98.173(b)(2)	Steelmaking
Site-specific factor, where the facility report includes reference to a basic oxygen process furnace process associated with the reported emissions	40 C.F.R. § 98.173 (b)(2)	Steelmaking
Site-specific factor for flaring associated with blast furnace processes or blast furnace gas	40 C.F.R. § 98.253(b)(1)	Blast furnace operations
Site-specific factor for flaring associated with coke oven processes or coke oven gas	40 C.F.R. § 98.253(b)(1)	Metallurgical coke production
Carbon mass-balance method for lime production	40 C.F.R § 98.193(b)(2)	Lime and dolime production

Source: Compiled by the USITC.

Note: Any emissions using the carbon mass-balance method for direct reduction furnaces were allocated to the rotary hearth furnace subdivision. The carbon mass-balance method for lime and dolime production is used for GHG emissions data reported under Subpart S of the GHGRP.

Certain processes described above, particularly those associated with CEMS measurements, can include fuel combustion or the use of feedstock that could be reported under either the GHGRP subpart C (fuel combustion emissions) or the GHGRP subpart Q (process emissions). The GHGRP allows for emissions from certain stationary combustion processes that are difficult to distinguish from process emissions (e.g., from a common monitored stack) to be reported under subparts C or Q.496F[493] This can result in ambiguity between what is reported under subpart C and what is reported under subpart Q. Notwithstanding this ambiguity, this report refers to emissions reported under subpart C as “scope 1 fuel combustion emissions” and those reported under subpart Q as “scope 1 process emissions.” The Commission considers both sets of emissions to be scope 1 emissions that can be linked to specific subprocesses for the purposes of product allocation, regardless of whether they are fuel combustion or process emissions.

The Commission used $S 1 P_{s u b p r o c}$ to calculate unit process-level process emissions ( $S 1 P U G H G_{p r o d u c t}$ ) described in equation E.3 in the overview section of this appendix. For subprocesses that only produced a single reference product, $S 1 P U G H G_{p r o d u c t}$ was considered equivalent to $S 1 P_{s u b p r o c}$ .497F[494] The steelmaking and production of calcined lime and dolime subprocesses produce multiple reference products, and the Commission calculated $S 1 P U G H G_{p r o d u c t}$ for products made from these subprocesses using equation E.4.

$\begin{matrix} S 1 P U G H G_{p r o d u c t} = S 1 P_{s u b p r o c} * \frac{O u t p u t_{p r o d u c t}}{O u t p u t_{s u b p r o c}} (E .4) \end{matrix}$

II.A.2. Use of Survey Data to Calculate Scope 1 Process Emissions for Certain EAF Facilities

For facilities reporting that they used EAFs to produce semifinished steel but did not have a corresponding GHGRP report for 2022, the Commission used equation E.5 to calculate process emissions associated with steelmaking for that facility. Equation E.5 is based on the GHGRP’s mass-balance approach for EAFs and decarburization vessels. The Commission’s main sources for the data used in this equation were questions in sections 5 and 6 of the questionnaire, which covered input and output data from EAFs and decarburization vessels as well as the carbon content of those materials.

$C O_{2} = \frac{44}{12} * (\begin{array}{l} I r o n * C_{I r o n} + S c r a p * C_{S c r a p} + F l u x * C_{F l u x} + E l e c t r o d e * C_{E l e c t r o d e} + C a r b o n \\ * C_{C a r b o n} - S t e e l_{E A F} * C_{S t e e l} + S t e e l_{D e c a r b} * (C_{S t e e l i n} - C_{S t e e l o u t}) \\ + F_{g} * H H V * E F * 0.001 - S l a g * C_{S l a g} - R_{E A F} * C_{R E A F} - R_{D e c a r b} * C_{R D e c a r b} \end{array}) (E .5)$

Table E.3 shows the mass-balance equations for EAFs, decarburization vessels, and for gaseous fuel combustion emissions from the GHGRP on which the Commission based its calculations in equation E.5.

