Framework v1 — built for manual, checked data entry

UK AI Energy Index

The UK AI Energy Index is a structured record of what artificial intelligence infrastructure is doing to Britain’s electricity system. It tracks AI electricity demand, the growth of UK data centres and the grid pressure that follows, indicator by indicator. Each entry below carries a named source, a stated update schedule, a confidence level and a public revision history, so you can judge how solid a number is before you use it.

Status: the index structure is live. Readings are compiled and published by hand, indicator by indicator, as each source is verified. We are deliberately not automating collection yet — an automated feed that quietly ingests a bad number is worse than an empty cell.

Indicators tracked

Published readings

Last updated

15 Aug 2026

Every figure carries its own source and collection date.

The three sub-indices

The UK AI Energy Index tracks how artificial intelligence and data-centre growth are affecting Britain’s electricity demand, physical infrastructure and power networks. It brings together available evidence on electricity consumption, data-centre development and regional grid pressure, and is designed to distinguish verified information from estimates, forecasts and undisclosed data.

AI Electricity Demand Index

2024 baseline

RisingPeriod: 2024 (DESNZ, published 30 June 2026)

The electricity required to support AI and data-centre activity in the UK: consumption, share of national demand, growth rates and published forecasts.

Last updated: 15 Aug 2026

View the full sub-index

Data Centre Growth Index

Baseline in development

UnknownReporting period: not yet set

The physical build-out of UK data centre capacity: operational sites, construction, approvals and proposals, with capacity recorded only where it is disclosed.

Last updated: Data not yet available

View the full sub-index

Grid Pressure Index

Baseline in development

UnknownReporting period: not yet set

Whether regional electricity networks can accommodate new data-centre load, based on published connection, constraint and reinforcement evidence.

Last updated: Data not yet available

View the full sub-index

Latest available UK indicators

Each field shows its value, geographic coverage, reporting period, source, classification, confidence and the date it was last reviewed. Where the published evidence does not support a figure, the field reads “Insufficient evidence” — nothing is estimated to fill a gap.

Latest available UK AI energy indicators
Indicator	Value	Coverage	Reporting period	Source	Classification	Confidence	Last reviewed
Estimated annual data-centre electricity consumption	Insufficient evidence	—	—	—	Data not yet available	Not assessed	—
Data-centre share of UK electricity demand	Insufficient evidence	—	—	—	Data not yet available	Not assessed	—
Operational UK data centres	Insufficient evidence	—	—	—	Data not yet available	Not assessed	—
Data centres under construction	Insufficient evidence	—	—	—	Data not yet available	Not assessed	—
Approved or proposed developments	Insufficient evidence	—	—	—	Data not yet available	Not assessed	—
New capacity in megawatts	Insufficient evidence	—	—	—	Data not yet available	Not assessed	—
Average grid-connection waiting time	Insufficient evidence	—	—	—	Data not yet available	Not assessed	—
Regions currently experiencing high or severe pressure	Insufficient evidence	—	—	—	Data not yet available	Not assessed	—

Read the methodology and full source register

Indicators and sources

AI electricity demand

Data-centre electricity consumption, Great Britain

Measured in TWh per year

high

Electricity consumed from the public grid by data centres in Great Britain, as estimated by DESNZ.

Latest reading — 2024 (estimate)

4.5 TWh per year

+9.8% on 2023

DESNZ workbook Table_2, Great Britain row. Exact workbook value 4.4595375820 TWh. DESNZ reports this as 4.5 TWh.

Period	Value	Basis
2024	4.5 TWh per year	Estimate · high
2023	4.1 TWh per year	Estimate · high
2022	3.8 TWh per year	Estimate · high
2021	3.4 TWh per year	Estimate · high
2020	3.2 TWh per year	Estimate · high

Source: Department for Energy Security and Net Zero (DESNZ)(Official statistics)
Last updated: 15 Aug 2026
Update schedule: Annual
How it is collected: Manual entry

Method: DESNZ statistical estimate of electricity consumed from the public grid by data centres, published in the Energy Trends June 2026 special feature article and its accompanying workbook (Table_2, TWh).

Caveats: Covers data centres generally. It is not an estimate of electricity used specifically by artificial intelligence. Great Britain only, so Northern Ireland is excluded. Measures electricity taken from the public grid, so any on-site generation is excluded.

Data-centre share of Great Britain grid electricity consumption

Measured in Percentage

high

Data-centre electricity consumption as a share of all electricity consumed from the grid in Great Britain.

Latest reading — 2024 (estimate)

1.8 Percentage

+5.9% on 2023

DESNZ workbook Table_3, Great Britain row. Exact workbook value 1.789199%, reported by DESNZ as approximately 2 per cent of the 249.2 TWh consumed from the grid in Great Britain.

