Hydrogen Steel Plants Could Cut Emissions 89 Percent With Efficiency Gains

The steel industry has long been one of the world’s most stubborn climate problems, responsible for roughly seven percent of global carbon dioxide emissions and about five percent of emissions across the European Union. Now, a new study published in the Journal of Industrial Ecology offers one of the most detailed pictures yet of what actually happens inside a steel plant when it swaps coal for hydrogen, and the findings carry a striking message: green steel is technically within reach, but only if engineers obsess over every last tonne of material and every megawatt-hour of electricity. Using a prospective plant-level material flow analysis of a representative integrated steel plant in Sweden, researchers from the Norwegian University of Science and Technology and Swerim AB quantified how the entire metabolism of a steelworks changes when the coal-fired blast furnace route is replaced by hydrogen-based direct reduced iron and electric arc furnace steelmaking.

The research team, led by Moritz Langhorst, modeled the transformation of a plant modeled on publicly reported production data from SSAB’s facility in Oxelösund, Sweden, producing around one million tonnes of liquid steel per year. In the conventional blast furnace-basic oxygen furnace system, the plant consumed roughly 1,325 kilotonnes of iron ore pellets and 579 kilotonnes of coal annually, generating about 1,570 kilotonnes of direct carbon dioxide emissions from iron and steelmaking alone. The rolling mill added another 70 kilotonnes. Crucially, the study also tracked the energy-rich process gases—coke oven gas, blast furnace gas, and basic oxygen furnace gas—that integrated plants routinely recycle as fuel, providing over 1 terawatt-hour per year of surplus energy for power generation and district heating.

When the researchers modeled the shift to the hydrogen-based route, the picture changed dramatically. Direct emissions from iron and steelmaking fell by 94 percent, and total greenhouse gas emissions across the system, including indirect emissions from electricity, dropped by 84 percent under Sweden’s low-carbon electricity mix. When hydrogen combustion replaced natural gas in the rolling mill’s reheating furnaces and heat treatment was electrified, the reduction climbed to 89 percent, cutting emissions intensity from 1.94 to 0.25 tonnes of carbon dioxide equivalent per tonne of plate. That is well below the near-zero emission threshold of 0.33 tonnes proposed by the International Energy Agency for primary steel production. But the transformation came at a steep energy price: electricity demand rose eighteen-fold, from roughly 260 gigawatt-hours to more than 4,600 gigawatt-hours per year, driven overwhelmingly by the alkaline electrolysers producing hydrogen on site.

This electrification shock is the study’s central tension. The additional 4.42 terawatt-hours of annual electricity demand would equal roughly 2.7 percent of Sweden’s entire electricity generation in 2023. In countries with carbon-intensive grids, the climate arithmetic collapses: under the average European electricity mix of 242.3 kilograms of carbon dioxide per megawatt-hour, the emission reduction shrinks from 84 percent to just 30 percent. The authors stress that the promise of hydrogen steelmaking therefore depends entirely on locating production in regions with abundant clean electricity—at least until Europe’s power sector approaches climate neutrality around mid-century. It is a caveat with real-world bite, given that several announced green hydrogen and green steel projects have already been delayed or cancelled across Europe.

The study goes further than previous analyses by treating the steel plant as a single, interconnected organism rather than a collection of independent processes. When the blast furnace disappears, so do the process gases that once fueled the rolling mill, forcing downstream operations to rely on external natural gas, hydrogen, or electricity. This structural coupling means that changes upstream ripple through the entire production chain, and it is precisely where the researchers found unexpected leverage. Material efficiency measures—reducing iron losses in the electric arc furnace, improving casting yields from 96.5 to 98 percent, and raising cutting yields in the rolling mill from roughly 91 to 95 percent—trigger cascading reductions that travel all the way back to the hydrogen demand of the direct reduction plant.

The numbers are compelling. Improving the electric arc furnace iron yield from 89.5 to 95.7 percent, a level typical of scrap-based operation, reduced direct reduced iron demand by nine percent and cut the hydrogen needed for direct reduction by 166 gigawatt-hours per year. Energy efficiency measures worked differently: upgrading electrolyser efficiency from 58.7 percent to a projected 2050 value of 69.94 percent cut the specific electricity demand for hydrogen production by 16 percent, while oxyfuel combustion in the reheating furnace trimmed hydrogen use for that process by 15 percent. Because energy efficiency measures act locally while material efficiency measures propagate upstream through the entire chain, the two strategies proved complementary rather than redundant.