Table E.3 GHGRP equations used for the Commission’s approach for calculating scope 1 process emissions from EAF facilities that do not report to the Greenhouse Gas Reporting Program (GHGRP)

GHGRP calculation	Description	Source
$C O_{2} = \frac{44}{12} * (\begin{array}{l} I r o n * C_{I r o n} + S c r a p * C_{S c r a p} + F l u x * C_{F l u x} \\ + E l e c t r o d e * C_{E l e c t r o d e} + C a r b o n * C_{C a r b o n} \\ - S t e e l_{E A F} * C_{S t e e l} + F_{g} * C_{g f} * \frac{M W}{M V C} * 0.001 \\ - S l a g * C_{S l a g} - R_{E A F} * C_{R E A F} \end{array})$	Mass-balance equation for steel produced in EAFs. GHGRP Equation Q-5.	40 C.F.R. § 98.173(b)(1)(v)
$C O_{2} = \frac{44}{12} * (S t e e l_{D e c a r b} * (C_{S t e e l i n} - C_{S t e e l o u t}) - R_{D e c a r b} * C_{R D e c a r b})$	Mass-balance equation for decarburization vessels. GHGRP Equation Q-6.	40 C.F.R. § 98.173(b)(1)(vi)
$C O_{2, F g} = F_{g} * H H V * E F * 0.001$	Subpart C gaseous fuel combustion equation (tier 1 approach). GHGRP Equation C-1.	40 C.F.R. § 98.33(a)(1)(i)
$C O_{2, F g} = \frac{44}{12} * (F_{g} * C_{g f} * \frac{M W}{M V C} * 0.001)$	Subpart C gaseous fuel combustion equation (tier 3 approach). GHGRP Equation C-4 and portion of GHGRP Equation Q-5 shown above.	40 C.F.R. § 98.33(a)(3)(iii); 40 C.F.R. § 98.173(b)(1)(v)

Source: Compiled by the USITC.

Note: Variables used in the equations in this table are described in table E.4.

Equation E.5 consolidates GHGRP equations Q-5 and Q-6, allocating all process emissions from EAFs and decarburization vessels to the steelmaking subprocess. In order to reduce the burden on surveyed facilities, the Commission substituted the portion of GHGRP equation Q-5 that mirrored GHGRP equation C-4 (a more complex approach to calculating the CO₂ emissions from fuel combustion, $C O_{2, F g}$ ) with the simpler GHGRP equation C-1 that is also used to calculate $C O_{2, F g}$ .498F[495] The question numbers in the Commission’s questionnaire under which information on the above inputs was gathered and the variables they correspond to within equation E.5 are noted in Table E.4.

Table E.4 USITC calculation variables, description, and questionnaire mapping for the process emission methodology for EAFs that do not report to the Greenhouse Gas Reporting Program (GHGRP)