Period	Value	Basis
2024	1.8 Percentage	Estimate · high
2023	1.7 Percentage	Estimate · high
2022	1.5 Percentage	Estimate · high
2021	1.3 Percentage	Estimate · high
2020	1.2 Percentage	Estimate · high

Source: Department for Energy Security and Net Zero (DESNZ)(Official statistics)
Last updated: 15 Aug 2026
Update schedule: Annual
How it is collected: Manual entry

Method: Taken directly from the DESNZ workbook, Table_3 (share of electricity consumed from the grid).

Caveats: Covers data centres generally, not AI workloads. DESNZ rounds this to 2 per cent in the article text; the workbook value for 2024 is 1.79 per cent.

Change in GB data-centre electricity consumption, 2020 to 2024 (absolute)

Measured in TWh per year

high

Absolute increase in data-centre electricity consumption in Great Britain between 2020 and 2024.

Latest reading — 2020 to 2024 (estimate)

1.3 TWh per year

First published reading — no earlier period to compare against yet.

Increase of 1.3 TWh between 2020 and 2024, as reported by DESNZ.

Source: Department for Energy Security and Net Zero (DESNZ)(Official statistics)
Last updated: 15 Aug 2026
Update schedule: Annual
How it is collected: Manual entry

Method: Calculated from the DESNZ workbook, Table_2: 2024 value minus 2020 value.

Caveats: A change between two DESNZ statistical estimates, not a metered measurement. Not AI-specific.

Growth in GB data-centre electricity consumption, 2020 to 2024

Measured in Percentage change

high

Percentage increase in data-centre electricity consumption in Great Britain between 2020 and 2024.

Latest reading — 2020 to 2024 (estimate)

41 Percentage change

First published reading — no earlier period to compare against yet.

DESNZ workbook Table_2, percentage change column. Exact workbook value 41.1736%, reported by DESNZ as 41 per cent.

Source: Department for Energy Security and Net Zero (DESNZ)(Official statistics)
Last updated: 15 Aug 2026
Update schedule: Annual
How it is collected: Manual entry

Method: Published in the DESNZ workbook, Table_2, percentage change column (0.4118, reported by DESNZ as 41 per cent).

Caveats: A historical DESNZ estimate of change, not a real-time reading and not AI-specific.

Data-centre electricity consumption, London

Measured in TWh per year

high

Electricity consumed from the grid by data centres in London (Inner and Outer London combined).

Latest reading — 2024 (estimate)

1.7 TWh per year

First published reading — no earlier period to compare against yet.

Inner London and Outer London rows of the DESNZ workbook, Table_2, added together. DESNZ reports approximately 1.7 TWh for London.

Source: Department for Energy Security and Net Zero (DESNZ)(Official statistics)
Last updated: 15 Aug 2026
Update schedule: Annual
How it is collected: Manual entry

Method: Sum of the Inner London and Outer London rows in the DESNZ workbook, Table_2.

Caveats: Describes total data-centre consumption in the region, not the location of AI workloads.

Data-centre electricity consumption, South East

Measured in TWh per year

high

Electricity consumed from the grid by data centres in the South East of England.

Latest reading — 2024 (estimate)

1.8 TWh per year

First published reading — no earlier period to compare against yet.

DESNZ workbook Table_2, South East row. Exact workbook value 1.7853198722 TWh.

Source: Department for Energy Security and Net Zero (DESNZ)(Official statistics)
Last updated: 15 Aug 2026
Update schedule: Annual
How it is collected: Manual entry

Method: Taken directly from the DESNZ workbook, Table_2, South East row.

Caveats: Describes total data-centre consumption in the region, not the location of AI workloads.

London and South East combined share of GB data-centre consumption

Measured in Percentage

high

Share of all Great Britain data-centre electricity consumption accounted for by London and the South East together.

Latest reading — 2024 (estimate)

77 Percentage

First published reading — no earlier period to compare against yet.

London and the South East together accounted for 77 per cent of Great Britain data-centre electricity consumption in 2024.

Source: Department for Energy Security and Net Zero (DESNZ)(Official statistics)
Last updated: 15 Aug 2026
Update schedule: Annual
How it is collected: Manual entry

Method: Calculated from the DESNZ workbook, Table_2: (Inner London + Outer London + South East) divided by Great Britain.

Caveats: A concentration measure for data centres generally, not AI workloads.

Data-centre electricity consumption, Slough

Measured in TWh per year

high

Electricity consumed from the grid by data centres in the Slough local authority area.

Latest reading — 2024 (estimate)

1.3 TWh per year

First published reading — no earlier period to compare against yet.

DESNZ workbook Table_2, Slough row. Exact workbook value 1.2934434192 TWh.

Source: Department for Energy Security and Net Zero (DESNZ)(Official statistics)
Last updated: 15 Aug 2026
Update schedule: Annual
How it is collected: Manual entry

Method: Taken directly from the DESNZ workbook, Table_2, Slough row.

