Udine Day One: Europe's Steelmakers Have the Technology — What They Lack Is Clean Power, Scrap Discipline and Time

The first day of the AIST/Danieli European Steel Forum in Udine — held 27–28 October per the trade-press account, though the organisers list a 27–29 October programme — had the shape of a confession rather than a celebration. Speakers from Arvedi, SSAB, Stahl-Holding-Saar, Salzgitter, Marcegaglia, Primetals, SMS group and the lime, cement and energy sectors converged on a message that would have been heretical five years ago: the metallurgy is largely solved, and the timeline is not.

Three constraints dominated every panel: the price and volume of clean electricity, the quality and availability of iron feedstock, and a regulatory framework that must deliver carbon costs without delivering deindustrialisation. Arbitration between those three is now the real substance of European steel strategy. A fourth thread — trade policy — ran through the day as a caution against using climate instruments as a blunt defensive shield.

The commitments are real, and the arithmetic is unforgiving

The headline messages were less interesting than the numbers behind them. SSAB's Martin Pei pledged to keep replacing blast furnaces with EAFs and "expand our opportunities to use sustainable inputs"; Stahl-Holding-Saar's Sebastian Börner, presenting Power4Steel, tied output directly to power supply: "the more clean energy we have available, the more green steel we will be able to produce." Arvedi's Andrea Bianchi was blunter still on the limits: carbon neutrality is difficult, carbon capture "still appears impractical", and yet "we will not give up."

Those are not empty sentiments, but the capex and energy implications are rarely laid out together. The table below assembles publicly disclosed figures for the three most-advanced integrated conversions represented on stage.

Sources: SSAB transformation pages and 2025 annual report; EU INCITE/JRC entry for Power4Steel; SHS press release of 28 September 2026; Primetals and Arvedi disclosures on the Cremona ESP line.

What the electricity numbers actually mean

SSAB's own disclosure is the most useful single anchor for the whole debate. Its Nordic electricity demand rises from about 1.6 TWh/y today to roughly 4.5 TWh/y for the EAF conversion alone — a 2.8-fold increase — with a further ~5 TWh/y if commercial-scale HYBRIT sponge iron runs. Call it 8–9.5 TWh/y in total for one company in one region.

The Saarland data adds a dimension that panels usually miss: the energy mix, not just the volume. Today's BF-BOF route at Dillingen and Völklingen consumes 23.3 TWh/y, of which 21.1 TWh is coal. Under a 50:50 DRI-to-scrap charge, total energy falls to 19.7 TWh/y — but 10.4 TWh of that is electricity for hydrogen electrolysis. Push the DRI share to 80:20 and total energy rises again to 27.5 TWh/y.

That is the central technical insight of the whole transition, and it was implicit in Börner's remark. Decarbonising an integrated works does not simply swap a fuel; it converts a thermal energy system into an electrical one, moving the fuel supply burden onto the grid and onto electrolyser capex that the steelmaker does not own. A plant can finance a DRI furnace and an EAF. It cannot will 10 TWh of electrolytic hydrogen into existence on its own balance sheet.

Which makes the hydogen contract sizes worth stating plainly. SHS's ten-year agreement with Verso Energy, beginning 2029, covers at least 6,000 t/y of green hydrogen. Salzgitter's SALCOS agreement with EWE, finalised in June 2026, covers 10,000 t/y. Against a DRI plant that at full hydrogen operation would need orders of magnitude more, both contracts are first steps, not supply security. They are also, read correctly, hedge instruments — commitments that keep the green-hydrogen option alive and the subsidy dialogue open while the plants start on natural gas.

The unglamorous lever: scrap ratio as a design variable

The most revealing detail in the Saarland scheme is not the capex but the choice of a 50:50 DRI-to-scrap charge. It is an economic optimisation, not a metallurgical constraint — higher DRI shares raise both capital and operating costs. That says something important about where European steel is actually heading.

Scrap is the backbone of the transition. Mitsui's Naota Furihata gave the day its most quoted figures: global DRI/HBI trade of 12.4 Mt against 79 Mt for scrap. The DRI trade is roughly a sixth of the scrap trade and is expected to remain broadly balanced through 2030, with outcomes hinging on high-quality scrap availability and the pipeline of new DRI-based projects.

Analytical note: pellet projections of this scale should be treated with caution, since they aggregate announcements rather than firm FIDs. History suggests a sizeable attrition rate.

