The criteria for choosing batteries are starting to change in a field far from the EV range race. General Motors (GM) has partnered with U.S. energy storage startup Peak Energy to co-develop next-generation sodium-ion battery cells designed specifically for grid storage, and has also invested in the company. The move aims to bring a new option to the stationary storage market, which has so far been dominated by lithium-ion batteries.
Notably, GM has stated clearly that it has no plans to put this battery in its own EVs. Unlike passenger cars, where energy density per unit of weight is paramount, stationary batteries installed at substations, data centers and similar sites face relatively loose weight and volume constraints. Instead, upfront costs, long-term operating costs, thermal stability and simpler cooling equipment largely determine the economics. With these differences in mind, GM and Peak Energy are working to develop a storage system that takes advantage of sodium-ion characteristics.
Against a backdrop of grid congestion and surging power demand from AI data centers, strengthening electrical infrastructure has become urgent in the United States. The two companies aim to simplify the structure and operation of stationary storage systems by combining lithium-free battery materials with passive cooling that uses no fans or pumps. Whether that advantage holds up in real installations, however, depends on verification of manufacturing costs, long-term durability and cooling performance.
Why Sodium-Ion for Stationary Storage Instead of EVs
In EV battery development, gravimetric energy density, which indicates how much energy can be stored per kilogram, has been a key performance metric. A heavier battery adds to a vehicle's overall weight, affecting efficiency, driving performance and vehicle design. By contrast, weight is not necessarily a decisive weakness for stationary batteries fixed to concrete foundations and the like. Constraints such as installation footprint and foundation load capacity exist, but for operators, the upfront cost per unit of storage capacity and long-term operation and maintenance costs are often more important.
Kurt Kelty, GM's vice president of batteries and sustainability, has explained that battery technology should be chosen to suit the application, and that sodium-ion is well suited to large-scale stationary storage. Because sodium-ion batteries use sodium, which is abundant on Earth, instead of lithium, their raw material procurement is less exposed to international prices of lithium carbonate and lithium hydroxide. The energy density of sodium-ion cells currently on the market is generally lower than that of lithium-ion batteries, but in stationary applications, where weight and space constraints are relatively loose, this drawback does not necessarily hinder adoption.
The mainstream chemistry in today's stationary storage market is lithium iron phosphate (LFP). LFP uses no cobalt or nickel, offers higher thermal stability than nickel-manganese-cobalt (NMC) batteries, and tends to be cheaper in materials, which is why it is widely used in large-scale storage facilities. Even so, LFP generates heat from internal resistance during charging and discharging, so depending on the facility's scale and operating conditions, temperature control is needed through air conditioning, chillers, coolant circulation piping and the like. Such cooling equipment not only consumes electricity but also adds to installation and maintenance costs.
Peak Energy's current product, the "GS-1.1," by contrast, uses a sodium iron phosphate-type cathode material (NFPP, sodium vanadium fluorophosphate-based) and is designed to operate the entire system with passive cooling. Passive cooling is a method of managing battery temperature through enclosure structure and natural heat dissipation, without pumps that circulate coolant, electric fans or compressors. It does not eliminate the heat generated by the cells' internal resistance; it works by combining cell characteristics that limit heat generation with an enclosure design that lets heat escape. In cold conditions, the system uses resistive heaters to keep the cells at an appropriate temperature.
Omitting forced cooling in stationary storage brings several economic advantages. Conventional cooling equipment consumes power while running, so part of the stored energy is used for cooling, lowering the system's overall round-trip efficiency. Moving parts such as pumps and fans also carry risks of wear and failure, and when a site is remote, there are costs for dispatching maintenance staff for inspections and repairs. Peak Energy aims to use the thermal characteristics of sodium-ion batteries to reduce such auxiliary equipment, cutting costs across the facility's entire life cycle, not just the manufacturing cost of the cells.
However, it would be a mistake to regard sodium-ion batteries as nonflammable or immune to thermal runaway. In an independent study published by a U.S. Department of Transportation-related agency, external heating tests on four commercially available sodium-ion cells manufactured in 2024 found thermal runaway and release of flammable gases in every sample. The severity of the reaction and temperature characteristics vary with battery chemistry. The NFPP-type material Peak Energy uses is said to be more thermally stable than layered-oxide types and others, but that does not eliminate the possibility of thermal runaway altogether. Operating passive cooling safely requires not only dissipating the heat generated in normal charging and discharging but also safety design that accounts for abnormal heat generation and gas release.
