Deep Green

Scale AI or minimise carbon emissions? Support UK enterprises with high-performance compute or reduce pressure on the energy grid? These are being framed as binary choices by MPs, who warn that the data centres needed to fuel AI growth zones will overload the grid and risk the UK’s net zero targets.
This isn’t the case. With the right colocation facilities, we can have it all: fewer emissions, reduced grid pressure, and a thriving AI infrastructure. Here’s how: by giving priority grid access to data centres in the UK that reuse their heat, so every electron is used twice.
Not all data centres are built the same
MPs are right to be concerned about the grid. Energy bills are rising. Data centre operators face up to eight-year delays for grid access. Geopolitics is exacerbating uncertainty, as oil supplies fluctuate and prices soar.
Meanwhile, there’s pressure for the UK to maintain its position as a global AI innovator. To achieve this, we need infrastructure: UK data centres where enterprises can run their AI workloads. This is especially crucial for sectors that require low latency inference workloads or where data sensitivity requires sovereign UK-based processing. Beyond competing globally, the UK simply needs more data centre facilities to support our increasingly digital world.
Then, there’s the third concern: rising emissions. The UK aims to source all its power from clean energy by 2030. MPs recently raised concerns that data centres could increase energy demand to the point that we won’t meet this target.
This concern stems from a misconception: that data centres in the UK are all energy-hungry, power-guzzling hyperscalers that drain resources from local communities. Yes, hyperscale facilities are putting pressure on grids. Yes, some data centres discard the heat they generate, wasting valuable resources. However, when data centres reuse that heat by distributing it to the community via district heating systems, they reduce emissions and grid strain, while still fuelling our country’s AI growth.
Heat reuse benefits local communities and UK enterprises
As the UK struggles with growing energy demands, we’re missing a massive opportunity: to use our electrons twice over. When data centre servers process workloads, they generate huge amounts of heat, which can be captured and distributed to the local community via district heating systems. These heat offtakers (such as residential or public buildings, swimming pools, or greenhouses) then reduce their heating bills, emissions, and energy requirements.
Heat reuse also has benefits for the colocation provider and its compute clients. When heat offtakers receive free heat, they return cooled water to the data centre in a circular system. This improves efficiency and reduces operating costs. Subsequently, UK enterprises that run their AI workloads in heat-reuse data centres cut compute costs and boost their sustainability performance. The solution for MPs is clear: give priority grid access to data centres in the UK that reuse heat, and you’ll see the environmental and economic benefits across the board.
AI growth zones boost both the local and national economy
The recent suggestion from MPs that we abandon AI growth zones misses a crucial point: UK enterprises can’t run all their AI workloads in US-based clouds or foreign colocation providers. We need data centres in the UK where enterprises can process sensitive information, such as NHS or financial records, to ensure data sovereignty. What’s more, inference workloads (for example, in self-driving vehicles or advanced manufacturing) require low latency to function smoothly. This means colocation providers must be located near an enterprise’s HQ to ensure no lag in performance.
When done right, AI growth zones can benefit the economy locally as well as nationally. For example, Manchester is an upcoming AI hub that’s relieving grid strain from the saturated South and attracting tech companies to the North. Buying compute from Manchester colocation providers means these tech companies must set up new HQs in the city to ensure low latency. In turn, this creates jobs for Manchester locals.
Data centres with heat-reuse schemes bring even more local benefits, reducing carbon emissions, grid strain, and heating costs for the community. An AI growth zone consisting of data centres that reuse heat would have a quadruple impact: supporting local employment, reducing the community’s heating costs, destressing the grid, and helping the UK move towards net zero.
UK policy must support data centres that reuse heat
From Denmark to France to the Netherlands, government policy supports data centres that reuse heat, whether by mandating heat recovery or streamlining grid planning for heat-reuse data centres. These are the kind of policies we need on British soil. The recent NPPF sets out recommendations for the future of data centres in the UK. However, without a policy that prioritises grid access for data centres that reuse heat, our grid will continue to come under strain as we scale AI workloads. There are huge quantities of energy being thrown away in UK data centres. For a net-zero future that encompasses AI growth, we have to use every electron twice. In this way, reducing grid pressure and scaling AI workloads go hand in hand. Want to see a heat reuse data centre in action? Check out our Urmston, Greater Manchester facility.



