GM Pivots Toward the Grid: Sodium-Ion Batteries and the AI Power Hunger
Table of Contents
The Automotive Pivot to Power Infrastructure
The race to stabilize the electrical grids supporting the AI boom has found an unlikely new protagonist: the American automaker. General Motors is officially expanding its battery ambitions beyond the chassis of electric vehicles, moving into the energy-storage system (ESS) market with a strategy that targets both the industrial sector and the massive power requirements of AI data centers.
While competitors like Ford have previously mentioned repurposing battery capacity for the grid, GM’s approach is more fundamental. The company is betting on a chemical shift. Through a new partnership with the startup Peak Energy, GM is developing a sodium-ion battery chemistry specifically tailored for stationary, grid-scale deployments. This move is significant because, outside of the Chinese market, Western automakers have largely remained hesitant to commit to sodium-ion production.
“The way we’re getting into the market is the easy way, through ESS,” Kurt Kelty, vice president of battery and sustainability at GM, noted. He argues that the performance characteristics of sodium-ion are a natural fit for the stationary market, where the trade-offs of the chemistry are less punishing than they are in a moving vehicle.
Why Sodium Matters Over Lithium
In a car, energy density is everything; a battery must be light and compact to maximize range. Sodium-ion batteries fail this test—they are heavier and bulkier than their lithium-ion counterparts for the same amount of energy. However, when the battery is sitting in a concrete vault at a data center or a factory, weight becomes irrelevant. What matters is cost, safety, and longevity.
Sodium is abundant and cheap, making these cells more economically viable for massive installations. More importantly, they are less prone to the thermal runaway issues that plague some lithium chemistries. Paul Menson, GM’s director of energy-storage commercialization, highlighted a critical engineering advantage: because sodium-ion is inherently more stable, Peak Energy’s systems can potentially strip out complex cooling and fire-suppression hardware. By removing these components, GM and Peak are effectively removing the most common failure points and maintenance costs associated with grid storage.
The Roadmap: From LFP to Trial Production
The transition to sodium-ion won’t happen overnight. GM’s sodium-ion cells aren’t expected to enter trial production at the company’s Battery Cell Development Center until 2028. To bridge the gap, GM is utilizing a tiered rollout. In the short term, the automaker will sell lithium iron phosphate (LFP) cells to LG Energy Solution for use in their energy-storage systems. This leverages GM’s existing Ultium joint venture with LG, ensuring they have a footprint in the ESS market while the newer chemistry matures.
GM has already committed $900 million toward commercializing new battery chemistries, a capital injection that includes a dedicated development center designed to shave a full year off the traditional commercialization timeline.
Closing the Loop with Redwood Materials
Parallel to the development of new cells, GM is deepening its relationship with J.B. Straubel’s Redwood Materials. This represents the ‘circular’ side of GM’s energy strategy. Redwood already processes scrap from GM’s factories and takes in old EV packs—roughly 10,000 units are currently in the pipeline.
The practical application of this is already visible in Nevada, where Redwood operates a 12 megawatt/63 megawatt-hour microgrid using second-life packs to support a Crusoe data center. GM is now bringing this technology in-house, purchasing a 7.2 megawatt-hour Redwood system for one of its Michigan plants. While a data center uses batteries to smooth out the erratic power spikes caused by high-performance GPUs, GM’s industrial application focuses on ‘peak shaving’—reducing power draws during expensive peak hours to lower utility bills and providing critical backup during outages.
As the energy demands of generative AI continue to strain regional grids, the ability for industrial giants like GM to manage their own power loads and sell the hardware to do so transforms the automaker into a critical player in the energy infrastructure landscape.
