Europe AI data centres seek cheaper, quicker energy and land

By Simon Jessop and Iain Withers

LONDON, Aug 19 (Reuters) - Europe's developers of data centres for artificial intelligence are seeking out less traditional locations away from major cities, where they can benefit from cheaper energy and land and quicker connection times, data analysis seen by Reuters shows.

The trend can mean investment in underprivileged areas where governments are trying to stimulate jobs and growth, but the developers also risk opposition from residents concerned about dwindling natural habitats and competition for power and water.

JLL, one of the world's largest commercial real estate and property services companies, shared data on plans for hyperscale data centres that are due to come online rapidly over the next two years.

The centres, which do not need to be as close to end-users, will be more than three times further away from major cities than those built in the prior three years.

JLL did not disclose the names of the companies developing the centres, although the world's largest companies, such as Meta, Google and Microsoft, are investing globally.

SITES IN CITIES BECOME SCARCE

Between 2026 and 2028 the average distance of the sites from a major hub will be 175 kilometres (109 miles) compared with 46 kilometres for projects delivered between 2022 and 2025 as power and land in cities such as London and Frankfurt become scarce.

Greenfield projects account for 39% of Europe's future pipeline, compared with only 8% of delivered projects, the data showed. At the same time, the share of pipeline projects in inner-city locations is expected to fall to 5% from 13%, with the rest in industrial or edge-of-city locations.

The shift is driven largely by AI training campuses, which require vast amounts of electricity, as well as needing water for cooling.

"The determining factor is increasingly where sufficient power can be secured, rather than simply where demand exists," said Assad Noori, JLL's head of data centres in Europe, the Middle East and Africa. "Data centres are being brought to where the power is, not the other way around."

JLL estimates the world's four largest hyperscale cloud providers will spend $725 billion in 2026, up 77% from $410 billion in 2025, mostly on AI computing and data-centre infrastructure.

By 2030, AI workloads could account for around half of global data-centre capacity, it added.

Other sources of data-centre analysis supported the trend shown by JLL.

Of nine proposed gigawatt-plus capacity data centres across Europe, just one is planned for near a major city - Paris - with the rest strewn across areas from rural Spain to northern Sweden, according to DC Byte data tracking early stage projects.

While the core markets of Frankfurt, London, Amsterdam, Paris and Dublin remain the largest, and continue to be in demand, they increasingly face land shortages, planning restrictions and lengthy waits for connections to the grid.

"Europe's core markets will remain critical because enterprise demand isn't going anywhere," said Martin Jensen, president of JLL's EMEA data centres division.

"Hyperscale AI infrastructure requires a completely different scale of power and land," he added.

According to JLL, powered land costs an average €2.36 million per megawatt of IT load in the core markets, €978,000 in cities regarded as secondary, including Copenhagen, Warsaw and Milan, and €512,000 in tertiary areas such as Bordeaux, where costs can be as low as €200,000.

Amsterdam remains the most expensive market at roughly €2.7 million per megawatt, followed by London at €2.6 million and Frankfurt at €2.5 million.

"London has already seen significant digital infrastructure development driven by cloud and has other asset classes competing for space leading to high land prices. Power is now constrained in the key cloud locations across the market," said Rupert Duckworth, associate director, EMEA data-centre advisory at real estate firm Savills.

(Reporting by Simon Jessop; editing by Barbara Lewis)