Greenhouse Gas Emissions Intensities of the Steel and Aluminum Industries at the Product Level

We began our work on border adjustments nearly a decade ago, because we did not expect that the approach taken in the American Clean Energy and Security Act to provide protection for domestic producers against imports from countries without greenhouse gas (GHG) control policies would be acceptable to the World Trade Organization (WTO). In 2018, together with Georgetown University Law School professors Jennifer A. Hillman and Mathew C. Porterfield, we developed a WTO-compatible Framework Flannery, Brian P., Jennifer A. Hillman, Jan W. Mares, and Matthew C. Porterfield, 2020. Framework Proposal for a US Upstream GHG Tax with WTO-Compliant Border Adjustments: 2020 Update. Washington, DC: Resources for the Future. for border adjustments in the context of a US domestic carbon tax. As a central concept of the Framework, we proposed a Greenhouse Gas Index (GGI) to account for the carbon dioxide equivalent emissions (CO2e) required to manufacture covered GHG-intensive products. For a given manufacturing facility or operation, e.g., to produce steel or petrochemicals, GGI accounts for GHG emissions occurring both from production operations, as well as the emissions required to produce GHG-intensive products purchased from suppliers of electricity, fuels used to generate thermal energy and raw materials. For many years, US facilities that emit more than 25,000 tonnes CO2e annually have determined and reported their GHG emissions to EPA. Key innovations in the Framework include the treatment of emissions from products acquired through the manufacturer’s supply chain (in a fashion similar to value-added taxes) and the design of straightforward procedures to allocate emissions from a facility to the GHG-intensive products it manufactures. Our approach, GGI, is consistent with standards developed by the International Organization for Standardization (ISO). In a consistent, comprehensive fashion, GGI applies to GHG-intensive products in all sectors of the economy, including those that produce aluminum, iron, and steel. The appendix to this report contains a list with links to our blogs and reports on border adjustments and the GHG intensity of products.

Over the past several years, we have interacted with experts from academia, national governments, international organizations, and importantly, with over a dozen sectoral trade associations. Based on these interactions, discussions, and our own relevant experience, we believe that our approach is feasible for industry and relevant to various applications that involve GHG-intensive products, including, for example, border adjustments, procurement policies, and corporate reporting. In particular, GGI could apply to products of US and foreign manufacturers in the aluminum, iron, and steel sectors.

This report updates, expands upon, and should be considered as a replacement for the testimony and submissions we provided the US International Trade Commission related to its hearing on December 7, 2023. It additionally serves as a modification and expansion of the modules for Iron, Steel, and Ferroalloys and for Alumina and Primary and Secondary Unwrought Aluminum in our 2022 report: The Greenhouse Gas Index for Products in 39 Industrial Sectors.

We are currently developing and will publish the estimated range of GGIs for many products in the modules for 39 industrial sectors in our 2022 report because of the US ITC hearing and anticipating further government interest in this subject.

The tables below provide estimated, illustrative low and high GGI values for representative basic products in the aluminum and steel sectors. Results are illustrative because manufacturers in the United States and around the globe utilize an enormous variety of processes, sources of energy, and raw materials in facilities with differing efficiencies to create similar products.

