Tag: electric insights

Richard Gwilliam, Future of Drax Power Station Director: “The UK needs to think differently about its electrical infrastructure if it wants to be a leader in AI”

20 February 2026

Opinion

UK electricity demand is changing after years of stability

Drax’s Electric Insights recently published its quarterly report into the state of the electricity market. Amongst the plethora of useful information one metric stood out: electricity demand in the UK is starting to surge.

The report stated that demand in Q4 of last year reached 72.3 TWh over the quarter, up nearly 4% year-on-year. That might not sound much but, Covid rebound aside, that increase represents the fastest uptick in electricity demand since 2011.

From the mid-2000s until the early 2020s, GB’s total electricity demand fell steadily driven by gains in energy efficiency, deindustrialisation and an increase in behind the meter generation like rooftop solar, but it appears that trend is coming to an end.

The electrification of the UK’s economy has long been predicted and the reasons are of no surprise. Electric vehicles reached roughly 27% of new car registrations in 2025 and as the EV fleet grows, charging adds steadily to residential and overall electricity consumption.

Heat pumps too are scaling up – although still a small proportion of heating overall – the trend towards the electrification of heating is continuing and beginning to influence winter power demand profiles.

Finally, whilst some energy-intensive industry demand has grown, partly due to policy support and investment, the rapid increase in new data centres, particularly those serving AI and cloud computing, are emerging as rapidly growing demand centres, adding material loads on the grid.

Data Centres are likely to have a profound effect on the UK grid

Whilst demand for EVs and heat pumps are set to grow, it’s likely the demand for compute capacity will put the most pressure on the grid.

Such is the anticipated change that data centres are arguably no longer just customers of the grid, their scale means they are system actors. They can influence investment signals, dispatch patterns and ultimately the carbon intensity of the electricity system. That means we need to plan for them as part of the system, not bolt them on afterwards.

The processes that underpin new grid connections to meet demand were not designed for growth at the pace or scale that AI is moving. We are already seeing this emerge with:

  • long lead times for new grid capacity
  • constraints around transmission and distribution access
  • rising connection costs

This is important to UK growth, productivity and global competitiveness

The Government identifies AI as a key driver of future productivity growth across the economy: while the UK AI sector currently contributes around £12 billion in GVA, government-cited modelling suggests that widespread AI adoption could raise productivity, adding tens of billions of pounds a year to economic output. Delivering this growth is contingent on access to domestic compute capacity, with the UK Compute Roadmap indicating a need for a substantial expansion in AI-capable data centre capacity by 2030.

On the face of it, the ongoing electrification of our economy and slow adaptation of the grid appears to be a real stumbling block to meeting our AI ambition. But there are options that could see us simply adapt existing infrastructure and re-use what we already have in a more efficient way.

Existing energy infrastructure is a strategic advantage

One of the UK’s greatest, and most under-used asset is the infrastructure we already have.  Sites like Drax Power Station were built to handle very large amounts of power safely and reliably. They already have:

  • substantial grid connections;
  • transformers and electrical infrastructure sized for industrial-scale operations;
  • established sites with an energy and industrial planning context; and
  • experienced workforces and supply chains.

Repurposing and reusing this infrastructure is the fastest and most credible way to support large new electricity demand without waiting years for the system to catch up.

Re-use and re-purposing will keep the UK in the race for AI

At the end of last year, we unveiled proposals to submit a planning application for an initial <100MW data-centre development at Drax Power Station – an important step towards boosting AI capacity at the site. The project looks to re-use existing infrastructure and capacity, creating a practical and expedient way of meeting demand without waiting a decade for network reinforcement.

Not only that, but it could help pave the way for a >1GW data centre facility on the site before the end of the decade.

The Government is responding but more pace is needed

In December, the UK Government published the terms of reference for a review of AI deployment in the electricity networks. It is good to see that as part of that the Government will attempt to grapple with regulatory constraints and begin to consider AI deployment at a spatial level.

This is an acknowledgement that we need to challenge convention and think differently about this issue, unpick regulations that were devised for a different time and rethink how we maximise efficiency from existing assets. Time is of the essence; left unchecked we’ll likely end up falling behind in the global AI race and be left with an energy system with baked in inefficiencies.

The AI revolution is an energy story

It’s becoming increasingly clear that we’re now at the end of a decades long trend of decline and stability in electricity demand and arguably our systems are not ready to cope. The countries that succeed will not be those that build the most servers the fastest, but those that integrate compute into their power systems.

The UK already has much of the infrastructure it needs. The challenge now is to use it differently. At Drax Power Station, we are focused on how an existing site – which has kept the lights on around the UK for 50 years – can evolve to meet new demands supporting long term economic growth and prosperity for the UK.

Read the full Q4 2025 Electric Insights report here.

Wind droughts show the need for low-carbon flexible generation

5 February 2025

Opinion

By Dr Iain Staffell, Imperial College London 

As our energy mix changes and a different weather challenge has been taking up the headlines, latest analysis from Electric Insights has revealed that the need for reliable low-carbon generation when the wind doesn’t blow and the sun doesn’t shine is becoming more important. Dr Iain Staffell took a look at the data.   

