AI Grid Impact Forecast Tool
A UK-focused, scenario-based forecast model that estimates how AI adoption, data centre growth and AI infrastructure expansion could affect electricity demand, grid pressure, carbon emissions and infrastructure needs over time. Educational and indicative — not an official grid forecast.
Important: These tools provide educational estimates only. Actual AI electricity use varies by model, data centre, cooling system, hardware, location and workload. Do not use these results as official energy, planning, investment or engineering advice.
Forecast assumptions
AI electricity demand over time (TWh/yr)
Annual electricity cost over time (£m/yr)
Cumulative AI demand over time (TWh)
Scenario comparison — demand by 2035 (TWh/yr)
What final-year demand is equivalent to
Scenario insights
- Under this scenario, AI-related electricity demand could reach approximately 121.5 TWh per year by 2035.
- The model crosses the selected grid stress threshold of 20 TWh in 2029.
- At this level, AI demand may require additional grid reinforcement, new generation capacity or dedicated energy infrastructure.
- Improving average PUE to 1.2 (Excellent) could reduce final-year demand by approximately 13.7 TWh.
- This scenario suggests AI infrastructure could become a major planning factor for the UK electricity grid, with a grid pressure rating of Severe.
What is AI grid impact?
AI grid impact describes how artificial intelligence increases electricity demand and places pressure on the power network. AI affects demand through data centres running dense clusters of GPU and accelerator hardware, the server loads they sustain around the clock, the cooling systems that remove their heat, and the supporting infrastructure needed to keep them online.
Why data centre growth matters
Large AI data centres can require substantial grid connections — often hundreds of megawatts each. Securing these connections can take years and may compete with other local demand such as housing, industry and electrification of transport and heating. Where and how quickly capacity is added has a direct influence on local infrastructure and the wider grid.
What is PUE?
Power Usage Effectiveness (PUE) compares the total energy a facility uses against the energy delivered to its IT equipment. A PUE of 1.0 would be perfectly efficient; 1.2 is excellent, 1.4 is good, 1.6 is average and 2.0 is poor. The higher the PUE, the more energy is spent on cooling and overheads rather than computing — so reducing PUE directly lowers total demand, cost and carbon.
Why forecasting is uncertain
AI electricity demand is genuinely hard to forecast. It depends on model efficiency, hardware improvements, AI adoption levels, government policy, energy pricing, where data centres are located and the availability of grid connections. Small changes in any of these can produce very different outcomes, which is why this tool presents scenarios rather than a single prediction.
Why this matters for the UK
AI infrastructure links directly to several UK priorities: AI Growth Zones, National Grid upgrades, electricity generation and renewables, planning and local infrastructure, and business energy costs. Understanding indicative demand trajectories helps businesses, local authorities and analysts ask better questions about grid readiness and energy strategy.
Limitations
- This is an educational scenario model, not a precise engineering tool.
- It does not use live grid connection data.
- It is not an official National Grid or government forecast.
- It should not be used as engineering, investment, planning or legal advice. Always seek qualified professional guidance for real decisions.
How this forecast model works
For each year, demand grows from the previous year by the annual growth rate, plus additional data centre demand calculated as MW added × utilisation × PUE × 8,760 hours, converted to TWh. Annual cost uses total demand in MWh × electricity price, and annual carbon uses total demand in MWh × the grid carbon factor. Cumulative demand sums total AI-related demand across the period, and the grid pressure score compares final-year demand against your grid stress threshold.
Related Tools & Resources
FAQ
Explore more AI energy tools
Estimate electricity, water, cost and grid impact across AI Energy Intelligence UK's suite of UK AI energy intelligence tools.
View all AI Energy Tools