Green Steel

Precise annual emissions of global steelmaking figures vary between references with year and the system boundaries considered. [1] cites the 2023 IEA World Energy Outlook report that steelmaking was responsible for 8% of CO₂ and 6% of GHG emissions in the year prior. [4] estimates direct iron-and-steel emissions at 2.6 GtCO₂ in 2019, equivalent to 7% of global energy-system CO₂. [3] places the industry at approximately 7-8% of anthropogenic greenhouse-gas emissions using a broader industry boundary. The script’s 8% figure is primarily drawn from [1] but also serves as a rounded overview of these overlapping estimates.

Carbon dioxide emissions produced by the transportation sector worldwide in 2024 reached an estimated 9 GtCO₂, out of a total estimated 41.6 GtCO2 (i.e. total anthropogenic CO₂ emissions from fossil fuel use and LULUCF [land-use, land-use change, and forestry], factoring in the man-made cement carbonation sink) ^[5,6]. Aviation (including international and domestic) accounted for a reported 9.9% of emissions from the transportation sector, which was equal to 2.14% of global carbon emissions (just under one billion metric tons). Shipping (including international, domestic and fishing) accounted for a reported 10.67% of emissions from the transportation sector, which was equal to 2.31% of global carbon emissions ^[5,6].

An earlier pre-pandemic comparison, using dedicated aviation and shipping studies against the same global CO₂ emissions total, provides a useful cross-check on this 2024 estimate and produces a broadly similar result. In 2018, global aviation emitted approximately 1.034 GtCO₂ ^[7], while shipping emitted approximately 1.056 GtCO₂ ^[8]. Combined emissions were therefore approximately 2.090 GtCO₂. Compared with the Global Carbon Budget’s total of 42.5 GtCO₂ for the same year ^[6], aviation and shipping together accounted for approximately 4.92%. An 8% share of emissions from the steelmaking sector would therefore be approximately 3.08 percentage points higher, or 1.63 times the combined aviation-and-shipping emissions. This compares annual CO₂ only and does not include aviation’s additional non-CO₂ warming effects.

At the time of writing, these represent the best available figures for comparing sectors directly, drawing wherever possible on recent, authoritative sources (e.g. shipping emissions provided by the International Maritime Organization). However, such comparisons require care, because published estimates may cover subtly different types or categories of greenhouse gases, activities and sources of emissions.

Firstly, different studies may report either CO₂ alone, all greenhouse gases, or carbon dioxide-equivalent (CO₂eq) values, which include gases like methane and nitrous oxide weighted by their global warming potential relative to that of CO₂. These metrics are not interchangeable or directly comparable without consistent standardization, and presenting one versus another can shift the apparent share of sector emissions significantly. Sources, studies, and organizational reports often do not make it immediately clear which specific metric (/type of emissions measure) has been used, and regularly state one when they actually mean another (e.g. stating ‘CO₂’ when they mean ‘CO₂eq’).

Secondly, the choice of global baseline matters: some comparisons are made against energy-related CO₂ only, while others include all anthropogenic sources such as land-use, land-use change, and forestry (LULUCF), which dictates the denominator and thus the sector’s calculated percentage.

International shipping emissions in particular are difficult to track because emissions occur largely in international waters. Historically, data often came from bunker fuel sales rather than direct monitoring of ship activity, introducing further uncertainty. IMO now combines fuel-sales, vessel-activity (AIS modeling) and voyage allocation methods, and estimates can vary according to which methods are used. A similar story can be seen with aviation data as well. Together with the aforementioned issues in figure publishing, this can make available global emission estimates range considerably, even when based on the same sector/activity.

1.892 Gt of steel is produced annually, of which 52% is used in construction (983.84 Mt of steel).

[14] indicates that global urban extent expanded by 9,687 km² per year between 1985 and 2015. This is equivalent to (9687/365 =) 26.539 km² per day.

The area of Manhattan is 59.1 km^2 A.: 59.1/26.539 = 2.2 days

However, it is worth noting that the current rate of growth may be even faster. The World Bank ^B estimates (as of April 2023) that 1.2 million km² of new urban built-up area is expected to be added to the world by 2030. This is equivalent to an average annual rate of urban expansion across the entire decade of the 2020s of roughly 120,000 km² per year, 328.77 km² per day or 13.698 km² per hour, which is the equivalent of the area of Manhattan being built every 4 hours.