Process emissions formula variable	Variable description	USITC questionnaire question
Iron	Annual mass of direct reduced iron or pig iron (if any) charged to the furnace	5.1.12a, 5.1.13a
C_Iron	Carbon content of the direct reduced iron expressed as a decimal fraction	5.1.12c, 5.1.13c
Scrap	Annual mass of ferrous scrap charged to the furnace	5.1.14a
C_Scrap	Carbon content of the ferrous scrap expressed as a decimal fraction	5.1.14g
Flux	Annual mass of flux materials (i.e., limestone, dolomite) charged to the furnace	5.1.8a
C_Flux	Carbon content of the flux materials expressed as a decimal fraction	5.1.8d
Electrode	Annual mass of carbon electrodes consumed	5.1.15a
C_Electrode	Carbon content of the flux materials expressed as a decimal fraction	5.1.15c
Carbon	Annual mass of the carbonaceous materials expressed as a decimal fraction	5.1.5b, 5.1.6a, 5.1.7a
C_Carbon	Carbon content of the carbonaceous materials expressed as a decimal fraction	5.1.5d, 5.1.6c, 5.1.7d
Steel	Annual mass of molten raw steel produced by the furnace	6.1.1a
C_Steel	Carbon content of the steel expressed as a decimal fraction	6.1.1c
Steel_Decarb	Annual mass of molten steel charged to the decarburization vessel	6.1d
C_Steelin	Carbon content of the molten steel before decarburization expressed as a decimal fraction	6.1f
C_Steelout	Carbon content of the molten steel after decarburization expressed as a decimal fraction	6.1f
F_g	Annual volume of the gaseous fuel used	5.1.4b
HHV	Annual average high heating value (HHV) of the fuel	Natural Gas high heating value from table C-1 of 40 C.F.R. 98, subpart C (MMBtu/scf)
EF	Fuel specific default CO₂emissions factor (EF)	Natural Gas Emissions factor from table C-1 of 40 C.F.R. 98, subpart C (CO₂/MMBtu)
Slag	Annual mass of slag produced by the furnace	6.3a
C_Slag	Carbon content of the slag produced by the furnace expressed as a decimal fraction	6.3c
R_EAF	Annual mass of air pollution control residue resulting from the EAF	6.2b
C_REAF	Carbon content of the air pollution control residue resulting from the EAF expressed as a decimal fraction	6.2d
R_Decarb	Annual mass of air pollution control residue resulting from the decarburization vessel	6.2b
C_RDecarb	Carbon content of the air pollution control residue resulting from the decarburization vessel expressed as a decimal fraction	6.2d

Source: USITC, Greenhouse Gas (GHG) Emissions Intensities Questionnaire: Facility-Level, 2024.

When certain information on the carbon content of materials was not known by a facility, the Commission used publicly available factors (table E.5) in equation E.5 to calculate process emissions.

Table E.5 Default carbon content values and sources

In metric tons of carbon dioxide equivalent per metric ton (mt CO₂e /mt) of material.

Formula variable	Default value (mt carbon/mt)	Sources
C_Iron	0.020 (direct reduced iron), 0.047 (pig iron)	IPCC, “2019 Refinement to the 2006 IPCC Guidelines for National Greenhouse Gas Inventories, Volume 3,” 4.31.
C_Scrap	0.01	IPCC, “2019 Refinement to the 2006 IPCC Guidelines for National Greenhouse Gas Inventories, Volume 3,” 4.31.
C_Flux	0.13 (dolime), 0.121 (lime/limestone), 0.23 (other)	IPCC, “2019 Refinement to the 2006 IPCC Guidelines for National Greenhouse Gas Inventories, Volume 3,” 4.31.
C_Electrode	0.82	EPA, OAR, “Inventory of U.S. Greenhouse Gas Emissions and Sinks,” October 22, 2024, 4 $-$ 87.
C_Carbon	0.91 (charcoal), 0.87 (petroleum coke)	IPCC, “2019 Refinement to the 2006 IPCC Guidelines for National Greenhouse Gas Inventories, Volume 3,” 4.31.
C_Carbon	0.73 (coking coal)	IPCC, “2019 Refinement to the 2006 IPCC Guidelines for National Greenhouse Gas Inventories, Volume 3,” 4.31.
C_Steel	0.04	EPA, OAR, “Inventory of U.S. Greenhouse Gas Emissions and Sinks,” October 22, 2024, 4 $-$ 87.
C_Steelin	0.04	EPA, OAR, “Inventory of U.S. Greenhouse Gas Emissions and Sinks,” October 22, 2024, 4 $-$ 87.
C_Steelout	0.01	Sullivan and Olson, “Building a Decarbonized Steel Sector,” July 7, 2023.
C_Slag	0.004	Moras et al., “Carbon Dioxide Removal Efficiency of Iron and Steel Slag in Seawater via Ocean Alkalinity Enhancement,” June 17, 2024.
C_REAF,C_RDecarb	0.15	EPA, “Air Pollution Aspects of the Iron and Steel industry” 43.

Source: Compiled by the USITC.