In terms of volumetric storage capacity and accumulated long-term operating experience, sodium-ion batteries fall short of improved existing lithium-ion batteries in some respects. Even so, the ability to reduce dependence on supply chains for lithium, cobalt and other materials, along with the potential to eliminate forced cooling equipment and thereby reduce auxiliary power and maintenance burden, could become a competitive advantage in stationary storage. That effect, however, will vary not only with cell characteristics but also with enclosure design, installation environment and charge-discharge operating conditions.
GM Handles Cell Development, Peak Energy Handles System Integration
The capital and technology partnership between GM and Peak Energy was formally announced by both companies on June 9, 2026. Some media outlets reported it as of June 10, but the official announcement date is June 9. Under the partnership, GM is responsible for cell research and development and manufacturing technology, while Peak Energy handles integration into stationary storage systems.
GM will use its advanced battery development facility in Warren, Michigan, the Wallace Battery Cell Innovation Center, to develop next-generation sodium-ion cells optimized for stationary use. The company plans to begin producing prototype cells at the center during 2026, and has also secured exclusive manufacturing rights for the jointly developed next-generation cells. Peak Energy, for its part, will incorporate the developed cells into its own storage platform and commercialize them as grid storage systems.
The publicly available specifications differ greatly between the current GS-1.1 and the next-generation cells to be co-developed. To avoid conflating the two, the information that can currently be confirmed is summarized below.
| Item | Peak Energy "GS-1.1" (current product) | Next-gen cell co-developed by GM and Peak | Status and notes |
|---|---|---|---|
| Cathode material | Sodium vanadium fluorophosphate (NFPP)-type | Undisclosed (sodium-ion) | Detailed composition of next-gen cell not disclosed |
| Cooling method | Passive cooling (no fans, chillers or pumps) | Aims to maintain and optimize passive cooling | Combination of cell heat characteristics and enclosure design is key |
| Usable storage capacity | 3.1MWh (per container) | Undisclosed | Published specification of current product |
| DC rated power | 775kW (at 4-hour discharge) | Undisclosed | Must be distinguished from inverter and AC-side specifications |
| DC round-trip efficiency | 96% (cell/rack side) | Undisclosed | Excludes losses from PCS, transformers, etc. |
| Manufacturing and supply structure | Integrates externally sourced cells into the system | GM holds exclusive manufacturing rights | Region, term and supply conditions of exclusivity not disclosed |
| Market stage | Pilot validation; accepting orders | Prototype production planned for 2026; mass production timing undecided | No operating track record for products using the co-developed cells |
The details of the "exclusive manufacturing rights" held by GM have not been disclosed. Conditions such as whether the rights are limited to North America, how long they last, and whether GM can sell the cells to system integrators other than Peak Energy are not public. As a result, all that can be said with certainty at this point is that GM holds exclusive rights to manufacture the jointly developed next-generation cells.
Backing this technology partnership financially is Peak Energy's Series B funding round, announced on October 21, 2026. GM Ventures, GM's corporate venture capital arm, led the round, and the total raised reached $108 million. The round was oversubscribed, with existing and new investors including LG Technology Ventures, Maverick Silicon, Infinite Capital, Eclipse and Xora Innovation participating. However, the ownership stakes of each investor and Peak Energy's valuation in this round have not been disclosed.
GM's move is also tied to the restructuring of its battery business. In March 2026, GM and its joint venture partner LG Energy Solution indicated a plan to convert EV cell production lines at some U.S. plants to produce LFP storage systems for stationary use. In other words, GM is expanding already commercialized LFP batteries into stationary applications while also developing sodium-ion batteries as a future option. Stationary storage is positioned not as a business that merely repurposes existing EV technology, but as a new field in which battery technologies are used selectively according to application.
Examining the Basis for "20% Lower Storage Cost" and "99% Availability"
Peak Energy claims that its sodium-ion storage platform can cut the levelized cost of storage (LCOS) by up to 20% compared with conventional LFP-based systems. LCOS is a metric that accounts for installation costs, operation and maintenance costs and the like to evaluate the cost per unit of electricity delivered over the system's lifetime. However, the published materials reviewed do not show absolute cost figures for the baseline or the comparison case. The company also says it can guarantee availability above 99% under long-term service agreements (LTSAs), and has presented an estimate that the spread of passive cooling could cut up to 2 terawatt-hours (TWh) of power losses per year across U.S. storage facilities.
