For most of the UK data-centre market's development, location has been driven by a familiar combination of connectivity, customer proximity, land and access to established digital-infrastructure clusters.
Power has always mattered. Increasingly, however, it is becoming one of the factors that determines the geography itself.
The shift is visible in several places at once. Capacity constraints are pushing development beyond established London clusters. Government policy is explicitly encouraging large electricity users to locate where they can make better use of the power system. And public and private capital is beginning to support data-centre investment in regions where energy availability would historically have been a weaker part of the location thesis.
The consequence is not that London is losing its position as the UK's dominant data-centre market. It is that location quality is becoming inseparable from power-system economics.
London still dominates, but the boundaries are moving
London remains by far the UK's largest data-centre market.
CBRE estimates that the London area accounts for more than 80% of national data-centre supply, reflecting the concentration of cloud availability zones, fibre connectivity, enterprise demand and established operator ecosystems around the capital. Its UK Data Centres Outlook 2026 also notes that capacity is increasingly being planned outside established clusters such as Slough and Hayes/Park Royal, where power is limited, towards locations including North London, Oxfordshire, Hertfordshire, the London Docklands and Essex where availability is greater.
Those advantages remain difficult to replicate.
But the geography within and around that market is already changing.
That is an important distinction.
The constraint is not necessarily reducing demand for the London market. Instead, it is expanding the geographic area over which developers must search for infrastructure capable of serving that demand.
In other words, electricity availability is beginning to alter what counts as a viable London-serving location.
Government is going further: using electricity economics to influence location
The more significant structural change is occurring beyond London.
The Government's AI Growth Zone programme explicitly treats access to power as part of industrial and regional policy. Early criteria sought locations capable of supporting at least 500MW of AI infrastructure, with preference for sites that already had substantial electricity connections, credible routes to additional capacity, and access to low-carbon generation and storage.
But the policy has progressed beyond simply helping data centres find electricity.
Government is now proposing to use the economics of the electricity system itself to influence where large data centres locate.
The reason is straightforward.
In parts of Britain, particularly Scotland and northern England, renewable generation can exceed the transmission system's ability to move that electricity to centres of demand. At those times, generators may be constrained and the system incurs costs to manage the imbalance.
Large, appropriately located electricity users can potentially change that equation.
A data centre consuming electricity in an area with surplus generation may reduce the amount of power that needs to be transported elsewhere and, under the right conditions, reduce system constraint costs.
The Government's Reformed National Pricing delivery plan therefore proposes targeted electricity-price support for eligible data centres within AI Growth Zones where their location creates measurable system benefits.
From April 2027, subject to legislative timetables and final scheme design, the Government has indicated potential electricity-cost reductions of up to £24/MWh in Scotland, £16/MWh in Cumbria and £14/MWh in the North East. The same policy is set out in Delivering AI Growth Zones.
The precise mechanism is still being developed. The strategic signal is nevertheless important.
The electricity system is beginning to assign economic value to where data-centre demand is located.
That changes the traditional location equation
Historically, electricity could be considered predominantly as an input to the location decision: is enough power available, at what cost, and when can it be delivered?
The emerging model introduces another question: does placing demand here create value for the wider power system?
That is a different way of looking at data-centre geography.
A location with significant generation but limited transmission capacity may previously have appeared disadvantaged relative to an established digital-infrastructure hub.
If large-scale demand can absorb locally available power, benefit from lower electricity costs and receive policy support for planning and connection delivery, some of that disadvantage can narrow.
This does not mean electricity will determine location by itself.
Fibre availability, latency requirements, customer geography, planning, land, water, workforce, supply chains and the ability to develop at scale all continue to matter. Different workloads also place different demands on location.
But electricity is moving from being one factor among many towards becoming a strategic variable capable of changing the relative competitiveness of regions.
Capital is beginning to follow the infrastructure thesis
The development of the Lanarkshire AI Growth Zone provides a useful early example.
In August 2026, the Government announced a £300 million financing package to support DataVita's expansion in Lanarkshire, including expansion of its existing data centre and construction of a second facility. The package includes a £202 million guarantee from the National Wealth Fund supporting lending from ING, ABN AMRO and Santander, alongside capital from the Scottish National Investment Bank and Siemens Financial Services. The Government's announcement positions the project explicitly around Lanarkshire's energy supply and strategic location within Scotland's electricity network.
The significance is not simply the size of the financing.
Lanarkshire has been positioned around the interaction between digital infrastructure and Scotland's energy system, with the project's energy requirements expected to be met overwhelmingly through Scotland's low-carbon generation mix.
That does not establish that every data-centre project in Scotland or northern England is inherently attractive.
It does demonstrate that the combination of power availability, system value, public policy and digital-infrastructure investment is becoming financeable at meaningful scale.
This is an important development for the market.
Power abundance alone will not create investable data centres
There is a risk of taking the argument too far.
Regions with abundant generation are not automatically attractive data-centre markets.
Electricity that exists somewhere on the system is not the same as electricity that can be delivered to a specific site, at the required voltage, within a commercially useful timetable and at a predictable cost.
The same distinction underpins our earlier analysis of deliverable power and our assessment of whether self-build can change the economics of time-to-power.
Nor does a lower electricity cost compensate automatically for weak fibre connectivity, unsuitable land, planning difficulty or insufficient customer demand.
For investors, developers and operators, this means the relevant asset is not simply a site in a region with abundant renewable generation.
It is a site where the energy advantage can be translated into an operational digital-infrastructure advantage.
That requires the power system and the data-centre development to work together.
A new dimension of site value
As this market develops, the concept of a "good data-centre location" may therefore become more differentiated.
Sites will still compete on connectivity, planning, land and access to customers. But they will increasingly also compete on:
- when power can actually be delivered;
- the scale to which the connection can grow;
- the cost and predictability of electricity;
- proximity to generation and network infrastructure;
- whether the site's demand can alleviate rather than exacerbate system constraints; and
- whether surrounding energy infrastructure can support future expansion.
That changes the investment lens.
A regional location may justify attention not simply because land is cheaper than in an established hub, but because its underlying power economics are structurally stronger.
Conversely, a site close to an established digital cluster may become less attractive if the cost or timing of obtaining sufficient electricity overwhelms its traditional locational advantages.
The comparison therefore becomes less about primary market versus secondary market and more about the quality of the complete infrastructure proposition.
The UK market is unlikely to decentralise evenly
None of this implies the disappearance of established clusters.
London's ecosystem, connectivity and concentration of customers give it structural advantages that power constraints alone are unlikely to erase.
A more plausible outcome is a more complex UK data-centre map.
Established markets will continue to attract capacity where power can be secured. Their boundaries will expand as developers search further from traditional clusters. At the same time, selected regional markets may become substantially more competitive where large-scale power, digital connectivity, planning support and customer requirements align.
AI Growth Zones add policy momentum to that process, but the underlying driver is broader.
The rapid growth of compute is colliding with an electricity system in which generation, network capacity and demand are not located in the same places.
That mismatch is beginning to affect capital allocation.
Power is therefore becoming more than an operating requirement for UK data centres. It is becoming part of location strategy itself. The strongest emerging markets are likely to be those that can turn an energy-system advantage into deliverable, commercially relevant digital infrastructure.