Why Hydrogen Won’t Win The Zero-Carbon Steel Race
Last Updated on: 29th June 2025, 11:19 am
Recent adjustments to my projections for global steel demand through 2100, reflecting a significant slowdown in Chinese infrastructure and cement consumption, have sharpened my economic focus on competing new steelmaking technologies. With lower growth trajectories for steel firmly established, every ton produced in the coming decades will increasingly face stringent scrutiny around cost, carbon intensity, and technological feasibility.
Four + one emerging steelmaking routes are central to this analysis: molten oxide electrolysis (MOE), electrified biomethane-based direct reduced iron (DRI) coupled with carbon capture and storage (CCS), hydrogen-based DRI, and flash ironmaking with either natural gas coupled with CCS or hydrogen. A sober economic assessment, grounded firmly in realistic assumptions about electricity costs, fuel pricing, and carbon policies, offers clear insights into their respective prospects.
As a note, all costs are ones I’ve assembled from multiple sources, adjusted for emerging sensible industrial electricity pricing policies, and in a simple spreadsheet model. They are less wrong than most, but for any specific solution in any specific location, they should be considered at best indicative of the future, with Northeastern Europe with industrial electricity prices post-spark gap sanity prevailing and Australia with a massive wind + solar + storage remote industrial site acting as examples, not a true technoeconomic assessment.
First, we must confront the persistent economic reality that hydrogen-based steelmaking has failed to achieve cost parity due to wildly optimistic projections of cheap green hydrogen from five to ten years ago. The widespread assumption that renewable electricity would rapidly drive electrolyzer and hydrogen costs to ultra-low levels has not materialized. Instead, real-world hydrogen prices remain stubbornly high, ranging from $5 to $8 per kilogram or higher in most developed countries, and still $3 to $4 per kilogram even in renewable-rich areas like Australia or Chile.
These costs effectively make hydrogen-produced iron significantly more expensive than other emerging low-carbon routes. Hydrogen’s substantial operational costs come largely from its fundamental inefficiencies: electrolytic hydrogen production, storage, and compression require large amounts of electricity, translating directly into high costs at current electricity prices. Assumptions of low hydrogen prices combined effectively free electricity from otherwise curtailed renewables with effectively free electrolysers combined with absent balance of plant capital costs combined with assumptions of high flexibility on the part of electrolysers, resulting in the illusion of low costs.
Each of those assumptions was at best dubious, and each has been heavily challenged by exposure to reality, and so real world green hydrogen costs have remained stratopheric. Hydrogen steelmaking emerges as economically uncompetitive and faces structural barriers rather than temporary or easily overcome market conditions.
Turning to molten oxide electrolysis (MOE), this technology offers an intriguing economic contrast. MOE uses electricity directly to split iron oxide into liquid iron and oxygen, bypassing the need for hydrogen, natural gas, or carbon-based fuels. The technology, notably advanced by Boston Metal, has moved past laboratory scale and into pilot demonstrations. It appears increasingly robust as a viable industrial method.
With roughly four megawatt-hours of electricity needed per ton of steel, MOE’s economics hinge predominantly on electricity costs and electrode longevity. Under realistic future assumptions — industrial electricity prices around five cents per kilowatt-hour in northern Europe and dedicated, firmed, blended renewable electricity around $0.03–$0.04 per kilowatt-hour in Australia at or near major mine sites — the resulting costs per ton are compellingly competitive. In regions abundant in wind and solar, particularly parts of Australia, Chile, or North Africa, MOE becomes the lowest-cost option among all zero-carbon pathways, with total costs around $170 per ton. In regions like Europe, where electricity prices remain slightly higher, MOE still fares well, though less decisively, at around $250 per ton.
Electrified biomethane-based DRI combined with carbon capture, meanwhile, occupies an economically powerful niche. This route starts with proven natural gas-based direct reduction technology, electrifies process heat to reduce methane feedstock consumption by 20% to 25%, substitutes fossil natural gas with biomethane, and incorporates full carbon capture on its concentrated, 95% pure, 30° to 60° Celsius CO₂ stream.
