Draft: This piece is part of a new series on zero carbon industrial processes and is currently an initial draft.


According to data from the IEA, Iron and steel production has the second-largest energy consumption of all industrial sectors, accounting for 22% of total industrial energy use and 31% of industrial direct CO2 emissions in 2012 [3].

2012 demand was 1546 Mt [3, fig24] and average intensity 20.7 GJ/t 2012 (14.3 GJ/t OECD) [3, p94] suggesting a total energy demand of:

1546 Mt x 20.7 GJ/t = 32 EJ

More recent data for 2014 suggests a total energy demand of 36 EJ out of a total industrial energy demand of 154.1 EJ or 23.3% [2].

Sustainable Materials without the hot air global steel product catalogue [4], note total demand here is significantly lower than 2012 demand figure above:

Mt %
Buildings Reinforcing bars 191 18.3%
Sheet products 134 12.9%
Structural sections 108 10.4%
Infrastructure Reinforcing bars 81 7.8%
Structural sections 36 3.5%
Pipes 24 2.3%
Hot rolled train rails 9 0.9%
Mechanical equipment Steel plate & hot rolled bar 55 5.3%
Tubes 30 2.9%
Hot and cold rolled coils 30 2.9%
Castings and wire rod 22 2.1%
Metal goods Plate, narrow strip, cast iron 67 6.4%
Hot rolled coil 40 3.9%
Hot rolled bar 27 2.6%
Cars and light trucks Body and structure 32 3.0%
Engine + Drive-train 21 2.1%
Suspension 11 1.1%
Wheels, fuel tank, Steering, braking 29 2.8%
Domestic appliances Cold rolled coil, welded rolled strip, etc 29 2.8%
Trucks and ships 28 2.7%
Electrical equipment High silicon content electrical steel 8 0.8%
Other 19 1.8%
Consumer packaging 9 0.9%
Total 1040

We can see that the majority of steel demand is for buildings and in particular reinforcing bars which also make up the majority of the steel demand in the infrastructure sector.

Steel production

Conventional blast furnace

70% of steel production currently takes place in conventional blast furnaces, fuelled primarily with coal [3].

Iron ore is pre-treated in a sinter or pellet plant and mixed with additives: limestone and dolomite. In the furnace the iron is reduced using coke and coal and resulting iron melted. The blast furnace is a continuously operated shaft furnace. The BF gas runs through a turbine to generate electricity (18 kWh/t) and a portion (1.536GJ/t) used to heat up the air going into the furnace.

Energy demand per ton of pig iron [5]
Coke 10.303 GJ
Coal dust 4.67 GJ
Net power demand 0.202 GJ
Power demand for N2 and O2 allocation 0.119 GJ
To hot-blast stoves
Natural Gas (NG) 0.213 GJ
Coke over gas (COG) 0.284 GJ
Oxygen steel furnace gas (BOF gas) 0.168 GJ
Total energy demand 15.95 GJ
Export of blast furnace gas 4.719 GJ
Net energy demand minus export gas 11.24 GJ

Cross check: We can see that this energy demand of between 11.24 GJ/t and 15.95 GJ/t depending on whether exported blast furnace gas is included or not. Is significantly (23% to 46%) below the global average energy intensity of steel production from source [3]. If we look closer at the OECD case where average intensity is lower, BOF makes up 56% of production and EAF 42% of production. The combination of both process intensities assuming no subtraction of exported blast furnace gas works out to being 10.3 GJ/t 28% below the 14.3 GJ/t OECD average, suggesting a discrepancy somewhere in my source data in both cases.

Electric arc furnace

Electric arc furnace steel production from scrap makes up 23% of steel production globally [3].

Materials that contain iron, such as scrap, are melted directly through the use of electrical power. The result is liquid steel rather than pig iron. The process is started with a natural gas burner.

Energy Demand per ton of Liquid Steel [5]
Electrical power 2.07 GJ
Natural gas 0.78 GJ
N2 8 kg (4.6MJel)
O2 50 kg (37.3 Mjel)
Coal 0.45 GJ
Total 3.34 GJ

We can see here that the energy required to recycle a ton of steel is 70-80% lower than that of a conventional blast furnace assuming electricity is from a primary energy source.

An electric arc furnace does not reduce iron ore and so is only useful as a secondary stage, to create new iron and steel iron ore reduction is still required. There are several alternative processes that may be used to do this including:

Blast furnaces with carbon capture The direct reduction of iron ore using hydrogen as the reduction agent (Circored Process). Other novel routes such as electrolysis reduction.

The next section will explore direct reduction via the circored process in more detail.

Circored Process [5]

The Circored process produces direct reduced iron briquettes from iron ore fines and uses pure hydrogen as a reduction agent. Iron ore fines are dried and heated through the combustion of natural gas at temperatures of up to 850-900C. Then it is reduced by hydrogen. The by product is water rather than CO2.

Fe203 + 3H2 → 2Fe + 3H20

The product of the circored process is not pig iron but a solid iron sponge in the form of briquettes or fines with 95% metallization which can then be added to the charge of an EAF to produce steel.

The solid iron sponge contains no carbon as hydrogen is used as the reduction agent, therefore an injection of carbon in the form of coal is required for metallurgical reasons.

