Electric recycling of Portland cement at scale

The study addresses how to decarbonize Portland cement—responsible for roughly 7.5% of anthropogenic CO2—by eliminating both process emissions from limestone decarbonation (~60%) and combustion emissions (~40%). Existing strategies (fuel switching, use of SCMs such as fly ash and slag, LC3 blends, alternative binders, and CCS) cannot alone deliver zero emissions at scale due to supply constraints, dependency on emitting industries, performance or scalability limits, and residual emissions. The paper proposes leveraging an existing high-temperature electric process—electric-arc furnaces (EAFs) used in steel recycling—by substituting conventional lime-dolomite flux with recovered or hydrated cement paste (already decarbonated), thereby reclinkering it as slag over molten steel. The hypothesis is that this co-production can yield Portland clinker meeting specifications, avoid sulfate/chloride issues due to higher EAF temperatures, reduce steel-recycling emissions by displacing lime flux, and, with clean electricity, enable zero-emissions cement at scale.

• SCM substitution (fly ash, ground granulated blast-furnace slag) reduces clinker content but depends on by-products of high-emission industries being phased out; LC3 blends (calcined clay + limestone) are promising but still require significant Portland clinker for activation, often limiting substitution >50%.
• Alternative clinkers/binders are under development but none yet at comparable scale without significant emissions; CCS for cement is being explored but usually targets process emissions and often without storage, leaving residuals.
• Prior work showed hydrated cement paste (HCP) can be reclinkered, but kiln-based approaches face challenges: sulfate-induced belite enrichment at expense of alite, volatilization and condensation of sulfates causing operational difficulties, and chloride contamination hindering reinforced applications.
• Recovery of cement paste (RCP) from demolition waste has been studied mainly to obtain high-quality aggregates; methods include thermo-mechanical beneficiation and various separation techniques, yet RCP is not available at scale, and fines may contain chlorides and additional silica from attached aggregates.
• Steelmaking fluxes (lime-dolomite) are decarbonated at high temperatures and carry process emissions akin to cement; EAF operation and slag chemistry literature define basicity and phase fields relevant to dephosphorization and desulfurization, providing a framework to target cementitious slag compositions.

• Process concept: Replace conventional lime-dolomite flux in oxidizing EAF steel recycling with flux based on hydrated or recovered cement paste (HCP/RCP), optionally adjusted with lime, alumina, and silica to reach target CaO–SiO2–Al2O3–Fe2O3 compositions. Over molten steel, the flux forms an oxidizing slag which, upon rapid cooling, yields Portland clinker phases.
• Experimental slag production: 28 slags produced using induction furnaces over clean steel under varying crucibles (carbon, magnesium; alumina lining initially but abandoned due to Al leaching) and oxidizing conditions. Flux inputs included HCP, RCP (from demolition waste), and additives (lime, silica sand, kaolin clay, reducing agents for specific trials). Representative raw meal/flux compositions are tabulated (Table 3).
• Cooling and stabilization: Air cooling at laboratory scale (estimated 10–20 K s−1) to stabilize alite. Slags were then ground for characterization and cement testing.
• Characterization:
  • XRF for oxide composition (Table 1; full data in Supplementary Table 1) to map slags in CaO–SiO2–Al2O3 and CaO–SiO2–Fe2O3 ternaries and compare with literature slag compositions.
  • XRD with Rietveld refinement to quantify phase assemblages (alite C3S, belite C2S, C4AF, C3A, gehlenite C2AS, etc.; Table 2) and validate alignment with thermodynamic phase fields.
  • Definition and use of effective lime-to-silica ratio (C/S)* (Methods) to predict silicate phase formation trends.
• Cement and mortar evaluation:
  • Selection of slags with compositions matching conventional clinkers were blended (including LC3-type blends with calcined clay and limestone) and made into mortar bars.
  • Isothermal calorimetry to determine setting times; assessment of workability (superplasticizer demand), bleeding, and early set.
  • Compressive strength development compared against commercial clinker/cement, considering grindability and fineness effects; undersulfated systems noted.
• Steel quality check:
  • Representative oxidizing slag compositions trialed with a high flux-to-steel mass ratio (~9:30 vs. ~1:30 typical) to conservatively assess sulfur transfer; post-melt steel analyzed by optical emission spectroscopy (Table 4).
• Sensitivity and adjustments:
  • RCP tended to have higher SiO2 due to attached aggregates; lime additions used to restore Ca/Si balance. Trials with chloride-saturated paste and various EAF slags combined with HCP and lime were included to study robustness.
• Scale and systems analysis:
  • Emissions and cost assessment comparing ordinary Portland cement, LC3-50, and Cambridge Electric Cement (CEC) variants under present and decarbonized electricity, with UK-focused cost assumptions and global emissions data (Fig. 3a; details in Supplementary Information).
  • UK material flow analysis and global scaling scenarios to estimate potential clinker and LC3 outputs and CO2 abatements (Fig. 3c–e).

