The Slow and Winding Path To Green Steel

Impact: 7% Global Warming Potential (GWP)

Technology Maturity: Scaled/Subsidized

This is the tenth article of a series on the climate technologies shown in my ClimateTech Market Map, with a deeper dive into technical maturity and potential to reduce global warming. Here I'll cover the maturity and potential impact of low and zero emissions technologies for primary steel production.

TL; DR - there are three major technical paths to achieve greener steel:

1. Adding carbon capture to legacy coal-based processes
2. Switching to green hydrogen based "direct reduction" processes
3. Switching to electrochemical processes based on molten oxide electrolysis or low-temperature alkaline/acidic electrolysis

None of these options are in a satisfactory state today. Carbon capture is proving stickier to implement at scale than originally thought. Green hydrogen based direct reduction is costly and requires high grade ores. Electrochemical methods have not yet scaled to industrial capacities (1M ton+ annual capacity).

However, of all these options, molten oxide electrolysis, (commercialized by a single startup - Boston Metal) looks like it has the best chance of being both cheap and capable of handling the wide range of ore requirements that a global solution needs.

The Global Steel Industry

Steel is one of the Big 5 CO? emitters along with electricity generation, transport, cement and buildings. 1.9 billion tons of steel was produced in 2023, a production volume that has more than doubled since 2000. A little over half of all steel in 2023 was produced by China, although India actually added the most new steel production. The map below shows relative total production globally.

Primary vs. Secondary Steel Production

Not all steel production is equally emissions intensive. An important distinction is the difference between primary and secondary steel production. Primary steel production uses mined iron ore as its input, whereas secondary steel production uses steel scrap (mostly shredded cars and large appliances). Although either type of steel can be used for most applications, secondary steel cannot be used for some high grade steels because of impurities in the input scrap.

Many steel-making processes exist, but today almost all primary steel is produced in coal-fueled integrated mills that pair a Blast Furnace (BF) with a Basic Oxygen Furnace (BOF). The Blast Furnace converts iron ore to pig iron, and the Basic Oxygen Furnace converts the pig iron to steel. In contrast, secondary steel production is overwhelmingly based on Electric Arc Furnaces, which use ~ 400kWh of electricity per ton of steel to directly melt the input scrap steel using electric arc discharges.

Primary steel produced by the BF/BOF process emits about 2.2 tons of CO? per ton of steel produced, whereas secondary steel emits about 0.3 tons of CO?.

Given its far superior energy and emissions behavior, a first thought is "let's use 100% secondary steel". However, secondary steel production is capped by regional availability of scrap. About 70% of total US steel production is now secondary production because of the easy availability of high quality scrap from cars and appliances. In contrast, only ~10% of Chinese steel production is currently secondary production because the first generation of mass-purchased vehicles and appliances are only now reaching end-of-life and being scrapped.

Because secondary production is already electrified, as the grid becomes greener, the already low emissions associated with secondary steel production will reduce even more. So we're far more interested in how to de-carbonize primary steel - which is a tougher problem.

Primary Steel Making Fundamentals

There are essentially just two steps in any primary steel-making process. The first step is to take iron oxide ore and strip away its oxygen atoms, leaving some form of elemental iron. This oxygen stripping process is called "reduction". The second step is to adjust the carbon and other material levels to produce the desired type of steel.

Reduction is the energy and emissions intensive step. Iron oxides are extremely stable compounds, so it takes a lot of energy to break the bonds between iron and oxygen. There are three fundamental ways to reduce iron oxide ore:

1. Combine it with a carbon monoxide source, which will strip its oxygen, producing CO? and iron. This is the process used in traditional blast furnace production using coking coal.
2. Combine it with a hydrogen source, producing H?O and iron. This is a "direct reduction" process that many traditional steel makers are planning to use to produce low emissions steel.
3. Use electricity or electro-chemistry to directly break the iron oxide bond, producing oxygen and iron.