Because the Commission considered all emissions calculated using equation E.5 to be from the steelmaking subprocess, these were further allocated to the reference products produced from steelmaking using equation E.5

II.B. Process Emissions for Aluminum

In line with the U.S. Environmental Protection Agency (EPA)’s Greenhouse Gas Reporting Program (GHGRP) methodology, the Commission considered process emissions associated with the production of covered aluminum products as emissions that “generally include emissions from chemical transformation of raw materials.”499F[496] As defined in chapter 3 (“Aluminum Process Emissions”), the process emissions emitted by facilities producing covered aluminum products consist of CO₂ and perfluorocarbons (PFCs) and only occur in the production of primary aluminum. CO₂ process emissions are released during the baking of the carbon anode and the consumption of the anode during electrolysis. PFCs are released when the levels of alumina within the pot fall below the level required for electrolysis, causing the voltage within the pot to spike (“anode effects”).500F[497] The information in this section will cover the use of EPA’s GHGRP data to estimate process emissions from primary unwrought aluminum production, and the steps taken to allocate these emissions to two subprocesses: anode baking and smelting of primary unwrought aluminum.

II.B.1. Scope 1 Process Emissions Reported Under the GHGRP

This investigation used only publicly available data from facility-level reports under EPA’s GHGRP to determine process emissions from primary aluminum production.501F[498] All operating U.S. aluminum smelters reported to the GHGRP in 2022.

The GHGRP provides calculation approaches to estimate process emissions using a number of techniques depending on which segment of process emissions are being measured. Like in the steel regulations explained above, these process emissions can be reported either through a mass-balance calculation method or by using a continuous emissions monitoring system (CEMS).502F[499] The Commission assigned emissions reported by facilities under subpart F to specific subprocesses used in this investigation ( $S 1 P_{s u b p r o c}$ ). Table E.6 shows the possible methods facilities used to report process emissions and the subprocess to which the Commission assigned those emissions.

Table E.6 Greenhouse Gas Reporting Program (GHGRP) emissions data reported under Subpart F and associated subprocess

GHGRP basis for calculating or measuring GHG emissions	Regulatory citation for calculation or measurement	Assigned subprocess
Continuous emissions monitory system (CEMS) or mass-balance equations for pitch volatiles and bake furnace packing material	40 C.F.R. § 98.63(d)(f)	Anode baking
CEMS or mass-balance equation for process emissions from anode consumption; calculations using measured and default values for anode effects	40 C.F.R. § 98.63(a),(b),(d),(e)	Smelting of primary unwrought aluminum

Source: Compiled by the USITC.

The Commission used ( $S 1 P_{s u b p r o c}$ ) to calculate unit process-level process emissions ( $S 1 P U G H G_{p r o d u c t}$ ) described in equation E.3 in the overview section of this appendix. The two subprocesses under which subpart F emissions were assigned each produced only a single reference product; therefore, $S 1 P U G H G_{p r o d u c t}$ was considered equivalent to $S 1 P_{s u b p r o c}$ .503F[500] The below sections describe the specific ways in which GHGRP reporters are to calculate process emissions in aluminum production.

II.B.1.a GHGRP Anode Baking Calculations

Anode baking of prebake carbon anodes releases CO₂ emissions both from pitch volatiles and bake furnace packing material.504F[501] For the baking of the carbon anodes, process emissions may be calculated by one of two approaches: either measurement by a CEMS monitoring system or the use of mass-balance equations.505F[502] The GHGRP regulation instructs that if these process emissions are measured by a CEMS on a stack with fuel combustion emissions from subpart C, those combined emissions must be reported under subpart F.506F[503] Under the Commission’s methodology, these emissions are all considered to be scope 1 emissions linked to primary unwrought aluminum. The mass-balance equation for estimating emissions from pitch volatiles takes into account the initial weight of the green anodes, the mass of hydrogen in the green anodes, the mass of the baked anodes, and the mass of waste tar collected.507F[504] The mass-balance equation for estimating emissions from the bake furnace packing material takes into account the packing coke (calcined petroleum coke) consumption rate per ton of baked anode production as well as the sulfur and ash contents of the packing coke.508F[505] Some values in this mass-balance equation (e.g., sulfur, ash, and hydrogen contents) can either be smelter specific or a default value from Table F-2 of subpart F.509F[506]

