Mercedes-Benz used its November 2026 sustainability update to announce two supply-chain moves. The one that drew attention was aluminium: a binding long-term offtake with Hydro plus membership of the First Movers Coalition. The steel item — a memorandum of understanding with HyIron, following a pilot that reduced Sishen iron ore fines directly with hydrogen in a rotary kiln — was quieter, and deliberately so. It is a technical validation agreement, not a purchase commitment.
That distinction matters. The pilot itself is genuinely interesting for reasons that have little to do with the car it ended up in.
What was actually demonstrated
Anglo American's Kumba Iron Ore supplied fines from the Sishen mine in the Northern Cape. HyIron reduced them with hydrogen in its proprietary rotary kiln, bypassing pelletisation entirely. The resulting DRI went into a steelmaking step and emerged as a demonstrator component for the Mercedes-Benz EQS.
The technical claim worth examining is the elimination of the pellet plant. Today fines are agglomerated into pellets — an induration step fired at roughly 1,250–1,350°C — before they can feed a shaft furnace. That step is energy-intensive, capital-heavy and typically coal- or gas-fired, and it is the reason why the upstream emissions of conventional hydrogen DRI are not zero even when the reductant is green. Pellets also constrain the feedstock base: a DRI plant is tied to a pellet supplier, and in the Atlantic market that has meant a small group of suppliers with limited leverage for buyers.
HyIron's route removes that link. According to the company's own description, ore is crushed and milled, magnetically beneficiated dry, and fed to an airtight rotary kiln where hydrogen strips the oxygen. Process water is recovered and returned to electrolysis. The commercial attraction is straightforward: a broader, cheaper feedstock base, and one less high-temperature, high-capex unit in the chain.
| Parameter | Conventional BF-BOF | Gas-based shaft DRI + EAF | HyIron rotary kiln + EAF (pilot) |
|---|---|---|---|
| Feedstock | Fines via sinter/pellets | Pellets (lump ore in some units) | Fines directly, no pelletising |
| Reductant | Coke/coal | Natural gas, increasingly H₂ | Hydrogen |
| Upstream agglomeration | Sinter + pellet plants | Pellet plant | None (dry beneficiation) |
| Stated Scope 1+2 (ironmaking) | ~1.8–2.0 t CO₂/t Fe (typical BF) | ~0.4–0.7 t CO₂/t DRI (gas, plant gate) | 10 kg CO₂/t DRI (HyIron claim) |
| Scale demonstrated | 1–5 Mt/y per furnace | 1–2.5 Mt/y per module | Pilot kiln (~0.5 t/h at Lingen) |
Figures for BF and gas DRI are indicative industry ranges for context; the HyIron figure is the company's own claim, not independently audited.
Reading the 10 kg figure correctly
The headline number — 10 kg of CO₂ per tonne of DRI on a Scope 1+2 basis — is attributed to HyIron and carried by Mercedes-Benz. It deserves careful handling for three reasons.
First, it is Scope 1+2 at the plant gate. It excludes mining and beneficiation at Sishen, the ~9,000 km ore voyage to Europe, hydrogen production losses upstream of the plant boundary if electricity is accounted separately, and the downstream EAF and casting steps. A full cradle-to-gate steel number would be materially higher — though the hydrogen input, if genuinely green, is the dominant remaining variable.
Second, it is a pilot figure. Heat losses, hydrogen utilisation and kiln availability all deteriorate with scale, and no continuous industrial operating record exists yet for this configuration.
Third, it says nothing about cost. Hydrogen at scale — even in Namibia or South Africa with strong solar resources — remains the largest single line in the DRI cost stack. A 10 kg CO₂/t product made with expensive hydrogen is a decarbonisation achievement, not yet a commercially competitive one.