These three methods can be hybridized (e.g. methane reduction uses both hydrogen and carbon) and used under various combinations of heat and pressure, producing a plethora of possible production routes. Birat et al. (2021), for example, enumerates 41 different production routes that use varying combinations of carbon, hydrogen, electricity, heat and pressure. Some of these are illustrated in the figure above.

Low Emission Production

Any of these production process can theoretically be de-carbonized. Coal/coke based processes can switch to bio-coal mixes or add a carbon capture system (CCS). Hydrogen-based processes can use green hydrogen from water electrolysis. And electricity based processes can use renewable electricity. All these options are estimated to be roughly equal in total energy intensity (+/-10%).

However, most of these de-carbonization options are immature. Bio-coal has limited production capacity and its organic feedstocks (crop residues etc.) have competing uses. CCS has multiple issues including scalability, cost and storage, many of which are extremely sticky to solve. Green hydrogen made from water electrolysis is currently 3-4x more expensive than "grey" hydrogen produced from high emissions steam methane reforming, although costs are falling slowly. And electricity based processes have yet to be proven in commercial scale plants (although electric arc furnaces for producing steel from pig iron or sponge iron are fully mature and cost effective.)

Here I'll cover just the technologies involved in hydrogen and electrical production processes, excluding biofuels and CCS.

Green Hydrogen Processes

Iron ore can be reduced to sponge iron (similar to pig iron) by "Direct Reduction" (DR): exposing it to hydrogen at high temperature (~1,000°C +) and pressure, producing water as a waste product. The industry already has experience with this type of process from methane and syngas-based DR plants, which are a small percentage (~4%) of global production today. Methane-based direct reduction is currently used for steel production where coal is expensive and methane is cheap - such as parts of South America and the Middle East.

Multiple steel-making capital equipment companies, such as MIDREX and Energiron, now offer direct reduction reactors fueled by hydrogen or hydrogen/methane blends (example below). These are vertical shaft furnaces based on previous methane DR models that require pelletized high grade ores (>67% iron content). This requirement for high grades is a significant drawback, since high grade ore only comprises ~4% of ore production globally, and critical regions like China have low grade ore resources.

Fluidized bed and other designs that can theoretically handle a wider range of ore qualities have been scaled to pilot stage by Metso and Mitsubishi Primetals but seem to be in cold storage for now. For example, HYFOR, a fluidized bed design by Primetals operated at pilot-scale from 2019-2022, but there has been no announcement of a date for a commercial scale plant. However, there are a number of industrial scale non-hydrogen fluidized bed reactors in operation. Posco (Korea), for example, operates two FINEX process plants with a combined capacity of 3.5M tons, and has announced an intention to deploy a hydrogen based process.

The sponge iron produced by a DR furnace is similar to the pig iron produced by a blast furnace, so its carbon level needs to be lowered and impurities removed. This is done by melting it with an Electric Arc Furnace and adjusting composition. A challenge with hydrogen DR is that residual hydrogen remaining in the output can cause brittleness in final steel.

Some prominent forthcoming green hydrogen DRI plants include:

• Baosteel (China): 1M ton capacity plant using Tenova Energiron shaft furnaces. The plant recently completed its first test at full capacity.
• Stegra (Sweden): 2.5M ton capacity plant using MIDREX shaft furnaces. Operations start is planned for 2026. (Formerly H2 Steel).
• Hybrit (Sweden): Currently in early development of its first industrial scale plant, after successful operation of a commercial scale pilot. The original target date for completion was 2030, but this date has been dropped from recent announcements.

Other companies planning plants include Salcos (Tenova furnace) and Blastr (Primesteel MIDREX furnace).

Electrochemical Processes

Because of the cost and ore quality restrictions of hydrogen direct reduction, interest is growing in electrochemical approaches.

Molten Oxide Electrolysis (MOE) dissolves iron ore in a molten electrolyte at high temperatures (~1,600°C) and uses electrolysis to separate Iron and Oxygen into two material streams at each electrode. The electrolyte in the original proof-of-concept by the founders of Boston Metal, a MOE startup, consisted of 85% Calcium and Aluminum Oxide, with the balance comprising Magnesium Oxide and Iron Oxide (magnetite: Fe?O?). The Boston Metal MOE process consumes ~4MWh per ton of steel and can handle a wide range of ore grades.