II.B.1.b GHGRP Primary Unwrought Aluminum Production Calculations

When primary aluminum is smelted, process emissions are generated both as the carbon anodes are consumed, and when anode effects occur. CO₂ emissions from anode consumption can be estimated using a mass-balance equation based on measurements of the net prebaked anode consumption rate per metric ton of aluminum produced, the ash and sulfur contents of the anodes, and the total mass of aluminum metal produced per year for all prebake cells.510F[507] Like in anode baking, process emissions for anode consumption during electrolysis may be calculated by either the use of this mass-balance equation or CEMS measurement. Similarly to anode baking, some values in this mass-balance equation for anode consumption (e.g., sulfur and ash contents) can either be smelter specific or a default value from Table F-2 of subpart F.511F[508]

Emission of PFCs resulting from anode effects are calculated by the GHGRP using given parameters on anode effect minutes and metal production, as well as a smelter specific or default slope coefficient for perfluoromethane (also known as carbon tetrafluoride, or CF₄) emissions to anode effect minutes.512F[509] Monthly totals of anode effect minutes per cell-day, aluminum production, and a slope coefficient (either smelter specific or, under certain circumstances, a default value) relating CF₄ emissions to the prior two variables are used to estimate CF₄ from anode effects. This estimate of CF₄is then combined with a mass ratio of perfluoroethane (also known as hexafluoroethane, or C₂F₆) to CF₄ to estimate C₂F₆ emissions.513F[510]

Beyond the sources of process emissions covered under the GHGRP, new research has revealed other potential sources of process emissions in the smelting of primary unwrought aluminum. While these other sources are not explicitly included in the GHGRP data the Commission uses for these estimates, these sources are described further in box E.1 for reference.

Box E.1 Other Types of Aluminum Process Emissions not Incorporated in the Commission’s Calculation

Emissions from low voltage anode effects: Depending on the level of the voltage change in the smelting pot, anode effects can be characterized as “low voltage” or “high voltage.” Although Subpart F the U.S. Environmental Protection Agency’s (EPA’s) Greenhouse Gas Reporting Program (GHGRP) does not define what should constitute high- or low-voltage anode effects, it is commonly understood that low voltage anode effects occur below the 6 $-$ 8 volt detection limit of most modern plant computers.a As low voltage anode effects and associated emissions are not currently detectable by most smelters, they were not required to be reported to the GHGRP in 2022 and are also not included within the estimates provided in this report.b A 2022 Federal Register notice on potential revisions to the GHGRP regulation noted that the EPA was considering an amendment to Subpart F to add reporting of emissions from low-voltage anode effects.c In April 2024, the EPA’s final rule on amendments to the GHGRP regulation indicated that it would not be making changes to the measurement methodology at this time.d

Cell startup emissions: Emissions are also released when a new or refurbished smelting pot (or cell) is brought online and operated at a higher voltage to bring the pot up to operating temperature.e This increased voltage can produce additional perfluorocarbon (PFC) emissions, designated as “cell start-up emissions.” As some operators do not begin measuring voltages and related emissions until the pot begins smelting aluminum, these emissions can go unrecorded.f As such, cell start-up emissions are also not explicitly included in the GHGRP reporting requirements and thus may often be excluded from the estimates provided in this report. Because no new potlines were built or old potlines restarted in 2022, cell start-up emissions were likely very small.g In 2022, the EPA requested comments on a potential amendment to Subpart F that would require the inclusion of cell startup emissions in future GHGRP reporting requirements.h As noted above, in April 2024, the EPA’s final rule indicates it will not make changes to Subpart F calculations at this time. i

^a Tabereaux, “Low Voltage Anode Effects and Unreported PFC Emissions,” October 2016, 631.

