The Waiting Game for Data Centre Capacity (And How UK Businesses Can Beat It)
This blog article from Pulsant looks at the current backlog in data centre compute demand, and how businesses can work around this.
UK data centre occupancy hit 91% in 2024, according to Arizton market data, and new capacity is not arriving fast enough to close the gap. Grid connection wait times for new projects now run between five and 15 years, reports Data Center Dynamics, and Savills has attributed the 11% year-on-year drop in new capacity delivery to power constraints rather than a lack of demand or investment. Rising wholesale energy costs are adding pressure to an already tight market.
With grid access constrained for years ahead and energy costs rising, the supply assumptions most IT strategies were built on no longer hold. The organisations responding most effectively are not waiting for the market to rebalance. They are finding out where capacity exists across the UK and planning around that reality.
Here, Mike Hoy, Chief Technology Officer at Pulsant, offers his insights.
Grid Constraint is Not a Temporary Problem
Around 80% of UK data centre stock sits in or around London, and CBRE forecasts that 2026 will be the fifth consecutive year in which take-up has exceeded new supply there, with vacancy expected to fall to 5.9% by year end. The government’s AI growth zones at Culham, Teesside, Newcastle and sites in Wales are intended to open up new development locations, but grid connection demand has already jumped to nearly three times peak national electricity consumption, as Ofgem’s data confirms, meaning sites that exist on paper are years away from having usable power.
UK data centres currently account for around 5.8% of national electricity consumption, a level at which political resistance to approving new facilities tends to harden, according to International Data Centre Authority research. UK industrial electricity prices run at roughly 125% above the EU median for large users, according to government energy price data, a gap the House of Commons Energy Security and Net Zero Committee has described as a severe competitive disadvantage for energy-intensive sectors. For data centre operators, a provider’s position on power procurement and grid access will increasingly shape both what they can offer and what it costs.
“The grid connection queue is the number that most businesses haven’t fully absorbed yet,”Mike Hoy, Chief Technology Officer at Pulsant, says. “When you tell an IT director that a new facility might not get power until the early 2030s at best, it reframes the whole conversation about where their infrastructure should actually sit.”
The Market has Moved Faster than Most Procurement Plans
New capacity in established markets is routinely contracted before it is built, which means businesses entering the market reactively are finding little to procure. Recent analysis of the delivery gap points to specialist labour shortages, supply chain pressure on cooling and power equipment, and component lead times running well beyond historical norms. Grid access is the primary constraint, but it compounds a pipeline already stretched from several directions, and the combined effect is that realistic delivery timelines are considerably longer than most IT planning cycles assume.
Providers operating outside the heavily constrained South-East face lower land costs, less competition for grid connections and, in some cases, access to renewable generation closer to site, all of which gives buyers a different conversation on pricing and availability. The concentration of demand in one part of the country has effectively made regional infrastructure a practical answer to a supply problem, rather than a secondary preference.
“A lot of companies have been going to the same markets because that is where they started, not because the infrastructure case points there,” says Hoy. “When you map workloads against where your users and data actually are, the case for regional distribution is usually stronger than expected. The question is whether that review happens before or after capacity becomes a problem.”
How Businesses are Adapting
CBRE’s mid-year 2025 review found that power constraints and rising development costs were already pushing enterprises towards regional facilities with existing connections and available space, and that shift has continued to accelerate. Businesses securing capacity now are operating on longer planning horizons, treating infrastructure procurement more like a property decision than a reactive IT purchase made when headroom runs low.
The more significant change for many organisations is architectural. Running everything through a single provider made sense when availability was not a constraint, but businesses are now examining whether that approach reflects a deliberate decision or simply an inherited one. Steady-state, compliance-sensitive or latency-critical workloads each have different infrastructure requirements, Distributing them appropriately across regional colocation and cloud typically reduces both cost and exposure to a single constrained market.
Hoy adds, “The businesses in the strongest position right now did two things: they planned further ahead than the immediate need, and they looked at regional colocation on its merits rather than as a fallback. Distributed infrastructure across the UK is not a workaround for the capacity shortage. For most enterprises, it is the more logical answer.”