1. Facilities and operations of US firms producing at least 25,000 tonnes per year of GHG emissions determine and report them annually to EPA. Similar reporting programs exist in many, but not all, nations that export to the United States.
2. Firms like CRU and international trade associations for steel and aluminum industries exist that collect and publish emissions intensity information on average for basic oxygen steel, EAF steel, and primary aluminum for various countries or groups of countries.
3. The procedures to determine GGI are similar to those used in other contexts to determine value-added taxes (VATs) for specific products. Here, they account for the cumulative emissions of GHGs required to create GHG-intensive products in a particular facility, including both those from operations of the manufacturer and from GHG-intensive products purchased from suppliers. Essentially, this is a cradle-to-gate approach for GHG emissions.
4. To the extent the US government wants to develop energy intensity data for exporters to the US, for the purpose of import tariffs, the relevant US associations could be expected to assist the Administrator in developing that information. Note that many domestic firms have operations outside the US which give them added perspectives concerning imports, and American firms are vulnerable to imports from countries with weak GHG control measures. Note also that exports to other nations from US firms are vulnerable to such competition.
5. All 13 trade associations that we talked to, including steel and aluminum, indicated that our GGI concept could be implemented by their members. The data to determine GGI are available. The procedures to determine GGIs simply involve accounting, albeit involving a great deal of information, much of it not currently publicly reported.
6. The steel and aluminum companies, labor unions, associations, suppliers, and communities and regions where they operate will insist that any carbon tax or other GHG-control policies that affect the competitiveness both of these businesses and their covered products must be imposed on similar imported products and rebated for covered exported US products.
7. It is important to focus on emissions associated with products of specific facilities and companies, because, as our research and the following tables demonstrate, GGI values of identical products produced in different facilities can vary significantly, not only for facilities of different companies, but even among facilities of the same company. Using an average value to characterize GHG emissions associated with products of an entire sector or for groups of products will disrupt competition both within a sector and between sectors. For example, products made from plastics, aluminum, or steel may compete in applications, e.g., in the automotive sector. For these reasons, it will be important to design metrics for emissions and allocation to products that are similar across all covered sectors.
8. We note that implementation of procedures to determine GGI would be facilitated if all manufacturers of GHG-intensive products (including electricity) were required to determine and report GGI values for their products to their customers and regulators, and that the information should be publicly available. In particular, this would simplify the treatment of GGI for purchases from suppliers. Otherwise, manufacturers would need to rely on their own estimates or third-party determinations of GGI for such purchases, rather than those developed by the suppliers themselves with direct information.
9. Results should be viewed as preliminary and illustrative for many other reasons. Because we do not have information for specific facilities, estimates are based on various averages for key industrial processes and emissions from suppliers. Also, the information derives from a variety of national, sectoral, and private sources using different methods and covering different time periods. The sources span two decades or more. During that time, many manufacturers have significantly improved the efficiency of their operations or may have increased emissions to satisfy requirements for cleaner or safer products. Nonetheless, the estimates do indicate how GGI values would be determined given appropriate information for a specific facility and give a sense of the anticipated range of values.
10. Besides using different suppliers and procedures to manufacture products, facilities also differ; for example: in size, age, maintenance, operating practices, and efficiencies that affect the GGI values of their products. For products like primary aluminum and electric arc furnace steel, the GGI of electricity they use has a major impact on the GGI of the product and on the range of GGIs of such products from different manufacturers. For products like secondary aluminum and electric arc furnace steel, the GGI is substantially impacted by the amount of scrap used. (Note that in the Framework we proposed, scrap is assumed to have a GGI of zero.)
11. Many of these same issues will affect the data that will be provided to the US ITC through its 2024 survey of individual producers. The essential takeaways are that there is no unique value for the greenhouse gas intensity of specific products and that values should be determined for products of specific facilities. Among other issues, our full reports (listed in the Appendix) describe approaches that might be used to define average or default values to be used for products imported from nations currently without detailed requirements for GHG reporting by firms and facilities to implement the regulations, especially during initial, start-up years.
12. Our detailed reports provide observations concerning other issues that may be relevant to the determination and use of GGI values for products. For example, there are several concerns surrounding the timeliness and availability of data in the United States and other nations. Currently, US facilities report GHG emissions annually in April following the inventory year. Information and procedures will need to be updated, likely annually. GGI values are not static; they will change as industrial processes, raw materials, procedures, technology, products, and markets evolve. Border adjustment procedures should be designed to promote continuous improvement. For example, they should include appeals processes for relevant participants to challenge declared GGI values that appear to be incorrect, incomplete, or fraudulent.
13. Lower GGI values contain estimates using the least GHG-intensive inputs and processes, e.g., using natural gas rather than coal for thermal energy, and using hydropower or nuclear energy rather than coal to generate electricity. The higher values for GGI are estimates using the most GHG-intensive inputs, e.g., using coal to produce electricity, more GHG-intensive raw materials and processes, and lower processing efficiencies.
14. Results provided in this report are illustrative, because manufacturers in the United States and around the globe utilize an enormous variety of processes, sources of energy, and raw materials in facilities with varying energy and materials efficiencies to create similar products. They are also Illustrative because the data used Is from different years and areas. Consequently, listed GGI values do not represent the full range of possible outcomes. As a step toward further characterizing the uncertainty in GGI, this report provides additional calculations of GGI ranges for the above products that incorporate some additional sources that affect the potential ranges. This also is a further update to our previous work. The text in Section D Includes data needed to calculate GGI ranges but does not include calculations of all ranges of GGIs in the tables.