^b USITC, hearing transcript, December 7, 2023, 261 (testimony of Matt Aboud, Century Aluminum); Alcoa Corporation, written submission to the USITC, December 7, 2023, 8.

^c 87 Fed. Reg. 36920, 37023 (June 21, 2022).

^d 40 C.F.R. § 98; 89 Fed. Reg. 31802, 31822 (April 25, 2024).

^e IAI, “IAI Good Practice Document on Measuring Perfluorocarbons,” December 2020, 9.

^f IAI, “IAI Good Practice Document on Measuring Perfluorocarbons,” December 2020, 9.

^g USGS, Mineral Commodity Summaries 2023: Aluminum, January 2023.

^h 87 Fed. Reg. 118, 37023 (June 21, 2022).

ⁱ40 C.F.R. § 98; 89 Fed. Reg. 31802, 31822 (April 25, 2024).

[482] The request letter states that the Commission should collect data for the scope 1 emissions used in its emissions intensity calculations from facility GHGRP reports to EPA when available. For facilities that do not report to GHGRP, the request letter notes that the Commission should use survey data to determine scope 1 emissions. See appendix A of this report for a copy of the request letter.

[483] The emissions from the use of these waste gases in on-site industrial processes are characterized as scope 1 fuel combustion emissions, as described in the energy emissions section below.

[484] Note that these fugitive emissions refer only to those emissions at steelmaking facilities and are distinct from those discussed in appendix F (“Fugitive Emissions Associated with Coal and Natural Gas Used in Steel and Aluminum Production”). The fugitive emissions referred to in the text above can occur during ladle metallurgy operations, desulfurization, hot metal transfer, sinter cooling, and the charging and tapping of furnaces. For example, fugitive emissions of blast furnace gas may be emitted during infrequent process upsets when gas is vented for a short time or from leaks in the ductwork through which these gases pass. EPA, OAR, “Technical Support Document for the Iron and Steel Sector,” August 28, 2009, 22.

[485] EPA, “GHGRP, Envirofacts GHG Query Builder,” accessed September 18, 2024. 40 C.F.R § 98.2(a)(2).

[486] EPA, OAR, “Technical Support Document for the Iron and Steel Sector,” August 28, 2009, 30.

[487] The GHGRP does not rely heavily on default emissions factors, an approach that differs considerably from other approaches that tend to rely heavily on default emissions factors. In the GHGRP proposed rule, the EPA argued that its low reliance on these default emissions factors was due to the very high uncertainty of these factors. It stated that default emissions factors would not provide site-specific estimates of emissions that reflect the differences in feedstocks, operating conditions, fuel combustion efficiency, variability in fuels, and other differences among facilities. It stated that a default emissions factor-based methodology is more widely used as a sector-wide or national total estimate and is not necessarily appropriate for facility-level estimates. EPA, OAR, “Technical Support Document for the Iron and Steel Sector,” August 28, 2009, 32.

[488] 40 C.F.R. § 98.173(c).

[489] 40 C.F.R § 98.173(b)(2), EPA, “Mandatory Reporting of Greenhouse Gases Iron and Steel (Final),” March 2012, 3.

[490] EPA, “Greenhouse Gas Reporting Program: Emission Calculation Methodologies,” July 2015; EPA, “Mandatory Reporting of Greenhouse Gases Iron and Steel (Final),” March 2012; EPA, “Reporting of Greenhouse Gases for Aluminum Production,” February 2018.

[491] 40 C.F.R. § 98.173(b)(1)(vi).

[492] EPA, “Mandatory Reporting of Greenhouse Gases Iron and Steel (Final),” March 2012, 3; 40 C.F.R. § 98.173(d).

