Steel’s Carbon Bill Is Coming Due
Steel built the modern world. It is in the bridges, the skyscrapers, the wind turbines, the hospital beds, the cars, the shipping containers carrying every item of traded goods across every ocean. It is also, by some measures, the single hardest industrial sector on the planet to decarbonise — and the one with the least room left to delay.
Every tonne of steel produced today generates, on average, 2.18 tonnes of CO₂. Multiply that across the 1.886 billion tonnes of steel the world produced in 2024, and you arrive at a sector responsible for approximately 4.1 billion tonnes of greenhouse gas emissions annually — somewhere between 7% and 8% of the entire global total. That figure does not come from a campaign group. It comes from the World Steel Association. The industry knows exactly what it is carrying.
What is changing, rapidly and with genuine financial consequence, is who pays for it.
The Free Ride Ends
For most of its history under European carbon pricing, the steel industry has operated in a comfortable exception. The EU Emissions Trading Scheme — the world’s largest compliance carbon market — allocated free allowances to energy-intensive industries deemed at risk of “carbon leakage,” the polite regulatory term for producers packing up and relocating to jurisdictions with no carbon price. Steel manufacturers received enough free allowances that, for the most efficient among them, the carbon price was essentially academic. In 2025, 79 of the 356 steel producers covered by the EU ETS remained largely unaffected by carbon pricing due to their above-average efficiency and the surplus of free allowances they received annually. The other 277 were already feeling the pinch.
That calculation is now unwinding, deliberately and by design. Free allowances for the steel sector will be phased down to zero by 2034, with the reduction beginning in earnest from 2026 onwards. At the same time, the Carbon Border Adjustment Mechanism — the EU’s landmark border carbon pricing policy — became fully operational on 1 January 2026, requiring importers of steel into the EU market to purchase certificates reflecting the embedded carbon cost of their products. The transitional reporting phase is over. The financial obligations are live.
The combined effect of these two mechanisms — shrinking free allocations and a border adjustment that levels the playing field for non-EU producers — represents the most significant structural shift in the economics of European steelmaking in decades. Carbon costs, once a peripheral line on the balance sheet, are becoming a central determinant of competitive position.
What a Tonne of Carbon Actually Costs a Steel Mill
The numbers are stark enough to warrant stating plainly. EU ETS allowances ranged between €60 and €80 per tonne through 2025, with analyst projections pointing to €91–93 in 2026 and a credible path towards €100 or beyond as the decade progresses. For a traditional blast furnace steelmaker emitting somewhere in the range of 1.8 to 2.5 tonnes of CO₂ per tonne of crude steel — depending on efficiency, feedstock, and energy mix — that translates to a carbon cost of roughly €110 to €250 per tonne of steel produced, once free allowances are fully phased out. Against an average European steel price that has historically ranged between €600 and €900 per tonne for flat products, that is not a rounding error. It is a margin-defining number.
The CBAM adds a further dimension for producers selling into the EU from outside. The mechanism imposes costs of roughly €20 to €80 per tonne on high-carbon steel imports, depending on their embedded emissions intensity. For producers in countries with no domestic carbon pricing — China’s national ETS currently trades at approximately ¥80–90 per tonne, barely 10% of EU carbon prices, and the steel sector was only brought into scope in 2024 — the CBAM effectively creates a tariff that mirrors the competitive disadvantage European producers have long complained about. Whether it will achieve its goal of driving global steel decarbonisation rather than simply redirecting trade flows is one of the more consequential open questions in international industrial policy.
Two Worlds, One Metal
The carbon pricing dynamic has carved the global steel industry into two increasingly distinct realities. In Europe and, to a growing extent, the United Kingdom — where the UK ETS mirrors EU allowance prices and the two systems are now moving towards formal linkage following the May 2025 announcement — carbon cost is an operational reality that reshapes investment decisions, procurement strategies, and long-term capital allocation. In China, which produces well over half the world’s steel, and in India, the Middle East, and Southeast Asia, the story is almost entirely different.
China’s position is particularly important and genuinely ambiguous. The 14th Five-Year Plan targeted peak steel production and sectoral emissions before 2030 and expanded ETS coverage to the steel sector in 2024. But the implementation is output-based rather than cap-based in the initial phase, running through 2026, which means the system is primarily familiarising companies with carbon accounting rather than imposing meaningful financial pressure. Chinese carbon prices remain a fraction of European levels. Meanwhile, more than 50 million tonnes of new conventional blast furnace capacity was announced for the 2024–2026 period — suggesting continued confidence in traditional steelmaking economics even as pilot green steel investments simultaneously accelerate. Baowu Steel’s 2-million-tonne hydrogen-based direct reduction project in Inner Mongolia and China’s lead in green steel pilot investment sit alongside this conventional capacity expansion in a posture that is best described as strategic hedging rather than committed transition.
