Ultium Cells Begins LFP Production at Spring Hill, Adding a Second Operational US Cell Source for Energy Storage

Ultium Cells, the LG Energy Solution and General Motors joint venture, will begin shipping lithium iron phosphate cells for stationary storage from its Spring Hill, Tennessee plant in the second quarter of 2026 following a $70 million retooling of a line that previously produced nickel-based cells for electric vehicles. The cells will be sold through LG Energy Solution Vertech into grid-scale, renewable, and data center storage projects in North America. Seven hundred workers furloughed in January are returning to the plant by April.

Spring Hill becomes the second operational US LFP cell line producing for the stationary storage market. LG Energy Solution’s 16.5 GWh Holland, Michigan facility converted from EV cells to LFP storage cells in 2025 and has been shipping commercial volume. American Battery Factory’s Tucson plant is ramping toward production. Taken together, the three sites transition domestic LFP supply for storage from a category of announcements into a category of shipments.

Spring Hill was an EV plant eight months ago. Ultium furloughed the Tennessee workforce in January 2026 after GM and LG reassessed EV cell demand. The $70 million retooling is modest by battery-plant standards precisely because the structural work of building a cell line already existed. Tom Gallagher, Ultium’s vice president of operations, told pv magazine the company could shift capacity “in rapid short order.” That speed is the point. The pivot from nickel chemistry for cars to iron phosphate chemistry for buildings took roughly a quarter of calendar time.

The same pattern is visible at Ford’s Kentucky site, Stellantis’s Belvidere plans, and LG’s own Michigan facility. Each conversion turns a stranded EV asset into a storage asset without the two-to-three year permitting and construction cycle a greenfield plant would require.

The procurement math for ITC adders changes when a second source ships. Under the Inflation Reduction Act as amended by the One Big Beautiful Bill Act, commercial storage projects qualifying for the 10 percent domestic content adder on top of the 30 percent Investment Tax Credit must source a rising share of battery components from non-Foreign-Entity-of-Concern suppliers. For 2026 placed-in-service projects, the applicable percentage under §45X and §48 safe-harbor guidance treats LFP cells as the constraining bill-of-materials line.

Through most of 2025, that constraint pointed at a single operational domestic cell line. A developer underwriting a commercial behind-the-meter portfolio in Q3 2025 was effectively betting a single supplier’s shipments and quality yield. Bankers responded by discounting the domestic-content adder in base-case financial models or by requiring alternative-supplier covenants that compressed margins elsewhere.

Two operational lines shipping into the same market permit a different assumption. A Q3 2026 deal can model dual-source supply, can build a second supplier into the contingency framework without a merchant-cell fallback, and can firm forward pricing to end customers who need the tax-credit math settled before signing.

Chinese cell prices reversed deflation in March. As covered in “The Cheapest Batteries in a Decade Were Last Quarters” on March 11, CATL, EVE, and tier-one Chinese LFP pricing turned upward for the first sustained move in three years, driven by capacity discipline in Fujian and raw-material cost pressure. Domestic US cells have historically carried a premium of roughly 15 to 25 percent over imported Chinese equivalents. A narrowing gap does not eliminate the premium, but it changes the procurement logic at the margin: buyers who had accepted the premium only for the projects needing the domestic-content adder now face a smaller penalty for extending domestic sourcing across a portfolio.

Spring Hill and Holland do not resolve this pricing question by themselves. Capacity at both sites combined remains a fraction of the 400-plus GWh of LFP storage deployed globally in 2025. What they resolve is the bankability question for the narrower set of buyers who require FEOC-compliant cells to capture the tax adder. That set is concentrated in commercial and industrial storage, where the adder matters more proportionally than in utility-scale where project scale absorbs the cost.

Commercial storage developers should revisit bankability models. Any 2025-vintage pro forma that discounted the domestic-content adder for single-source risk is now conservatively stale. Developers with H2 2026 pipeline can reasonably underwrite the adder at full value against a dual-source cell supply assumption. That shifts the discounted cash flow by approximately 3 to 5 percentage points at typical project IRRs, material enough to push borderline deals past underwriting and to tighten forward pricing to end customers signing in the third quarter.

The remaining supply-chain constraints on commercial storage sit further downstream. Transformer lead times at 80 weeks, switchgear at 52 weeks, and specialist installation labor in dense urban jurisdictions remain the binding constraints on 2026 deployment velocity. Cells moved from binding constraint to manageable input sometime between Holland’s startup and Spring Hill’s Q2 shipment.

A single factory produces the signal but not the story. Ultium’s pivot is meaningful less as its own event than as confirmation that a dozen similar retoolings announced in 2024 and 2025 are clearing the execution threshold. Stellantis, Ford, KORE Power, and SK On have each described conversion programs in public filings. Spring Hill is the one that has a line running in Q2. Whether the others ship on their stated schedules is the question that will determine whether 2027 domestic LFP capacity exceeds domestic demand, constrains it, or tracks it.

For now, the procurement conversation for commercial storage buyers has two phone numbers instead of one.

Sources