When deployed in regions with plentiful waste biomass, accessible geological or offshore CO₂ storage, and robust nearby demand for captured CO₂, such as the Netherlands’ industrial greenhouses, the economics become remarkably favorable. The table above is with $100 / ton sequestration costs, but in fact Tata’s steel plant has a 5 million ton per year offtaker for its captured CO2 and would likely see revenue per ton instead of a cost per ton.
Negative-emission credits for permanently storing biogenic CO₂ would reduce net costs as well. Tata could also leverage North Sea geological storage infrastructure, securing significant carbon offset revenues. Despite higher biomethane costs compared to fossil natural gas, often around $20 per gigajoule, biogenic CO₂ credits transform the economics, making biomethane DRI competitive at around $245 per ton even under higher CCS cost scenarios. Wherever significant anthropogenic methane emissions from biomass waste and inexpensive sequestration sites converge, and carbon pricing valorizes the negative emissions, electrified biomethane DRI with CCS emerges as economically dominant.
Flash ironmaking, which rapidly heats finely ground iron ore concentrate using pure oxygen combustion, occupies yet another interesting economic niche. Flash reactors significantly simplify traditional steelmaking plant complexity, eliminating the need for pelletization or sintering of iron ore and reducing capital expenditure. When fired with natural gas and combined with CCS, this approach becomes attractive in regions endowed with abundant natural gas reserves and geological storage.
For instance, Australia’s relatively low natural gas costs (approximately $6 per gigajoule) and well-characterized CCS potential position flash ironmaking competitively. Even with relatively high CCS costs of around $100 per ton of CO₂ captured, total steel production costs can remain below $200 per ton. Conversely, flash ironmaking with hydrogen combustion, despite producing zero direct CO₂ emissions, struggles with the same high hydrogen pricing that plagues pure hydrogen DRI, pushing costs substantially higher and making it less economically viable compared to its natural gas counterpart or the simpler MOE approach in renewable-rich regions.
Each of these technologies also varies notably in terms of technology readiness levels (TRL), presenting differing risks and timelines. Electrified biomethane DRI coupled with CCS enjoys a relatively high TRL, building directly on existing gas-based DRI systems that have operated successfully for decades. This reduces technological and integration risks significantly, allowing near-term scale-up in suitable locations.
Molten oxide electrolysis, while progressing rapidly through pilot-stage demonstrations, remains at a moderate TRL, likely to reach 6 in 2026 with current pilot plant construction. Key engineering challenges, particularly around electrode durability and materials stability under extremely high temperatures and high-temperature oxygen waste streams, persist. Nevertheless, rapid advancements by companies like Boston Metal indicate these hurdles are solvable, though scale-up risks remain noteworthy.
Flash ironmaking is similarly at moderate TRL, having shown successful pilot results. Its primary uncertainty centers around scaling reactor designs and integrating continuous oxygen supply and carbon capture into large-scale operations. It is based on flash copper smelting, a proven technology, so the development and technology risks are relatively low, with only economics and scalability to steel volumes being concerns.
Hydrogen-based DRI technology itself is technically mature; operational demonstrations exist worldwide. Yet, despite high technical readiness, its fundamental economic challenges severely constrain practical deployment.
Regional conditions will dictate the choice between these technologies more decisively than ever. In biomass-rich, CCS-ready regions like the Netherlands, electrified biomethane DRI will likely become the technology of choice, transforming an environmental liability, anthropogenic methane emissions, into a valuable resource.
In renewable-rich, biomass-poor regions with limited or expensive carbon sequestration options, such as large parts of Australia, MOE’s simplicity and direct electrification will likely dominate. In areas with abundant natural gas and affordable geological CCS, flash ironmaking coupled with carbon capture holds promise. Once again, flash ironmaking’s CO2 stream, like biomethane direct reduction’s, is fairly pure and concentrated, a key requirement for any CCS solution. Pure CO2 streams plus local, cheap sequestration that doesn’t require piping dense phase CO2 through populated areas is the combination to look for.
