Commercial manufacturing, planned in partnership with Horien Salt Battery Solutions, is targeting an initial 2 GWh facility with 70% domestic content and commercial shipments in 2027
Decision Focus
Inlyte Energy, a California-based iron-sodium battery startup, announced two concurrent pilot programs on May 22, 2026. The first targets a Tier IV colocation facility in Bern, Switzerland, operated by NTS Colocation. The second places its first integrated utility demonstration system at Southern Company’s test site in Wilsonville, Alabama. Neither installation is commercial yet. What makes this worth attention is the specific operational angle: Inlyte is not positioning against lithium-ion for short-burst UPS duty. It is positioning directly against diesel generators — the backup power asset that most data center operators know they need to replace and have not yet found a credible way to do so.
90-Second Brief
In recent days, inlyte Energy is targeting the data center market with an iron-sodium battery chemistry it claims offers a 7,000-cycle lifespan, no fire or thermal runaway risk, and installation density reaching up to one gigawatt-hour per acre. Two live pilots are planned for 2026: a 600 kWh installation at a Swiss Tier IV facility and a utility-scale demonstration with Southern Company in Alabama. U.S. Commercial manufacturing, planned in partnership with Horien Salt Battery Solutions, is targeting an initial 2 GWh facility with 70% domestic content and commercial shipments in 2027.
What Is Really Happening?
The diesel generator replacement problem in data centers is real and structurally unresolved. Operators rely on diesel and gas turbine backup because those assets can sustain output across prolonged outages — hours to days — and handle the volatile power demand profile that short-duration lithium systems cannot economically match. The trade-off has been tolerated for decades: carbon emissions, noise complaints, fuel logistics, and growing community resistance set against mounting ESG commitments.
Microsoft’s publicly stated pledge to eliminate diesel generators from its data centers by 2030 is cited in the source article as a leading indicator of where operator intent is heading. Whether the broader market moves on that timeline is uncertain, but directional pressure exists across hyperscalers and large colocations alike. What the market currently lacks is a technically mature, commercially available alternative that matches diesel’s duration and reliability without lithium’s fire risk or the cost structure that has prevented large-scale deployment.
Inlyte’s pitch targets this gap directly. Iron-sodium chemistry is inherently stable: the company states that cells that fail internally do not propagate to thermal runaway, and that modules continue operating through internal cell failures. That characteristic, if confirmed at scale, changes the site planning calculus for dense urban facilities where fire-suppression requirements, insurance costs, and community approvals around lithium-ion storage are already adding friction.
The density claim — up to 1 GWh per acre — is relevant to operators constrained on footprint. Urban colocation providers and campus-constrained hyperscaler sites have been unable to scale behind-the-meter storage without land that competes with IT density. That figure is a company projection, not an independently audited result, and should be treated as such until pilot data is available.
Why It Matters for Global Heads of Data Center Energy
The first practical implication is procurement timing. If Inlyte’s 2026 pilots produce credible performance data — particularly around discharge duration, depth-of-discharge consistency, and cell-failure behavior under load — that data will enter circulation before commercial shipments are available in 2027. Energy procurement teams that wait for a product launch to begin evaluation will be behind. The Swiss pilot is explicitly designed to test integration with existing facility systems, meaning NTS Colocation will generate the first independent operational dataset on this chemistry in a live Tier IV environment.
The second implication is supply chain positioning. Inlyte’s U.S. manufacturing target — 2 GWh initial capacity, 70% domestic content — is relevant given current trade policy and the Inflation Reduction Act’s domestic content incentives. Operators structuring behind-the-meter storage procurement through IRA bonus adders have a policy-driven reason to prefer domestic supply chains. If Inlyte’s 2027 commercial shipment target holds, the product could qualify. The dependency chain — Horien Salt Battery Solutions for production, Ervin Industries for iron powder formulation — is early-stage, and slippage in either partnership would push timelines.
The third implication touches interconnection strategy. Behind-the-meter storage capable of sustaining load across multi-hour or multi-day outages reduces exposure to grid instability without adding interconnection queue complexity. For operators in markets with constrained interconnection timelines — PJM, ERCOT, MISO — that is an indirect but material value.
Forward View
If both 2026 pilots proceed on schedule and generate clean performance data, the immediate consequence is that a new supplier enters procurement shortlists for diesel replacement evaluations in 2027. That is a narrow but specific market signal worth tracking. If the NTS deployment demonstrates that iron-sodium integrates manageably with existing Tier IV power management architecture, European colocation operators facing stricter emissions regulations will have early justification for a procurement evaluation ahead of their U.S. counterparts.
On the manufacturing side, the path from a 2 GWh initial facility to the capacity levels that would make Inlyte relevant at hyperscaler scale is long. Operators should watch whether additional capital rounds or strategic partnerships close before or alongside commercial shipments. A startup at pilot stage with a 2027 first-commercial target carries execution risk that lithium-ion incumbents with operating supply chains do not.
What Is Still Uncertain
Performance at pilot scale has not been independently validated. The 7,000-cycle lifespan and 1 GWh per acre density figures are company estimates, not third-party results. The Southern Company utility demo is described as lasting at least a year, meaning utility-scale performance data will not be available before 2027 at the earliest. Cost per kWh delivered is not disclosed in the source, making it impossible to assess economic competitiveness against diesel or lithium-ion on a total cost of ownership basis. The manufacturing partnership with Horien Salt Battery Solutions is early-stage, and whether the 2027 commercial shipment target is achievable depends on execution across multiple new relationships. None of the pilot outcomes are guaranteed.
One Question for Your Team
If Inlyte’s Swiss Tier IV pilot produces credible multi-hour discharge data by mid-2027, does your current backup power roadmap have a decision gate to evaluate iron-sodium alternatives before your next diesel generator procurement cycle closes?
Sources
- Pv-magazine-usa — Inlyte Energy plans pilot projects to prove its iron-sodium battery technology (Link)