Caveats: A local-authority figure, shown separately from the regional and Great Britain totals. Not AI-specific.

Slough share of GB data-centre electricity consumption

Measured in Percentage

high

Share of all Great Britain data-centre electricity consumption accounted for by Slough.

Latest reading — 2024 (estimate)

29 Percentage

First published reading — no earlier period to compare against yet.

Slough accounted for about 29 per cent of Great Britain data-centre electricity consumption in 2024.

Source: Department for Energy Security and Net Zero (DESNZ)(Official statistics)
Last updated: 15 Aug 2026
Update schedule: Annual
How it is collected: Manual entry

Method: Calculated from the DESNZ workbook, Table_2: Slough divided by Great Britain.

Caveats: A concentration measure for data centres generally, not AI workloads.

AI-attributable electricity demand

Measured in TWh/year

low

Estimated annual electricity consumed by AI workloads running in UK data centres.

No published reading yet — this indicator is defined and scheduled, but the first value is still being compiled by hand.

Source: NESO Future Energy Scenarios(System / network operator)
Last updated: 15 Aug 2026
Update schedule: Annual
How it is collected: Manual entry

Method: Taken from published scenario and data centre demand figures, then narrowed to the AI-attributable share using the assumptions stated on each reading. Not a metered figure.

Caveats: No authoritative UK or Great Britain source reviewed to date separates electricity consumed by AI workloads from electricity consumed by other data-centre workloads.

Data-centre share of Slough grid electricity consumption

Measured in Percentage

high

Share of all electricity consumed from the grid in Slough that is consumed by data centres.

Latest reading — 2024 (estimate)

65 Percentage

First published reading — no earlier period to compare against yet.

DESNZ workbook Table_3, Slough row. Exact workbook value 65.1642%, reported by DESNZ as approximately 65 per cent.

Source: Department for Energy Security and Net Zero (DESNZ)(Official statistics)
Last updated: 15 Aug 2026
Update schedule: Annual
How it is collected: Manual entry

Method: Taken directly from the DESNZ workbook, Table_3, Slough row.

Caveats: A local grid measure for Slough only. It says nothing about the share elsewhere in Great Britain and is not AI-specific.

GB data-centre electricity consumption (2030 projection)

Measured in TWh per year

medium

Projected annual electricity consumption of all data centres in Great Britain in 2030 under NESO's Clean Power 2030 pathway.

Latest reading — 2030 (Clean Power 2030 pathway) (estimate)

22 TWh per year

First published reading — no earlier period to compare against yet.

NESO projects GB data-centre electricity demand rising to 22 TWh a year by 2030, around a fivefold increase on its baseline.

Source: National Energy System Operator (NESO)(Official statistics)
Last updated: 11 Aug 2026
Update schedule: When the source publishes
How it is collected: Manual entry

Method: Recorded directly from NESO's published pathway figure. No adjustment or re-modelling is applied.

Caveats: Scenario-dependent forecast derived from planning and grid connection registers, which typically overstate eventual build-out. Not AI-specific.

GB total annual electricity demand

Measured in TWh per year

medium

Total annual electricity demand across Great Britain, used as the denominator for share calculations.

Latest reading — 2023

263 TWh per year

First published reading — no earlier period to compare against yet.

NESO Clean Power 2030, Annex 1: Electricity demand and supply analysis, section 2.1.1.

Source: National Energy System Operator (NESO)(Official statistics)
Last updated: 12 Aug 2026
Update schedule: Annual
How it is collected: Manual entry

Method: NESO Clean Power 2030, Annex 1, section 2.1.1.

Caveats: Great Britain, not the United Kingdom.

Quarterly AI electricity-demand growth

Measured in Percentage change

indicative

Quarter-on-quarter change in electricity demand attributable to AI workloads in the UK.

No published reading yet — this indicator is defined and scheduled, but the first value is still being compiled by hand.

Source: —(Official statistics)
Last updated: 15 Aug 2026
Update schedule: Quarterly
How it is collected: Manual entry

Method: No published UK series separates AI workloads at quarterly frequency. The indicator is defined so that it can be populated if such a series is published.

Caveats: No authoritative UK or Great Britain source reviewed to date separates electricity consumed by AI workloads from electricity consumed by other data-centre workloads.

AI-specific peak-power requirement

Measured in MW

indicative

Peak electrical power drawn by AI workloads in the UK, as distinct from annual consumption.

No published reading yet — this indicator is defined and scheduled, but the first value is still being compiled by hand.

Source: —(Official statistics)
Last updated: 15 Aug 2026
Update schedule: When the source publishes
How it is collected: Manual entry

Method: No UK source publishes peak power separately for AI workloads. The indicator is defined so that it can be populated if such data is published.