[493] The GHGRP subpart Q regulation specifies that facilities must report under subpart C emissions from, among other facility units, byproduct recovery coke oven battery combustion stacks, blast furnace stoves, boilers, process heaters, reheat furnaces, annealing furnaces, flame suppression, ladle reheaters, and other miscellaneous combustion sources. Subpart C can also be used by facilities reporting combinations of fuel combustion and process emissions from common stacks that are monitored using CEMS. Subpart Q includes mass-balance based emissions reporting provisions that require facilities to include carbon from fuels used in EAFs, BOFs, decarburization vessels, direct reduction furnaces, taconite indurating furnaces, nonrecovery coke oven batteries, and iron sinter units. Some facilities also report subpart Q emissions using CEMS that includes a mix of fuel combustion and process emissions. 40 C.F.R. § 98.33; 40 C.F.R. § 98.172; 40 C.F.R. § 98.173; U.S. industry representatives, interviews by USITC staff; Subject matter expert, email message to USITC staff, December 21, 2023.

[494] The subprocesses with process emissions under subpart Q that produced only a single reference product were metallurgical coke production, iron sinter production, liquid pig iron in rotary hearth furnaces, and blast furnace operations. See also table E.2 for an exhaustive list of all subprocesses.

[495] The use of GHGRP equation C-1 to capture the fuel consumption portion of the Commission’s equation E.5 allowed for surveyed facilities to report only their fuel consumption in questionnaire question 5.4 (if not otherwise captured in section 3 of the questionnaire). If equation E.5 had instead used GHGRP equation C-4 for the fuel consumption portion, facilities would have had to report the average annual carbon content and average annual molecular weight of fuel consumed in steelmaking in question 5.4. The Commission made this decision for the sake of simplification.

[496] EPA, “Greenhouse Gas Reporting Program: Emission Calculation Methodologies,” July 2015, 1. The GHGRP also considers fugitive emissions in its definition. See appendix F (“Fugitive Emissions Associated with Coal and Natural Gas Used in Steel and Aluminum Production”) for more information on fugitive emissions.

[497] IAI, “Perfluorocarbon (PFC) Emissions,” 2024.

[498] EPA, “GHGRP, Envirofacts GHG Query Builder,” accessed September 18, 2024.

[499] Note that some facilities may vent emissions from anode consumption during electrolysis or anode baking of prebake cells (which are described under subpart F) through the same stack as emissions from any fuel combustion units (described under GHGRP subpart C). If these emissions are measured by a CEMS, the GHGRP regulation notes that emissions from this stack should be reported under subpart C. 40 C.F.R. § 98.63(g). Regardless of where they are reported, these emissions are still linked to specific processes for the purposes of product allocation.

[500] The subprocesses with process emissions under subpart F each correspond with a single reference product. Anode baking with carbon anodes (a material input for primary unwrought aluminum) and smelting of primary unwrought aluminum with primary unwrought aluminum (see also tables E.1 and E.6).

[501] Prebake carbon anodes are anodes that are produced prior to being used in the smelting process, whereas Søderberg anodes are baked in the smelting pot during the smelting process. In the United States, all producers use prebake carbon anodes. Pitch volatiles, typically made from methane, tar, and hydrogen, are used to heat-treat green anodes during the anode baking process. For more information on anode baking, see chapter 2 (“Primary Unwrought Aluminum Production”).

[502] A CEMS measures the concentration and rate of gas or particulate matter being emitted from a production facility. Analogous to steel, CEMS devices will often collect GHG emissions information covering both process and fuel combustion emissions.

[503] 40 C.F.R § 98.63(g); Table F-2 to Subpart F of Part 98, Title 40.

[504] Carbon anodes are called “green anodes” before they are baked. 40 C.F.R § 98.63(f).

[505] EPA, “Reporting of Greenhouse Gases for Aluminum Production,” February 2018, 2. See also 40 C.F.R § 98.63(f).

[506] Table F-2 to Subpart F of Part 98, Title 40.

[507] 40 C.F.R § 98.63(e).

[508] Table F-2 to Subpart F of Part 98, Title 40.

[509] Anode effects occur when an insufficient supply of alumina to the smelting pot causes a rapid spike in voltage in the pot, leading to the emission of gases containing PFCs. These effects are measured by duration in minutes. 40 C.F.R § 98.63(b).

[510] 40 C.F.R § 98.63(b).