The 15th Five-Year Plan, currently under development for the 2026–2030 period, will be revealing. If Beijing commits to steel-specific carbon constraints with genuine price pressure, it will send a seismic signal through global steel markets. If it extends the dual-track approach, the divergence between European and Asian steel economics will continue to widen — and the CBAM will become an increasingly contested instrument in global trade politics.
The Green Steel Gamble
Into this landscape steps a technology that the industry has been discussing for years and is now, finally, beginning to build: hydrogen-based direct reduction of iron, or H2-DRI, paired with electric arc furnaces. The concept is straightforward enough — replace coking coal in the ironmaking process with green hydrogen, eliminating the process carbon emissions at source — but the engineering and economic challenges have been formidable enough to keep it largely in the pilot phase until recently.
That is changing. In Boden, Sweden, Stegra — formerly known as H2 Green Steel — is advancing towards a 2026 start-up of what would be the world’s first commercial-scale green steel mill powered entirely by renewable energy and hydrogen DRI, with an initial production capacity of 2.5 million tonnes per year. The HYBRIT joint venture between SSAB, LKAB, and Vattenfall has successfully demonstrated the hydrogen-based process at pilot scale and is moving towards demonstration-scale production, with trial heats already used by Volvo for truck components. In Germany, both Thyssenkrupp and Salzgitter are developing hydrogen DRI plants on timelines running from 2026 to 2028, representing combined investments in the range of €5–8 billion. Stegra has pre-sold volumes to Mercedes-Benz, BMW, and other premium manufacturers — a signal that the downstream market for certified low-carbon steel at a price premium is real, at least at the high end of the automotive supply chain.
The economics, however, remain genuinely contested. The green steel route is not cheap. Green hydrogen — produced by electrolysing water using renewable electricity — currently costs roughly three times more per kilogram than grey hydrogen produced from natural gas. Most analysts project green hydrogen reaching $2.00 to $2.50 per kilogram by 2028–2032 in the most favourable locations — those combining high renewable energy yield with access to DR-grade iron ore — declining further towards $1.50–$2.00 by 2035. At current prices, the premium for green steel over conventional blast furnace steel is estimated in the range of €100–$200 per tonne, depending on energy costs and hydrogen prices. Consumers in premium automotive and white goods manufacturing are willing to pay this. Commodity construction steel buyers are considerably less so.
There is also a more pointed academic challenge to the subsidy logic underpinning European green steel investment. Research published in 2025 using a general equilibrium cost-benefit framework found that, based on a case study of a large-scale hydrogen steel plant in northern Sweden, baseline scenarios indicated significant social losses, with the authors raising the spectre of a repetition of the 1970s European steel crisis driven by subsidised overcapacity in an uncompetitive technology. This is a minority view, and it carries the methodological caveats of any general equilibrium model applied to a fast-moving technology landscape. But it is a useful corrective to the narrative of green steel as an obvious inevitability. It is a bet, not a certainty — and the size of the bets being placed is very large.
The Role Carbon Credits Play — and Don’t Play
Given the scale of the industry’s emissions, there is an obvious temptation to ask whether carbon credits could play a more direct role in bridging the gap between today’s blast furnace economics and tomorrow’s hydrogen ambitions. The answer is: somewhat, at the margins, and with significant caveats.
Steel companies operating under the EU ETS primarily deal in compliance allowances — the regulated permits that correspond directly to their measured emissions — rather than in voluntary carbon credits. The question of whether and how the voluntary carbon market can complement the compliance architecture is one the industry is increasingly considering, particularly for residual emissions that prove technically or economically impossible to eliminate through direct process change. Carbon capture, utilisation, and storage fitted to existing blast furnaces or basic oxygen furnaces can capture CO₂ at costs in the range of $40–$120 per tonne — significantly below the durable removal credit prices of $187 per tonne for biochar and $320 or more for direct air capture, but requiring substantial capital investment in infrastructure that most steel companies have other calls on.
