Why Renewables Are No Good for AI – and No Good for Britain

1. Introduction: The Collision of AI and Net Zero

Artificial Intelligence (AI) is being hailed as the new industrial revolution. From revolutionising healthcare to transforming defence, logistics, and manufacturing, AI promises a future of rapid change and national competitiveness. But all this depends on one factor above all else: power.

Data centres—the factories of the digital age—are the engines of AI. They demand vast, constant, high-quality electricity to keep servers humming 24 hours a day, 365 days a year. A hyperscale data centre can consume as much electricity as a medium-sized British city.¹

And yet, Britain’s energy policy, dominated by the Net Zero agenda, is committed to building an energy system based on renewables that cannot deliver constant power. Wind and solar may tick boxes for carbon targets, but they cannot run AI—and if they cannot run AI, they cannot run Britain.

Instead of facing this truth, the government, DESNZ (Department for Energy Security and Net Zero), and the Climate Change Committee (CCC) cling to ideology, enforcing carbon budgets and building intermittent renewables while ignoring physics. The result is an energy strategy that is unfit for purpose—for AI, for industry, and for ordinary households.

2. Why Data Centres Cannot Rely on Renewables

2.1. AI’s hunger for power

Unlike homes or even factories, data centres cannot scale back operations when power is scarce. AI servers run at full capacity around the clock, with zero tolerance for outages. Even a momentary dip in supply can cause catastrophic data corruption or lost computations.²

2.2. Intermittency by design

Solar: UK solar output peaks in summer midday—useless in December when demand peaks on dark winter evenings.

Wind: Britain’s wind turbines can generate 60% of national demand one day and less than 5% the next. “Dunkelflaute” (dark doldrums) can last days or weeks, during which renewables produce almost nothing.³

Capacity factors: On average, UK wind farms deliver just 30–35% of their rated output, solar 10–12%. Nuclear, by contrast, runs at 70–90%, gas plants at 80–90%.⁴


For AI, which requires a constant power supply at 100% utilisation, this variability is unacceptable.

3. The Storage Illusion

Renewables advocates argue that storage—batteries or hydrogen—can cover intermittency. But reality tells a different story.

Batteries: Most UK grid batteries provide 1–2 hours of backup at full discharge. Even the longest (Hornsdale in Australia, Dinorwig pumped hydro in Wales) provide only hours, not days.⁵ AI centres require 8,760 hours per year.

Hydrogen: Round-trip efficiency losses mean up to 70% of input energy is wasted. Costs remain astronomical.⁶

Scale problem: To cover just one week of UK winter demand with lithium-ion batteries would require more lithium than is mined globally in a year.⁷


Storage is not a solution at scale. It is an expensive mirage that papers over the unreliability of renewables without ever eliminating the need for firm generation.

4. If Renewables Are No Good for AI, They’re No Good for Britain

This is the critical link. The government knows AI cannot run on renewables, which is why five major data centre developers have already approached National Gas about connecting directly to the UK gas pipeline to build on-site generation.⁸

But if AI—the future of the economy—cannot run on renewables, why should we believe that renewables are suitable for:

Hospitals, where ventilators, dialysis machines, and surgical theatres must never lose power.

Transport, where railways, airports, and logistics depend on steady electricity.

Homes, especially in winter, when solar is weakest and demand is highest.

Industry, where production lines cannot stop and start with the wind.


The truth is stark: what doesn’t work for AI doesn’t work for ordinary people either.

5. Westminster’s Pipedreams: The CCC and DESNZ

Instead of confronting these realities, Westminster doubles down.

The Climate Change Act 2008 forces ministers to comply with legally binding “carbon budgets” written by the Climate Change Committee (CCC), a quango of unelected officials and academics.

The CCC assumes that massive battery systems, hydrogen technology, and international interconnectors will be built—despite no evidence they can scale affordably or in time.

DESNZ (Department for Energy Security and Net Zero) is tasked not with keeping the lights on, but with delivering Net Zero. The clue is in the name: security comes second, Net Zero first.


This is not energy policy rooted in engineering—it is policy dictated by ideology.

6. The Consequences of Folly

6.1. Higher bills

Households are paying for two parallel systems: renewables plus fossil backup. Standing charges alone have risen by over 500% since 2008.⁹

6.2. Grid instability

In January 2025, Britain narrowly avoided a blackout when high demand met low wind output and limited backup availability.¹⁰

6.3. Investment flight

Data centre operators are already warning that if Britain cannot provide reliable power, they will move investment to the US or Gulf states, where nuclear and gas provide certainty.

6.4. Wasted countryside

Solar farms and BESS projects sprawl across prime farmland, destroying habitats, while still needing fossil backup. The carbon “savings” are largely illusory.

7. A Reliable Path Forward

Britain must abandon the fantasy that renewables can carry the load of an industrial economy. A rational, engineering-led path includes:

Gas as a bridge: Secure domestic supply to guarantee stability during the transition.

Nuclear revival: Deploy Rolls-Royce SMRs across the UK for 24/7 carbon-free baseload.

Rooftop solar & local generation: Empower homes and businesses with technologies like Power Roll—without paving over farmland.

Grid reform: Prioritise synchronous stability and firm generation instead of intermittent capacity.


This approach recognises reality: Britain needs firm, constant, reliable power—not pipedreams.

8. Conclusion: From Ideology to Engineering

If renewables are no good for AI, they are no good for Britain. The same stability, certainty, and reliability demanded by the next industrial revolution is demanded by hospitals, households, and industry.

Yet Westminster, shackled by the CCC and DESNZ, continues to gamble on intermittent renewables, blind to the physical limits of the grid. The price is paid by households through higher bills, by businesses through instability, and by the country through lost competitiveness.

It is time to end the folly. Britain needs a power system rooted in reality, not ideology. Firm, constant, reliable power—gas, nuclear, and smart local generation—not fantasies from quangos.

The future of AI, and the future of Britain, depends on it.

References & Footnotes

1. Financial Times, “Data centre developers explore linking up to UK gas pipelines”, Aug 2025.


2. Uptime Institute, Data Centre Resilience Report 2024.


3. National Grid ESO, Future Energy Scenarios 2024.


4. BEIS Energy Statistics 2023.


5. Guardian, “Dinorwig: the mountain that keeps Britain’s lights on”, May 2025.


6. Wired, “The False Hope of Hydrogen for Power Storage”, 2023.


7. Nature Energy, “The material limits of lithium-ion battery scaling”, 2022.


8. Financial Times, Aug 2025, op. cit.


9. Ofgem data on electricity standing charges, 2008–2024.


10. National Grid ESO, Winter Outlook Report 2025.