Caveats: No authoritative UK or Great Britain source reviewed to date separates electricity consumed by AI workloads from electricity consumed by other data-centre workloads.

GB peak electricity demand

Measured in GW

medium

Peak electricity demand recorded across Great Britain.

Latest reading — 2023

58 GW

First published reading — no earlier period to compare against yet.

NESO Clean Power 2030, Annex 1, section 2.1.1.

Source: National Energy System Operator (NESO)(Official statistics)
Last updated: 12 Aug 2026
Update schedule: Annual
How it is collected: Manual entry

Method: NESO Clean Power 2030, Annex 1, section 2.1.1.

Caveats: A capacity measure in gigawatts; not comparable with annual consumption in TWh.

GB data-centre electricity demand, 2035

Measured in TWh per year

medium

FES 2025 pathway projections of Great Britain data-centre electricity demand in 2035.

Latest reading — By 2035 - FES 2025 pathways

30 to 41 TWh (Ten Year Forecast 33; Hydrogen Evolution 30; Electric Engagement 41)

First published reading — no earlier period to compare against yet.

NESO Grid Code Development Forum, 5 November 2025, slide 10.

Source: National Energy System Operator (NESO)(System / network operator)
Last updated: 12 Aug 2026
Update schedule: Annual
How it is collected: Manual entry

Method: Pathway values are reported individually; no average is calculated across pathways.

Caveats: Scenario projections, not predictions.

GB data-centre electricity demand, 2050

Measured in TWh per year

medium

Published long-range projections of Great Britain data-centre electricity demand in 2050.

Latest reading — By 2050 - Clean Power 2030 (November 2024)

62 TWh per year

NESO Clean Power 2030, Annex 1, section 2.1.4.

Period	Value	Basis
By 2050 - Clean Power 2030 (November 2024)	62 TWh per year	Measured · medium
By 2050 - FES 2025 pathways	51 to 71 TWh (Hydrogen Evolution 51; Electric Engagement 71; no Ten Year Forecast value)	Measured · medium

Source: National Energy System Operator (NESO)(System / network operator)
Last updated: 12 Aug 2026
Update schedule: Annual
How it is collected: Manual entry

Method: FES 2025 pathway values and the earlier Clean Power 2030 figure are reported separately.

Caveats: Scenario projections, not predictions.

GB total annual electricity demand, 2030

Measured in TWh per year

medium

NESO Clean Power 2030 projection of total Great Britain electricity demand in 2030.

Latest reading — 2030 (Clean Power 2030 pathway)

287 TWh per year

First published reading — no earlier period to compare against yet.

NESO Clean Power 2030, Annex 1, section 2.1.2 - demand growth of 11% to 287 TWh in 2030.

Source: National Energy System Operator (NESO)(Official statistics)
Last updated: 12 Aug 2026
Update schedule: Annual
How it is collected: Manual entry

Method: NESO Clean Power 2030, Annex 1, section 2.1.2.

Caveats: Forecast, not an observation.

GB data-centre electricity demand - NESO modelled estimate

Measured in TWh per year

medium

NESO current estimate of annual electricity demand from data centres connected in Great Britain.

Latest reading — NESO modelled estimate as at November 2025 (estimate)

7.6 TWh per year

First published reading — no earlier period to compare against yet.

NESO Grid Code Development Forum, 5 November 2025, slide 10, citing FES 2025.

Source: National Energy System Operator (NESO)(System / network operator)
Last updated: 12 Aug 2026
Update schedule: Annual
How it is collected: Manual entry

Method: Taken verbatim from the NESO Grid Code Development Forum pack, 5 November 2025, slide 10, sourced to FES 2025.

Caveats: A modelled estimate covering data centres generally, not AI workloads. Its methodology differs from the DESNZ historical series, so it is not the next comparable annual observation after DESNZ 2024 and no growth rate is calculated between the two.

GB data-centre electricity consumption (superseded NESO modelling input)

Measured in TWh per year

medium

Annual electricity consumption of all data centres in Great Britain, used as the present-day baseline in NESO's Clean Power 2030 analysis. This measures consumption, not capacity, and covers all data centres rather than AI workloads specifically.

No published reading yet — this indicator is defined and scheduled, but the first value is still being compiled by hand.

Source: National Energy System Operator (NESO)(Official statistics)
Last updated: 15 Aug 2026
Update schedule: When the source publishes
How it is collected: Manual entry

Method: Recorded directly from NESO's published modelling assumption. No adjustment, apportionment or estimation is applied by AI Energy Intelligence.

Caveats: A modelling input rather than a measured statistic. NESO did not state a precise base year. Great Britain, not AI-specific. Superseded as the principal baseline by the DESNZ 2024 official estimate and retained here as historical context only.

Data-centre share of GB electricity demand (superseded internal calculation)

Measured in Percentage

medium

Data-centre electricity demand expressed as a share of total Great Britain electricity demand.

