AI Data Centres Need Big Batteries But Lithium Isn’t Fit-For-Purpose

The biggest constraint facing AI data centre expansion is storage and the unique way they use power, leading to unpredictability in drawing electricity from the grid. Traditional generation sources like coal and nuclear may still be necessary due to the volatile energy demand of AI data centres. Data centres require buffering capacity to reconcile intermittent supply with intermittent demand, with lithium-ion batteries degrading rapidly under constant cycling. Flow batteries offer a solution by allowing energy and power capacity to scale independently, enabling relentless cycling without degradation typical of lithium-ion systems. The world's renewable build-out is accelerating, but most large-scale batteries remain short-duration systems, highlighting the need for innovative storage solutions for AI-driven data centres. Investing in storage technologies designed for high-cycle operation can help harness renewable energy and prevent the reliance on fossil fuel backup, ultimately allowing AI data centers to contribute to a clean energy future.

https://cleantechnica.com/2026/05/06/ai-data-centres-need-big-batteries-but-lithium-isnt-fit-for-purpose/