“Dunkelflaute” must surely be an early contender for the 2025 Oxford Dictionary word of the year.  A German word meaning “dark doldrums”, it is used in the energy world to describe a dark, cold, calm spell of weather during which very little energy can be generated with wind or solar power.

In December and January, Britain has faced two spells of so-called Dunkelflaute.  The first, hitting around the 12 December, saw wind – the largest source of energy in the UK last year overall – drop to 6% of total supply.  In response, gas power stations ramped up to their highest output ever recorded, supplying more than 73% of Britain’s electricity and sending power prices soaring.  Wind output dropped suddenly again in the New Year causing prices to hit £2,900/MWh (40 times their average) on 8 January.

This winter has again demonstrated some of the challenges we must address in reaching a clean power system by 2030.  The combination of a long cold snap and low wind speeds left Britain’s power system relying heavily on natural gas and imports, drawing down the nation’s gas storage to ‘concerningly low’ levels, and coming close to generation falling short of peak demand.  Options for low-carbon flexibility are urgently needed – both investing in new technologies and maintaining existing sources – as electricity supply and demand become more dependent on the weather.

Daily average electricity mix in Britain during mid-December, highlighting the Dunkelflaute period, and the difference between output from dispatchable technologies which we control, and those that are driven by the weather or foreign power markets.

Gas was not the only technology to help during the shortfall.  Biomass and hydro plants increased their output by 40% and 60% on the peak day (12 December) compared to the weekends before and after.  While this helped meet the shortfall of wind, the impact was muted as Britain has relatively little capacity of either technology.  In previous years, coal power stations would have also helped to meet demand, but the last one closed in September.  Pumped hydro and batteries helped meet the evening peak on the 12th, but these only supply power for a few hours, and so cannot help with multi-day shortages.

Interconnection with neighbouring countries also provides flexibility, but on the 12th when we most needed them, imports from abroad fell by half relative to the surrounding days.  Britain’s neighbours were suffering from the same wind drought, as weather systems are often the size of continents.  More power could have flowed into Britain, but only if our prices rose high enough.  This exposes a key problem with relying on interconnection to solve capacity shortages, which leaves countries competing for limited supply of power at the same time.

Altogether, this leaves gas as the only large-scale source of flexibility in the country.  This is a risky proposition on three fronts: affordability, energy security, and our climate goals.

The cost of our gas dependence: We are still reeling from the gas price crisis.  Gas is very much the ‘crutch’ of the grid, and British electricity is more strongly swayed by gas prices than in any other European country, as we have so few alternatives for flexible generation (no coal, limited hydro and biomass, and less storage than neighbouring countries).  Gas sets the electricity price in 98% of hours, despite meeting only a third of electricity demand. That means Britain’s electricity prices track almost perfectly with gas prices, leaving consumers particularly vulnerable to price shocks, as seen during the recent gas price crisis.

The change in electricity and natural gas prices on Britain’s wholesale markets over the last decade, indexed to the 2010–19 average.  Gas prices increased by over 50% between February and December last year, dragging electricity prices up with them.

Energy security at risk: Relying so heavily on a single technology in times of system stress is leaving all our eggs in one basket.  Capacity was tight on 12 December and 8 January, causing NESO to issue rare Capacity Market Notices, a ‘blackout prevention system’ used to encourage generators to prepare extra capacity just in case.   Britain’s last coal plant has retired, all bar one nuclear plant is coming towards their end of life, and it is unclear if biomass will continue operating beyond 2027.  This all comes just as peak electricity demand is expected to grow from electric vehicles, heat pumps, AI, and data centres.  Unless more capacity is built or existing capacity has its lifetime extended, Capacity Market Notices will be increasingly likely in future.

The carbon challenge: Gas is the most polluting fuel remaining on the grid.  In just five years, government aim to run a clean power system, meaning just 5% of electricity produced from fossil fuels, down from over 25% today.  These plans include retaining almost all the current gas capacity to cover rare but intense periods of low renewable output.  Put together, this means gas plants will see fewer operating hours in the future, just as coal plants did over the last decade.  Either they will need to charge more for their output to cover costs, or the system needs to move more towards paying for availability than for output (e.g. capacity payments).

Phasing out gas will largely be achieved by scaling up wind and solar power, but that further intensifies the challenges posed by weather variability.  Both the CCC and NESO recognise that a balanced approach is needed, using all the tools at our disposal – flexible low-carbon generation, long-duration energy storage, interconnectors and a continued (but increasingly limited) role for gas.  Looking ahead, policy frameworks envisage the arrival of more low-carbon dispatchable power from 2030 onward.  This includes power stations equipped with carbon capture and storage (CCS), hydrogen, and long-duration storage.  All of these play little or no role in today’s power system, so the task now is to define a clear strategy for scaling and deploying these resources at pace, while avoiding cost escalation to consumers due to all the new investments.  By planning for Britain’s future energy needs and taking strategic action now, government, industry and investors can break free from paying for volatile gas expensive imports, and seize the opportunity of clean, stable, and lower cost electricity.

Read the full article here or in the Q4 2024 Electric Insights report, coming soon.

This article was written by Dr Iain Staffell, Senior Lecturer at Imperial College London, as part of the Electric Insights project. Drax does not guarantee the accuracy, reliability or completeness of this content.