AI Data Center Growth Drives Energy Storage Boom
The rapid expansion of AI data centers is fueling significant growth in the energy storage market. Companies like Redwood Materials are scaling up operations to meet the massive power demands from both data centers and the manufacturing facilities that support the AI hardware supply chain.
- Global electricity consumption from data centers is projected to more than double from 415 TWh in 2024 to approximately 945 TWh by 2030, an amount greater than Japan's total current electricity consumption. One estimate suggests data center power demand could increase by as much as 165% by the end of the decade compared to 2023. - A single query on a large language model like ChatGPT can consume roughly ten times more electricity than a standard Google search, highlighting the energy intensity of AI computations. The training of sophisticated AI models is also highly energy-intensive; for instance, the training for GPT-4 required about 30 megawatts of power. - The global energy storage market is forecast to grow from an installed base of 0.54 terawatts in 2026 to 1.52 terawatts by 2031, with a compound annual growth rate of 23.05%. In 2025 alone, annual energy storage deployment is expected to reach 92 gigawatts, a 23% increase from 2024. - Redwood Materials, founded by Tesla co-founder J. B. Straubel, is expanding its capacity to produce battery materials from recycled sources, aiming for 100 GWh of production by 2025, enough for one million EVs. The company's recycling process uses 80% less energy and reduces CO₂ emissions by 70% compared to traditional mining and processing of battery materials. - Beyond the data centers themselves, the manufacturing of advanced semiconductors for AI applications is energy-intensive. AI is also being applied within manufacturing and supply chains to optimize energy use, predict maintenance needs, and improve demand forecasting, which can reduce waste and overall energy consumption. - Apple integrates AI into its supply chain for predictive analytics, demand forecasting, and inventory optimization. This strategy includes investments in custom silicon to support on-device AI processing, which can be 100 to 1,000 times more energy-efficient per task than cloud-based AI. - Redwood Materials has launched a new business line, Redwood Energy, to deploy large-scale energy storage systems using both new and repurposed second-life EV batteries. One of their initial projects powers a small data center with a 63 megawatt-hour storage system composed of 792 used EV battery packs, demonstrating a circular economy approach to meeting the energy needs of the digital economy.