A. Introduction

Our hearing statement and submissions to the US ITC included a list of ranges of GGIs for four aluminum and steel products. A further development and explanation of those ranges is provided here. In the course of preparing this updated report, the GGI ranges for some of the products and determinations deviate slightly from those in the submissions:

Below are estimated ranges of GGI for steel and aluminum products. Lower estimates assume the least GHG-intensive inputs and processes, e.g., using natural gas rather than coal for thermal energy, and hydro rather than coal for electricity. Higher estimates assume lower efficiencies and more GHG-intensive inputs and processes:

Two sets of tables and data sources follow. Section B contains estimated high and low GGIs to produce some aluminum and steel products. The estimates include emissions both from operations of the manufacturer and those associated with products purchased from suppliers, including contributions from GHG-intensive raw materials, electricity, thermal energy and GHG process emissions. Section C indicates the percentage contribution from major inputs to production across the GGI ranges estimated for aluminum and steel products. Section D provides detailed discussions and sources for the data used in this analysis. Part of the text comes from the modules for various products in the RFF report The Greenhouse Gas Index for Products in 39 Industrial Sectors (September 27, 2022 https://www.rff.org/publications/working-papers/the-greenhouse-gas-index-for-products-in-39-industrial-sectors/ ), with the rest of the background information coming from other sources. Some of the copied text from the modules in the RFF report has been modified to use ranges of data that were available instead of averages.

Data used in this report comes from available studies published over the past 20 years and more. Thus, the estimates are surely not accurate in 2024, after all the years of efforts to improve efficiencies. The averages of these ranges of GHG emissions are probably higher than averages that would be determined based on current data, because the data used in these ranges comes from many different years. However, if the product has been heavily regulated over the past 20 years, in ways that required more energy use, e.g., to reduce emissions or improve safety, the averages of current GHG emissions may be higher than the ones provided here.

B. Ranges of GGIs from the Production of Aluminum and Steel

The following tables provide an indication of the possible ranges of GGIs for aluminum and steel and the various raw materials, thermal energy, and electricity to manufacture them. Where the underlying data was in a range, one extreme was used for the lowest data point while the other extreme for the highest data point. Similarly, where electricity was used, hydro was assumed for the lowest data point case and coal was assumed for the highest data point case. Where data is available for thermal energy, generally natural gas was assumed for the lowest data point, and fuel oil or coal was used for the highest data point. Where a range of efficiency or usage of raw material is available, the lowest of the range or usage was assumed for the lowest data point and the highest of the range or usage was assumed for the highest data point.

Bauxite Mining

Alumina, based on 2 and 3 tonnes of bauxite

Anode

Raw materials (assumes raw material is 100 percent pet coke with 90.5 percent C)

• Petroleum coke: 676 kg raw material /tonne anode
• Hard coal pitch: 148 kg raw material/tonne anode
• Recycled anode buts: 214 kg raw material/tonne anode
• Total carbon from raw material: 1.038 tonnes C/tonne anode

Electrolysis

Inputs to Aluminum GGI (tonnes CO2e/tonne aluminum)

GGIs for Secondary Aluminum and its Raw Materials (tonnes CO2e/tonne)

Range of GGIs Based on GHG Emissions from Production of Basic Oxygen Furnace (BOF) Steel

Range of GGIs Based on GHG Emissions from Production of Electric Arc Furnace Steel Based Totally on Scrap

Range of GGIs Based on GHG Emissions from Production of Electric Arc Furnace Steel Based on 75 Percent Scrap and 25 Percent Basic Oxygen Steel

C. Contributions (in Percent) to GGI of Aluminum and Steel Products from Various Raw Materials, Electricity, and Thermal Energy

The following results present the percent contribution to the high or low GGI for the named product of each of the major contributors to the GGI of the principal product.