• Clinker mineralogy from EAF-derived slags:
  • In compositions within the conventional kiln operating region (Al2O3 < ~6% in CaO–SiO2–Al2O3), XRD/Rietveld shows alite (C3S) + belite (C2S) together exceed ~70% by mass in most slags, meeting normative Portland clinker criteria (66.7% by mass of C3S+C2S).
  • The effective lime-to-silica ratio (C/S)* strongly predicts formation of alite vs. belite; higher ratios favor alite. EAF high temperatures favor alite stabilization and keep sulfates/chlorides in the gas phase, overcoming prior reclinkering limitations.
  • In belite-forming and gehlenite zones, C2S and C2AS appear as expected from phase diagrams.
• Composition control and contaminants:
  • XRF shows minimal interaction with molten steel when using carbon or magnesium crucibles; alumina linings caused Al leaching and were avoided.
  • RCP contains elevated SiO2 (from adhered aggregates). After lime correction to restore Ca/Si, RCP-based flux reclinkers comparably to pure HCP.
  • Scrap-derived SiO2/Al2O3 can be balanced by additional lime in the flux.
• Cement performance:
  • Setting time by calorimetry was ~240–280 min (±20 min) with no flash set or bleeding; workable mortars with small superplasticizer doses for certain mixes (graphite-crucible slags and LC3).
  • Strength development of cements (pure and LC3 blends) was similar to those made with commercial clinker despite undersulfation and minor contamination during tapping. Higher alite content correlated with higher early strength; higher belite with lower early strength.
  • Grindability expected to be comparable to commercial clinkers with industrial grinding.
• Steel quality compatibility:
  • Despite an exaggerated flux-to-steel ratio (~9:30), only very small sulfur entered steel (Table 4: S ~0.07–0.08% mass), readily reducible below 0.05% with standard downstream desulfurization, indicating suitability for EAF dephosphorization/desulfurization operations.
• Emissions and cost implications:
  • Replacing lime flux with decarbonated cement paste eliminates process emissions for the clinker fraction and, with emissions-free electricity, enables zero process and combustion emissions for cement; it also reduces emissions of steel recycling by cutting lime flux production.
  • Economic analysis suggests competitive production costs for CEC relative to Portland cement and LC3 under UK cost assumptions; details in Supplementary Information.
• Scale potential:
  • UK: Co-production could yield about 2.2 Mt yr−1 of new clinker; when blended as LC3-50, ~4.5 Mt yr−1 of cement, potentially meeting UK demand when combined with material efficiency strategies. If EAFs increase slag production (e.g., double), output could reach up to ~10 Mt yr−1.
  • Global: With anticipated EAF expansion and processing all resulting slag into LC3, about 1.4 Gt yr−1 of electric cement and ~2 Gt CO2 abatement are possible. With additional dedicated EAFs (or FAFs) for cement, up to ~2.4 Gt yr−1 of electric cement and ~3 Gt CO2 abatement (≈80% of sector emissions otherwise expected in 2050).

The findings validate that hydrated or recovered cement paste can function as an effective EAF flux, producing slags that, upon rapid cooling, yield Portland clinker mineralogy with high alite content and performance comparable to commercial cement. This directly addresses the core challenge of eliminating process and fuel emissions from cement manufacturing: the decarbonation is already embodied in used cement paste, and electrification via EAFs removes combustion emissions. The higher EAF temperature mitigates sulfate and chloride issues that hampered prior kiln-based reclinkering, while routine compositional tuning (lime additions) accommodates variable RCP and scrap contamination. The industrial symbiosis with steelmaking leverages existing large-scale, electrified infrastructure, lowering capital barriers and simultaneously reducing steel-recycling emissions by displacing lime flux. Cement compositions can be tuned to manage higher Fe content demanded by slag rheology and to target LC3 blends, enabling substantial clinker substitution without compromising performance. Systems analyses indicate that UK and global deployment could materially decarbonize the cement sector, with potential to meet future demand when combined with material efficiency. Remaining constraints include the availability and quality of RCP separation, EAF capacity, and optimization of slag volumes and compositions within steel quality requirements, but the process aligns well with existing standards and operations, supporting rapid scaling.

Using existing industrial-scale equipment, Portland cement can be recycled into Portland cement in an all-electric route by substituting conventional EAF lime flux with recovered or hydrated cement paste and reclinkering it as slag over molten steel. The process can operate either in co-production with steel or in dedicated EAFs and is fundamentally a material substitution within established processes and standards, enabling rapid scale-up. With emissions-free electricity, it offers a zero-emissions alternative to conventional cement production while also reducing emissions in steel recycling. Performance of the resulting clinker and LC3 blends matches commercial benchmarks, and global deployment could abate the majority of sectoral emissions by mid-century. Future work should refine large-scale RCP separation, manage scrap-derived impurities, optimize slag production within steelmaking constraints, and scale industrial trials.

• Feedstock variability: RCP contains variable and often elevated silica (from adhered aggregates) and may contain chlorides; composition must be corrected (e.g., with lime), and chloride management must be ensured for reinforced applications.
• Scrap contamination: SiO2 and Al2O3 from scrap can shift slag chemistry; careful flux adjustment is required.
• Experimental constraints: Laboratory trials used induction furnaces over clean steel and specific crucibles; alumina linings caused Al leaching. Minor contamination occurred during tapping. Results need confirmation in full-scale EAF campaigns.
• Sulfation/undersulfation: Tested cements were undersulfated; industrial sulfate adjustment and grinding optimization may be needed to fully match commercial performance.
• System capacity: Near-term scale limited by EAF availability and willingness to increase slag volumes; UK scenarios assume domestic recycling of all scrap and specific slag-to-steel ratios (~14%).
• Energy source: Achieving zero emissions depends on decarbonized electricity; residual emissions persist if grid power is carbon-intensive.
• RCP supply chain: Industrial-scale, high-quality separation of RCP from demolition waste is not yet widespread; logistics, quality control, and costs require development.
• Iron content and rheology: Co-production may yield higher Fe content in clinker due to slag rheology needs; composition must be tuned to meet cement standards.
• Environmental controls: Although existing EAF off-gas scrubbing is expected to suffice, real-plant assessments are needed to confirm air pollutant profiles with RCP flux.