No published reading yet — this indicator is defined and scheduled, but the first value is still being compiled by hand.

Source: National Energy System Operator (NESO)(Our own modelling)
Last updated: 15 Aug 2026
Update schedule: Annual
How it is collected: Manual entry

Method: Calculated by AI Energy Intelligence from two figures published in the same NESO document. No cross-source division is performed.

Caveats: Calculated by AI Energy Intelligence from NESO figures with mismatched reporting periods. Superseded by the official DESNZ 2024 share and retained only as history.

Data centre infrastructure

GB connected data-centre capacity

Measured in GW

medium

Electrical capacity of data-centre facilities already connected to the Great Britain electricity system.

Latest reading — As at November 2025 (estimate)

2.4 GW

First published reading — no earlier period to compare against yet.

NESO Grid Code Development Forum, 5 November 2025, slide 10.

Source: National Energy System Operator (NESO)(System / network operator)
Last updated: 12 Aug 2026
Update schedule: Annual
How it is collected: Manual entry

Method: Taken verbatim from the NESO Grid Code Development Forum pack, 5 November 2025, slide 10.

Caveats: Connected capacity, not IT load and not contracted future capacity. NESO does not state the capacity definition.

GB colocation data-centre capacity

Measured in GW

medium

Maximum rated IT load of colocation data centres in Great Britain, as estimated by DSIT. Excludes enterprise data centres, and measures IT load rather than facility load, grid-connection capacity or campus capacity.

Latest reading — Autumn 2024 (estimate)

1.6 GW

First published reading — no earlier period to compare against yet.

Parliamentary Question answer reproduced in the DSIT research and analysis page published 1 May 2025; the answer was first published 11 March 2025.

Source: Department for Science, Innovation and Technology (DSIT)(Official statistics)
Last updated: 11 Aug 2026
Update schedule: When the source publishes
How it is collected: Manual entry

Method: Recorded directly from the DSIT total. No adjustment or extrapolation to a UK-wide or all-facility figure is applied.

Caveats: Colocation only, so it understates the total estate. Modelled estimates, rounded. Great Britain only: Northern Ireland is excluded.

UK AI-capable data-centre capacity requirement (2030)

Measured in GW

medium

Forecast requirement for AI-capable data-centre capacity in the UK by 2030, as set out in the UK Compute Roadmap. This is a forecast requirement, not installed capacity.

Latest reading — By 2030 (UK Compute Roadmap forecast)

6 GW

First published reading — no earlier period to compare against yet.

The UK Compute Roadmap forecasts a need for at least 6 GW of AI-capable data-centre capacity by 2030.

Source: Department for Science, Innovation and Technology (DSIT)(Official statistics)
Last updated: 11 Aug 2026
Update schedule: When the source publishes
How it is collected: Manual entry

Method: Recorded directly from the published forecast. No comparison is drawn with recorded capacity figures, because the definitions and geographies differ.

Caveats: The roadmap does not define whether "AI-capable capacity" means IT load, facility load or grid-connection capacity. UK-wide, so it is not comparable with Great Britain colocation figures.

Total contracted connection demand in the GB queue

Measured in GW

medium

Volume of data centre and large-load projects awaiting a transmission or distribution connection.

Latest reading — As at November 2025

125 GW

First published reading — no earlier period to compare against yet.

NESO Grid Code Development Forum, 5 November 2025, slide 10, citing Energy Networks Association connections data.

Source: National Energy System Operator (NESO)(System / network operator)
Last updated: 12 Aug 2026
Update schedule: Quarterly
How it is collected: Manual entry

Method: Reported by NESO citing Energy Networks Association connections data.

Caveats: Total contracted connection demand across all technologies and sectors, not a data-centre-only figure.

Colocation data-centre IT capacity - London

Measured in MW

medium

Maximum rated IT load of colocation data centres in the London ITL1 region.

Latest reading — Autumn 2024 (estimate)

1,048 MW

First published reading — no earlier period to compare against yet.

DSIT, Estimate of Data Centre Capacity: Great Britain 2024 - regional table.

Source: Department for Science, Innovation and Technology (DSIT)(Official statistics)
Last updated: 12 Aug 2026
Update schedule: When the source publishes
How it is collected: Manual entry

Method: DSIT internal dataset of colocation data centres, estimated from rack densities and floor area, summed by ITL1 region.

Caveats: Colocation facilities only; enterprise data centres are excluded. ITL1 regions differ from the nine index regions - Slough and the Thames Valley falls within South East.

Colocation data-centre IT capacity - South East

Measured in MW

medium

Maximum rated IT load of colocation data centres in the South East ITL1 region.

Latest reading — Autumn 2024 (estimate)

128 MW

First published reading — no earlier period to compare against yet.