D. Sources of Data Used to Develop GGIs for Aluminum and Steel and Their Raw Materials

The following text is primarily copied verbatim from the modules for aluminum, steel, and coal in the RFF report The Greenhouse Gas Index for Products in 39 Industrial Sectors (September 27, 2022). However, some of the values taken from the modules have been revised to reflect greater variations in ranges of data. In addition, here we augment the previous citation list with additional sources of data.

D.1. GHG Emissions from and Contribution to GGIs for Primary Unwrought Aluminum

Determination of the GGI for alumina is made below by using 2019 data from the country group of North America (which includes the United States and Canada) from the International Aluminium Institute. https://www.world-aluminium.org/statistics/ The data include the number of tonnes of alumina for which data is reported and the energy used, expressed in terajoule (TJ), from coal, oil, gas, electricity, and other sources used to manufacture the alumina. For North America, 1.392 million tonnes of metallurgical alumina are reported with the following sources of energy: one terajoule of fuel oil (which is small enough that it is neglected in this subsequent analysis), 5,710 TJ from natural gas, and 3,115 TJ consumed to produce electricity.

The contributions to the GGI of alumina are as follows https://international-aluminium.org/statistics/metallurgical-alumina-refining-fuel-consumption/ :

From hydro or coal for electricity: (3,115 TJ/1.392 million tonnes alumina) (277.78 MWh/TJ) (0 or 1.0 tonnes CO2e/MWh) = 0 or 0.622 tonnes CO2e/tonne alumina;

From natural gas or fuel oil for thermal energy: (5,710 TJ/1.392 million tonnes alumina) (0.9478 Btu/1,000 joule) (0.0532 or 0.0733 tonnes CO2e/MBtu gas) = 0.207 or 0.285 tonnes CO2e/tonne alumina;

Thus, the contribution to the GGI for alumina from electricity and thermal energy derived from hydro and natural gas is (0 + 0.207) tonnes CO2e/tonne alumina = 0.207 tonnes CO2e/tonne, while from coal and fuel oil the contribution is 0.622 + 0.285 = 0.907 tonnes CO2e/tonne alumina).

Bauxite is the raw material for alumina. It is mined and then processed into alumina. Between two and three tonnes of bauxite are required per tonne of alumina, as well as 5 kWh electricity and 0.0015 tonnes fuel oil per tonne of bauxite to mine the bauxite. See: Aluminum for future Generations; Efficiency. https://bauxite.worldaluminium.org/refining/energy-efficiency/) and Aluminum for Future Generations; Refining Process. Mining and Refining – Process (world-aluminium.org) The link to the originally cited source for this footnote is now marked as compromised. An equivalent source for this generic information is: https://en.wikipedia.org/wiki/Bayer_process

The contribution to the GGI of bauxite from mining is as follows:

• From coal for electricity: (5 kWh/tonne of bauxite) (1.00 tonnes CO2e)/(1000 kWh) = 0.005 tonnes CO2e/tonne bauxite;
• From fuel oil = (0.0015 tonnes fuel oil/tonne bauxite) (3.50 or 3.82 tonnes CO2e/tonne fuel oil) = 0.00525 or 0.00573 tonnes CO2e/tonne bauxite;
• Thus, the contribution to GGI from mining bauxite is (0.0 or 0.005 + 0.00525 or 0.00573) tonnes CO2e/tonne bauxite = 0.00525 or 0.0107 tonnes CO2e/tonne of bauxite.

The contribution to the GGI of alumina from mining bauxite is as follows:

• (0.00525 or 0.0107 tonnes CO2e/tonne bauxite) (2 or 3 tonnes bauxite/tonne alumina) 0.0105 or 0.0321 tonnes CO2e/tonne alumina.

