AI datacentres plan on-site power stations on huge scale despite climate risks, research reveals
Developers of AI datacentres plan to build on-site gas power stations on a massive scale, creating 181GW of capacity across the US, UK and Europe, research reveals.
The planned construction would generate 345m metric tonnes of carbon dioxide equivalent annually in the US alone, equivalent to emissions from 72.8m fossil fuel cars. This occurs as global leaders meet at the 31st UN climate summit in Turkey, where electrification from clean sources is a central discussion point.
Massive Energy Demand
By 2035, electricity demand from AI datacentres is predicted to almost quadruple to 2,354 terawatt-hours, equivalent to 7% of global electricity demand. These facilities require enormous power for servers, cooling, and operations.
The extraordinary economics around AI are making speed to power one of the most important priorities for defining constraints on datacentre developers.
Temporary Solution Concerns
While developers claim gas-fired power would be temporary during grid connection delays, experts warn the economics could encourage continued operation, creating long-term fossil fuel dependency. In the UK, sites like Wapseys Wood in Buckinghamshire and Quest Park in Bedfordshire are seeking on-site gas power due to grid connection waits.
Chancellor John Healey has approved temporary on-site fossil fuel stations for datacentres as part of a datacentre covenant, a move criticized by environmental groups.
Private gas power stations for datacentres should be a red line for any government serious about climate action, so it's shocking to see the chancellor give them his explicit backing.
Sustainable Alternatives
The BloombergNEF report emphasizes that ensuring new datacentres connect to the grid is the more sustainable option. Faster grid connections are crucial to limiting emissions and preventing fossil fuel lock-in. The report notes that the climate outcome depends not only on electricity usage but on how quickly grids can connect facilities and whether cleaner generation is added alongside demand.
Related
