Stargate AI project highlights need for coordinated infrastructure planning

This week’s launch of Stargate – a major AI investment bringing OpenAI’s Sam Altman and Nvidia’s Jensen Huang to Blyth, Northumberland – showcases a new AI data centre partnership with UK company Nscale. But it also highlights a growing tension: how do we balance national AI ambitions with local planning realities?

The challenge: managing data centre resource demands

Data centres are essential infrastructure for the UK’s digital economy and AI leadership. However, they require significant electricity and water resources. A single facility can use as much power as a town of 90,000 people.

This creates planning challenges that need careful management:

• Grid capacity: Network operators face mounting connection requests but reinforcement takes years
• Water resources: Cooling systems need access to reliable water supplies alongside housing and other local needs

Managing energy demand and grid capacity

Modern data centres need 50 to 100 MW of constant power. For example, a proposed UK facility from Amazon will consume as much electricity as the town of Burnley.

Distribution Network Operators (DNOs) are managing increasing connection requests. Grid reinforcement takes several years – often longer than developers are willing to wait. Some developers now build their own power generation to reduce reliance on grid connections.

The planning challenge

Local planners need to coordinate grid capacity allocation between industry, data centres and other key infrastructure priorities including housing, transport electrification and heat pumps to maximise local benefit.

Water resources: planning for shared needs

Data centres use water for cooling, especially in high-processing facilities. Many rely on evaporative cooling that draws from local freshwater supplies.

Careful planning is needed in water-stressed areas where population growth and climate change are already tightening supply. Both data centres and power stations require access to the same water sources.

​Our Energy Water Nexus project highlighted the looming conflict that will arise as industrial users and consumers compete for the same limited water resources in the same places. As water availability dries up in south-east England – an area the Environment Agency predicts will face serious water stress by the 2030s – developers will be forced to look further afield. This means expanding outwards into other regions where water is more available, but grid and land constraints may persist.

Balancing national priorities with local needs

Local authorities face a complex balancing act: the need to support critical national infrastructure while ensuring local communities have access to essential resources for housing, development and decarbonisation.

Local decisions get overturned

Several councils have rejected data centre applications due to environmental concerns or lack of community benefit. Many decisions get overturned on appeal by the Secretary of State.

Result: Local authorities feel bypassed, despite understanding their communities best.

Housing developments stall

Water constraints are forcing councils to scale back affordable housing projects. UK government rules require new housing to have “no net increase in water demand” – but councils lack funding for water offset schemes.

Water scarcity could cost the UK economy £25 billion over five years, mainly from stalled housing and infrastructure projects.

Similarly, limited electricity network capacity is a barrier to UK housebuilding, restricting the delivery of new homes and net zero targets.

Great British Energy changes priorities

Proposals for Great British Energy to offer a service that accelerates grid connections may help to prioritise projects with local economic value and quality jobs.

This could help housing and clean energy projects access constrained grid capacity. But where does this leave hyperscale data centres?

The Stargate opportunity

Stargate positions the UK as a global AI leader. The partnership brings together:

• UK energy resources
• Nvidia’s chips
• OpenAI’s technology
• UK sovereign AI capabilities

But repeating this kind of success will depend on solving the local planning puzzle.

Next steps for authorities

While the government can fast-track construction of data centres, it must work with regions to ensure that the national race to become an AI superpower doesn’t trample the needs for housing, transport and local power plans.

To balance these needs, authorities should take a strategic approach to energy planning at the local level which considers:

1. Water impact assessments: Require data centre developers to prove water availability will not hinder local priorities;
2. Local area energy goals: Ensure data centre power demand is coordinated with key decarbonisation initiatives, like the electrification of heat, required to transition local energy systems to Net Zero;
3. Grid capacity transparency: Work with DNOs to understand what grid capacity remains for local priorities;
4. Community benefit requirements: Ensure data centre developments deliver measurable local economic value, not just national strategic benefit;
5. Regional co-ordination: Work with neighbouring authorities to understand cumulative impacts across wider areas.

Prime Minister Keir Starmer’s AI ambition and the investments promised by tech firms this week are significant for the UK. With thoughtful local planning, we can ensure they benefit communities as well as the national agenda.