Two July Lithium Fires in Commercial Buildings Came From Stored Cells, Not Storage Systems

Charles County, Maryland firefighters traced a July 13 commercial building fire on Ray Drive to lithium-ion batteries stored on a shelf. The county hazardous materials team was called in to secure the cells. Eighteen days later, on July 31, firefighters in La Salle, Illinois responded to a fire at Batteries + Things, a retail store, and issued what the local newspaper described as a rare hazmat alert.

Neither incident involved a certified stationary energy storage system. Both occurred inside commercial occupancies, and both produced a hazardous materials response.

The scope gap. NFPA 855 addresses the installation of stationary energy storage systems: the engineered, listed, thermally managed assets that a contractor mounts, wires and commissions. Loose cells in a back room and retail battery inventory are a different exposure, sitting under general commercial storage and hazardous materials provisions rather than under the storage standard. The two July fires fell into the category the storage codes do not reach. The requirements applying to the category they do reach are tightening.

The new test. UL Solutions states that UL 9540A is the only fire and explosion test method referenced in the 2026 edition of NFPA 855. The sixth edition of UL 9540A, published March 13, 2026, revised Section 10 to incorporate an explicit large-scale fire test method aligned to Annex G.11 of NFPA 855.

For indoor installations, the sixth edition evaluates enclosure design, separation distances and the effectiveness of building-based suppression under an assumed post-deflagration condition, rather than crediting deflagration venting as the primary control. That assumption raises the evidentiary bar for any product marketed for placement inside an occupied building, and it disadvantages designs whose safety case rests on venting geometry rather than on cell and module behavior. Manufacturers holding reports written to the fifth edition face a re-test obligation as jurisdictions adopt the 2026 code.

None of that testing regime touches a pallet of retail stock.

Permitting posture. The gap matters because the audience for a test report and the audience for an incident response are the same people. Fire marshals and building officials who authorize commercial battery installations are also the officials whose departments run the hazmat calls. It is reasonable to expect that response experience shapes how those officials weigh lithium risk in commercial buildings, and a dispatch protocol cannot distinguish between a listed system with a large-scale fire test report and a shelf of loose cells, because at the moment of the call the burning material is unidentified.

If that reflex hardens into slower permitting for listed systems, the cost falls on a line item that the commercial storage sector is counting on to shrink.

Soft costs. On Latitude Media’s Catalyst podcast on July 30, Tim Hade of Voltus described a commercial and industrial storage cost curve moving from roughly $800 per kilowatt-hour all-in toward roughly $510. Hardware supplies part of that: Hade put battery pack costs down about 40 percent over eighteen months and system-level costs down about 30 percent.

The larger share comes from transaction and soft costs, the bucket covering development, interconnection, permitting and customer acquisition. Hade put those at roughly $200 per kilowatt-hour today, falling toward $25 to $50 per kilowatt-hour as automated development tooling replaces manual work, a reduction he characterized as roughly 90 percent over three years. Installation costs, on the same account, halve over five years. The two soft-cost figures do not reconcile precisely: a 90 percent cut from $200 lands below the stated $25 to $50 range.

Either way, the projection assigns most of the improvement to the non-hardware portion of a project. Permitting sits inside that portion. A forecast that assumes development friction falls by an order of magnitude is a forecast that assumes authorities having jurisdiction treat commercial battery installations as routine work.

The demand case. Hade framed the revenue side as four streams: time-of-use arbitrage, demand-charge management, ancillary services and capacity, with PJM capacity prices up sharply over the past three years. He offered a customer-side threshold as well: a project is worth a conversation at a 5 percent reduction in annual electricity spend and compelling above 10 percent.

The addressable base is large and largely untouched. Voltus, announcing its June acquisition of Brightfield AI, put commercial and industrial customers at 60 percent of United States electricity consumption while hosting under 1 percent of installed battery storage capacity. That gap is the segment’s opportunity and also the reason it has little accumulated regulatory precedent to lean on. Each jurisdiction is working through indoor commercial storage close to first principles.

Limits of the record. Two fires are not a dataset. One occurred at a store whose inventory consists of batteries, which is not a generalizable commercial exposure. Neither event produced new information about the failure behavior of a listed lithium iron phosphate system.

What the two incidents do illustrate is a classification problem. No national dataset separates stationary storage failures from consumer and inventory cell fires inside commercial occupancies. Absent that separation, the sector cannot point to incident statistics to demonstrate that certified systems are not the source of the commercial-building fire record, and officials evaluating a permit application have no ready reference distinguishing the two.

The sixth edition of UL 9540A applies a more demanding large-scale test to installed systems intended for indoor use. Inventory storage in the same buildings remains governed by general provisions, and the incidents in Maryland and Illinois both came from that side of the line.

Sources