DSIT, Estimate of Data Centre Capacity: Great Britain 2024 - regional table.

Source: Department for Science, Innovation and Technology (DSIT)(Official statistics)
Last updated: 12 Aug 2026
Update schedule: When the source publishes
How it is collected: Manual entry

Method: DSIT internal dataset of colocation data centres, estimated from rack densities and floor area, summed by ITL1 region.

Caveats: Colocation facilities only; enterprise data centres are excluded. ITL1 regions differ from the nine index regions - Slough and the Thames Valley falls within South East.

Colocation data-centre IT capacity - Wales

Measured in MW

medium

Maximum rated IT load of colocation data centres in the Wales ITL1 region.

Latest reading — Autumn 2024 (estimate)

154 MW

First published reading — no earlier period to compare against yet.

DSIT, Estimate of Data Centre Capacity: Great Britain 2024 - regional table.

Source: Department for Science, Innovation and Technology (DSIT)(Official statistics)
Last updated: 12 Aug 2026
Update schedule: When the source publishes
How it is collected: Manual entry

Method: DSIT internal dataset of colocation data centres, estimated from rack densities and floor area, summed by ITL1 region.

Caveats: Colocation facilities only; enterprise data centres are excluded. ITL1 regions differ from the nine index regions - Slough and the Thames Valley falls within South East.

Colocation data-centre IT capacity - South West

Measured in MW

medium

Maximum rated IT load of colocation data centres in the South West ITL1 region.

Latest reading — Autumn 2024 (estimate)

53 MW

First published reading — no earlier period to compare against yet.

DSIT, Estimate of Data Centre Capacity: Great Britain 2024 - regional table.

Source: Department for Science, Innovation and Technology (DSIT)(Official statistics)
Last updated: 12 Aug 2026
Update schedule: When the source publishes
How it is collected: Manual entry

Method: DSIT internal dataset of colocation data centres, estimated from rack densities and floor area, summed by ITL1 region.

Caveats: Colocation facilities only; enterprise data centres are excluded. ITL1 regions differ from the nine index regions - Slough and the Thames Valley falls within South East.

Colocation data-centre IT capacity - North West

Measured in MW

medium

Maximum rated IT load of colocation data centres in the North West ITL1 region.

Latest reading — Autumn 2024 (estimate)

52 MW

First published reading — no earlier period to compare against yet.

DSIT, Estimate of Data Centre Capacity: Great Britain 2024 - regional table.

Source: Department for Science, Innovation and Technology (DSIT)(Official statistics)
Last updated: 12 Aug 2026
Update schedule: When the source publishes
How it is collected: Manual entry

Method: DSIT internal dataset of colocation data centres, estimated from rack densities and floor area, summed by ITL1 region.

Caveats: Colocation facilities only; enterprise data centres are excluded. ITL1 regions differ from the nine index regions - Slough and the Thames Valley falls within South East.

Colocation data-centre IT capacity - East

Measured in MW

medium

Maximum rated IT load of colocation data centres in the East ITL1 region.

Latest reading — Autumn 2024 (estimate)

44 MW

First published reading — no earlier period to compare against yet.

DSIT, Estimate of Data Centre Capacity: Great Britain 2024 - regional table.

Source: Department for Science, Innovation and Technology (DSIT)(Official statistics)
Last updated: 12 Aug 2026
Update schedule: When the source publishes
How it is collected: Manual entry

Method: DSIT internal dataset of colocation data centres, estimated from rack densities and floor area, summed by ITL1 region.

Caveats: Colocation facilities only; enterprise data centres are excluded. ITL1 regions differ from the nine index regions - Slough and the Thames Valley falls within South East.

Colocation data-centre IT capacity - Scotland

Measured in MW

medium

Maximum rated IT load of colocation data centres in the Scotland ITL1 region.

Latest reading — Autumn 2024 (estimate)

30 MW

First published reading — no earlier period to compare against yet.

DSIT, Estimate of Data Centre Capacity: Great Britain 2024 - regional table.

Source: Department for Science, Innovation and Technology (DSIT)(Official statistics)
Last updated: 12 Aug 2026
Update schedule: When the source publishes
How it is collected: Manual entry

Method: DSIT internal dataset of colocation data centres, estimated from rack densities and floor area, summed by ITL1 region.

Caveats: Colocation facilities only; enterprise data centres are excluded. ITL1 regions differ from the nine index regions - Slough and the Thames Valley falls within South East.

Colocation data-centre IT capacity - North East

Measured in MW

medium

Maximum rated IT load of colocation data centres in the North East ITL1 region.

Latest reading — Autumn 2024 (estimate)

17 MW

First published reading — no earlier period to compare against yet.

DSIT, Estimate of Data Centre Capacity: Great Britain 2024 - regional table.