To evaluate these figures, it is necessary to separate values that can be confirmed as product specifications from estimates based on certain assumptions and from contractual guarantees. Peak Energy's current GS-1.1 houses batteries with 3.1 megawatt-hours (MWh) of usable capacity in an enclosure equivalent to a 40-foot standard container. Its DC rated power is listed as 775 kilowatts (kW) at a 4-hour discharge, and it supports discharge durations of 4 to 8 hours. The operating voltage range is 750–1500V, the ambient temperature range is −40 to +55°C, and the published design life is 20 years with a warranted 7,300 cycles.
The "96% DC round-trip efficiency" listed in the product specifications also requires an understanding of its scope. This is the round-trip charge-discharge efficiency on the battery's DC side, and it does not include all losses from the power conversion system (PCS) and transformers needed to connect to the grid, standby power of control equipment, or the power consumed by heaters used in cold conditions. In actual storage facilities these losses and auxiliary power are added, so the system-wide round-trip efficiency measured on the AC side (AC-AC efficiency) is generally lower than the DC-side value. The 96% figure therefore cannot be taken directly as the effective efficiency when connected to the grid.
The claim of up to 20% lower LCOS also leaves conditions to be verified. From the public materials, it cannot be adequately confirmed whether the LFP storage system used as the comparison is air-cooled or liquid-cooled, which period's cell prices were used, or how charge-discharge cycles, electricity prices and equipment service life were set. LFP battery prices also fluctuate, so if the comparison date or operating conditions change, the economic advantage of sodium-ion batteries could change as well. The 20% figure should be taken as the company's own estimate, not a guarantee that the same savings will be achieved in every project.
The "availability above 99%" is likewise not an average uptime measured by a third party across many commercial installations. It is a contractual service level (SLA) that Peak Energy offers in long-term service agreements, guaranteeing uptime of 99% or more. A design that reduces failure factors by cutting moving parts for cooling underpins that guarantee, but how much availability actual installations maintain over the long term will need to be confirmed through future operating results.
The figure of up to 2TWh in annual power loss reduction is also not an amount of electricity actually saved at present. It is an estimate that totals the effect of replacing LFP storage facilities that use forced cooling with passive-cooled sodium-ion batteries, based on the scale of future adoption Peak Energy envisions. Whether a reduction of that scale is realized across the entire U.S. storage market depends not only on the volume of sodium-ion batteries deployed but also on actual operating conditions and the cooling power of existing facilities.
From Demonstration to Mass Production: Building Manufacturing Capacity for 2027
Trial operations using actual units have also begun to verify Peak Energy's claims about performance and economics. In summer 2025, the company announced that it had shipped megawatt-hour-class test systems for a joint pilot program involving nine major U.S. utilities and independent power producers (IPPs). According to product materials, at a demonstration site in Watkins, Colorado, operation connected to the commercial power grid has continued since August 2025. Such demonstrations are important for confirming the behavior of the entire system in real operating environments, which cannot be understood from laboratory cell evaluations alone.
On March 11, 2026, Peak Energy also announced a partnership with RWE Americas, the U.S. arm of major renewable energy developer RWE. The plan is to install Peak Energy's storage system at RWE's technology validation lab in eastern Wisconsin. However, this is an effort to validate the technology at a specific facility, and does not mean that large-scale commercial deployment has been decided across the territory of MISO (Midcontinent Independent System Operator), which oversees the grid in the U.S. Midwest.
To move from the pilot stage to commercial mass production, Peak Energy is also building out its own manufacturing base. On July 8, 2026, the company announced that it would build a storage system manufacturing plant with 183,000 square feet of floor space in Sacramento, California. The plant is planned to have a production capacity of up to 4 gigawatt-hours (GWh) per year, with capital investment of up to $71 million expected. It has also received a $10.5 million tax credit award under California's CalCompetes program for business job creation, and plans to create 239 direct jobs in the region over the next 18 months.
