Hydrogen-based direct reduction plants have demonstrated large-scale technical feasibility, but project schedules for 2026 have slipped toward 2027 and beyond as green hydrogen costs, infrastructure, and reactor optimization remain unresolved [S2][S7].
The principal pathway uses hydrogen instead of carbon monoxide to reduce iron ore, followed by electric arc furnace steelmaking, a configuration that can cut direct carbon dioxide emissions by roughly 90–95% compared with conventional coal-based blast furnace and basic oxygen furnace production [S1][S2].
Commercial Schedule and Plant Reality
As of 2026, hydrogen-DRI start-ups are slipping to 2027 and beyond, while scrap-based electric arc furnace routes retain an earlier deployment advantage [S7]. This schedule shift does not invalidate the metallurgy, but it changes procurement, financing, and engineering risk for plant owners.
H2 Green Steel is developing a fully integrated hydrogen-DRI steel plant in northern Sweden, with renewable power and hydrogen production intended to form part of the integrated concept [S2]. The project illustrates the direction of travel from stand-alone pilots toward integrated plants combining power, electrolysis, hydrogen handling, ironmaking, and steel production.
Industrial trials have confirmed that hydrogen reduction is technically feasible at large scale and can achieve metallization rates comparable with traditional direct reduction processes [S2]. Commercial realization has not followed because renewable hydrogen remains expensive, supporting infrastructure requires substantial capital, and unresolved process issues include water-vapor control, pellet quality, and reactor design [S2].
Process Chain and Plant Sizing
A hydrogen-DRI steel plant requires four linked operating blocks: renewable electricity and electrolysis, hydrogen storage and distribution, the DRI shaft furnace, and hot-charged DRI processing in an electric arc furnace [S1]. The hydrogen chain must be balanced against the furnace because direct reduction is continuous while wind and solar output varies by hour.
A representative plant configuration requires 50–100 MW of electrolyzer capacity to produce 20,000–40,000 Nm³/hr of hydrogen [S1]. That hydrogen flow is sized to feed a single DRI shaft furnace producing 1–2 million tonnes of DRI per year, making electrolysis the largest single investment in the green-steel chain [S1].
Storage commonly covers 4–24 hours of furnace operation, with hydrogen held in vessels at 30–70 bar or in underground caverns before distribution through dedicated piping [S1]. This buffer reduces production stops caused by variable renewable generation, but it adds pressure-boundary equipment, controls, and hydrogen-specific material risks.
Emissions, Metallurgy and Energy Basis
Hydrogen reduction releases water vapor rather than carbon dioxide as the principal reaction product, and renewable hydrogen can reduce carbon monoxide emissions by as much as 90% according to the reviewed life-cycle studies [S2]. A separate process comparison places emissions at 1.85 t CO₂ per tonne of steel for coal-based BF-BOF, approximately 1.0 t for natural-gas DRI/EAF, and 0.1–0.2 t for the described green-hydrogen DRI/EAF route [S1].
Metallurgical results support hydrogen as an efficient reductant for iron oxides, with trial-derived metallization rates reaching those of established direct reduction processes [S2]. That evidence separates the core reaction chemistry from the unresolved plant-integration problem: the difficult engineering sits in continuous hydrogen supply, water-vapor management, feed quality, and reactor configuration [S2].
The energy basis also determines the emissions outcome. A plant cannot claim green-steel performance from hydrogen chemistry alone because electrolyzers consume substantial electricity, and the carbon intensity of that electricity affects the full production chain [S1][S2]. The 90–95% reduction cited for hydrogen DRI/EAF applies to a renewable-powered configuration, not an arbitrary hydrogen supply [S1].
Options Compared by Plant Decision Criteria
Hydrogen DRI is suited to integrated primary steelmaking where operators need a low-carbon iron source, access to renewable power, capital for electrolysis and storage, and tolerance for early-stage operating risk. It is not suited to brownfield sites without firm power, hydrogen buffering, iron-ore logistics, and an EAF able to process hot DRI. [S1]
For a plant comparing routes, the decision matrix is straightforward. Coal-based BF-BOF carries an emissions benchmark of 1.85 t CO₂ per tonne of steel and requires coke, coal, and established bulk-material infrastructure [S1]. Natural-gas DRI/EAF reduces that figure to about 1.0 t CO₂ per tonne while retaining carbon-based reducing chemistry [S1]. Hydrogen DRI/EAF reaches 0.1–0.2 t CO₂ per tonne in the renewable configuration, but demands 50–100 MW of electrolysis, 4–24 hours of storage, and continuous process integration [S1].