Obviously, it will be important to obtain accurate data on the materials and processes used to produce alumina. (see footnote 5)

For fuel oil in this report, we use GGI = 3.18 tonnes CO2e/tonne fuel oil, based on our analysis that cumulative GHG emissions from oil production and refining to manufacture fuel oil are 10 percent to 20 percent greater than that from its carbon content alone (See Table 1 “Greenhouse Gas Index for Products in 39 Industrial Sectors: Petroleum Refinery Products” Sept. 2022, Flannery, Mares).

Table 1 of the introduction to the report Greenhouse Gas Index for Products in 39 Industrial Sectors (Sept. 2022, Flannery, Mares) provides estimates of reference CO2 (not CO2e) emissions from electricity and thermal energy derived from fossil fuels (using data from the US EPA). For thermal energy derived from natural gas and coal, respectively, Table 1 lists GGI as 0.0532 and 0.0935 tonnes CO2/tonne. For electricity derived from natural gas and coal respectively, Table 1 lists GGI as 0.42 and 1.00 tonnes CO2/MWh. Table 2 of the same report (using results from a 2011 report published by the World Nuclear Association), lists estimates of the wide range of GHG emissions from various sources of electricity generation around the world. For electricity generated from coal, they cite a range 0.756 to 1.310 tonnes CO2e/MWh. For electricity generated from natural gas, they cite a range 0.362-0.891 tonnes CO2e/MWh.

These estimates for natural gas and coal do not take into account emissions required to produce, process, and transport these fuels. In particular, for natural gas lost or consumed in production and transport, estimates range in amounts from 0.005 MCF/MCF to 0.03 MCF/MCF or even higher according to various sources.

Since pet coke is over 75 percent of the anode raw material, this analysis assumes it is 100 percent of the raw material. CRS 2013 Petroleum Study Table 3 indicates carbon range for pet coke is 89 to 92 percent. This analysis assumes the carbon content of pet coke is 90.5 percent, which is used here with (44 tonnes CO2/12 tonnes carbon)(1.038 tonnes pet coke/tonne anode)(0.905 tonnes carbon/tonne pet coke) or 3.44 tonnes CO2/tonne anode.

The EIA coal study indicates that bituminous coal has a carbon content of 45 to 86 percent carbon and represents 45 percent of 2021 production while sub-bituminous coal has a carbon content of 35 to 45 percent and represents 46 percent of such production. “Coal explained” (US Energy information Agency) This range of carbon content was not used in the GGI analyses for coal above.

Electrolysis of primary aluminum requires 12 to 16 MWh/tonne aluminum. “Quantifying the Carbon Footprint of the Alouette Primary Aluminum Smelter (2022)

D.2. GHG Emissions from and Contribution to GGIs for Unwrought Secondary Aluminum

Unwrought secondary aluminum is covered by NAICS Code 331314. Unwrought secondary aluminum has essentially no raw materials other than scrap aluminum and sometimes primary aluminum as well. Thus, the only component of its GGI is from the energy used to smelt the scrap and potentially some primary aluminum. However, since reports of amounts of primary aluminum used in secondary aluminum have not been obtained, estimates of GGI for secondary aluminum assume that no primary aluminum was used in making the secondary aluminum. If even a small percent of primary aluminum is used with scrap aluminum to make secondary aluminum, this will have a significant impact on the GGI.

For this analysis, we omit the contribution to CO2e(TOT) from purchased argon, chlorine, quicklime, nitrogen, oxygen, and caustic soda involved in making secondary aluminum. Assume all secondary aluminum is made in a remelting furnace—1,047 kg of aluminum scrap is needed to produce 1,000 kg of aluminum—and that unwrought secondary production is almost all natural-gas fired. The sources of data for secondary aluminum are the following: Dai, Q. et al. 2015. Updated Life-Cycle Analysis of Aluminum Production and Semi-Fabrication for the GREET Model. Argonne National Laboratory. https://publications.anl.gov/anlpubs/2015/10/121291.pdf, and The Environmental Footprint of Semi-Finished Aluminum Products in North America: A Life-Cycle Assessment Report. https://www.aluminum.org/sites/default/files/2022-01/2022_Semi-Fab_LCA_Report.pdf

The total energy consumption to remelt scrap into secondary aluminum is 110.3 kWh of electricity and 4,789 MJ of thermal energy from natural gas per tonne of secondary aluminum.

