The battery energy storage system (BESS) fire protection market is heading into a multi-year build cycle: SNS Insider values the 2025 base at USD 5.30B with a 10.50% CAGR to USD 14.40B by 2035 [S5], Pragma Market Research places 2025 at USD 6.8B with a 10.9% CAGR through 2031 [S3], and DataIntelo reports USD 3.8B in 2025 scaling to USD 9.1B by 2034 [S4].
IDTechEx widens the lens to the combined thermal management and fire protection envelope, projecting the segment to exceed USD 25B by 2036 at a 12.4% CAGR and to absorb more than 20% of total BESS project cost by 2035 [S2]. Drivers are converging: NFPA 855 and UL 9540A are now underwriting prerequisites, lithium-ion off-gas behaviour defeats conventional water-based systems, and pre-fire multi-gas detection is moving specification earlier in the engineering cycle.
Why lithium-ion BESS fires break conventional suppression logic
US EPA guidance states plainly that lithium battery fires are extremely difficult to extinguish and may reignite hours or days after the initial event, while off-gases from a venting cell create health hazards for first responders long before any visible flame [S1]. The Gateway Energy Storage fire in San Diego on 15 May 2024 flared repeatedly for seven days inside a 15,000-cell NMC lithium-ion container, and the Moss Landing, California incident on 16 January 2025 triggered a 24-hour evacuation of about 1,200 residents [S1]. The US EPA also notes that, despite headline events, BESS failure incidents per gigawatt-hour deployed have trended down since 2020 as cell quality and design improved [S1].
The behaviour that defeats legacy agents is a self-sustaining exothermic chain reaction inside the cell, propagating module-to-module once a neighbouring cell vents; once that propagation is established, the standard fire triangle model no longer applies, and the suppression job shifts from extinguishing visible flame to interrupting cell-to-cell heat transfer and starving the released electrolyte vapour of oxygen [S3][S5]. Stat-X's whitepaper on BESS suppression lays out the same problem from the agent vendor side, framing suppression strategy around aerosol, clean-agent, and water-mist options matched to the BESS hazard profile [S8].
Market sizing: which 2025 base and which CAGR actually holds up
Three published 2025 base values diverge because the analysts scope the segment differently. DataIntelo counts USD 3.8B in 2025 with a 10-year climb to USD 9.1B by 2034 [S4]. SNS Insider counts USD 5.30B in 2025 to USD 14.40B by 2035 at 10.50% CAGR, with the US alone at USD 1.26B in 2025 heading to USD 3.43B by 2035 at 11.80% CAGR [S5]. Pragma Market Research counts USD 6.8B in 2025 with 10.9% CAGR through 2031, anchored to NFPA 855 and UL 9540A adoption [S3].
For engineering spec planning, the most useful reference is IDTechEx's combined thermal management and fire protection number: more than 20% of total BESS project cost by 2035, dominated by active system value (liquid cooling loops, sensors, fire suppressants, venting), while passive materials such as mica and ceramic barriers continue to provide the proven low-cost flame-retardant baseline [S2]. Foton Energy's strategic guide echoes the cost-share projection, flagging that thermal management and fire protection will exceed 20% of total BESS cost by 2035 and that older sites are being retrofitted as insurance underwriters tighten [S6].
Spec map: NFPA 855, UL 9540A, and what each actually requires
NFPA 855 governs the siting, spacing, and maximum stored energy of stationary storage installations in the US, while UL 9540A is the test method that defines cell-, module-, unit-, and installation-level fire propagation behaviour and underwrites the suppression agent choice [S3][S5]. SNS Insider specifically names NFPA 855 (2023 revisions) and UL 9540A as the certification anchors driving suppression and detection demand in 2026 [S5], and Pragma Market Research describes UL 9540 as a de facto baseline for system-level certification [S3].
For BESS fire detection and suppression specification, the working stack is multi-gas electrochemical sensing for early off-gas detection, thermal imaging or heat-flux monitoring at the rack level, and a primary suppression agent chosen from clean agents, high-pressure water mist, or inert gas systems, all wired into the Battery Management System and SCADA layer for automated shutdown and isolation [S3][S5]. For broader BESS hazard framing, our fire safety reference lays out the regulatory layer that the BESS segment sits inside. Where the chosen agent is a water-mist or clean-agent system, the rest of the suppression envelope follows standard gas fire suppression engineering practice, scaled to the BESS cubic volume and cell count.