But across nearly all conceivable contexts, hydrogen-based steelmaking remains stubbornly uneconomic under realistic assumptions. Its persistent inability to achieve the low hydrogen costs once predicted sharply curtails its viability, positioning it as the least competitive option in nearly every realistic scenario. While it might pencil out in theory in renewables rich areas like Chile and Australia, any solution which doesn’t have to deal with the complexity and challenges of hydrogen will be simpler and likely cheaper and more robust.
This economic clarity has direct implications for policy and investment strategies. Policymakers and industrial investors must carefully avoid the alluring yet economically dubious promises of cheap hydrogen-driven steelmaking. Instead, they should strategically invest in regionally tailored steel decarbonization pathways aligned closely with local resource strengths. For biomass and CCS-equipped regions, incentives for biomethane-based steelmaking make clear economic and environmental sense. Renewable-rich regions should prioritize infrastructure and incentives for direct-electrification approaches like MOE. Natural gas-rich locations, especially with accessible geological storage, may sensibly explore flash ironmaking to leverage existing infrastructure and resources effectively.
This is more bad news for hydrogen maximalists and pipeline operators. Green steel was the only major growth area for hydrogen demand in my update of my hydrogen demand projection through 2100 from two years ago. Approximately 40 million tons per year were earmarked for that use case, at 56 kg per ton of new steel and a major subset of the roughly 400 million tons of new steel I saw required per year when scrap steel was providing the majority of the market. But with all of the alternatives being cheaper, even if two are currently lower technology readiness level, I don’t see any demand for hydrogen green steel. Too complex, too expensive, outcompeted, just as in transportation and energy storage.
This new projection is much more aligned with outcomes of the scenario planning workshops for TenneT, where early indications are that the Netherlands’ current hydrogen demand will drop by 80% in the future, and be fulfilled with current industrial byproduct hydrogen and an autothermal reformer under construction. Where hydrogen is needed and carbon pricing valorizes negative emissions, thermolysis of biomethane produces one ton of hydrogen and four tons of solid carbon, which can be buried for permanent atmospheric carbon drawdown.
Once again, anthropogenic biomethane from our food, agricultural and forestry systems is a major greenhouse gas emitter, larger than methane emissions from all of the oil and gas industry. We have to minimize it through obvious measures such as ungulate dietary supplements and vaccines which reduce methane production, and we have to capture it at whatever point sources remain. Turning waste biomass into biomethane in industrial biodigesters for industrial feedstocks like green steel and hydrogen production is, I’ve now accepted, a requirement. More on that in an upcoming piece, along with an analysis of what that likely means systemically, where it’s problematic.
While bad news for hydrogen maximalists and pipeline operators, it’s good news for the rest of us. Hydrogen manufacturing today is a greenhouse gas emitter on the scale of all of global aviation, a major climate problem. If we had to fix all of that, we would require absurd scales of renewables and electrolyzers running constantly. If hydrogen demand grew, it would eat all green electricity for decades. And if hydrogen became a distributed commodity, its significant leaks and indirect global warming potential would eliminate a significant portion of the purported benefits.
As we adjust our expectations downward for future steel demand globally, achieving economically rational steel decarbonization becomes paramount. It is vital to base technology choices not on optimistic projections or theoretical hydrogen cost curves, but on realistic economic and technological assessments.
Hydrogen-based steelmaking, despite its strong policy push and early momentum, has not delivered on its economic promise and appears unlikely to do so in the foreseeable future. Instead, practical regional strategies — leveraging biomethane, CCS where viable, renewable electricity, and limited natural gas — will define the steel industry’s transition toward sustainable, economically viable zero-carbon production. The future of steelmaking will be diverse, pragmatic, and firmly grounded in economic realities rather than hopeful illusions.
Sign up for CleanTechnica's Weekly Substack for Zach and Scott's in-depth analyses and high level summaries, sign up for our daily newsletter, and follow us on Google News!
Whether you have solar power or not, please complete our latest solar power survey.
Have a tip for CleanTechnica? Want to advertise? Want to suggest a guest for our CleanTech Talk podcast? Contact us here.
Sign up for our daily newsletter for 15 new cleantech stories a day. Or sign up for our weekly one on top stories of the week if daily is too frequent.
CleanTechnica uses affiliate links. See our policy here.
CleanTechnica's Comment Policy