The more immediate carbon credit dynamic for steel is not about buying offsets. It is about the price of carbon allowances reshaping the competitive economics of different production routes in real time. At €80–100 per tonne, the ETS price makes the blast furnace route progressively more expensive relative to the electric arc furnace route, which produces steel from scrap and carries a fraction of the direct emissions intensity. EU modelling suggests that secondary steel production through EAF — which currently runs at approximately 54 million tonnes per year in Europe — could increase to 68 million tonnes by 2035 as scrap availability grows and the carbon cost advantage of the route compounds. This is not green steel in the hydrogen sense, but it is a lower-carbon pathway that the carbon price is actively incentivising, and it is happening faster than the headline debate about hydrogen DRI would suggest.
The Competitive Fracture
What carbon pricing is creating, in effect, is a fracture line running through the global steel industry. On one side, producers operating in jurisdictions with meaningful carbon prices face rising costs, shrinking free allocations, and investment decisions that must account for a carbon price trajectory pointing towards €100 and beyond. On the other, producers in lower-regulation environments face none of these pressures — at least for now. The question of how durable this divergence proves to be is probably the central strategic question in global steel for the rest of the decade.
The CBAM is the EU’s answer to that question, and its full operationalisation in 2026 marks the point at which the theory becomes testable. If it succeeds in imposing equivalent carbon costs on imported steel without triggering a wave of WTO challenges and trade retaliation, it will have achieved something unprecedented in the history of border carbon pricing. If it becomes embroiled in political controversy — particularly as US trade policy under the current administration has shown limited sympathy for multilateral environmental mechanisms — it could face significant implementation headwinds.
For the UK, the strategic picture is complicated by the need to maintain alignment with EU carbon pricing to avoid the CBAM applying to British steel exports to European markets, whilst simultaneously managing the domestic political economy of an industry that employs tens of thousands and sits at the heart of several of the UK’s most economically fragile communities. The Port Talbot closure — and the broader question of whether British steelmaking has a viable long-term future in electric arc form — is the sharpest domestic expression of a tension running through the entire global industry.
Where the Investment Is Going
The capital flows tell a reasonably clear story, even if the precise destinations are still being negotiated. The global direct-reduced iron market, a proxy for the direction of new steelmaking investment, was valued at $32.4 billion in 2025 and is projected to reach approximately $64.77 billion by 2034 — an 8% compound annual growth rate driven by the shift towards lower-carbon EAF-based steelmaking and the growing availability of hydrogen-ready DRI technology. The number of hydrogen DRI pilot rollouts increased by 28% between 2024 and 2025 alone.
The critical dependency is green hydrogen. The IEA’s roadmap for steel sector decarbonisation requires widespread adoption of hydrogen-based routes in the 2030s, but this depends on green hydrogen reaching cost-competitive levels — around $2 per kilogram or below — across a broad range of geographies. That in turn depends on the buildout of renewable energy capacity and electrolyser manufacturing at a pace that most independent analysts currently regard as ambitious relative to the actual installation rate. Research modelling competitive locations for green hydrogen steel production finds the most favourable sites located near the Tropics of Cancer and Capricorn — characterised by superior solar resources, high-quality iron ore, and lower labour costs — rather than in the European industrial heartlands where most of the capital is currently being deployed. This geographic mismatch between where competitive green steel can be produced and where the political will to support the industry is concentrated may prove to be the single most consequential structural challenge in the sector’s transition.
The Decade Ahead
Steel decarbonisation is simultaneously more urgent, more technically feasible, and more economically uncertain than the public narrative tends to suggest. The urgency is real: an industry responsible for 7–8% of global emissions cannot be exempt from the transition to net zero, and the policy mechanisms designed to make that transition financially unavoidable are now operational. The technical feasibility is genuine: hydrogen DRI works at pilot scale and is moving into commercial demonstration. The economic uncertainty is also genuine: green steel at scale requires a hydrogen cost trajectory, a carbon price trajectory, a scrap availability trajectory, and a premium steel demand trajectory to all move in the right direction simultaneously over a fifteen-year investment horizon.
What carbon pricing has achieved, most concretely, is shifting the question from whether steel decarbonises to when and at what cost. The companies best positioned for the decade ahead are those that have begun treating carbon as a structural input cost rather than a compliance afterthought — those that have locked in long-term power purchase agreements, assessed their ETS exposure with free allowance phase-down modelled in, and made honest assessments of which production assets are viable in a €100-per-tonne carbon world and which are not.
The free ride that European steelmakers have enjoyed for much of the ETS’s existence is ending by regulatory design. The CBAM means that the producers outside Europe who have never paid a carbon price are about to encounter one at the point of sale. And the technology that might ultimately resolve the tension between the industry’s economic centrality and its emissions profile is beginning, finally, to leave the laboratory.
The bill is arriving. The question is whether the steel industry pays it on its own terms or has the terms set for it.