Combined, all modeled efficiency measures reduced the plant’s hydrogen demand by ten percent and its electricity demand by twenty percent—savings of 6.3 kilotonnes of hydrogen and 920 gigawatt-hours of electricity per year. In a world where green hydrogen remains scarce, these percentages matter enormously. The study notes that without efficiency measures, the modeled plant would require up to 63.3 kilotonnes of hydrogen annually, while only about 31 kilotonnes of electrolytic hydrogen were produced across the entire EU, EFTA, and the United Kingdom combined in 2023, out of nearly 8 million tonnes of total hydrogen production that was over 90 percent fossil-based. Every tonne of hydrogen saved through smarter material flows directly expands the number of plants that can realistically decarbonize within the constrained supply expected this decade.

The researchers are careful to acknowledge the limits of their model. It does not include techno-economic assessment, and without carbon pricing or financial support, hydrogen-based direct reduction remains uncompetitive with conventional routes. Yield assumptions for the electric arc furnace fed with direct reduced iron may prove optimistic, since oxide gangue in the DRI limits achievable yields, and electrolyser efficiency projections for 2050 have already been revised downward. Still, the model’s emissions estimate for the reference case closely matched SSAB’s reported 2023 figures, and the qualitative findings proved robust across variations in electricity mix and scrap share. The authors also emphasize that scrap-based electric arc furnace recycling, which would cut electricity demand by over 70 percent compared to hydrogen-based production, remains limited by the availability of low-impurity scrap needed for high-quality flat products.

What emerges from this work is less a prediction than a planning instrument. By quantifying material, energy, and carbon flows across an entire production chain under different decarbonization strategies, prospective plant-level material flow analysis gives steelmakers, grid planners, and policymakers a common physical baseline for site-specific roadmaps. It reveals where emissions migrate as the transition proceeds—the rolling mill’s share of plant emissions jumps from four percent to nearly thirty percent—where hydrogen scarcity will bind hardest, and where operational improvements yield outsized systemic returns. As Europe’s steel industry confronts the largest industrial transformation since the invention of the blast furnace, the message of this study is clear: the technology for near-zero-emission steel exists, but its success will be decided in the details of yields, electrolysers, and every kilowatt-hour in between.

Beyond the headline figures, the study’s methodological choice deserves attention. Material flow analysis has a long pedigree in industrial ecology, where it has been used to map the flows of substances through economies and industrial systems for decades. What distinguishes this application is its prospective character: rather than auditing an existing facility, the framework is designed to simulate structural transformations that have not yet occurred, embedding process-level detail within a plant-wide accounting boundary. This allows the researchers to capture interactions that fall through the cracks of both regional-scale models, which lack operational resolution, and single-process studies, which isolate technologies from their surroundings.

The Swedish setting is not incidental. The country combines an unusually clean electricity grid with pioneering industrial initiatives in hydrogen-based ironmaking, making it a plausible early adopter. Yet even under these favorable conditions, the modeled plant’s electricity appetite rivals that of a mid-sized city, underscoring why grid capacity and electrolyser siting have become central concerns for industrial planners. The finding that the rolling mill’s relative share of plant emissions rises sharply after decarbonization of ironmaking also illustrates a broader principle: as the dominant emission sources are eliminated, previously marginal processes become the next targets, requiring successive waves of intervention rather than a single technological switch.

The distinction between the two families of efficiency measures carries practical implications for how decarbonization investments are sequenced. Material efficiency gains, such as yield improvements, propagate upstream and shrink the entire hydrogen production system needed at the front of the plant, while energy efficiency gains act locally but aggregate into substantial electricity savings. Because hydrogen production dominates the new electricity load, improvements in electrolyser performance translate almost directly into reduced pressure on the power system. For policymakers designing support schemes, the analysis suggests that funding yield optimization and electrolyser development together yields compounding benefits that neither achieves alone, effectively stretching a constrained green hydrogen supply across more tonnes of decarbonized steel.

Subject of Research: Plant-level material flow analysis of the transition from blast furnace steelmaking to hydrogen-based direct reduction and electric arc furnace steelmaking

Article Title: A prospective plant-level material flow analysis to assess systemic efficiency in the transition to hydrogen-based steelmaking

Article References: Langhorst, M., Billy, R. G., Song, X., & Müller, D. B. (2026). A prospective plant-level material flow analysis to assess systemic efficiency in the transition to hydrogen-based steelmaking. Journal of Industrial Ecology. https://doi.org/10.1007/s44498-026-00166-1

Image Credits: AI Generated

DOI: 10.1007/s44498-026-00166-1

Keywords: hydrogen steelmaking, green steel, material flow analysis, decarbonization, electric arc furnace, direct reduced iron, resource efficiency, electrolysis, steel industry emissions, energy efficiency, Sweden, industrial ecology