How much hydrogen will be needed to replace coal in making steel?

About 7-9% of the world’s emissions arise from the manufacturing of steel. It is the world’s most polluting industry. Hydrogen could entirely replace the massive use of coal, although the transition will be expensive. However it is probably the only realistic way that steel can get to net zero, a conclusion that seems increasingly shared within the industry.

This note looks at the likely costs of making steel without significant emissions. It assumes that hydrogen is made using renewable electricity and briefly assesses how much new wind or solar capacity will be required to allow the industry to get to ‘net zero’. Making hydrogen from steel only takes place today in tiny quantities so the figures in this article cannot be definitive but I thought it would be helpful to give a sense of scale. Corrections are very welcome.

The basic numbers

The world makes about 1.8 billion tonnes of steel a year. This number is expected to rise to possibly double this level by 2050, although there is a very wide range of forecasts. 

Steel use in developed countries will not rise substantially, if at all. A modern economy typically requires a stock of about 12 tonnes of steel per person to provide the buildings, cars and other infrastructure required. Most OECD countries are already at this level. A decade of rapid building has given China a large fraction of the circa 12 tonnes per person required. 

But total steel sales of about 1.8 billion tonnes a year only provides about a quarter of a tonne per person globally. Although we will probably see improvements in the efficiency of steel use, replacing some metal with wood or carbon fibre, the world is very far from sating its needs.

‘New’ steel versus recycled metal. 

About three quarters of all steel made today comes from the processing of iron ore. Coal is burnt in a blast furnace to ‘reduce’ the ore, that is extract its oxygen leaving metallic iron. The remainder is almost all made from the recycling of existing steel in electric arc furnaces. 

 ‘New’ steel                 1.35 billion tonnes

Recycled steel             0.45 billion tonnes

Total                            1.80 billion tonnes

 Some processes in the manufacture of ‘new’ steel can be improved. New plants use less coal than ones that are fifty years old. But the processes employed today will always need very large amounts of coal.[1]

 Each tonne of ‘new’ steel typically requires about 0.77 tonnes of coal, meaning that the industry as a whole uses just over 1 billion tonnes a year.

The energy value of the type of coal used for steelmaking is about 8 megawatt hours (MWh) per tonne. So each tonne of ‘new’ steel has typically required about 6 MWh in the process of getting from iron ore to a finished steel product, such as coil used for making the exteriors of cars. 

The coal energy needed for steel-making is therefore

1.35 billion tonnes of steel times 6 MWh = about 8,000 Terawatt hours (TWh) = as a comparison, about one third of global electricity consumption

 By contrast, recycled steel uses much less energy per tonne. One source suggests about 0.67 MWh per tonne of finished product. 

Using hydrogen instead

A small quantity of steel is made today using what is called the ‘direct reduction’ process and the technology is mature. A synthesis gas (hydrogen and carbon monoxide) made from methane (natural gas) is burnt in a large chamber to extract or ‘reduce’ the iron ore to metal.

The first experiments in large scale direct reduction using pure hydrogen are now being carried out at the SSAB steel works in Sweden. These experiments will give us more accurate data on the amount of hydrogen needed. 

Direct reduction using hydrogen will almost certainly be more energy efficient than using coal. From reading around the subject, I guess that a tonne of finished ‘new’ steel will require about 3 MWh of hydrogen, considerable less than the 6 needed for coal-based processes. However the process of making the hydrogen will incur some additional energy losses in the electrolyser, taking the amount of electrical energy required up to between 4 and 4.5 MWh per tonne of steel. Let’s assume the figure is about 4.25 MWh.

 Amount of electricity required to create the hydrogen to make all the world’s ‘new’ steel at today’s production levels = 1,350 million tonnes times 4.25 MWh = 5,700 Terawatt hours or about one quarter of world electricity production.

If the hydrogen is all made from renewable electricity, how much extra wind or solar capacity will be required?

 If the average new wind turbine has a capacity factor of 40% (low for offshore, probably about right for onshore) then the world would need about 1600-1650 gigawatts of extra turbines. This is well over two and a half times the currently installed amount of wind power globally. The figure for solar PV would be roughly twice this level.

What weight of hydrogen will be required?

 Figures for the world’s current hydrogen production vary between sources but most indicate that about 70-80 million tonnes of the gas are made each year. None is currently used for making steel.

A tonne of ‘new’ steel will need about 90 kilogrammes of H2 (with an energy value of about 3 MWh). 

 1,350 million tonnes of steel, each requiring 90 kg will use about 122 million tonnes of hydrogen, or about 50% more than current world production.

What about the capacity of electrolysers?

If we assume that the electrolysers work every hour of the year, then we will need about 650 gigawatts of capacity. This compares to less than 1 gigawatt installed globally at present. 

Conclusion

A swing to hydrogen as the fuel and reducing agent for steel production will involve a major transition. Very large amounts of new renewable capacity will be required if ‘green’ hydrogen is used. The electrolyser manufacturing industry will need to expand by several orders of magnitude. And, of course, the steel industry will have to invest billions in the new plants required. Most sources suggest that for the main steel firms to make the transition voluntarily that they will have to see a mixture of low power prices (say below $40 a MWh) and a reasonable carbon tax (at around $50 a tonne). These figures seems entirely attainable to me.

 

 

 

 

 

 

 

 

 

 

 

 

 


[1] Today’s plants use a blast furnace (BF) in which coal is used to reduce ore to liquid iron. The iron is then turned into steel in a basic oxygen furnace (BOF). The BF-BOF process is now used to make a very large fraction of all ‘new’ steel.