Contributions to CO2e(TOT) are as follows:

• Electricity: (110.3 kWh/tonne aluminum) (0.42 tonnes CO2e/1,000 kWh) = 0.046 tonnes CO2e/tonne aluminum, if natural gas is used for electricity;
• Thermal energy: (4,789 MJ) (1 MBtu/1,055 MJ) (0.0532 tonnes CO2e/MBtu gas) = 0.241 tonnes CO2e/tonne aluminum, if natural gas used for remelting;
• GGI for unwrought secondary aluminum, excluding energy consumption and GHG emissions related to scrap raw material: (0.046 + 0.241) tonnes CO2e/tonne aluminum = 0.287 tonnes CO2e/tonne aluminum, if natural gas is used for electricity and remelting;
• GGI for unwrought secondary aluminum, excluding energy consumption and GHG emissions related to scrap raw material: (0.11 + 0.424) tonnes CO2e/tonne aluminum = 0.534 tonnes CO2e/tonne aluminum if coal is used for electricity and gas for remelting.

Scrap is assumed to have no GHG emissions per tonne and the GHG emissions from use of small amounts of gases, quicklime and caustic soda are not known or considered. Since some secondary aluminum has primary aluminum as an input, this analysis assumes one case has no primary and the other has 25 percent primary.

D.3. GHG Emissions from and Contribution to GGIs for Basic Oxygen Steel

We begin this section with an example that illustrates the major sources of GHG emissions that contribute to the determination of GGI for raw steel from a basic oxygen furnace (BOF). The raw material for BOF steel is usually pig iron from a blast furnace. That, in turn, is further processed to raw steel with coke from converted coal, and limestone that decomposes into CO2 and calcium oxide when heated.

We use less recent data from various sources and use information from the 2022 report Steel Climate Impact: An International Benchmarking of Energy and CO2 Intensities that provides more uniform, up-to-date estimates and comparisons of CO2 emissions in 2019 from manufacturing steel in the United States and fifteen other countries using both the basic oxygen and electric arc furnace. See: Ali Hasanbeigi, Steel Climate Impact: An International Benchmarking of Energy and CO2 Intensities, April 7, 2022. https://www.ourenergypolicy.org/resources/steel-climate-impactan-international-benchmarking-of-energy-and-co2-intensities/.

The following example illustrates how a manufacturer would compute its GGI for BOF raw steel. It describes contributions to CO2e(TOT) from iron ore mining and processing; pelletizing iron ore; bituminous coal mining; use of coke, oxygen, and limestone; and from use of fossil fuel for thermal energy and electricity for the blast furnace and basic oxygen furnace. Contributions to CO2e(TOT) are as follows:

