16MAY

OpenAI puts a number on UK electricity gap

4 min read

13:06UTC

OpenAI cited UK industrial electricity at 'more than four times' US, Finnish, Norwegian, and Swedish rates as a primary driver of the Cobalt Park pause, the first time a hyperscaler has named energy cost as a co-equal barrier alongside the grid queue.

← Back to OpenAI cuts $800bn; rivals double down Jump to analysis ↓

IndustryDeveloping

Key takeaway

OpenAI's 4x figure makes UK energy cost a permanent siting blocker until the merit-order pricing structure is reformed.

OpenAI publicly attached the multiplier 'more than four times' to UK industrial electricity rates relative to US, Finnish, Norwegian, and Swedish equivalents on 27 April 2026, naming the differential as a primary driver of the Cobalt Park pause in North Tyneside . The figure aligns with IEA price data showing UK industrial rates near $0.20-0.22/kWh against US $0.06-0.07/kWh and Finnish $0.07-0.09/kWh. For a 100 MW campus running at typical AI-training utilisation, that gap means roughly $100 million in additional annual electricity cost compared with a US site, before any consideration of compute density or cooling overhead.

OpenAI quantified what hyperscalers had previously said off the record. Until this week, the UK's headline data-centre problem ran through the grid queue: 50 GW of contracted demand against 45 GW of national peak . Hyperscalers had been treating energy cost as a secondary issue subordinate to interconnection wait times, on the working assumption that NESO and the AI Growth Zone designations would ease the queue. OpenAI's intervention separates the two and quantifies the second. With a queue solution, the UK still loses the campus economics by a factor of four.

UK industrial electricity pricing is shaped 60 to 70 per cent by gas-marginal merit-order pricing, plus carbon and network charges. None of those components have been reformed since the AI boom began, and the DESNZ and NESO policy responses available within the next twelve months do not change the merit-order structure. The implication is that the UK's data-centre future is becoming a talent-and-research base rather than a compute base; the capital that would have built Cobalt Park, Nscale's Glasgow expansion, and the Nottinghamshire AI Growth Zone is now redirecting into Nordic, Aragón, and Abu Dhabi tracks. The 50 GW grid queue persists, but it has been joined by a structural cost ceiling no rule cycle can lift.

Deep Analysis

In plain English

OpenAI explained publicly that electricity for businesses in the UK costs more than four times what it costs in the US, Finland, Norway, and Sweden. For a large data centre using 100 megawatts of power, roughly what a mid-sized city hospital district uses, that difference adds up to about $100 million in extra costs every year compared to a US location. This is the first time a major tech company has put a specific number on the energy cost gap and tied it directly to a decision not to build in the UK. The gap exists because UK electricity bills for businesses include several layers of charges for subsidising renewable energy and maintaining the grid that do not appear in Nordic or US tariffs at the same scale.

Deep Analysis

Root Causes

UK industrial electricity rates have three distinct structural components that do not appear in Nordic or US rates at comparable magnitude.

Network charges account for roughly £60-70/MWh of UK industrial tariffs in 2026 because the UK's Contracts for Difference (CfD) subsidy mechanism for renewables is recovered through network tariffs rather than direct taxation. The CfD mechanism has underwritten approximately £90 billion of offshore wind investment since 2014 and its cost recovery is embedded in the tariff structure until the early 2030s.

Balancing costs (the cost of keeping the grid stable when intermittent generation and demand are mismatched) add another £15-25/MWh in a market with 35 per cent wind and solar penetration but limited grid-scale storage. The UK's balancing mechanism is more expensive per MWh than any comparable European market because the GB grid runs at higher residual demand variability than the larger continental ENTSO-E synchronous zone.

The third component is the absence of a co-payment mechanism for transmission reinforcement: in the US, large loads trigger a generator-style interconnection study that assigns upgrade costs to the triggering party, capping unexpected network cost socialisation. In the UK, transmission upgrade costs are socialised across all users before cost recovery flows into network tariffs, which means existing users partly subsidise new entrants' connection costs, inflating the aggregate tariff.

What could happen next?

Consequence
The UK falls off the global top-10 shortlist for new hyperscale greenfield sites as long as the 4x cost gap persists, redirecting an estimated 2-4 GW of announced UK data centre pipeline to Finland, Spain, and the Gulf before 2028.
Opportunity
Ofgem's 2026 transmission tariff review and NESO's direct connection agreement process could, if they produce a 30-40 per cent reduction in network charges for large greenfield users, narrow the gap to 2-2.5x and restore UK competitiveness for mid-scale operators not requiring the Nordic price point.

Source Landscape

This story draws on neutral-leaning sources

Primary parallel: The 2010s wave of aluminium smelter closures in the UK followed the same industrial electricity cost logic on a sector that requires roughly 15,000 kWh per tonne of aluminium produced. Alcan's Lynemouth smelter (closed 2012) and Rio Tinto Alcan's Anglesey operation (idled 2009, closed permanently 2013) both cited UK electricity costs at 2-3x European continental rates as the primary driver.

The government's response, carbon price support exemptions for energy-intensive industries, came too late and was too narrow to reverse the closures. Both Alcan and Rio Tinto Alcan exhausted government relief options before closing; the question for UK data centre policy is whether Ofgem's 2026 tariff review moves faster than the aluminium precedent suggests.

Counter-parallel: Ireland followed the same cost trajectory through the 2000s but reversed it for the data centre sector specifically: EirGrid's Large Energy User (LEU) framework, introduced in 2014, gave hyperscale operators a direct-connection tariff that bypassed distribution network charges, reducing effective rates by 30-40 per cent.

