A Norwegian Battery-Room Burn Test Found IG-541 the Only Suppression System to Stop Module-to-Module Spread

SINTEF and RISE Fire Research ran a series of large-scale fire suppression experiments inside a mock-up battery room for the Norwegian SafeBESS project. Four suppression systems were compared: sprinklers, high-pressure water mist, low-pressure water mist, and the inert gas IG-541. Two further tests let the fire burn without suppression in order to study venting.

One of the four stopped fire spreading between modules.

The room. The test enclosure measured 3 by 6 by 3 metres and held a multi-module cabinet built on NMC pouch cells. That detail governs how far the results travel: a suppression and ventilation design validated against one cell chemistry and one module format transfers to another only by argument. The thermal-runaway gas volumes an LFP room produces are a separate measurement, not an inference from this one.

The suppression result. According to SINTEF’s account of the tests, IG-541 extinguished the flames and stopped propagation between modules. The water-based systems cooled the modules well. In the low-pressure mist test, hot spots survived and fire spread from module to module.

The gas. Extinguishing the flames did not clear the hazard. SINTEF reports that flammable gas kept accumulating in the room even after IG-541 had put the fire out. The implication for a specifier is that suppression and explosion prevention are being measured as separate outcomes in the same test, and a system that scores well on the first does not thereby score on the second.

SINTEF’s own framing makes the operational version of the conflict plain: inert gas and forced ventilation work against each other, because one requires the room sealed and the other requires air moving through it. Whoever designs the room has to decide how long it stays sealed and how the accumulated gas is removed afterwards without igniting it.

The ventilation result. In the tests that examined venting, ventilation prevented an explosion. The required airflow, the researchers say, remains an open question. They are finalising guidelines for battery rooms in buildings, and the minimum airflow figure is not yet settled.

That leaves a gap in the middle of an otherwise usable data set. A mechanical engineer sizing a fan for a battery room has a demonstrated qualitative result (ventilation worked) without the quantitative threshold that would let the result be specified.

What this does to a building-scale project. For an owner placing 50 kWh to 1 MWh of lithium-ion inside an occupied building, the reading is that the suppression system, the gas detection system and the ventilation system form a single scope of work that has to be priced together rather than line by line. NFPA 855’s 2026 edition requires a hazard mitigation analysis for nearly every installation and leans on NFPA 69 explosion prevention, which puts the gas question in front of the specifier regardless of which suppression system is chosen. The SafeBESS tests are the first full-room data set that addresses suppression performance and gas accumulation together, and the airflow number that would complete the ventilation side of that scope is still outstanding.

Where this leaves the certification argument. Vendor fire-test disclosure in the commercial storage market is built around the single enclosure or the cabinet array, which is the unit a manufacturer controls and can certify. UL 9540A large-scale test reports describe what happens to and around a product. The room, its ventilation rate, its hold time and the interaction between the two belong to the building. A cabinet-level test result transfers into a specific room only with a project-specific analysis behind it, which is the gap a hazard mitigation analysis under NFPA 855 is meant to close.

That distinction matters commercially as the sixth edition of UL 9540A approaches its mandatory date in January 2027 and vendors begin publishing margin-against-limit numbers: wall temperature rise, ceiling steel-beam temperature, adjacent-module voltage after an event. Those are cabinet-level and array-level figures. They answer a different question from the one SafeBESS was built to answer, which is what the room itself has to do.

What remains unpublished. SINTEF states that the SafeBESS guidelines for fire safety of battery rooms in buildings are being finalised. Until they appear, the citable material is the test account itself: four suppression systems compared in a single room, one of them stopping propagation, gas accumulating regardless, ventilation preventing an explosion at an airflow rate the project has not yet quantified.

For buyers specifying indoor storage before those guidelines publish, the practical consequence is that the explosion-prevention and ventilation portions of an indoor battery room design currently rest on engineering judgment rather than on a published threshold, and that judgment has to be documented in the hazard mitigation analysis that the code already requires.


Sources