AI datacentres plan on-site power stations on huge scale despite climate risks, research reveals

Detailed summary

181GW of gas-fired capacity proposed for sites in US and Europe because of delays in connecting to national grids, says reportDevelopers of AI datacentres plan to build on-site…

181GW of gas-fired capacity proposed for sites in US and Europe because of delays in connecting to national grids, says report

Developers of AI datacentres plan to build on-site climate-destroying gas power stations on a huge scale to meet their energy demands, research reveals.

Datacentres planned across the US, the UK and mainland Europe have proposals to build fossil fuel plants on-site to create 181GW of gas-fired capacity because of delays to connect to national grids, according to the findings. This is almost four times the total peak electricity demand of Great Britain.

The research by BloombergNEF says this “speed to power” will – if the planned rollout happens – emit about 345m metric tonnes of carbon dioxide equivalent a year in the US, where 90% of on-site gas-fired power stations are planned. This is equivalent to annual greenhouse gas emissions from about 72.8m fossil fuel cars.

The report was published on Wednesday as global leaders gathered in Turkey for the 31st UN climate summit, where increasing electrification – ideally from clean sources – is a central part of discussions. However, a number of key European leaders, including from France and Germany, will not be at Cop31, instead heading to Dublin where EU leaders will hold their first dedicated AI summit on Friday.

Mark Daly, the head of datacentres and emerging tech at BloombergNEF, said: “The extraordinary economics around AI are making speed to power one of the most important priorities for defining constraints on datacentre developers.

“The climate outcome will depend not only on how much electricity these facilities use, but on how quickly grids can connect them and whether power systems add cleaner generation alongside that demand.”

By 2035, the electricity demand of AI datacentres was predicted to almost quadruple to 2,354 terawatt-hours – equivalent to 7% of global electricity demand – the report said.

AI datacentres use a massive amount of electricity to power servers and cool equipment, and to…

Opening of the article, excerpted from The Guardian Business. To read the rest, open the article on the publisher's site.