Scrap-based EAF is the schedule comparator rather than a direct low-carbon iron substitute [S7]. It can deploy without building a new hydrogen-DRI island, but it depends on scrap availability, scrap quality, and competing demand for recycled steel feedstock. The OECD analysis also cautions that hydrogen-based DRI expansion can worsen excess capacity where transition markets are not managed carefully, so low-carbon intent alone does not guarantee a viable commercial project [S4].
Operating Constraints and Failure Modes
Water-vapor management is a central reactor-design constraint because hydrogen reduction changes the reaction products and heat balance relative to carbon-monoxide reduction [S2]. Pellet quality must remain consistent enough to maintain permeability, gas flow, and metallization, while reactor design must tolerate the hydraulic and thermal behavior of a high-hydrogen gas stream [S2].
Electrolyzer degradation adds a separate reliability layer. Cell voltage rises, efficiency falls, and membranes deteriorate over service, with stack replacement cycles of 60,000–90,000 hours requiring long-range maintenance planning [S1]. Deionized water quality, cooling systems, power electronics, and gas purification are therefore plant-critical systems rather than auxiliary utilities [S1].
Hydrogen handling introduces a material failure mode absent from many conventional steel systems. Embrittlement can affect carbon-steel piping, valve seats, and pressure vessels, so equipment selection, inspection, and maintenance cannot simply replicate a coal-fed steel-plant playbook [S1]. Storage at 30–70 bar and 4–24 hours of furnace coverage increase the consequence of leaks, control faults, and loss of buffer capacity [S1].
Conversion Economics and Infrastructure
Conversion of existing natural-gas DRI furnaces offers a more practical first step than building a fully integrated green-steel complex, but technical feasibility and economic viability depend on furnace configuration, hydrogen supply, and retrofit costs [S3]. The pathway retains an existing ironmaking asset while replacing part of the reductant and energy basis, reducing greenfield scope.
Green hydrogen cost remains the central commercial constraint despite demonstrated metallurgy [S2]. Electrolyzer capital, renewable power contracts, storage, hydrogen compression or transport, water-vapor control, and dedicated piping all enter the delivered hydrogen cost. A technically successful shaft furnace does not remove those upstream costs.
China provides a relevant deployment signal because it was on track to exceed its 2025 green hydrogen deployment target by the end of 2024 and had led global hydrogen production and deployment at that point [S5]. The implication is not that Chinese steel projects automatically meet European low-carbon criteria, but that electrolyzer scale and renewable hydrogen availability can improve the commercial conditions required for H-DRI [S5].
Standards, Procurement and Tracking Signals
Hydrogen-DRI procurement should separate metallurgy performance from carbon accounting. The 0.1–0.2 t CO₂ per tonne benchmark and 90–95% reduction figure apply to defined green-hydrogen process configurations, while the 90% life-cycle reduction cited for renewable hydrogen concerns carbon monoxide and requires a consistent system boundary [S1][S2]. Equipment and operating decisions should use the same feedstock, power, and output basis.
Owners evaluating a 2027 or later start-up should request 4-hour and 24-hour hydrogen-continuity cases, 50–100 MW electrolysis output at 20,000–40,000 Nm³/hr, and a DRI design rate of 1–2 million tonnes per year for the reference configuration described in the plant guide [S1]. Vendors should also document stack service life against the 60,000–90,000-hour replacement band, storage conditions at 30–70 bar, and water-vapor control in the shaft-furnace design [S1][S2].
Two signals will separate announced projects from operating assets: the start-up date moving from 2026 into 2027 or beyond, and the completion of an integrated green-steel complex rather than another isolated demonstration [S2][S7]. Project trackers should also record firm electrolyzer capacity, renewable-power linkage, storage duration, pellet specification, hot-DRI charging capability, and an independently traceable emissions calculation. Those six data fields expose whether the plant has moved beyond a slide-level DRI concept.