However, the figure of 4GWh per year is the production capacity if the plant is completed as planned and runs at full capacity, not current output. The new plant is scheduled to begin operations and ship its first products in the first quarter of 2027. Actual manufacturing capacity at present is limited, and moving from small-volume production of pilot products to large-scale continuous production will require stabilizing assembly processes, quality control and improved production efficiency. Planned facility capacity must be distinguished from the quantity of products that can actually be shipped to customers.
Peak Energy plans to use the $108 million raised in the Series B to ramp up the Sacramento plant, serve existing customer projects, and develop and launch products for data centers. According to the company, orders from existing customers exceed $1 billion, and it positions the data center market, where power demand is growing with the spread of AI, as another key growth area. However, the figure of more than $1 billion is the order scale as described by the company, not revenue already booked or the value of products already shipped. Delivery terms and cancellation conditions for each contract also cannot be confirmed, so it cannot be judged that the full amount is established as future revenue.
To turn orders into continuing revenue, the company must proceed with the planned 2027 startup of its new plant as scheduled and stably ship products that meet the required quality and performance. Delivery timing could change because of installation delays, manufacturing problems or supply chain disruptions, so going forward, actual production volumes and the progress of shipments to customers, not just the start of plant operations, will be important factors in assessment.
Four Checkpoints That Will Decide Sodium-Ion Storage's Competitiveness
To judge how competitive the GM–Peak Energy partnership can be in the stationary storage market, it is necessary to evaluate currently announced plans and estimates separately from performance that has actually been verified. Four elements are especially important: the specifications of the next-generation cell, thermal management during long-term operation, suitability for data center applications, and establishment of mass production.
The first checkpoint is the actual performance of the next-generation cells to be prototyped at GM's Wallace Battery Cell Innovation Center. Peak Energy's current GS-1.1 is believed to use externally sourced cells, but details of the cells GM will co-develop, such as cathode composition, gravimetric and volumetric energy density, cycle life and internal resistance, have not been disclosed. How much of an advantage the cells' mass-production yield and material costs will offer over LFP batteries, whose prices keep falling, is also important. The abundance of sodium as a raw material alone does not guarantee that the finished cell will be cheaper to manufacture.
Second is thermal management performance and durability over long periods of use in harsh environments. With passive cooling, the question is whether cell temperatures can be kept within a safe range even in summer, when outside temperatures exceed 40°C, or under continuous charge-discharge conditions. In a structure that does not actively move heat through forced cooling, uneven temperature distribution within the enclosure or modules could cause some cells to run hot and degrade faster. If capacity retention, changes in internal resistance and temperature distribution after long-term operation at the demonstration sites in Colorado and Wisconsin can be confirmed, the practicality of passive cooling could be evaluated more concretely. Whether third-party measurement results will be published is of particular interest.
The third checkpoint is suitability as a power source for data centers. Storage used for data center backup power or peak demand shaving requires not only high reliability but also the ability to handle sudden load changes. However, grid storage systems designed for charging and discharging over several hours and uninterruptible power supplies (UPS) that compensate for voltage fluctuations on a millisecond scale require different output characteristics and response speeds. A design for long-duration discharge like the GS-1.1 is not necessarily suited to every data center use. If specific customer names, deployment scale and intended operating methods are disclosed going forward, it will become clearer which segment Peak Energy is targeting.
Fourth is the startup of the Sacramento plant, scheduled for the first quarter of 2027, and the establishment of mass-production quality. A new plant may take time for equipment tuning, parts procurement and stabilizing assembly processes, and may not immediately reach its planned capacity. Stably producing thousands of large storage systems with passive cooling structures per year requires a setup capable of carrying out the necessary inspections while keeping quality variation between units small. Whether the more than $1 billion in orders can be converted into actual sales depends on building out this production system and shipping products as planned.
GM's investment in sodium-ion batteries does not signal a scaling back of its EV battery business. It should be seen as diversification of the battery business: using batteries suited to applications demanding high energy density in EVs, and a different technology emphasizing cost, thermal management and long-term reliability in stationary storage. The passive-cooled sodium-ion storage system that GM and Peak Energy are pursuing has the potential to lower stationary storage operating costs by reducing complex cooling equipment. For that advantage to be firmly established, however, cell manufacturing costs, long-term durability and actual grid-connected efficiency must be verified, and the system must be shown to reduce total cost of ownership compared with LFP batteries. The release of prototype cell specifications and demonstration data, along with the production ramp-up planned for 2027, will be important milestones in judging the technology's commercial strength.