Nippon Steel's 100% hydrogen DRI run matters less for its tonnage than for the ore grade it used

Nippon Steel's 4 November announcement that its Hasaki pilot shaft furnace produced more than 1,000 t of direct reduced iron using 100% hydrogen and blast-furnace-grade pellets has been reported largely as a tonnage milestone. The more consequential claim is buried in the raw material: the ore.

Commercial shaft furnaces — the Midrex and Energiron fleet that produced roughly 100 Mt of DRI worldwide in 2025, almost all of it on natural gas — are designed around high-grade, low-gangue DR pellets, typically above 67% Fe. Nippon Steel's pilot ran on what the company calls "blast furnace grade" pellets, a materially cheaper and more abundant feedstock. If that can be made to work at scale, it changes the feedstock map for green iron, not just the reductant.

What was actually demonstrated

Two technical obstacles were cited. First, hydrogen reduction is endothermic, so the furnace loses heat that natural-gas reduction partly supplies through the exothermic shift reaction; Nippon Steel says it applied "optimal thermal compensation technology" to hold furnace temperature. Second, lower-grade ore powders during reduction, degrading gas permeability and upsetting the reaction profile — a problem the industry usually sidesteps by buying premium pellets. The company claims proprietary operating techniques suppressed that powdering.

Both claims are plausible and consistent with the direction of the GREINS programme, which describes external heating of the reducing gas as the key departure from conventional shaft furnace practice, and with NEDO's earlier reporting that hydrogen preheating was still being optimised. But Nippon Steel has disclosed no metallisation degree, no Fe grade, no gangue or fines-generation data and no energy consumption per tonne. Without those numbers, the result establishes operability, not competitiveness.

A note on the arithmetic

The two headline numbers do not describe a single campaign. At a nameplate 1 t/h, a 300-hour run yields roughly 300 t; producing 1,000 t requires something close to 1,000 operating hours. The release refers to more than 300 hours of continuous operation and more than 1,000 t cumulative — the latter presumably accumulated across campaigns since the March 2026 start-up, including the 72-hour continuous run that NEDO reported earlier. This is an inference, not a company statement, but it means the durability test is the 300-hour run and the tonnage figure reflects cumulative throughput.

The real significance: feedstock flexibility

Green iron's cost problem has been framed almost entirely around hydrogen price and electricity. Feedstock has been treated as a given — which is why the DR-grade pellet market carries a premium and why projects in ore-rich but pellet-poor regions have struggled.

If hydrogen-only shaft furnaces can tolerate blast-furnace-grade pellets, three things follow:

• Supply chain relief. Blast-furnace pellet supply is far deeper and more diversified than DR-grade, and it is produced in large volumes in Australia, Brazil and Canada. It also reduces exposure to the dr-grade pellet squeeze that has accompanied the wave of announced H2-DRI projects.
• Cost floor. Iron ore is the single largest input cost in DRI; a lower-grade, lower-priced pellet significantly shifts the economics of a hydrogen route that currently looks expensive against blast furnace ironmaking.
• Lessons for ore chemistry. Powdering and sticking behaviour are grade- and gangue-specific, so a result on one pellet source is not transferable to all. This is the point that most deserves press scrutiny, and it is the point the company has said least about.

The downstream problem nobody has solved at scale

Hydrogen-only DRI is essentially carbon-free, which is a benefit in the furnace and a problem afterwards. Conventional DRI carries 1.5–2.5% carbon, which lowers its melting point and helps it dissolve in a BOF or EAF. Carbon-free DRI melts at higher temperatures, is more prone to reoxidation and requires either a purpose-built EAF, a hot-briquetted iron route plus a melter, or substantial process redesign in existing steelmaking.

NEDO has flagged this explicitly. JFE Steel targets a DRI-optimised commercial EAF in FY2028, Nippon Steel in FY2029–30. The Hasaki result therefore validates the reduction step, not the route to liquid steel. Nippon Steel's three-track Carbon Neutral Vision 2050 — large-scale EAF, hydrogen injection into blast furnaces, and hydrogen DRI — is a hedge structure, and it is not yet clear which track gets the capital.

Why 2040 is the honest timeline

The company says it will advance scale-up development from FY2027, aiming to establish implementation technology around 2040. GREINS' stated goal is a medium-sized furnace — between one-fifth and one twenty-fifth of commercial scale — demonstrating at least 50% CO2 reduction versus the blast furnace by 2030. The gap between a 2030 medium-scale demonstration and 2040 implementation is a decade of engineering and, more importantly, of hydrogen infrastructure.

Analysis: The binding constraint on Japanese hydrogen DRI is not the shaft furnace. It is the cost and volume of clean hydrogen. Japan's domestic renewable resource is limited, and the country's hydrogen strategy leans heavily on imported ammonia and hydrogen carriers. Each tonne of DRI made with hydrogen-only reduction requires roughly 55–60 kg of H2 at stoichiometric minimum, and considerably more in practice once heating and recycling losses are included. With imported clean hydrogen likely to land in the range of $4–6/kg in the 2030s, the hydrogen cost embedded in a tonne of DRI could sit in the hundreds of dollars — well above the cost of metallics from a blast furnace. That arithmetic, not furnace metallurgy, will determine whether 2040 is achievable.

A useful comparison: European projects such as Stegra and the German H2-DRI schemes face the same hydrogen cost problem but sit alongside a growing renewable power base. Japan's choice is harder, which makes the ore-grade flexibility more valuable, not less.

What to watch

1. Disclosure of metallisation, fines rates and specific energy consumption for the 300-hour campaign. Without them the result cannot be benchmarked against Midrex or Energiron performance.
2. Which pellet sources were tested. A result on one low-gangue BF pellet does not generalise to high-alumina or high-phosphorus ores.
3. The scale-up decision in FY2027 — specifically the capacity of the next furnace and whether it feeds a melter or an EAF.
4. Hydrogen procurement. Any signal on whether Nippon Steel is contracting imported hydrogen or ammonia for a demonstration plant tells you more about the 2040 target than any furnace test.

Bottom line

Nippon Steel has demonstrated something technically non-trivial: a hydrogen-only shaft furnace can run for days on ore that the DRI industry normally rejects. That is a genuine advance and one of the more credible results to come out of the Japanese green steel programme. But the announcement is a proof of operability, not of economics, and the company has kept the data that would allow anyone to judge the difference. The 1,000-tonne figure will date quickly; the feedstock flexibility, if it holds at scale, will not.

References

• Nippon Steel, Successful Test Production of Direct Reduced Iron from Low-Grade Iron Ore Using 100% Hydrogen, 4 November 2026.
• Nippon Steel (Japanese original), 水素100%による低品位鉄鉱石からの直接還元鉄の試験製造に成功.
• SteelOrbis, Nippon Steel successfully produces over 1,000 mt of DRI using 100% hydrogen and low-grade iron ore, 4 November 2026.
• NEDO Green Innovation Fund, Test Blast Furnace Achieves 45% Reduction in CO2 Emissions Using Hydrogen.
• GREINS (Green Innovation Fund hydrogen steel consortium), Technology for Direct Reduction of Iron Ore Using Hydrogen.
• Nippon Steel, Progress of Technology Development for Producing Direct Reduced Iron Using Hydrogen (IR presentation), March 2025.