The feedstock problem has a second, harder edge that SMS group's Thiago Campos put precisely: high-grade iron ore is getting harder to source, and the industry must develop routes that process lower-grade material effectively. "Decarbonization is not just about the energy transition; it also presents a challenge in terms of raw material quality." This is the quiet bottleneck. DRI plants are sensitive to gangue, and the premium for DR-grade pellets is set to become a strategic cost line in the same way electricity is — a point that deserves far more attention than it got on the day.

Marcegaglia's warning: policy design is where this is won or lost

Antonio Marcegaglia's intervention deserves to be read as the day's most consequential policy argument rather than a grumble from a reroller. Three points, in ascending order of importance:

• Emissions have fallen while European production has fallen too — the transition is, so far, correlated with deindustrialisation rather than decoupled from it.
• The ETS should stay a compliance instrument and resist becoming a financial one, and targets must be technologically achievable.
• CBAM, properly construed, offsets the additional decarbonisation cost borne by European producers; if it becomes a protectionist wall, the effect is to shut in a market that leaks steel via finished goods and components, while closing export opportunities for EU mills.

That last point is the one that should preoccupy strategists. An isolated European market with a carbon cost on primary steel, but weak coverage of embodied steel in downstream imports, simply relocates value added rather than emissions. EUROFER's parallel push to extend CBAM to downstream steel products, and the European Parliament's own positioning on the mechanism, show that Brussels is aware of the gap. Marcegaglia's warning is not that the mechanism should be abandoned but that its scope design is the difference between a level playing field and a self-inflicted supply shock for European manufacturing.

The same logic applies to the safeguard measures he welcomed. Coordinated and pragmatic implementation, in his formulation, means avoiding a market so protected that European OEMs buy imported components instead of domestic coil.

The energy panel was a value-chain panel in disguise

The second session, on carbon-free energy, included Unicalce (lime) and Alpacem (cement) — not charity invitations. As Unicalce's Gabriele Imparato noted, roughly a third of global lime output goes to the steel industry; cement and steel also compete for the same infrastructure projects. These industries will bid for the same scarce volumes of affordable clean power and hydrogen, and they will be covered by the same CBAM architecture. Their decarbonisation timelines interact through shared infrastructure — pipelines, grid connections, storage.

The session's conclusions were deliberately unromantic. Imparato's "technology does not appear to be the problem; we need a system that gives us a clear direction in terms of regulation and investment" and Danieli Green Metal's Gabriele Guastaferro's point that market-ready technology must be adopted widely enough to become an industry standard both point to the same gap: deployment economics. TotalEnergies' Christian Koulic supplied the useful framing — the energy trilemma of affordable, sustainable and reliable supply, a reminder that reliability has been the neglected leg in European policy debate.

What to watch, and what would falsify the optimism

Ron Ashburn's "period of realignment" framing — security of supply alongside efficiency, overcapacity driving tariffs, decarbonisation and AI creating new demand — sets up the year ahead. The useful test is not whether European mills keep announcing EAF and DRI projects, which they will, but whether four measurable conditions converge:

1. Power contracts at industrial scale. The Verso and EWE deals are the right shape but a fraction of the required volume. Watch for contract sizes in the tens of thousands of tonnes per year of hydrogen, or for a decision to run DRI plants on a much higher natural-gas share for longer than planned.
2. The 2029–2030 start dates. Völklingen's EAF in Q2 2028 and Dillingen's DRI in Q3 2029 are fixed points. Any slippage is a direct read on grid connection and hydrogen availability, not on engineering.
3. Feedstock economics. DR-grade pellet premiums and the progress of lower-grade ore processing technologies will determine whether DRI costs converge or diverge.
4. CBAM scope. Whether the mechanism extends credibly to downstream steel content will decide whether European producers face a carbon cost advantage or a leakage penalty.

The Udine consensus — investment continues, timelines lengthen — is coherent, but it rests on a bet that clean energy will arrive on schedule. That is the one variable steelmakers cannot control, and the only one they cannot hedge except by keeping a foot in both the DRI and scrap routes. Which is, perhaps, exactly what a 50:50 charge ratio is.

Sources and notes on evidence

This article draws on a single trade-press account of the first day (SteelOrbis, 28 October 2026) for all programme details, quotations and panel composition; conference-specific claims should be read accordingly. Corporate and EU figures used for the capex, energy and feedstock analysis come from the primary sources listed below. The 2.8 Mt/y figure quoted by Arvedi in Udine should be read as ESP line output, below Primetals' stated 3 Mt/y upgrade target and the ~4 Mt/y capacity of the Cremona works as a whole. Dates of the forum differ between the trade press (27–28 October) and the organisers (27–29 October); the organisers' version is treated as authoritative.