• Iron ore mining and pelletizing: The 2002 report Energy and Environmental Profile of the US Mining Industry indicates that 94,400 Btu are required to extract and process the ore. See: ITP Mining: Energy and Environmental Profile of the US Mining Industry: Chapter 4: Iron ) December 2002.This results in a contribution (94,400 Btu/short ton) (1.1 short tons/tonne) (0.0733 tonnes CO2e/MBtu) = 0.0076 tonnes CO2e/tonne ore. Iron ore pelletizing: A 2013 report by Lawrence Berkeley Laboratory indicates that the energy consumed by pelletizing iron ore is 2.1 GJ/short ton. 12 Hasanbeigi, Ali, and Lynn Price. 2013. Emerging Energy-Efficiency and Carbon Dioxide Emissions-Reduction Technologies for the Iron and Steel Industry. https://china.lbl.gov/sites/default/files/guidebooks/6106e-steel-tech.pdf This results in a contribution of (2.1 GJ/short ton iron ore) (1.1 short ton/tonne) (947,800 Btu/GJ) (0.0733 tonnes CO2e/MBtu from fuel oil) = 0.160 tonnes CO2e/tonne iron ore. Iron ore total: The total contribution from mining and pelletizing iron ore is 0.168 tonnes CO2e/tonne ore.
• Bituminous Coal Mining: The coke needed is assumed to be produced from bituminous coal mined underground. A decades-old study for the US Department of Energy indicates that 420,000 Btu/short ton coal are required for an underground coal mine and beneficiation of the coal, while 77,300 Btu/short ton are required for surface mining plus 94,500 Btu for beneficiation. ITP Mining: Energy and Environmental Profile of the US Mining Industry. December 2002: Chapter 2: Coal Assuming the coal is from underground and that oil products are the major energy source, the contribution to CO2e(TOT) = (420,000 Btu/short ton coal) (1.1 short ton/tonne) (0.0733 tonne CO2e/MBtu) = 0.034 tonnes CO2e/tonne coal.
• Coke: Approximately 1.5 tonnes of coal (which we assume is 65 percent carbon) See: Bowen, B.H. and M.W. Irwin Coal Characteristics. 2008 produce 1 tonne of coke https://www.corsacoal.com/about-corsa/coal-in-steelmaking/ . The report cited in footnote 7 indicates that 5.5 to 6.5 GJ thermal energy from natural gas are required to produce a ton of coke, which we assume to be 90 or 93 percent carbon. How Steel is Made", American Iron and Steel Institute. These result in a maximum contribution to CO2e(TOT) from coke of (1.5 tonnes coal/tonne coke) (0.034 tonnes CO2e/tonne mined coal) + (6.5 GJ/short ton coke) (1.1 short tons/tonne) (947,800 Btu/GJ) (0.0733 tonnes CO2e/MBtu from fuel oil) + ( 0.93 tonnes carbon/tonne coke) (44 tonnes CO2/12 tonnes carbon) = 3.96 tonnes CO2e/tonne coke.
• Oxygen: The module on industrial gases estimates that the GGI for oxygen (O2) is 0.525 tonnes CO2e/tonne O2, based on producing 95 percent O2 using the PSA process with electricity derived from coal as fuel.
• Limestone: Limestone is composed primarily of calcium carbonate (CaCO3); it disintegrates in the blast furnace to CO2 and calcium oxide that becomes part of slag. We assume that the contribution from limestone is 44 tonnes CO2 per 100 tonnes limestone = 0.44 tonnes CO2/tonne limestone (see also the module on cement).

A 1986 EPA publication and World Steel Raw Materials publication indicate that the production of 1 tonne of iron requires 1.4 to 1.6 tonnes of ore, 0.5 to 0.65 tonnes of coke, and 0.25 tonnes of limestone. See: AP 42, Fifth Edition, Volume I Chapter 12: Metallurgical Industry October 1986.

According to Britannica, 110 cubic meters of oxygen are required per tonne of BOF raw steel. https://www.britannica.com/technology/steel/Basic-oxygen-steelmaking So, the high GGI for BOF raw steel would be determined as follows:

GGI = CO2e(TOT) per tonne BOF raw steel =

(1.6 tonnes ore/tonne raw steel) (0.168 tonnes CO2e/tonne ore)

+ (0.65 tonnes coke/tonne raw steel) (3.96 tonnes CO2e/tonne coke)

+ (0.25 tonnes limestone/tonne raw steel) (0.44 tonnes CO2/tonne limestone)

+ (110 cubic meters O2/tonne BOF raw steel) (0.0014 tonnes O2/cubic meter O2) (0.525 tonnes CO2e/tonne O2)

= (0.269 + 2.58 + 0.11 + 0.081) tonnes CO2e/tonne BOF raw steel

= 3.04 tonnes CO2e/tonne BOF raw steel.

According to a 2018 report, cryogenic oxygen facilities making oxygen at 95 percent purity use 175 to 225 kWh tonne O2 and the pressure swing absorption process uses 525 kWh/tonne O2. See: Gas Technology Institute. 2018. Emerging and Existing Oxygen Production Technology Scan and Evaluation. GTI Project Number 22164. Des Plaines, Illinois.

Coke is a porous, hard black rock of concentrated carbon (contains 90 to 93 percent carbon). How Steel is Made, American Iron and Steel Institute.