Low Temperature Electrolysis (LTE) dissolves powdered iron ore in a concentrated electrolyte (e.g. sodium hydroxide) at relatively low temperatures (60-300°C). Once in solution, a combination of reactions occur that crystallize iron as a thick plating on the electrode. The resulting iron plates can be extracted, melted, alloyed and recast using an EAF to create final commercial steel product.

LTE has the benefit of being compatible with intermittent renewables since a cell can be put into sleep-mode indefinitely by reducing the current to 10% of normal. It can also theoretically process low grade ores, potentially including ultra-cheap waste-grade ore that can't be processed by blast furnaces.

Electrochemical approaches can be stymied by poor faradaic efficiencies (aka: high diversion to unwanted side reactions), poor energy efficiencies and short electrode lifetimes. However, LTE seems to perform favorably on these axes.

AccelorMIttal's SIDERWIN project reported 91.4% faradaic efficiency at respectable current densities of 0.1 amps/cm2 using high grade inputs in an alkaline electrolyte (1). Its pilot-scale plant reported an average of 2.7MWh of electricity consumption per ton of iron production - which is more energy-efficient than MOE (2). However, lower-grade ores did not perform well at these current densities, and excess non-conductive forms of aluminum, silica, magnesium and titanium dramatically reduced conversion efficiency when present.

Both large steel makers and startups are commercializing LTE, but no commercial scale plants are yet in operation. ArcelorMittal and John Cockerill plan to open their Volteron plant by 2027 with a demonstration-scale capacity of 40-80,000 tons. Fortescue, the fourth largest global producer of iron ore, has also announced an LTE plant.

Iron can also be produced with Molten Salt Electrolysis, but no companies are commercializing this approach for iron making.

Other Processes

Plasma processes for reducing iron oxides have been explored extensively in the research lab, both standalone and hybridized with DR pre-processing. Thermal plasma (high temperature ionized gas) can be generated by passing an electric arc through a mix of molecular hydrogen and argon. The resulting ionized hydrogen can reduce iron oxides to iron at room temperature. Plasma systems are theoretically highly energy efficient and capable of processing a wide range of ores, but managing ultra-high temperature materials requires a plethora of cooling and control systems, and constant maintenance and the technology has yet to progress beyond lab-scale prototypes. The Susteel project, for example, scaled a plasma process up to a 90kg capacity, but was discontinued at that stage.

Laser-based thermal decomposition is another early stage technology. Thermal decomposition of iron oxides without the use of a reducing agent like carbon monoxide or hydrogen, requires temperatures above 3,400°C, which no conventional fuel-burning process can reach. However lasers can easily reach these temperatures. Limelight Steel is a startup currently developing its first proof-of-concept using this process.

Economics and Policy

The cost behavior of industrial scale green steel production is well characterized for green hydrogen DR processes, because we know the cost of water electrolysis and DR furnaces.

General estimates are that European steel made with hydrogen DR will be ~25% more expensive than legacy processes - or roughly the same cost as a legacy process with CCS. The cost premium is directly driven by the high cost of green hydrogen.

The Global Efficiency Intelligence consultancy estimates that hydrogen DR can be directly cost competitive with unabated legacy processes at ~$1.20 per kg of H? (3). But as I've previously written, while it's plausible (but optimistic) to achieve ~$2/kg for H? by 2030, $1/kg looks infeasible by then - always excluding the heavy use of zero cost curtailed renewable electricity.

In Europe, the competitiveness of DR steel can be helped by carbon taxes on unabated legacy production. These taxes currently run at ~€70 per ton but without further subsidization, green steel will likely remain a luxury input, used only when its sustainability marketing value is high (e.g. for Volvo high end EVs).

It's also worth noting that steelmakers currently get free emissions credits in Europe and tapering of these free credits isn't scheduled to start until 2026. Whether tapering actually happens given the political influence of European steelmakers is a coin-toss.