The UK equivalent (direct connection agreements via NESO) exists but is processed through a 50 GW queue that makes it inaccessible in realistic timescales for new entrants. Ireland's LEU framework worked because EirGrid had spare transmission capacity to offer; NESO does not.

The $100 million annual cost gap OpenAI cited, confirmed by IEA data, understates the full financial impact for a hyperscaler building at the 500 MW or 1 GW scale that defines Tier 1 campuses. At 500 MW, the gap reaches $500 million per year, roughly 1.4 per cent of Microsoft's 2026 AI workload capex budget, or the equivalent of one medium data centre campus in electricity overpayment annually.

For UK-listed real estate investment trusts and infrastructure funds with data centre exposure, Land Securities, Segro, Digital Infrastructure Vehicle, the OpenAI statement depresses the addressable market for UK data centre investments at hyperscale.

colocation operators serving enterprise customers (Equinix LD8, Global Switch, CyrusOne) face a different calculus: their tenants are less price-sensitive than hyperscalers and are not shopping the same greenfield location market, so the 4x figure has less direct effect on existing colocation tenancies.

Consensus view: Energy Systems Catapult analyst Jess Ralston assessed that the 4x multiplier is structurally accurate for industrial tariffs in 2026: UK business rates include network charges (roughly £60-70/MWh), Climate Change Levy exemptions that were partially removed in 2023, and balancing costs that have risen with the penetration of intermittent renewables.

The gap to Nordic rates is structural, not cyclical, it reflects infrastructure investment decisions made over two decades and cannot be narrowed by short-term policy adjustment.

Counter-view: Aurora Energy Research's Great Britain power team argues that the 4x figure conflates Cobalt Park's current network tariff with what a Purpose-built data centre with a direct connection agreement (DCA) and a BESS (battery energy storage system) co-location arrangement could achieve.

Their modelling suggests a greenfield UK data centre with an optimised connection and storage strategy could reach an effective industrial rate of $0.12-0.14/kWh, still higher than the US, but 2x rather than 4x, which changes the competitiveness calculation for co-located renewable generation.

Key tension: Whether OpenAI's 4x figure reflects a permanent UK cost disadvantage that no policy intervention can correct in the near term, or a current-tariff snapshot that a new DCA and renewables integration deal could substantially reduce, which is the central policy question that Ofgem's 2026 transmission tariff review is attempting to answer.

Sources:The Next Web

Mentions:OpenAI →Stargate →NESO →International Energy Agency →Aragón →DESNZ →Finland →Ofgem →North Tyneside →EirGrid →Norway →Nscale →Abu Dhabi →Glasgow →Sweden →

First Reported In

Update #2 · Maine veto, Seattle freeze, $725bn capex

The Next Web· 6 May 2026

Read original →

Causes and effects

Caused by

OpenAI pauses Cobalt Park Stargate site

OpenAI's Cobalt Park pause (ID:2785) was extended in this window by OpenAI attaching the specific 4x energy-cost figure, turning a vague regulatory complaint into a quantified site-selection barrier.

Occurred 23 Apr 2026

Read story →

This Event

OpenAI puts a number on UK electricity gap

For a 100 MW campus, the UK-US rate gap means roughly $100 million a year in additional electricity cost; the figure drops the UK off the global shortlist for new greenfield builds.

Led to

UK Gate 2 grid offers begin issuing

OpenAI's quantification of UK industrial power at four times US and Nordic rates forced the electricity discount concession that is now part of the Gate 2 AI Growth Zones offer.

Occurred 13 May 2026

Read story →

Different Perspectives

European energy regulators and climate advocates

GE Vernova's 100 GW gas-turbine backlog, driven by AI data-centre demand, puts IEA net-zero pathways under pressure: 15-27 GW of onsite gas is forecast for US data centres by 2030. The Amazon Boardman $20.5m pollution settlement gives environmental litigation a financial template it previously lacked.

Irish Commission for Regulation of Utilities (CRU)

The CRU-compliant Pure DC behind-the-meter template gives operators a replicable European consent pathway outside the UK queue. Ireland's existing hyperscaler density and the CRU framework's behind-the-meter provisions make it the lowest-friction large-load jurisdiction in Europe for 2026 approvals.

Nordic operators (Equinix-CPP-atNorth, Aikido Technologies)

Equinix's CPP-atNorth acquisition and Aikido's AO60DC floating-wind pilot at METCentre Norway offer hyperscalers a consented, low-carbon supply chain that bypasses both US moratorium risk and European grid queues. Norway's renewable surplus and Fingrid connection windows make the Nordic corridor the clearest alternative supply chain in the current environment.

G42 and Abu Dhabi sovereign funds

G42 and Khazna are tracking Stargate UAE phase 1 for Q3 2026 commissioning with a sovereign anchor tenant. The model insulates the project from community opposition and planning litigation, making Abu Dhabi the furthest-advanced non-US build on the current verified green-light map.

UK Government and NESO

NESO began issuing Gate 2 Phase 1 transmission connection offers this week against a 116 GW applications backlog, with electricity discounts and HV self-build rights attached. The UK is using grid access as an industrial-policy instrument to attract compute investment redirected from US jurisdictions with active moratoriums.

US hyperscalers and OpenAI

The four big hyperscalers raised collective 2026 guidance to ~$725bn while OpenAI compressed its own commitment by 57% to $600bn and pivoted to leased compute; the divergence shows capital allocation cycles running independently of AI developer demand, with hyperscalers betting that transformer-slot scarcity in 2027 is riskier than current community opposition.