At the end of July 2026, ArcelorMittal restarted blast furnace No. 1 at Fos-sur-Mer, which had been idle since September 2023. BF No. 2 had returned in December 2025 after a fire-related stoppage. With the furnaces also brought back at Asturias in Spain and Dabrowa Gornicza in Poland, approximately 7 million tonnes per year of pig-iron capacity have returned to the European market.

Northwest European hot-rolled coil (HRC) was assessed at €725/t ex-works on 31 July, €55/t higher than a month earlier. The restart therefore follows a visible improvement in domestic steel pricing, but it arrives while European demand remains weak. EUROFER's June outlook expected only modest growth in 2026 after three consecutive years of contraction.

The commercial case for restarting the furnaces rests partly on a market that has become more protected. The EU's new steel import system took effect on 1st July, reducing annual tariff-free quotas to 18.3 million tonnes and applying a 50% duty above quota. For buyers of imported coil, a lower quota and a much larger penalty make a domestic offer more attractive even when end-user consumption has not recovered strongly.

ArcelorMittal can therefore improve utilisation without assuming that automotive, construction and manufacturing demand will return to previous levels immediately. The risk appears later: if higher European output runs ahead of actual consumption, inventories can rebuild and weaken the price increase that supported the restart.

Returning BF-BOF capacity carries an ETS cost

The furnaces at Fos-sur-Mer, Asturias and Dabrowa Gornicza produce iron through the blast furnace/basic oxygen furnace (BF-BOF) route. Conventional BF-BOF production generally emits around 1.8-2.3 tonnes of CO2 per tonne of crude steel, although the result depends on the installation, feedstocks, energy recovery and accounting boundary.

Those emissions remain covered by the EU ETS. Free allocation limits the carbon cost borne by an efficient installation today, but the allocation for steel and other CBAM sectors is being withdrawn under the legislation currently in force.

The Commission's July ETS review proposal would make that transition slower: it proposes returning 15% of the allocation removed through the CBAM factor from 2028 and extending the phase-out to 2038. The proposal is not yet law, but it would reduce the near-term carbon-cost increase for operating European blast furnaces compared with the current schedule.

The policy logic is to preserve cash for industrial investment while electricity, renewable hydrogen and carbon infrastructure remain expensive or unavailable. It also reduces the immediate incentive to replace a working blast furnace: a slower withdrawal can support competitiveness and decarbonisation finance at the same time, but only if retained margins are converted into investable low-emission projects.

CBAM and the steel quota reward different attributes

The CBAM adds a carbon cost to imported steel, adjusted for the free allocation still available to comparable EU production. Its effect depends on the import's embedded emissions, the EU allowance price and any eligible carbon price already paid in the country of production.

CBAM can therefore distinguish between a high-emission and a low-emission imported tonne. An exporter using an electric arc furnace (EAF), low-emission electricity or gas-based direct reduced iron can retain a carbon-cost advantage over BF-BOF steel when the importer has accepted emissions data.

The tariff quota does not make that distinction. It restricts tonnes according to product and origin, irrespective of production route. Once a quota is unavailable, the 50% duty can outweigh the carbon advantage of a low-emission supplier.
This distinction explains why the current policy mix can improve the economics of conventional EU production before it creates a liquid market for low-emission steel.

Low-emission steel demand remains early

The International Energy Agency's 2026 steel assessment calls for demand policies capable of producing firm offtake commitments for near-zero and low-emission steel. Announced corporate demand has grown, particularly among automotive companies, but it remains small beside conventional steel consumption and is often expressed as a target rather than a binding purchase.

The Commission reached a similar conclusion in its ETS review analysis. It found that the carbon price and current regulation are not sufficient on their own to support many industrial decarbonisation investments, while demand for green products remains low or uncertain. The analysis cites estimates placing low-carbon steel procurement costs around 33-70% above conventional production, depending on the technology and scenario.

ArcelorMittal, thyssenkrupp Steel and voestalpine have asked for competitive power, affordable renewable hydrogen and stronger lead markets before committing further capital. Their position reflects producer interests, but the underlying constraint is visible across the market: public support can reduce project cost, while firm offtake is still needed to cover the operating gap after commissioning.

The Fos-sur-Mer restart is therefore evidence of a stronger short-term case for European steelmaking. If the returned furnaces operate at high rates while demand remains modest, the next test will be whether domestic HRC prices can absorb the additional output. If they do, ArcelorMittal gains cash and time. Whether that supports another cycle of conventional production or finances replacement capacity will depend on the investment decisions and customer commitments that follow.