When we move to electrochemical processes, estimates are more nebulous. The earliest assessment of long-term electrochemical costs by the ULCOS project in the 2000's, estimated that the potential long term (2050) cost of steel from electrochemical methods ranged from €330 to €540 per ton (nominal cost) - cheaper than the long term cost of most CCS or direct reduction methods and similar to current market prices for unabated legacy steel.

More recent estimates of electrochemical costs by various authors, have estimated the capex range for MOE to be anywhere from $764 to $4,901 per ton of annual capacity and LTE capex costs from $611 to $1,076 per ton of annual capacity.

We can do our own back of the envelope SWAG on costs, by assuming a $1,000 capex per ton, a ten year life for the capital equipment and a $40 per MWh electricty cost. This would yield:

• $100 per ton capital costs
• $160 per ton electricity costs (4MWh x $40 per MWh)

Adding labor and balance of plant on top of that would likely get us in the $400 per ton cost range, without making heroic assumptions about ultra-cheap renewables. Although this is a SWAG - that would be excellent news for electrochemical steel.

Beyond that, ultra-cheap solar could turbocharge cost-competitiveness: with the $15 per MWh electricity generated in the American West, Morocco or the Middle East, electrochemical approaches could easily produce the world's cheapest steel.

Policy Supports

Policy support for steel is a complex area and would require its own article. In the US, the Inflation Reduction Act introduced massive tax credits for green hydrogen production, carbon capture and also allowed government procurement to favor green steel. Europe's emission credit scheme is the main tool to incentivize greener steel, but heavy support has been provided for hydrogen production. Billions of euros of assistance has been provided for pilot projects through Horizon Europe, the European Innovation Fund and the Research Fund for Coal and Steel.

Relevant Startups

From recent history, novel steel making processes take 10-15 years and hundreds of millions of dollars to develop and upscale to commercial production capacity. So this segment is a challenging one for startups, requiring access to large pools of mid-stage and growth capital.

Boston Metal (US) was founded in 2013 as a spinout from MIT and has raised $400M+ in funding to commercialize MOE as a steel manufacturing process. The company's key innovation is a chrome-alloy based anode that can resist corrosion at very high temperatures. It has currently scaled its process to a commercial scale (25 kiloamps) pilot, and plans to open its first industrial scale plant in 2026 in Brazil. The company claims that its process can be competitive with traditional steel-making processes at $40 per MWh at industrial scale. (4)

Electra (US) has raised $110M+ to commercialize its acidic LTE process. Acidic electrolysis is generally avoided in favor of alkaline because of lower faradaic efficiencies, but the company has developed a novel cell design that avoids the oxidation state cycling that normally causes this inefficiency. The company began operation of its pilot plant this year.

Element Zero (Australia) is an early stage startup commercializing alkaline LTE for production of various metals, including steel. They are currently operating a lab-scale prototype and working on their first intermediate scale pilot.

Limelight Steel (US) is a seed-stage startup planning to use lasers to directly thermally decompose iron ore.

Conclusions

Commodity industries tend to tip quickly to whatever process demonstrates the lowest cost. Today that means the overwhelming dominance of the BF/BoF process for regions where coal is available, and methane based direct reduction for regions where it is not. But today's technology mix is not stable. If CCS continues to elude scale-up, and MOE and LTE processes achieve their cost targets at scale, then the industry could rapidly convert over to these processes.

What seems clear is that the International Energy Agency's forecast of an equal blend of alternative processes co-existing into the future is highly unlikely. Rapid changeovers in steel processes have occurred in the past, dictated by economics. From 1960 to 1970 - in just a single decade - the production share of Open Hearth furnaces decreased from 90% to 30%, in favor of Basic Oxygen Furnaces which went from essentially nothing to 60% production share.

Open Hearth furnaces simply couldn't compete on marginal cost, and billions of dollars of capital expenditures on Open Hearth furnaces became stranded assets. The same fate may befall the trillions dollars of capital expenditure represented by the world's legacy blast furnaces.