Source: Department for Science, Innovation and Technology (DSIT)(Official statistics)
Last updated: 12 Aug 2026
Update schedule: When the source publishes
How it is collected: Manual entry

Method: DSIT internal dataset of colocation data centres, estimated from rack densities and floor area, summed by ITL1 region.

Caveats: Colocation facilities only; enterprise data centres are excluded. ITL1 regions differ from the nine index regions - Slough and the Thames Valley falls within South East.

Colocation data-centre IT capacity - Yorkshire and the Humber

Measured in MW

medium

Maximum rated IT load of colocation data centres in the Yorkshire and the Humber ITL1 region.

Latest reading — Autumn 2024 (estimate)

16 MW

First published reading — no earlier period to compare against yet.

DSIT, Estimate of Data Centre Capacity: Great Britain 2024 - regional table.

Source: Department for Science, Innovation and Technology (DSIT)(Official statistics)
Last updated: 12 Aug 2026
Update schedule: When the source publishes
How it is collected: Manual entry

Method: DSIT internal dataset of colocation data centres, estimated from rack densities and floor area, summed by ITL1 region.

Caveats: Colocation facilities only; enterprise data centres are excluded. ITL1 regions differ from the nine index regions - Slough and the Thames Valley falls within South East.

Colocation data-centre IT capacity - West Midlands

Measured in MW

medium

Maximum rated IT load of colocation data centres in the West Midlands ITL1 region.

Latest reading — Autumn 2024 (estimate)

15 MW

First published reading — no earlier period to compare against yet.

DSIT, Estimate of Data Centre Capacity: Great Britain 2024 - regional table.

Source: Department for Science, Innovation and Technology (DSIT)(Official statistics)
Last updated: 12 Aug 2026
Update schedule: When the source publishes
How it is collected: Manual entry

Method: DSIT internal dataset of colocation data centres, estimated from rack densities and floor area, summed by ITL1 region.

Caveats: Colocation facilities only; enterprise data centres are excluded. ITL1 regions differ from the nine index regions - Slough and the Thames Valley falls within South East.

Colocation data-centre IT capacity - East Midlands

Measured in MW

medium

Maximum rated IT load of colocation data centres in the East Midlands ITL1 region.

Latest reading — Autumn 2024 (estimate)

9 MW

First published reading — no earlier period to compare against yet.

DSIT, Estimate of Data Centre Capacity: Great Britain 2024 - regional table.

Source: Department for Science, Innovation and Technology (DSIT)(Official statistics)
Last updated: 12 Aug 2026
Update schedule: When the source publishes
How it is collected: Manual entry

Method: DSIT internal dataset of colocation data centres, estimated from rack densities and floor area, summed by ITL1 region.

Caveats: Colocation facilities only; enterprise data centres are excluded. ITL1 regions differ from the nine index regions - Slough and the Thames Valley falls within South East.

Announced UK data centre capacity

Measured in MW

low

IT capacity announced, consented or under construction across UK data centre projects.

No published reading yet — this indicator is defined and scheduled, but the first value is still being compiled by hand.

Source: Operator and developer announcements(Industry body)
Last updated: Not yet published
Update schedule: Quarterly
How it is collected: Manual entry

Method: Built by hand from named project announcements and planning records. Each project is counted once, at its most recently stated capacity.

Caveats: Announcements are marketing documents as much as engineering ones; announced capacity regularly differs from what is eventually energised.

Cost and market

UK wholesale electricity price

Measured in GBP/MWh

high

Reference wholesale electricity price used as the cost baseline across our calculators.

No published reading yet — this indicator is defined and scheduled, but the first value is still being compiled by hand.

Source: Ofgem data portal(Regulator)
Last updated: Not yet published
Update schedule: Monthly
How it is collected: Manual entry

Method: Period average taken from the published regulator dataset stated on each reading.

Caveats: Wholesale prices are volatile within a period; a period average will not match any single settlement period.

Efficiency and intensity

Data centre energy efficiency (PUE)

Measured in ratio

low

Reported power usage effectiveness for UK data centre estates — how much total power is drawn per unit of IT power.

No published reading yet — this indicator is defined and scheduled, but the first value is still being compiled by hand.

Source: IEA — Energy and AI(Industry body)
Last updated: Not yet published
Update schedule: Annual
How it is collected: Manual entry

Method: Averaged from published operator and international agency reporting, with the sample size recorded on each reading.

Caveats: PUE is self-reported, defined inconsistently, and best-performing operators are the most likely to publish.

Public and business adoption

UK business AI adoption

Measured in % of businesses

high

Share of UK businesses reporting that they use artificial intelligence technologies.

No published reading yet — this indicator is defined and scheduled, but the first value is still being compiled by hand.

Source: ONS Business Insights and Conditions Survey(Official statistics)
Last updated: Not yet published
Update schedule: Monthly
How it is collected: Manual entry

Method: Taken directly from the published survey release for the stated period.