Agent comparison: clean agent vs water mist vs aerosol vs immersion
No single agent covers every BESS failure mode, and the choice is set by UL 9540A test results, not by marketing literature. Clean agents (Novec 1230, FM-200 class chemistries) are effective on the incipient flame stage and leave no residue, but require tight envelope integrity and do not cool the cell mass. High-pressure water mist provides the best sustained cooling duty on a propagating rack fire, addressing the re-ignition problem EPA flags, but adds weight, floor load, and water damage exposure that some operators refuse in indoor C&I sites [S3][S8]. Aerosol-generating compounds (Stat-X class systems) deliver condensed-phase potassium-based knockdown on a small footprint, and are commonly used as a complementary agent inside enclosures rather than the primary bulk-volume system [S8].
Immersion cooling fluids, originally developed for thermal management, are being re-positioned as a suppression layer because the dielectric fluid both removes heat and physically excludes oxygen from the cell, attacking the runaway condition at the chemistry layer rather than the flame layer; Etica Ag's 2025 review documents this dual-use trajectory [S7]. Across the four options, the criteria an engineer actually ranks are: cell-cooling duty (W per cell), oxygen displacement for the chosen enclosure volume, UL 9540A test outcome for the specific cell format, weight and floor-load for indoor sites, and water or residue tolerance for the asset owner's operations team. By those criteria, water mist leads on re-ignition control, clean agents lead on footprint and asset cleanliness, immersion cooling leads on thermal-runaway prevention, and aerosol leads on retrofit simplicity inside existing enclosures.
Who BESS fire protection is for, and who should not retrofit
Mandatory specification applies to any new US grid-scale BESS subject to NFPA 855, any installation seeking UL 9540 system listing for financing or insurance, and any European site falling under the evolving EU BESS regulatory framework that IDTechEx flags as still fragmented relative to the US baseline [S2]. C&I sites between 250 kW and 1 MW typically fall under local fire code adaptations of NFPA 855 with spacing and sprinkler requirements, and benefit most from pre-fire gas detection retrofits that satisfy insurer monitoring demands [S3][S5].
Residential battery storage below the threshold fire codes define as BESS (typically single-family wall-mount or rack-mount systems under NFPA 855's residential limits) is generally not in scope for the full suppression architecture; for these sites, the reference point is the broader fire extinguisher selection logic for lithium-ion incidents. Existing grid-scale installations that pre-date NFPA 855 2023 are the retrofit pool, and SNS Insider names older BESS sites as the recurring-revenue segment for the next several years [S5].
Detection, controls, and BMS integration
The 2026 specification direction is unambiguous: pre-fire gas detection is moving upstream of suppression activation. Multi-gas electrochemical sensors tuned for the specific off-gas signature of lithium-ion cell venting (hydrocarbon solvents, CO, H2, HF where present) trigger at sub-visible-venting conditions, and SNS Insider flags AI-driven pre-fire gas detection as the dominant US demand driver for 2026 [S5]. Pragma Market Research lists multi-gas early-detection sensors, thermal imaging or heat-flux monitoring, and automated response platforms integrated with the Battery Management System and SCADA as the complete modern BESS detection stack [S3].
The fire safety control point on a BESS site still has to land somewhere readable by a human during an incident, which is where the broader detection and notification discipline described in our electrical fire monitor reference comes in. For a BESS site, that monitor is increasingly an addressable panel receiving inputs from the multi-gas sensor array, the BMS state-of-health stream, and the thermal cameras, with hard-wired output to a manual release station and to the fire department connection so first responders see the same display EPA recommends communities coordinate with [S1].
Failure modes, limits, and what to plan around
Three failure modes drive most BESS loss events and each maps to a different suppression limitation. Cell-internal short circuits from manufacturing defects or dendrite growth initiate runaway that cannot be suppressed at the gas-phase level; only cell-level thermal management (immersion, sufficient cell spacing, ceramic barriers) and the BMS cell-balancing function can lower the probability, which is why mica and ceramic passive barriers remain a baseline material demand [S2]. Mechanical damage from shipping, seismic events, or enclosure compromise initiates runaway at the module level, where only water mist or full-flooding immersion provides the heat-removal duty to prevent module-to-module propagation [S3][S7].
Overcharge or external fire exposure initiates runaway at the rack level and is the case where the standard fire door and fire hydrant discipline remains relevant: containment boundaries and firefighter water supply for the surrounding structure reduce the probability of external fire reaching the BESS enclosure. EPA also flags the water-management problem downstream: a BESS fire can release gases, contaminated suppression water, and damaged cells that require specialised disposal, and the agency required the Gateway site to run extensive environmental monitoring during battery handling [S1]. The implication for spec writers is that any BESS water-mist or sprinkler system must be paired with a contaminated-water containment plan, because the firewater run-off itself is a regulated waste stream.
Insurance underwriting bulletins from major carriers will continue to set the practical floor for pre-fire gas detection, since that requirement is now appearing as an exclusion clause in property policies for sites without certified detection [S3][S5].
See also our earlier report, Conductivity vs pH for strong-acid concentration control.