All in on AI, but what about sustainability?

EXECUTIVE SUMMARY:

This article discusses the growing tensions between the rapid development of Generative AI (GenAI) and the data centre industry’s decarbonisation goals. Once being recognised as a leader in energy efficiency and proactive climate change, the sector is now facing insatiable electricity demand, with intensive AI workloads projected to more than double global consumption within the decade. The article highlights mounting pressure on national grids, especially in the UK, where high costs and limited capacity raise concerns about future competitiveness in an already competitive field. To regroup growth with sustainability, the article states that near-site renewable solutions, private-wire generation, and integrated energy reuse will be necessary in meeting rising demand while ensuring long-term decarbonisation progress.

The data centre industry at one point in time was considered an industry leader amongst most sectors, striving for decarbonisation results, balancing the seemingly impossible tasks of cutting emissions whilst reaching the growing global demands from cloud, IoT, and other sources of digital growth.

Given historical trend data available, internet traffic and data centre workloads skyrocketed throughout the 2010s, as sources from the IEA shows that data centre power consumption had plateaued over that same time period.

Periods such as lockdown from 2020-2021, where internet traffic functionally doubled, data centre operators managed to limit the industry’s energy consumption to just 1% of global demand, this clearly showed that all parties involved in the system were working together to minimise overconsumption and waste whenever possible, seemingly showing real, tangible efforts being made to implement decarbonisation protocols.

Even when countries were beginning to ease restrictions, giving people more options that that of the internet, data centre engagement was still high all the whilst governments and policymakers making bold net zero commitments, far exceeding the standards set by the Paris Climate Accords and embracing green power initiatives for a decarbonised future for digital infrastructure.

Unfortunately, 2025 has shown that these ideals are no longer the current norms.

Are we ready for the AI boom without compromising on decarbonisation goals?

Generative Artificial intelligence (GenAI) became a prominent topic at the end of 2022, with the first commercial large language models (LLMs) being released to the public.

As time has gone on, GenAI has pushed data centre demand past the natural order, demanding many orders of magnitude more computing power than the cloud, or something as simple as the Google search. Prior to GenAI becoming popular, the average data centre required between three and six kilowatts (kW) per rack. The latest and greatest of this generation of Nvidia AI servers require roughly 130kW to 250kW. The next iteration may need as much as four times that.

The consequence of this being, The IEA’s latest annual electricity report estimates that, data centers consumed 460TWh in 2022, this figure could rise to more than 1,000TWh by 2026.

The growing demand for electricity at this scale is already putting pressure on overstressed power grids, and the solutions becoming more and more of a reality, for the biggest data centre markets are regressive for 21st century ideals. AI has a major driving force for the additional fossil fuel consumption and, even with lighter, less consumptive models like DeepSeek, the big push from Silicon Valley has been to double down, build bigger and “better” models, whilst promising that AGI in the future will promise transformative effects desperately needed to overcome the climate crisis.

All across the sector, tech firms, operators, analysts and agencies have in conjunction with each other, spent the past two years walking back, underplaying and entirely abandoning their decarbonisation goals, as it becomes clear just how big a gap exists between net zero and where the industry needs to go in order to grow.

Many companies have turned to using carbon credits, which as a credit system is flawed at its best and at its worst, a covert method of greenwashing, in order to close the sustainability gap made by AI.

Within the US, AI growth has remotivated the country’s gas-fired power industry, with energy companies planning new natural gas-fired power generation, at a rapidly growing rate, not seen in decades. Jed Dorsheimer, group head of the energy and sustainability sector at investment bank William Blair, told Bloomberg last September that he’s seeing gas accounting for well over half of new energy generation — a far cry from the reality a few years ago, when “there was the expectation that solar and wind would be able to solve our additional generation needs.”

Put simply, the data centre industry needs to do better than what is currently the status quo, needing to develop superior methods of reaching these power demands for a GenAI future without deserting decades of progress towards decarbonising the sector.

Can the ageing UK grid keep up with increased investment?

The British Government has been supporting various industries within the field, not just data centres, being deemed as critical national infrastructure, but also aiming to turn the UK into an “AI superpower” with many regulation changes, intended to create AI development zones and attract more data centre operators to the country.

The National Grid has issued a warning that the further push towards a great reliance on GenAI will cause the UK’s total electricity demand to spike by 500% over the coming decade. The UK, along with many European countries, is being challenged with major bandwidth issues when it comes to grid capacity as well as delivering the power needed for data centres. With all these issues compounding each other, it is very difficult to truly understand the full limits that we are not equipped to deal with.

Power demand continues to be the most pressing issue facing the data centre industry for the upcoming years to come, even without the impact imposed on by GenAI.

The UK is not only ill-equipped to handle the kind of monuments rise in demand, but the levels of data investment Sir Keir Starmer is needing to attract is unlikely to come true given the fact that powering and cooling large data centres in the UK costs roughly four times as much as in they do in the US, let alone in the Nordic regions.

Simply put, the UK data centre sector needs to grow significantly in terms of investment and capacity, mitigate its environmental impact, reduce the existing strain it’s placing on the grid and for the future, and find a way to cut electricity prices over four-fold.

There is no one solution that will resolve this problem, but the need for nearby energy solutions are an essential piece of the puzzle.

What type of near-site energy solutions should the data centre industry implement?

Solar Photovoltaics (PV) is the cheapest form of energy in the UK today. It’s unsubsidised and, in many regions, including Spain and parts of Europe, wind and solar are cheaper than oil and gas. The cost differences have shifted; renewables are now cheaper to set-up and more valuable to sell power. However, the key challenge is to build up these energy sites at scale. It can provide immense power, but only intermittently, as without a UPS or backup power reserve, there will be no “constant” clean source of power.

Private-wire solutions can support between 20% and 30% of a hyperscale facility’s power demands using options such as near-site solar, not only reducing their carbon impact, but also insulating data centres against energy price spikes, grid failures, and higher consumption during the hottest months of the year when cooling systems are under the heaviest load and PUE is prone to creep past optimum levels.

It cannot solve the entire problem, but it is an important milestone to achieve to tackle this multi-pronged problem.

Holistic approaches are needed for more efficient systems to not only provide electrical generation, but power stores and more importantly, redistribution methods to utilise excess heat and power. By integrating these near-site generators, it can provide a circular economy that more importantly tackles the decarbonisation crisis of our time.

These near-site renewables have the potential to take strain off the grid in many ways, lessening strain on power requirements during peak heat seasons, and returning excess heat to district heating systems. Excess heat can also be rerouted to power industrial facilities, pipelined into a community-wide heating network, supporting multiple industries and reducing the need for additional energy.

For the UK data centre industry, if the goals are realistic and all parties are wanting to meet these targets that are demanded of GenAI, whilst keeping sustainability an honest aspiration and not critically damaging the national power infrastructure, it needs tangible steps towards that goal that solve problems today.