AI data centers have ignited interest in a new "clean energy" trend! Meta and Google are betting on long-term energy storage, capable of storing energy for up to 100 hours.

AI data centers have ignited interest in a new "clean energy" trend! Meta and Google are betting on long-term energy storage, capable of storing energy for up to 100 hours.

The electricity demands of artificial intelligence data centers are bringing a long-dormant clean energy sector back into the spotlight. Tech giants like Meta and Google are increasing their investment in long-term energy storage, opening up markets for previously costly and slow-to-commercialize new energy storage technologies through large-scale procurement and project collaborations.

On September 18, Bloomberg reported that since the beginning of 2025, at least 150 gigawatt-hours of long-duration energy storage projects have been announced, approved, or financed in the United States, more than five times the total planned installed capacity at the end of 2024. Most of the new demand comes from AI data centers.

Unlike traditional lithium-ion batteries, which typically only provide a few hours of energy storage, long-term energy storage can store electricity for tens of hours or even days. This makes it more suitable for meeting the ever-increasing electricity demand of AI data centers and can also help renewable energy sources such as wind power and solar power overcome the time mismatch between power generation and consumption.

The involvement of tech giants has also brought about previously scarce commercial validation. Google is partnering with Italian company Energy Dome to deploy a 10-hour carbon dioxide storage system in Arizona; Meta has signed an agreement with California-based Noon Energy to install a system with a maximum storage capacity of 100 hours. With increasing orders, long-duration energy storage startups are expected to secure more funding and expand production capacity.

Long-duration energy storage encompasses various technological approaches, enabling the sustainable storage and release of electricity for hours or even days. Historically, this field has been hampered by high costs and a lack of large-scale validation, hindering its commercialization. However, the expansion of AI infrastructure is accelerating its adoption. Cully Cavness, co-founder and president of Crusoe, stated that long-duration energy storage is becoming a crucial tool for expanding AI infrastructure.

Tech giants are betting on different technological paths: Google is partnering with Energy Dome to achieve 10 hours of energy storage using carbon dioxide; Crusoe is partnering with Energy Vault to store electricity using gravitational potential energy; and Meta's solution with Noon Energy can extend the energy storage time to 100 hours.

Amanda Yang, head of clean and renewable energy at Meta, said, "Four hours of energy storage is far from enough." Lucia Tian, director of advanced energy technologies at Google, said that some long-term energy storage technologies do not rely on key minerals such as graphite, which helps reduce dependence on related supply chains .

Orders from tech giants open funding windows for startups

For long-duration energy storage startups, the significance of large data center customers lies not only in the scale of their purchases, but also in helping new technologies gain market recognition.

Aric Saunders, executive vice president of Noon Energy, said that AI data centers now account for about half of the company's planned deployments, compared to almost zero two years ago. He said that while other customers are also increasing, few companies can match the speed and financial strength of their AI data center purchases, "They are the accelerators of this industry."

Tom Sisto, co-founder and CEO of XL Batteries, stated that the purchase order from data center developer Prometheus Hyperscale provides the company with market validation. Such orders also help alleviate financing difficulties for innovative hardware companies that need to expand production.

According to Yiyi Zhou, an energy storage analyst at BloombergNEF, as these technologies are first implemented in AI data centers, actual operational data and large-scale production are expected to drive down costs and further expand to other industries.

Risk warning and disclaimerInvesting involves risk; please exercise caution. This article does not constitute personal investment advice and does not take into account the specific investment objectives, financial situation, or needs of individual users. Users should consider whether any opinions, views, or conclusions in this article are suitable for their specific circumstances. Any investment decisions made based on this information are at your own risk.