Global ironmaking resource requirements ↓
When can hydrogen
match carbon?
An interactive learning model. Change the assumptions.
Find the price combinations that bring two ironmaking routes into balance.
The two routes, on equal terms
Per tonne of pure liquid iron · 1600 °C
Includes the selected model carbon charge.
Electrolysis + electric heat
At your electricity and coal-equivalent prices.
Explore the price boundary
Where does the hydrogen route become cheaper?
What you pay for
Resource inputs + model CO₂ charge · CNY/t Fe
Before equipment, oxygen supply, O&M and upstream coal emissions.
What changes emissions
Model CO₂ · tonnes per tonne of Fe
Carbon: direct CO₂. Hydrogen: electricity-related CO₂. This is not a complete lifecycle CO₂e inventory.
The reaction is only part of the bill.
Iron must reach the same product state in both routes. The model also tracks the heat leaving with the gases.
External reduction heat
Reaction reference + hot iron + exit gas · GJ/t Fe
This is the heat load before the selected heat-use efficiencies. Electrolysis electricity is shown separately above.
Every tonne of pure Fe starts here
The two alternatives do not supply each other. Oxygen cost and coproduct credit are excluded here.
A transparent benchmark.
A starting point for better questions.
Use the equations to understand the result. Use the boundary to understand what the result can claim.
Balance the atoms
The stoichiometry fixes reductant demand: about 161.3 kg of carbon or 54.15 kg of hydrogen for a tonne of pure iron. Carbon burned for heat is additional.
Account for every state
Enthalpy includes chemical formation, temperature and phase. Cold solid iron and hot liquid iron cannot be used interchangeably in a fair energy comparison.
ΔH is a heat balance. ΔG describes reversible work or equilibrium under specified conditions. Neither alone is a practical furnace electricity demand.
Turn resources into costs
A is tonnes of carbon-equivalent coal per tonne Fe; W is kWh/t Fe. E is model CO₂ in t/t Fe. The selected carbon charge is a uniform analytical price, not an estimate of current ETS payments.
Find the crossing
If hydrogen is initially more expensive and emits less, a positive carbon charge can close the gap. If it is more expensive and emits at least as much, a nonnegative carbon charge cannot make it cheaper.
A negative crossing is retained and explained. It is not silently replaced with zero.
Read the system boundary before using these numbers
Included
- One tonne of pure Fe at the selected state
- Pure hematite and graphite, cold feeds at 298.15 K
- CO₂ or water-vapour tail gas at the specified state
- Graphite combustion heat, electrolysis and electric heat
- Scenario electricity emissions and a uniform model CO₂ charge
Needed for an engineering model
- FeO reduction equilibrium and gas circulation
- Heat recovery constrained by temperature and heat quality
- Real coal, gangue, air or oxygen supply
- Equipment cost, O&M and auxiliary electricity
- Capacity utilisation, storage and renewable-power variability
Interpret the result carefully
This is a state-explicit material and enthalpy benchmark. It does not prove furnace feasibility, simulate an industrial process, or establish a universal minimum cost. Pure iron is not finished liquid steel.
The heat-use sliders rescale useful heat within this benchmark. They do not represent a detailed furnace or heat-exchanger design.
Inspect the data. Reproduce the result.
Thermochemistry: NIST Chemistry WebBook and JANAF, Chase (1998). Phase offsets and fit ranges are retained. No temperature extrapolation.
05 / FROM ONE TONNE TO THE WORLD
Enough resources.
Enough supply?
What would it take to produce the world’s primary iron using coal, natural gas or renewable hydrogen?
2025 production baseline · Sources checked 19 October 2026 · Adjustable screening scenarios
Adjust production and process assumptions
These values are sensitivity inputs; not every combination represents a feasible plant. They do not change the thermodynamic cost-parity model above. Gas consumption already includes process and fuel gas; H₂ consumption is fresh make-up, not recirculating gas. Shaft electricity includes electric gas heating and auxiliaries.
ALL COAL / BLAST FURNACE
—million tonnes of dry coal / year
Coal for coke manufacture + injected coal. Not pure carbon or finished coke.ALL NATURAL GAS / DRI
—billion standard m³ / year
—ALL RENEWABLE H₂ / DRI
—million tonnes of H₂ / year
Fresh hydrogen for reduction; the selected shaft heat is supplied electrically.POWER FOR THAT H₂ ROUTE
—TWh of renewable electricity / year
—Ironmaking, before steelmaking. The BF produces liquid hot metal; the DRI routes produce solid DRI. This is a resource-scale comparison, not an equal-temperature energy-efficiency ranking. Mining, ore preparation, DRI melting, steelmaking, rolling, hydrogen transport/storage and power-network losses are outside this estimate. The model above handles a different, matched-product thermodynamic boundary.
ANNUAL FLOW / THE SIZE OF THE TASK
Compare annual requirements with today’s supply scale.
100% means an amount equal to an entire year of the stated global output. That output already serves other users. A ratio below 100% does not establish a surplus; a ratio above 100% indicates a larger required supply scale, not a permanent resource limit.
| Comparison | 2025 reference | Scenario / reference |
|---|
All-renewable electricity includes hydropower and other renewable sources. Wind + solar is a narrower definition. These denominators overlap and must not be added. Coal figures compare dry process coal with aggregate reported coal tonnage across grades and moisture bases: an approximate scale comparison.
HYDROGEN / A MANUFACTURED CARRIER
A global switch requires a much larger green-hydrogen industry.
