The rapid growth of data centers driven by the increasing use of artificial intelligence (AI) is leading to a surge in power demand, with estimates suggesting a 160% increase in power demand from data centers by 2030 compared to 2023 levels. This increased demand is expected to have a significant impact on the environment, with a scenario in which 60% of the increased demand is met by thermal sources like natural gas resulting in an expected emissions increase of 215-220 million tons globally, equivalent to 0.6% of the world’s energy emissions.
To meet this increased demand, a mix of power sources is needed, including renewables, natural gas, and nuclear energy. While renewables have the potential to meet some of the increased power needs, they do not produce power consistently enough to be the only energy source for data centers. Schneider, a digital infrastructure analyst at Goldman Sachs Research, notes that nuclear is the preferred option for baseload power, but the difficulty of building new nuclear plants means that natural gas and renewables are more realistic short-term solutions.
The article highlights the challenges of developing new nuclear power plants, including the scarcity of specialized labor, permitting issues, and the difficulty of sourcing sufficient uranium. However, by the 2030s, new nuclear energy facilities and advancements in AI could start to bring down the overall carbon footprint of AI data centers.
In addition to nuclear energy, companies are also investing in renewable energy sources, with 40% of new capacity built to support increased power demand from data centers expected to be from renewables. The cost of renewable energy is increasingly competitive with natural gas, with onshore wind and solar energy costs of $25-26 per megawatt hour in the US, compared to $37 per megawatt hour for combined cycle natural gas.
The article concludes that a “mix” approach is needed to meet the increasing power demands from data centers, with a combination of power sources required to provide reliable and cost-effective energy.