D.4. GHG Emissions from and Contribution to GGIs for Electric Arc Furnace Steel Based Totally on Scrap

The GGI for electric arc furnace (EAF) raw steel is based on the electricity used to manufacture the steel, which can be generated from coal, natural gas, or non-fossil sources. Each such source will result in a different GGI for the steel. The scrap input does not impact the GGI. If feedstocks include HBI, pig iron, or iron, their GGIs will contribute to the GGI for the electric arc furnace raw steel. As in the previous section, we begin with an example that illustrates the major sources of GHG emissions that contribute to the following information about EAF steel: On a global basis, scrap is about 75 percent of metal inputs, direct reduced iron and hot briquetted iron (DRI and HBI) provide about 15 percent, with the balance 10 percent being pig iron and hot metal. The graphite electrode, which is calcined at the rate of 3,000 to 5,000 kWh/tonne coke, is consumed at the rate of about 1.1 kg/tonne iron, and 391 kWh/tonne steel and 35 kWh/tonne thermal energy provided by natural gas. For calcining the electrode see: “Electric Arc Furnace Steelmaking “ by Jeremy A.T. Jones, Nupro Corporation, Steelworks, American Iron and Steel Institute, 2008 Other reports indicate that the electricity consumed for EAF steel ranges between 360 kWh/tonne and 600 kWh/tonne. See https://www.britannica.com/technology/steel/Electric-arc-steelmaking, and Electric Arc Furnace Energy Consumption, see also http://heattreatconsortium.com/metals-advisor/electric-arc-furnace/electric-arc-furnace-energy-consumption/#:~:text=Energy%20consumption%20varies%20from%20350,burners%20uses%20475%20kWh%2Fton. If the EAF steel is produced solely from scrap, the scrap would not contribute to the CO2e(TOT) (see Section 4.3 of Framework Proposal for a US Upstream GHG Tax with WTO-Compliant Border Adjustments: 2020 Update (Report 20-14 October 2019 https://www.rff.org/publications/reports/framework-proposal-us-upstream-ghg-tax-wto-compliant-border-adjustments-2020-update/ ) for a discussion of the treatment of scrap).

Contributions to GGI based on 600 kWh/tonne steel electricity use and thermal energy and electricity derived from coal occur as follows:

• Electrode: (1.1 kg graphite/tonne EAF) (1 tonne graphite/1000 kg graphite) (1 tonne carbon/tonne graphite) (44 tonnes CO2/12 tonnes carbon) = 0.004 tonnes CO2e/tonne EAF raw steel;
• Electricity: (600 kWh/tonne EAF steel) (1.0 tonnes CO2e/1,000 kWh) = 0.600 tonnes CO2e/tonne EAF;
• Electricity (5,000 kWh/tonne coke) for calcining electrode 0.0007 tonnes CO2e/tonne EAF;
• Thermal energy: (35 kWh/tonne EAF steel) (1.0 tonnes CO2e/1000 kWh) = 0.035 tonnes CO2e/tonne EAF raw steel;
• Other non-scrap feedstocks (HBI, pig iron, iron): These are assumed to amount to 25 percent by weight of the EAF raw steel. We assume that the GGI for these feedstocks is the same as that for BOF raw steel: based on coal (0.25 tonnes other/tonne EAF) (3.04 tonnes CO2e/tonne other) = 0.758 tonnes CO2e/tonne EAF raw steel.

GGI for EAF based on 25 percent basic oxygen steel = CO2e(TOT) per tonne EAF raw steel;

= (0.75)(0.640) + 0.758 = 0.48 +0.758 = 1.24 tonnes CO2e/tonne EAF raw steel

Note that the contribution from non-scrap feedstock is significant.

Electrodes are calcined three times to convert needle coke to electrodes.

Assume coke raw material has the same GGI as coke for basic oxygen steel (although it is based on pet coke and the other is based on coal) and the same electricity use as EAF steel. This analysis includes electricity for calcining electrodes for EAF steel, whereas Working Paper 22-16 M6 Greenhouse Gas Index for Products does not.