Additional comparisons may be drawn: If 983.84 Mt of steel is used annually in construction (see above), this is equal to 2.7 Mt daily. In comparison, the Golden Gate Bridge in San Francisco, USA, contains ~75,000 metric tons of steel ^C, meaning ~36 Golden Gate Bridges could be constructed from daily construction steel. Alternatively, the Empire State Building contains ~51,700 metric tons of steel ^D,E, meaning ~52.2 Empire State Buildings could be constructed from daily construction steel.

^A Overview of New York City | NYMTC

^B Urban Development Overview | World Bank Group

^C Design & Construction Stats | Statistics & Data | Golden Gate

^D ASCE Metropolitan Section | Empire State Building

^E Empire State Building | World's Tallest Towers

94% of ore mined is iron ore and 98% of iron ore is used for steel. Therefore, 98% of 94% is equal to 92%. This is a derived mass-based estimate, and not a measured statistic.

[39] provides a maximum potential emissions reduction of up to 85%; however this may not be universally applicable, and the source does not clearly specify a complete system boundary for its estimate. In practice, emissions savings may depend heavily on the share of scrap metal in the electric arc furnace, electricity mix and the extent of the wider system boundaries considered. Nevertheless, WorldSteel ^F reports similarly substantial differences between ore- and scrap-recycling-based steelmaking routes: a global average of 2.34 tCO₂/t of crude steel for the traditional blast-furnace/basic-oxygen-furnace route, and 0.69 tCO₂/t for scrap-metal/electric-arc-furnace route, equivalent to an average reduction of approximately 70.5%.

^F Sustainability Indicators Report 2025 | World Steel Association

Leakage means CO₂ escaping from the intended containment - through a well, fault, fracture or caprock - and potentially reaching another geological formation, groundwater, the seabed or atmosphere. Storage failure can include leakage, but can also mean excessive pressure, unexpected migration or loss of injectivity that prevents the site operating as intended without necessarily releasing CO₂ into the atmosphere. Neither is an expected outcome from a properly selected and managed site. EU rules require proposed sites to demonstrate no significant leakage risk and to maintain monitoring and corrective-measure plans.

Replacing ‘coal in the blast furnace’ refers to replacing the conventional coal-based blast-furnace reduction stage itself with gas-based direct reduction, which typically takes place in a separate shaft/direct-reduction furnace; it does not mean a one-for-one substitution of fossil gas inside the same blast furnace.

Other methods of DRI vary significantly in their emission outputs. [60] suggests that coal-based DRI can reduce emissions by 38% compared to the traditional blast-furnace/basic-oxygen-furnace method. Fossil gas, however, with its usual mix of CH₄, H₂, and CO, reduces emissions by 61%. Replacing fossil gas with H₂ entirely can reduce emissions by 97%. However, these are not universal estimates, and exact emissions savings will depend not only on the key reductant used in the process, but also the rest of the DRI configuration, the energy mix and system boundary. Notably, emissions from the method of production of H₂ prior to use in steelmaking have not been included in the hydrogen-DRI estimates. Green hydrogen uses low-carbon electricity, but emissions may still be embodied in transport & infrastructure (see below). Despite the use of fossil gas in DRI releasing fewer emissions compared to coal, fossil fuels are still involved so emissions are inevitable.

Green hydrogen is produced by using renewable electricity to split water into hydrogen and oxygen through a process of electrolysis. This eliminates the emissions from producing hydrogen from fossil fuels. When the hydrogen enters the steelmaking process at the iron-reduction stage to separate iron from its ore, it reacts with iron oxides in the ore and combines with oxygen to produce water vapor rather than CO₂.

However, the wider process may still cause some emissions. These indirect, associated emissions may arise from infrastructure used in production, transport and storage of green hydrogen, or from electricity drawn from a grid that’s energy mix is not yet completely carbon free. In practice, hydrogen described as ‘green’ may therefore not always be produced exclusively with renewable electricity throughout the entire process.

Green-hydrogen steel can thus come very close to zero emissions, without necessarily reaching absolute zero carbon emissions. In reality, steel made using renewable hydrogen produces approximately 0.08 kg CO₂ per kg of steel ^[67].