Scale is the unanswered question
HyIron's reference points are modest. The Lingen pilot in Germany, run under the GEiSt project with Benteler as steel partner and RWE supplying energy, is rated at roughly 500 kg of DRI per hour — about 4,000 t/y at continuous operation. The Oshivela plant in Namibia, inaugurated in April 2025, started at a planned 15,000 t/y. HyIron has outlined a staged ramp to 200,000 t/y and, eventually, million-tonne scale.
For perspective, a single mid-size European EAF consumes on the order of 1 Mt of metallics a year. The gap between a validated kiln and a plant that can supply an automotive supply chain is roughly two orders of magnitude. Rotary kilns do scale — the coal-based SL/RN processes proved it — but heat transfer, temperature uniformity and hydrogen containment in a rotating vessel become progressively harder engineering problems with diameter.
Mercedes-Benz's own framing is honest about this: the MoU covers "technical validation towards industrial scale-up and potential next steps." That is a research and derisking mandate. The aluminium agreement announced alongside it — binding, with stepped carbon-intensity thresholds through 2030 — is the template for what a real supply commitment looks like. Steel is not there yet.
The missing link: nobody named the steelmaker
A notable gap in the public record is that no steelmaker, EAF operator, forger or component supplier is identified as the party that converted the pilot DRI into steel and then into the EQS demonstrator. Given that Mercedes-Benz does not make steel, someone in the chain did. The omission is worth flagging because converter availability is precisely the bottleneck in automotive green steel procurement: DRI is not a saleable automotive product, and the EAF-plus-downstream route for a component as specification-sensitive as a suspension or structural part requires qualification work that a pilot batch does not resolve.
HyIron's Namibian and German projects both involve steel partners — Benteler at Lingen — but the Mercedes announcement does not say whether a European or South African converter handled the pilot heat. Anyone modelling the supply chain should treat that step as undisclosed.
Why the ore source matters as much as the process
Sishen is a hematite deposit with a relatively high iron grade and low gangue — useful for direct reduction, and Kumba has separately been developing direct-reduction-grade lump products. But the announcement's inclusion of IRMA75 certification for Sishen and Kolomela is not incidental. Mercedes-Benz has been an IRMA member since 2021, and its procurement narrative explicitly links responsible sourcing to emissions. Kumba states these are the only iron ore mines in Africa at that level.
The pairing is deliberate: the argument to regulators and investors is that lower-emission iron need not come at the expense of mining standards. For Kumba, a fines-to-DRI outlet would be strategically valuable — it would convert a blast-furnace-dependent product stream into a direct-reduction feedstock without the pelletising investment, and open a premium European customer base as the EU's carbon border mechanism raises the cost of high-embodied-carbon imports.
What to watch
The MoU's value will be judged on a small number of observable milestones rather than on the announcement itself:
- Kiln scale-up evidence. Continuous operating data from Oshivela's ramp, and whether HyIron commits capex to a unit large enough to matter to a steelmaker.
- Hydrogen economics. The delivered cost of green H₂ at the reduction site, and whether South Africa's and Namibia's hydrogen programmes move from pilot to bankable supply.
- A named converter. Until a steelmaker is attached to the chain, the route is a materials-science result rather than a supply chain.
- Mercedes-Benz's next move. Whether steel follows aluminium into a binding offtake with carbon-intensity thresholds, and on what timeline. The 2039 value-chain target — a 70% reduction against 2019 — leaves room, but flat steel is a far larger and more fragmented spend category than the aluminium volumes involved.
For the metals industry, the interesting signal is not that an automaker tested green iron. It is that a fines-fed, pellet-free reduction route has cleared a technical gate with an OEM's engineering function behind it. That is a credible first step. The second step — a kiln big enough to feed an EAF, running on hydrogen that someone has agreed to pay for — is the one that determines whether this becomes a supply chain or stays a demonstrator.
Source & methodology +
MetalMate separates original-source evidence from summary and analysis. Verify the original source before procurement, legal or engineering decisions.
- Article status
- Published · Version 1
- Source date
- 12 November 2026
- MetalMate publication
- 14 November 2026
- Verification
- Automated evidence checks; not human reviewed

