Batteries Become Cheaper Than Gas Turbines for Data Centers
Global study shows four-hour battery storage undercuts peaking gas plants as AI power demand spikes equipment costs
Four-hour battery energy storage systems have become less expensive than open-cycle natural gas turbines across every global market, according to a comprehensive report by research firm Wood Mackenzie. Surging demand from hyperscale artificial intelligence data center developers has created severe supply shortages for natural gas turbines, driving up equipment prices and making battery storage the more economical choice for meeting peak electricity loads.
Key details
The Wood Mackenzie study surveyed 43 energy markets across every continent, concluding that four-hour duration battery energy storage systems are now cheaper than open-cycle gas turbines globally. In regions like the Middle East and Africa, battery storage is projected to be 33% cheaper than gas peaking plants by 2035, while energy storage costs in China are already 55% below those of neighboring countries.
The shift is driven largely by massive procurement backlogs in gas turbine manufacturing. AI data center developers attempting to bypass multi-year electrical grid interconnection queues have bought up available natural gas turbines, extending lead times to two to four years for simple open-cycle units and into the early 2030s for combined-cycle turbines. This purchasing rush has dramatically increased turbine prices. In contrast, utility-scale solar remains the cheapest source of new power generation globally, protected in North America by tax credit provisions under federal legislation for projects beginning construction before the end of 2027.
Why this matters
Data centers powering frontier AI models require continuous baseload power alongside flexible capacity to handle sudden GPU compute spikes. Historically, utilities and data center operators relied on natural gas peaking plants to handle rapid load fluctuations. As turbine prices surge due to supply constraints, utility-scale battery storage offers a lower-cost, faster-to-deploy alternative for balancing grid supply and stabilizing high-density AI data center workloads.
Context
As hyperscalers expand compute capacity, grid constraints have forced operators to explore behind-the-meter generation and energy storage. While natural gas was initially seen as the fastest bridge fuel to meet multi-gigawatt power requests, supply chain bottlenecks for turbines are changing project economics. The widening cost advantage of battery storage allows utilities and developers to integrate renewable solar capacity with battery backup, reducing reliance on fossil-fuel peaking plants.
Risks and open questions
While four-hour battery systems undercut open-cycle gas turbines on cost, four hours of storage duration may be insufficient to maintain continuous 24/7 operations during extended grid outages or multi-day periods of low renewable output. Additionally, battery supply chains face potential price volatility from international tariffs and raw material restrictions, which could temporarily narrow the cost gap in specific regional markets.
What happens next
Utility planning commissions and data center developers are expected to re-evaluate proposed gas peaking projects in favor of large-scale battery storage installations paired with solar. As battery costs continue to fall and gas turbine waitlists stretch further into the decade, energy analysts project that battery storage will increasingly replace natural gas peakers as the primary tool for managing grid stress from AI infrastructure.
Source: TechCrunch Published on AI Usage Global, author: AUG Bot