Circored process, energy demand per 1.03 ton of HBI
Electrical power 0.46 GJ (128 kWh)
Natural gas for heat provision 5.62 GJ (1560 kWh)
H2 provision 8.31 GJ el (58.17 kg H2) 2306 kWh el
+ EAF, energy demand per ton of liquid steel
Electrical power 2.07 GJ (575 kWh)
Natural gas 0.78 GJ (217 kWh)
N2 8kg (4.6MJel) = 1.3 kWh
O2 50kg (37.3MJel) = 10.4 kWh
Coal 1.024 GJ

The total energy input not including the coal which is required for metallurgical reasons is 17.3 GJ/t, 8% higher than the blast furnace process where export blast furnace gas energy is not subtracted. There is perhaps an important point here that alternative processes may not be lower energy alternatives, the key here is that a large share of the energy can be supplied from renewable electricity lowering carbon emissions rather than overall energy intensity.

The above figures use natural gas for heat provision. Natural gas is methane and can be produced from renewable sources as explored in the page here on sabatier enhanced anaerobic digestion.

Using methane produced by the AD + sabatier process to produce the 6.4GJ of natural gas above (5.62 GJ +0.78 GJ) would require 6.4GJ of biomass (rotational grasses) and 3.2 GJ of hydrogen, which would in turn require 3.8 GJ of electricity for electrolysis.

The total electricity demand of the full circored process per ton of steel is therefore: 0.46+8.31+2.07+0.0046+0.0373+3.8 = 14.7 GJ/t (4078 kWh/t) and the total biomass demand is 6.5 GJ/t (1805 kWh/t, 404 m2/t/year @ 0.51W/m2).

The electricity demand for the circored reduction only per ton of steel is: 0.46+8.31+3.3 = 12.07 GJ/t (3353 kWh/t) and the total biomass demand is 5.62 GJ/t (1561 kWh/t, 350 m2/t/year @ 0.51W/m2).

The electricity demand for the EAF only per ton of steel is: 2.07 + 0.0046 + 0.0373 + 0.464 = 2.58 GJ/t (716 kWh/t) and the total biomass demand is 0.78 GJ/t (217 kWh/t, 48 m2/t/year @ 0.51W/m2).

Fabrication scrap and recycling

The final energy intensity of a kg of steel in a particular product is complicated by the proportion of the EAF feedstock from steel scrap and whether it is fabrication scrap or post-consumption scrap.

Lets say we have a product that contains 1kg of steel and that during manufacture 30% of the input steel was machined away, our 1kg steel product requires 1.43 kg of input steel. The 0.43 kg of scrap can however be returned directly to the EAF reducing the amount of iron required from the Circored process for the production of steel for further products in the line.

In a perfect waste recovery scenario our 1kg steel product could also be returned to the EAF at end of life for production of the next product, which could in theory negate the need for the Circored process part all together, however steel stocks are still growing globally and so the demand for steel continues to exceed the availability of scrap steel.

Global scrap use 2014: 585 Mt, Global steel demand: 1665 Mt, Recycling rate 35% [6]. A large part of this recycled steel will likely be from fabrication scrap.

If we imagine a future where our 1kg steel product generates 0.43kg of fabrication scrap and that on average 0.5kg of steel is returned at end-of-life.

Then the demand for new steel for our 1 kg product via the Circored process would be 1.43 kg – 0.43 kg – 0.5kg = 0.5kg while the full 1.43 kg would need to pass through the EAF per kg of product.

The energy demand to manufacture our 1kg steel product would therefore be:

Circored part:

0.5kg x 1.03 of HBI = 0.515 kg HBI.
0.515 kg x 3.353 kWh/kg = 1.73 kWh electric
0.515 kg x 0.35 m2/kg = 0.180 m2 for 1 year of rotational grasses

EAF part:

1.43 kg x 0.716 kWh/kg = 1.02 kWh electric
1.43 kg x 0.048 m2/kg = 0.067 m2 for 1 year of rotational grasses


Electricity: 1.73 kWh + 1.02 kWh = 2.75 kWh
Land area: 0.247 m2 for 1 year of rotational grasses

If our product was a car with 960kg of steel with the same scrap and end-of-life recycling rate as above, we would need 2640 kWh of electricity and 237.12 m2 for a year to manufacture. If the car has a lifetime of 15 years. Then the impact of the steel part of the car’s manufacture on our household energy consumption would be 0.48 kWh/d of electricity per household and 15.8 m2 of land for biomass.


[1] Sustainable Materials without the hot air p27 → IEA Energy technology perspectives report (2008a) Table 16.4

[2] http://www.iea.org/etp/explore/ http://www.iea.org/etp/etpmodel/industry/

[3] http://www.iea.org/publications/freepublications/publication/energy-technology-perspectives-2015.html

[4] Sustainable Materials without the hot air p30 & p31

[5] Power-to-Steel: Reducing CO 2 through the Integration of Renewable Energy and Hydrogen into the German Steel Industry

[6] http://bdsv.org/downloads/weltstatistik_2010_2014.pdf

Further research

Steel production processes: Ulcored, Hisarna and Ulcowin and Ulcolysis, developed by ULCOS, are electricity-based process concepts that produce iron using electrolysis reduction systems.

In this Chapter:

Copied to Clipboard