The rapid deployment of artificial intelligence is putting pressure on electricity grids as they try to absorb large, continuous and localised loads.
The International Energy Agency says a typical AI-focused data centre already consumes as much electricity as 100,000 households.
New facilities could require 20 times more, placing them “on par with the energy consumption of small countries”.
A paper released by the OECD this week says the use of AI will increasingly be based the capacity of electricity systems to manage large power drains without conflicting with other usages.
Ruben Maximiano and Wouter Meester from the OECD Economics Department argue that AI is often discussed as though it operates independently of physical systems.
“In practice, AI depends on vast amounts of electricity,” their paper says.
“Its future will be determined not only by advances in algorithms and computing power, but also by kilowatt-hours – by the ability of electricity systems to deliver power reliably and at scale.”
“The importance of energy to AI roll-out is visible in corporate energy sourcing strategies,” the paper says.
“Big Tech companies now account for the majority of Corporate Power Purchase Agreements (PPAs) in Europe.
“Yet the scale and speed of AI deployment are already outpacing what traditional PPAs can guarantee.”
The paper says AI “hyperscalers” are turning to direct investment in generation, including solar, wind and nuclear, to secure long-term supply.
“While global electricity demand from AI remains moderate (expected to reach 3 percent globally by 2030 and 4.5 percent in the EU) its impact is highly concentrated,” the OECD paper says.
“Data centres cluster in locations offering robust fibre connectivity, favourable cooling conditions, low electricity prices, and fast, reliable grid access.
“This concentration amplifies pressure on local grids and exposes the limits of existing planning and connection frameworks.”
The full article can be found here.