Caveats: Survey-based and self-reported; respondents define "using AI" differently and the question wording changes between waves.

Methodology

How the index is built

Every indicator is defined first: what it measures, in what unit, from which named source, on what schedule, and which direction counts as more pressure on the system.
A reading is entered for a stated period. Readings are marked as measured or estimated, and carry the source used for that specific figure.
Readings are normalised to a comparable score before being weighted into an edition. Weights are published on each indicator card above.
An edition is a dated release with its own methodology version. Older editions are kept as published, not silently rewritten.
Any change to a published figure is logged in the revision history below, with the old value, the new value and the reason.

Confidence levels

High: Published by an official source, a regulator or a system operator, directly measuring the thing we describe.
Medium: Credible published figure that needs interpretation — for example a national statistic used as a proxy, or a figure covering a slightly different scope.
Low: Derived from partial reporting, industry disclosure or a small sample. Directionally useful, not quotable as fact.
Indicative: Our own modelled estimate under stated assumptions. Treat as a scenario, not a measurement.

Full site-wide approach: research methodology and editorial standards.

Update schedule

Each indicator follows the cadence of its underlying source rather than a fixed publishing calendar. Where a source slips, the indicator keeps its last published reading and the delay is recorded rather than back-filled with an estimate.

Indicator	Frequency	Next review	Collection	Confidence
GB data-centre consumption (DESNZ)	Annual	—	Manual entry	medium
GB share (DESNZ)	Annual	—	Manual entry	medium
Change 2020-2024 (TWh)	Annual	—	Manual entry	medium
Growth 2020-2024	Annual	—	Manual entry	medium
London	Annual	—	Manual entry	medium
South East	Annual	—	Manual entry	medium
London and South East share	Annual	—	Manual entry	medium
Slough	Annual	—	Manual entry	medium
Slough share of GB	Annual	—	Manual entry	medium
AI demand	Annual	—	Manual entry	low
Share of Slough grid	Annual	—	Manual entry	medium
GB consumption 2030	When the source publishes	—	Manual entry	medium
Connected DC capacity	Annual	—	Manual entry	medium
GB total demand	Annual	—	Manual entry	medium
Quarterly AI demand growth	Quarterly	—	Manual entry	indicative
AI peak power	When the source publishes	—	Manual entry	indicative
GB peak demand	Annual	—	Manual entry	medium
GB colocation capacity	When the source publishes	—	Manual entry	medium
AI-capable capacity need 2030	When the source publishes	—	Manual entry	medium
DC demand 2035	Annual	—	Manual entry	medium
DC demand 2050	Annual	—	Manual entry	medium
GB total demand 2030	Annual	—	Manual entry	medium
Connection queue	Quarterly	—	Manual entry	medium
London	When the source publishes	—	Manual entry	medium
DC demand (FES 2025)	Annual	—	Manual entry	medium
South East	When the source publishes	—	Manual entry	medium
Wales	When the source publishes	—	Manual entry	medium
South West	When the source publishes	—	Manual entry	medium
North West	When the source publishes	—	Manual entry	medium
East	When the source publishes	—	Manual entry	medium
Scotland	When the source publishes	—	Manual entry	medium
North East	When the source publishes	—	Manual entry	medium
Yorkshire and the Humber	When the source publishes	—	Manual entry	medium
West Midlands	When the source publishes	—	Manual entry	medium
East Midlands	When the source publishes	—	Manual entry	medium
Announced capacity	Quarterly	—	Manual entry	low
GB data-centre consumption	When the source publishes	—	Manual entry	medium
DC share of demand	Annual	—	Manual entry	medium
Wholesale price	Monthly	—	Manual entry	high
Efficiency (PUE)	Annual	—	Manual entry	low
Business adoption	Monthly	—	Manual entry	high

Revision history

Every change to a published indicator, reading, edition or to the methodology itself is recorded here. Corrections are also handled under our corrections policy.

createdmethodology·10 Aug 2026· methodology v1.0
UK AI Energy Index framework v1.0 published — indicator definitions, sources, update schedules and confidence levels.
Initial release of the index structure. Readings are compiled manually and published indicator by indicator as each source is verified.

Cite or share this index

AI Energy Intelligence, UK AI Energy Index, accessed 27 August 2026. https://aienergyintelligence.co.uk/uk-ai-energy-index

Share on X Share on LinkedIn

Get UK AI Energy Index updates

We email a short note whenever an indicator is updated or a new edition is published, including what changed and why. No daily noise.

Go deeper with the reports

The index tracks the numbers. Our reports set out what they mean for UK grid capacity, data centre build-out and business electricity costs — with the same sourcing standards.

Browse reports & research →How we source data

Continue your research

Reports & research →UK AI Growth Zones →UK data centres →AI energy calculators →Latest AI energy news →