IEA reports almost 1 Mt of low-emissions H₂ in 2025. This broader category includes fossil production with carbon capture and is not a green-hydrogen total. It is used only to show the order of magnitude; it is neither spare supply nor a renewable-only denominator.
ELECTRICITY / WHAT NEEDS TO BE BUILT
Translate annual electricity into generation capacity.
An annual-energy equivalent assuming all output is usable. This is the dedicated capacity needed to supply the scenario from scratch, not a calculated net addition to the world’s fleet. Hourly matching, curtailment, storage and network constraints can raise the requirement. A capacity factor does not guarantee round-the-clock power.
Explore GreenEnergy →GEOLOGICAL STOCK / A DIFFERENT QUESTION
Reserves are not annual supply.
| Resource | Published reserve benchmark | One scenario-year relative to that stock |
|---|
Coal and gas have large global geological stocks relative to one year of this scenario. But aggregate coal reserves include many grades unsuitable for coke making. Location, mine and pipeline capacity, transport, environmental constraints and delivered cost can be binding long before geological depletion.
Green hydrogen and renewable electricity have no comparable underground reserve. Renewable energy flows can be harvested repeatedly; the constraints are deployable sites, equipment, grids, water, storage, permitting and reliable affordable supply. These reserve comparisons are not depletion dates, and the older coal estimate is not presented as a 2025 stock.
WHAT THIS MEANS FOR ECONOMICS
Availability sets the scale.
Delivered cost decides competitiveness.
Large reserves do not guarantee cheap delivered energy. A small current green-hydrogen industry does not rule out a competitive project in a region with inexpensive renewable power. New supply can be built, but its capital cost, utilisation and delivery infrastructure must be paid for.
All route totals are alternative requirements, not additions to today’s consumption. Existing ironmaking already uses coal, gas and electricity. A net global transition forecast would subtract those displaced uses and account for growth, scrap recycling and competing demand.
Return to the price-parity model ↑Sources, equations and parameter provenance
Production scale
worldsteel’s World Steel in Figures 2026, pp. 18–19 reports 2025 pig iron of 1,271.2 Mt and DRI of 153.0 Mt. These are product masses, not pure Fe or crude steel. The default proxy assumes 94.5% contained Fe in pig iron and 92% total Fe in DRI. These composition assumptions are adjustable; DRI total Fe includes iron remaining in oxides. MIDREX’s product brochure gives 90–94% total Fe as a typical DRI range. Scrap-derived steel is not counted as newly reduced iron.
The separate MIDREX 2025 statistics report 151.0 Mt DRI. This page retains the internally consistent worldsteel series; users can substitute another baseline.
Process inputs
- Coal: 550 kg dry coal/t hot metal. Educational central value within the 538–568 kg range for typical PCI operation in JRC Iron and Steel BREF (2012), Table 6.25, p. 340. It includes coke-making coal and injected coal; upstream sintering/pelletising and auxiliary electricity are excluded. This dated technical benchmark is not a measured 2025 global average.
- NG: 9.5 GJ/t DRI; H₂: 58.4 kg/t DRI; H₂-DRI shaft electricity: 0.86 MWh/t DRI. Based on IEAGHG 2024-02, Table 8 and footnote 18, pp. 33–34. H₂ is rounded from 650 Nm³ × 0.0899 kg/Nm³. Shaft electricity includes 750 kWh/t for gas preheating and 110 kWh/t auxiliaries; electrolysis and EAF melting are separate. These are published engineering-model assumptions.
- Electrolysis: 55 kWh/kg H₂. US DOE PEM targets, system efficiency, 2022 status (not the 2026 target). Includes electrolyser-system electricity; site-specific transport/storage is excluded.
- Gas conversion: 34 MJ per standard m³ (LHV). Adjustable illustrative heating value for gas at the OPEC standard-volume convention (approximately 15.6°C). Actual gas composition changes the conversion. Standard m³ here must not be confused with 0°C Nm³. Energy demand is also displayed.
Annual supply and stocks
- IEA Coal Mid-Year Update 2026: approximately 9,100 Mt coal production in 2025. Its demand chapter reports 1,174 Mt metallurgical-coal demand. The latter is a market-size comparator, not an available-supply estimate.
- OPEC ASB 2026 gas production: 4,341.52 billion standard m³ marketed in 2025. Gas reserves: 219.61 trillion standard m³ at end-2025.
- Ember Global Electricity Review 2026: 10,730 TWh renewable electricity in 2025. Technology totals: wind 2,715 + solar 2,778 = 5,493 TWh (sum calculated here).
- EIA world proved recoverable coal reserves: about 1,161 billion short tons at end-2021. Converted using 1 short ton = 0.90718474 metric tonnes, approximately 1,053 Gt. Older stock benchmark, not a current estimate.
Equations and boundaries
Alternative hot metal = F / pig-iron Fe fraction
Alternative DRI = F / DRI Fe fraction
Coal = alternative hot metal × coal intensity
Natural gas energy = alternative DRI × gas intensity
H₂ = alternative DRI × fresh H₂ intensity
Electricity = H₂ × electrolyser electricity + alternative DRI × shaft electricity
Generation capacity = annual electricity / (8,760 hours × capacity factor)
The model’s full-hematite-to-pure-Fe stoichiometric reference is about 54.15 kg H₂/t Fe. It is not substituted for industrial plant demand or described as a universal lower bound for incompletely metallised DRI. Source years and route boundaries differ and remain visible. No ore-quality, regional-deliverability, price-response or hourly power-system optimisation is performed.