We are in the midst of a technological revolution. From asking ChatGPT to help with a simple question to requesting your favourite song from Alexa, we live in an AI world.
The backbone of the new digital economy is data centres: secure buildings packed with servers that store data, run applications, and keep modern digital services running. These data centres depend on a reliable, secure and consistent source of energy, but developers trying to connect to the electricity grid are regularly hitting a brick wall.
As the Department for Energy Security and Net Zero said in their recent consultation on ‘Accelerating electricity network connections for strategic demand’[1], projects are being held back by long lead times to build new (electricity) network infrastructure and a connections process that is overwhelmed by unprecedented growth in demand for new connections. The queue for demand connections to the electricity transmission network jumped by a staggering 460% in the first 6 months of June 2025[2], mirroring the challenges that have been seen with the electricity generation connections queue as well.
With so much pressure on the electricity grid, the reality is that getting a data centre – or indeed anything major source of demand – connected to electricity could take up to 15 years. This simply will not be palatable for data centre developers and is a bottleneck which threatens to undermine economic growth and hold back the UK in the global AI race.
That’s why data centre developers are increasingly turning to the only alternative they have for a quick connection that can supply them with a significant amount of energy – the gas network.
Figures recently released by FEN reveal how gas network companies considered a total of 113 enquiries regarding data centre gas connections in 2024 and 2025[3]. Demand has risen exponentially over this 24-month period, with the number of enquiries in 2025 around three times higher than 2024.
Gas networks are in the process of considering 46 applications, while seven have accepted their quote for connection. These numbers may sound small, but we’re talking about a lot of energy. Those seven applications equal a combined capacity of 15.4 terawatt-hours: that’s enough energy to heat 1.3 million British homes or the equivalent of Denmark’s entire annual gas demand.
With gas connection times between six and 24 months, compared to up to 15 years for electricity, is it any wonder that data centres are turning to on-site gas generation to produce power to run and to cool these facilities? After all, securing a gas connection is generally quicker than constructing a data centre, meaning that connecting to the network shouldn’t result in delays.
Britain is no international outlier here: rising gas demand in the United States has been linked to the rise of data centres, which already account for around 4–5% of U.S. electricity demand. They are expected to add up to 2 - 4 billion cubic feet per day to gas consumption by 2030, while Europe’s first gas-connected data centre has been commissioned in Ireland.
There’s no question that the future of our economic success as a country depends on finding more effective ways of powering data centres quickly and reliably. How to do so while simultaneously decarbonising our energy system is a profound challenge, and one that senior politicians are grappling with. The Environmental Audit Committee (EAC) has recently closed a call for evidence on the sustainability of data centres[4], while DESNZ has further consultations planned.
The good news is that this is a challenge that the gas networks can help to meet – and it doesn’t even need to be the trade off between economic growth and decarbonisation which many people assume it would be.
In our response to the EAC’s inquiry, we have provided a series of archetypal examples to demonstrate this point. Each one builds on the other to show more and more sophisticated ways in which the gas network can be used to quickly connect up data centres (with all of the economic opportunity that comes with them) while also keeping us on track with or even advancing our decarbonisation ambitions as a nation.
These archetypes are:
Archetype 1: Immediate deployment via natural gas. This archetype assumes a baseline model where data centres are unable to connect to the electricity network so instead connect into the gas network and use natural gas to fuel on-site generation.
While these sites would be fuelled by natural gas, the alternative is not renewable electricity given long grid connection delays. Instead, developers are likely to build overseas, often using gas anyway, as seen in the United States, or electricity systems using coal. The result if the UK were to block connections to the gas network is no climate benefit, but a loss of UK jobs and economic value.
Archetype 2: Transition to green gases. This archetype assumes the same as Archetype 1, but with the gas needs for the site met by low carbon, green gases rather than fossil fuels, making the site fully compatible with Clean Power and Net Zero ambitions. In the longer term, if carbon capture and storage technology was fitted to larger gas-fired generation on the major data centre sites then this could even make the sites carbon negative in their energy use.
Archetype 3: Hybrid flexibility. This archetype assumes the data centre is able to get an electricity connection, but it isn’t sufficient to meet the entire energy needs of the site – hence also securing a gas connection. The developers can switch between the different vectors to support the energy system and as a revenue driver.
Archetype 4: Green gas data centres as “energy hubs”. This archetype assumes the same as Archetype 3, but with the capacity of the electricity connection improving over time until it can meet the entire energy needs of the site. This then leaves the gas-fired generation on site as an incredibly useful asset for the wider energy system, providing the flexibility that a clean power system will need to manage periods of low renewables output. There is even the opportunity to integrate a third vector – heat – where waste heat from both the gas-fired generation and the data centre could be used to cool the data centre (reducing energy demand) and/or used to connect into a local heat network to heat homes and businesses.
It is clear that the UK is unable to deliver data centre growth without gas; electricity alone cannot meet the scale or speed required. The gas networks stand ready to propel the UK forward in the global AI race, but the right support from Government is required.
We’ve set out in our response to the EAC that to enable sustainable and scalable data centre growth, ministers should formally recognise gas as a strategic enabler alongside electricity and adopt a whole-system planning approach across DESNZ, Ofgem and NESO. Regulatory and planning barriers to gas connections must be removed, while long-term policy certainty is needed to support biomethane and hydrogen. Carbon accounting frameworks, including the UK ETS, should properly reflect the benefits of low carbon gases, and greater co-location of data centres with energy and heat infrastructure should be encouraged. Modelling by the CCC and NESO must also be updated to reflect rising demand.
We must embrace a whole-systems approach that allows rapid deployment now while enabling a transition to low carbon gases over time. Only then will we ensure that the UK can reap the economic benefits of AI while decarbonising our energy system in parallel.
[1] More details can be found here: https://www.gov.uk/government/consultations/accelerating-electricity-network-connections-for-strategic-demand/accelerating-electricity-network-connections-for-strategic-demand-accessible-webpage
[3] Data centre gas connections enquiries surge amid electricity grid queues