- General Motors led a $108 million Series B for Peak Energy, which makes sodium-ion grid storage systems. The company reports about $1.1 billion in customer orders, roughly ten times the money it has just raised, mostly from a 4.75 gigawatt-hour agreement with Jupiter Power and a 1.5 gigawatt-hour deal with Energy Vault.
- It has not finished a commercial plant. Peak ships small volumes from a San Francisco pilot line and expects to complete its first commercial facility in the first quarter, with systems rolling off shortly after.
- The strategic rationale is domestic supply rather than battery chemistry. Chief executive Landon Mossburg said the aim is to catalyse the supply chain as much as possible, adding that it is already catalysed in China, and GM Ventures president Kevin McCabe cited abundant, domestically sourced materials.
- The market is small but projected to grow enormously. BloombergNEF forecasts sodium-ion demand reaching 257 gigawatt-hours by 2035 from just over 1 gigawatt-hour in 2025, by which point it could account for almost 10% of stationary storage.
What Happened?
Peak says its technology tolerates higher operating temperatures than lithium-ion cells, giving cost and safety advantages for data centres, grids and independent power producers. GM is developing sodium-ion cells at its Michigan battery labs for Peak’s later use, following a partnership announced in June. Bloomberg notes the market remains far from mature, that sodium-ion cells are still more expensive than lithium-ion, and that both face the same trade pressures with manufacturing concentrated in China and elsewhere in Asia. Peak investor Greg Reichow of Eclipse Capital argued lithium-ion is cheap because it operates at very high scale, and that sodium-ion’s fundamental materials and transformation costs should ultimately be lower.
Why It Matters?
An order book ten times the size of the funding round is unusual and cuts both ways. It demonstrates genuine customer demand, which is more than most companies raising money in this sector can show, and Peak is weeks from completing a plant rather than years. It also means a great deal has been promised by a company that has not yet manufactured at commercial scale. The cautionary comparison is Nscale, which carries more than $103 billion of contracts having built under 1% of its targeted capacity, and whose flagship site is six months late. Contracts are a claim on future delivery, and delivery is where these businesses succeed or fail. The real investment case is supply chain location rather than battery performance. Sodium is abundant and sourceable domestically, whereas lithium-ion depends on inputs concentrated in a handful of countries. That places this alongside JPMorgan deploying $230 billion toward US economic security, a $7.4 billion Korean zinc smelter in Tennessee, and hyperscalers contracting copper from mines that will not produce until 2029. Capital is flowing toward materials the US can obtain without geopolitical exposure, and sodium-ion is a battery chemistry chosen substantially for that property. Reichow’s cost argument deserves the scepticism it invites. Saying lithium is cheap only because of scale, and that sodium will ultimately be cheaper, is a claim about a future state with no present evidence, made by an investor in the company. The honest position today is that sodium-ion costs more and is bought for reasons other than price.
What Next?
Completing the first commercial facility in the first quarter is the milestone that determines whether the order book converts. Watch deliveries against the Jupiter Power and Energy Vault agreements, which together account for most of the $1.1 billion. BloombergNEF’s 2035 projection implies growth of more than 250 times from a tiny base, so early manufacturing execution across the sector is what establishes whether that trajectory is plausible. For GM, this is a modest diversification into a growth adjacency while its core business contends with global capacity utilisation near 60%.
Affected Tickers and Coins: GM, NRGV
Source: Bloomberg