Blue and gray hydrogen have more defined associated emissions. Gray hydrogen is generally produced from fossil gas using a process called steam methane reforming (SMR), which splits hydrogen from the methane in the fossil gas mixture, and releases CO₂. This is currently the most common and cheapest way to produce hydrogen. Blue hydrogen is made through the same basic process, but combined with carbon capture and storage (CCS) technology to prevent most of those production emissions escaping into the atmosphere. Because the capture process tends not to catch everything, blue hydrogen still produces some atmospheric emissions.

When either form of hydrogen is used to separate iron from its ore, the reduction reaction itself still produces only water rather than CO₂. The main difference lies in the upstream emissions created earlier when the hydrogen was initially produced. Smaller emissions can also remain elsewhere in the steelmaking process, including from electricity and carbon used in the Electric Arc Furnace.

According to [67], steel made using blue hydrogen produces approximately 0.22 kg CO₂ per kg steel, assuming that 85% of the emissions from producing hydrogen are captured. This compares with 1.87 kgCO₂ for conventional coal-based steel and 0.38 kgCO₂ for coal-based steel fitted with carbon capture itself, for which a separate capture rate of 80% is assumed ^[67].

More than 95% of hydrogen used worldwide is made using fossil gas, mostly without any form of carbon capture. If this conventional gray hydrogen is used in hydrogen-based DRI steelmaking, [67] estimates emissions of approximately 0.7 kg CO₂ per kg steel ^[67]. This is substantially lower than conventional coal-based steel production, but considerably higher than that for steel made using green hydrogen.

In these calculations, most of the remaining emissions are those that have evaded CCS, because it does not capture all the CO₂ produced during conventional steelmaking or blue-hydrogen production.

The potential emissions saving of more than 95% therefore describes a best-case green-hydrogen DRI route rather than a universal result that applies to every form of hydrogen-based steel ^[60,67]. [67] estimates approximately 0.08 kg CO₂ per kg steel for its renewable-hydrogen case; compared with 1.87 kgCO₂/kg for conventional coal-based steel, this equates to a reduction of 95.7% ^[67]. Ultimately, the outcome depends on hydrogen production, electricity carbon intensity, raw material and scrap inputs, and the wider emissions/system boundaries involved.

Steel is not pure iron: it is an alloy that normally contains a small, carefully controlled amount of carbon, which helps give it the required properties. Consequently, even when hydrogen replaces carbon during the first, iron-reduction stage, some carbon is still needed when the iron is turned into steel.

Steelmakers also use carbon for several technical jobs inside an Electric Arc Furnace that are not covered in detail in the script. These include adjusting the final carbon content (including carburization), helping remove remaining oxygen (deoxidation) and slag formation (When steel melts, impurities float to the top and form a liquid waste layer called slag. Injecting carbon into the mix reacts with the iron ore to produce bubbles of carbon monoxide gas, which makes the slag expand into a foam. This foamy layer of slag helps to protect the furnace from the intense heat and improves furnace performance by trapping heat inside the molten metal, reducing necessary energy input).

Scrap steel serves primarily as a recycled source of iron and pre-made steel (metallic feedstock) rather than a substitute for purposeful carbon input, although the carbon already contained in the composition of the scrap contributes to the furnace’s final carbon balance. EAF steelmaking may therefore still require additional carbon to be added.

Biogenic waste residues can potentially supply an additional source of carbon, but they must first be processed into a suitable form of biocarbon or biochar. Suitably prepared waste plastics and end-of-life tires can also act as carbonaceous reducing or slag-foaming agents or help to provide energy, helping to replace some of the conventional coke or coal that would otherwise be used.

However, most plastics and much of the carbon contained in tires originally came from fossil fuels, so using them can still release fossil CO₂. The environmental benefit therefore depends on the waste material used (feedstock composition), how it is processed (pretreatment), how much newly mined (virgin) fossil carbon it replaces, furnace performance and which parts of the lifecycle are considered. Using waste can reduce demand for conventional fossil carbon, but does not automatically make the process zero-carbon.

Steel produced via the hydrogen method is being trialed across a range of countries:

[76] Germany, [77] Austria, [78] The Netherlands, [79,80] UK, [81] South Korea