Almost thirty years after the Kyoto protocol, it's alarming that so little progress has been made to date to de-carbonize one of the world's largest emitting industries. However, the potential for electrochemical startups to revolutionize the steel industry gives optimism for the future.

The ClimateTech Series

The ClimateTech Market Map

1: Decarbonizing Nitrogen Fertilizer with Electrochemical Processes

2: Anti-Methane Livestock Vaccines Are Nascent

3: Intro to Enhanced Geothermal Energy (Fracked Geothermal)

4: Abating Nitrous Oxide Emissions from Agriculture

5: Decarbonizing Concrete is Hard

6: Decarbonizing Cooking Doesn't Matter (Much)

7: Biochar = Cheap & Reliable Carbon Removal

8: Is There a Path to $1/kg Green Hydrogen?

10: The Slow and Winding Path to Green Steel

11: Is Enhanced Rock Weathering Effective?

12: Scaled Offshore Wind: Cheap & Getting Cheaper

13: Animal Free Dairy: Are We There Yet?

The Climate Change Impacts Series

1: Climate Change Alters Ecosystems

References

(1) SIDERWIN, 2021. SIDERWIN introductory webinar. Available online.

(2) SIDERWIN, 2023. SIDERWIN final report webinar. Available Online.

(3) Hasanbeigi, A. et al. 2024. Green Steel Economics. Global Efficiency Intelligence. July 2024. Available online.

(4) Volts, 2024. Volts Podcast with Boston Metal.

Bibliography

Allanore, A., Yin, L. and Sadoway, D.R., 2013. A new anode material for oxygen evolution in molten oxide electrolysis. Nature, 497(7449), pp.353-356.

Birat, J.P., Patisson, F. and Mirgaux, O., 2021. Hydrogen steelmaking, part 2: competition with other net-zero steelmaking solutions–geopolitical issues. Matériaux & Techniques, 109(3-4), p.307.

Bloomberg NEF, 2024. "Scaling Up Hydrogen: The Case for Low-Carbon Steel". Available online.

Braun, T., Wallace, C., Pham, Q., Nijhawan, S. and Alexander, C.L., 2024. Electrochemistry in Action: Iron and Steel Manufacturing. The Electrochemical Society Interface, 33(2), p.38.

Humbert, M.S., Brooks, G.A., Duffy, A.R., Hargrave, C. and Rhamdhani, M.A., 2024. Economics of Electrowinning Iron from Ore for Green Steel Production. Journal of Sustainable Metallurgy, 10(3), pp.1679-1701.

IEA, 2020. Iron and Steel Technology Roadmap. Available online.

Peplow, M. 2021. "Can Industry Decarbonize Steelmaking?" Chemical & Engineering News, Vol. 99 Issue 22. Available online.

Roland Berger, 2020. The Future of Steelmaking - How the European Steel Industry Can Achieve Carbon Neutrality. Roland Berger Focus. May 2020.

Satritama, B., Cooper, C., Fellicia, D., Pownceby, M.I., Palanisamy, S., Ang, A., Mukhlis, R.Z., Pye, J., Rahbari, A., Brooks, G.A. and Rhamdhani, M.A., 2024. Hydrogen Plasma for Low-Carbon Extractive Metallurgy: Oxides Reduction, Metals Refining, and Wastes Processing. Journal of Sustainable Metallurgy, pp.1-50.

Souza-Filho, I.R., Springer, H., Ma, Y., Mahajan, A., da Silva, C.C., Kulse, M. and Raabe, D., 2022. Green steel at its crossroads: Hybrid hydrogen-based reduction of iron ores. Journal of Cleaner Production, 340, p.130805.

Spreitzer D., et al., 2022. HYFOR - Hydrogen-based Fine Ore Reduction. Mitsubishi Heavy Industries Technical Review Vol. 59 No. 2. Available Online.

World Steel Assocation, 2024. World Steel In Figures 2023. Available Online

Zheng, J., Zhou, Y., Yu, R., Zhao, D., Lu, Z. and Zhang, P., 2019. Survival rate of China passenger vehicles: A data-driven approach. Energy Policy, 129, pp.587-597.