Physical and fiscal reality mean EU ‘green steel’ programme is bonkers

No 'green dream' in China: Workers monitor the blast furnace at the Bao steel mill., where steel is powered by coal. 'If Europe wishes to retain the integrated steel industry needed for infrastructure and armaments, it will rely on fossil fuels to power it.' (Photo by Ryan Pyle/Corbis via Getty Images)

Share

The EU Commission’s proposal to triple data centre capacity over the next five to seven years is intended to help Europe catch up with China and America in the race to unlock the potential of AI.  The draft “Cloud and AI Development Act” has prompted a fierce and predictable pushback from environmental groups, who point to data centres’ voracious consumption of electricity and water. Barcelona hosts a cluster of data centres that are contributing to water shortages and a depleting water table.  Frankfurt’s array of data centres already takes one-third of the region’s electrical capacity, with many new data centres installing gas-fired generators on site due to the lack of new grid connections.   Innovation in what promises to be a transformative technology is running into local opposition organised by a range of NGOs.  

It’s not terribly surprising that the eco-left would mobilise in opposition to data centres hosting “scary” AI, many owned and operated by their favorite villains, American billionaires Mark Zuckerberg and Jeff Bezos.  The EU’s favoured precautionary principle aids its desire to regulate AI into innovation-stifling submission.  Yet the impact of tripling data centre capacity in Europe on precious water and electrical resources will be dwarfed by the demands of a pet project of the Greens: The decarbonisation of the European steel industry.  

The termination of free carbon credits to the steel industry in 2034 is intended to end the use of coal-fired blast furnaces for the reduction of iron ore into the pig iron needed to make steel.  In its place, the EU has approved over €9 billion in grants to replace this coal with hydrogen to make “green steel.”  Where will this hydrogen come from?  The cheapest and most readily available source is natural gas, but every ton of natural gas produces only a quarter ton of hydrogen while the steam process used creates up to twelve tons of carbon dioxide.  Europe’s answer is electrolysis:  Windmills and solar panels generate electricity, which power electrolysers, which produce clean hydrogen, which in turn fuel the blast furnaces needed to make the pig iron needed for steel production.  Problem solved, at least to the satisfaction of the policy wonks in Brussels.

Even a cursory examination of this scheme reveals grave difficulties.  Blast furnaces run 24 hours a day for years at time, while solar and wind produce electricity only when the weather cooperates. EU rules set to take effect in 2030 will require the hydrogen producing electrolysers to match their power consumption with renewable generation on an hourly basis, which is likely impossible given the wild swings in power sourced from wind and solar.  In order to achieve hourly matching, there will need to be expensive battery buffers between generation and electrolysis sufficient to bridge days (and nights) when renewable sources fall far short of their topline generating capacity.  Buffers will also be needed between electrolysers and blast furnaces as hydrogen production will not reliably match the precise demands of iron production.  Storing hydrogen, quite literally the smallest molecule on the periodic table, is a demanding and hazardous enterprise (recall the Hindenberg) requiring massive tanks or sealed underground vaults three times larger than the equivalent energy volume of natural gas.

Electrolysers are themselves an expensive and complex technology that won’t easily scale to a level sufficient to replace coal.  Sixty percent of the world’s electrolysers are made in China, meaning that EU environmental rules will (once again) grant China a dominant role in a key industrial technology.  They are also voracious consumers of electricity needed to crack water into hydrogen and oxygen.  According to analysts at the International Energy Agency (IEA), global production of hydrogen today is just 70 megatons, with 99 per cent sourced from natural gas and coal.  They estimate that replacing this with electrolysers will require an annual electricity demand of 3600 terawatt hours, or roughly the EU’s entire annual electricity consumption. Bear in mind that the full replacement of coal in EU blast furnaces will need 70 megatons of hydrogen, or today’s entire global production and the EU’s entire electricity production. Hydrogen production also uses water at a rate of nine litres for each kilogram of hydrogen generated, meaning those 70 megatons of hydrogen will consume an astonishing 63 billion litres of water.   

The Green Dream of clean, hydrogen-powered steel production will soon collide with physical and fiscal reality. The EU’s forced march to green steel depends on wildly unrealistic assumptions.   The cost and difficulty of building a fleet of electrolysers sufficient to generate the necessary volumes of hydrogen is beyond the capacity of European industry. A massive escalation in Europe’s power generation and distribution will break national budgets, as will the battery and hydrogen storage facilities needed to smooth the path from variable renewable supply to constant blast furnace demand.  The dim prospects of mastering all these moving parts suggests that if Europe wishes to retain the integrated steel industry needed for infrastructure and armaments, it will rely on fossil fuels to power it.

The EU does recognise that the transition to green steel will raise European production costs by double digits, widening the competitive disadvantage with Asian steel makers.  Like wind and solar, the green steel sector will thus depend on public subsidies to survive. The Carbon Border Adjustment Mechanism may serve as a tariff barrier to cheap foreign steel, but has no means of verifying the low carbon claims by overseas manufacturers.  Neither China nor India will permit EU inspectors the access needed to verify whatever representations they make to access the European steel market. The upshot will be an acceleration in the de-industrialisation of Europe as her manufacturers seek cheap energy and steel in the rest of the world.  While the EU foresees the end of coal in steel production in a mere eight years time, the IEA offers a more realistic assessment: even under optimistic assumptions, only 8 per cent of steel production will rely on electrolyzed hydrogen by 2050.  

Data centres may be a convenient boogeyman for environmental NGO fundraising, but it is their own green steel fantasy that will push Europe toward massive increases in electricity demand and the depletion of precious water resources.  Taxpayers will pay the price as electrolysers drive up power bills from an overstressed grid, steel-dependent industries flee the continent, while expensive and scarce hydrogen fails, ever, to displace cheap and abundant fossil fuels.  The precise technical term for this particular policy programme is “completely bonkers.”