Lithium-ion battery fire - Revision history

Jan: Remove links to pages that do not exist; retain displayed text (wiki-wide link audit).

2026-11-02T21:34:03Z

Remove links to pages that do not exist; retain displayed text (wiki-wide link audit).

← Older revision		Revision as of 22:34, 2 November 2026
Line 76:		Line 76:
	How often electric cars catch fire is largely unknown, because there are no reliable datasets.<ref name="NIPV-54">Van der Graaf et al. (2021/2023), p. 54.</ref> The report judges that a parked electric car is unlikely to catch fire more often than a fossil-fuelled car. Charging, however, is expected to raise the probability, because it adds an activity, the technology is still in its teething phase and existing electrical installations may not be safe enough.<ref name="NIPV-28">Van der Graaf et al. (2021/2023), p. 28.</ref> Ageing probably raises the internal resistance of the cells, which increases the heat they produce. An older battery also holds a lower maximum charge, however, so its thermal runaway is less intense and releases fewer flammable gases.<ref name="NIPV-28" />		How often electric cars catch fire is largely unknown, because there are no reliable datasets.<ref name="NIPV-54">Van der Graaf et al. (2021/2023), p. 54.</ref> The report judges that a parked electric car is unlikely to catch fire more often than a fossil-fuelled car. Charging, however, is expected to raise the probability, because it adds an activity, the technology is still in its teething phase and existing electrical installations may not be safe enough.<ref name="NIPV-28">Van der Graaf et al. (2021/2023), p. 28.</ref> Ageing probably raises the internal resistance of the cells, which increases the heat they produce. An older battery also holds a lower maximum charge, however, so its thermal runaway is less intense and releases fewer flammable gases.<ref name="NIPV-28" />

	The NIPV report covers only lithium-ion passenger cars in Dutch indoor car parks, and "lithium-ion" itself covers many kinds of lithium batteries that respond differently to fire.<ref name="NIPV-18" /> It warns that test results vary greatly with the amount of fuel, the battery capacity and the plastics in the car, so single studies cannot simply be generalised.<ref name="NIPV-20" /> Battery packs can also be combined into systems of 10 to 400 kWh, for example in trucks and electric ships. Stationary installations such as a [[battery energy storage system]] lie outside the report's scope.<ref name="NIPV-21" /><ref name="NIPV-18" /> Open questions include the exact process of thermal runaway, the conditions under which packs re-ignite and the best detection and suppression systems.<ref name="NIPV-55">Van der Graaf et al. (2021/2023), p. 55.</ref>		The NIPV report covers only lithium-ion passenger cars in Dutch indoor car parks, and "lithium-ion" itself covers many kinds of lithium batteries that respond differently to fire.<ref name="NIPV-18" /> It warns that test results vary greatly with the amount of fuel, the battery capacity and the plastics in the car, so single studies cannot simply be generalised.<ref name="NIPV-20" /> Battery packs can also be combined into systems of 10 to 400 kWh, for example in trucks and electric ships. Stationary installations such as a battery energy storage system lie outside the report's scope.<ref name="NIPV-21" /><ref name="NIPV-18" /> Open questions include the exact process of thermal runaway, the conditions under which packs re-ignite and the best detection and suppression systems.<ref name="NIPV-55">Van der Graaf et al. (2021/2023), p. 55.</ref>

	== See also ==		== See also ==

Mark: Revise against the FSE wiki writing guidelines: corrections after independent fact-check (revised by Claude Opus 5.5)

2026-11-02T18:32:41Z

Revise against the FSE wiki writing guidelines: corrections after independent fact-check (revised by Claude Opus 5.5)

Mark: Revise against the FSE wiki writing guidelines: verified and corrected citations, added sources, See also, links, style (revised by Claude Opus 5.5)

2026-11-02T17:52:24Z

Revise against the FSE wiki writing guidelines: verified and corrected citations, added sources, See also, links, style (revised by Claude Opus 5.5)

Show changes

Sanne: Add diagrams fse-lithium-ion-battery-fire-v1.svg, fse-lithium-ion-battery-fire-v2.svg, fse-lithium-ion-battery-fire-v3.svg (level-1 visual plan)

2026-11-02T07:24:48Z

Add diagrams fse-lithium-ion-battery-fire-v1.svg, fse-lithium-ion-battery-fire-v2.svg, fse-lithium-ion-battery-fire-v3.svg (level-1 visual plan)

← Older revision		Revision as of 08:24, 2 November 2026
Line 22:		Line 22:

	The heat from one cell can raise the temperature of adjacent cells and cause a chain reaction. The geometry of the pack and the ease of heat transfer influence that propagation. A fire involving one cell can therefore develop into a much larger battery-pack event.<ref name="NIPV-22" />		The heat from one cell can raise the temperature of adjacent cells and cause a chain reaction. The geometry of the pack and the ease of heat transfer influence that propagation. A fire involving one cell can therefore develop into a much larger battery-pack event.<ref name="NIPV-22" />
			[[File:fse-lithium-ion-battery-fire-v1.svg\|thumb\|upright=1.5\|alt=One hot cell heats neighbours, releasing gases that accumulate and can ignite or explode.\|Thermal runaway can propagate and produce flammable gases. Based on Van der Graaf, Hessels, and Rosmuller (2021), p. 22 and p. 27.]]

	Pressure can build in a pouch, cartridge or other enclosure and release gases if the enclosure or pressure valve fails. Mechanical damage can also eject burning parts and fragments from the pack. Those hazards differ from a fire that spreads only by ordinary surface flame.<ref name="NIPV-22" />		Pressure can build in a pouch, cartridge or other enclosure and release gases if the enclosure or pressure valve fails. Mechanical damage can also eject burning parts and fragments from the pack. Those hazards differ from a fire that spreads only by ordinary surface flame.<ref name="NIPV-22" />
Line 46:		Line 47:

	Hydrogen fluoride is a particular concern in the cited vehicle research. NIPV reports a source estimate of about 20 to 200 mg HF per Wh, implying 2 to 20 kg for a 100 kWh pack depending on state of charge. Another cited comparison found about 1.8 times as much HF for electric vehicles as for conventional vehicles.<ref name="NIPV-26" />		Hydrogen fluoride is a particular concern in the cited vehicle research. NIPV reports a source estimate of about 20 to 200 mg HF per Wh, implying 2 to 20 kg for a 100 kWh pack depending on state of charge. Another cited comparison found about 1.8 times as much HF for electric vehicles as for conventional vehicles.<ref name="NIPV-26" />
			[[File:fse-lithium-ion-battery-fire-v2.svg\|thumb\|upright=1.5\|alt=Battery smoke travels through air and deposits on clothing, walls, floors and residue.\|Battery-fire contamination includes airborne and deposited hazards. Based on Van der Graaf, Hessels, and Rosmuller (2021), pp. 26–27 and pp. 39–40.]]

	Those numbers belong to the cited tests and assumptions and must not be used as a universal emission factor. Battery chemistry, state of charge and test arrangement affect releases. The report itself emphasises the differences between studies and the uncertainties in those results.<ref name="NIPV-26" /><ref name="NIPV-28" />		Those numbers belong to the cited tests and assumptions and must not be used as a universal emission factor. Battery chemistry, state of charge and test arrangement affect releases. The report itself emphasises the differences between studies and the uncertainties in those results.<ref name="NIPV-26" /><ref name="NIPV-28" />
Line 56:		Line 58:

	Putting out visible flames is different from stopping thermal runaway. The battery's closed construction can prevent water from reaching the cells directly, so surface application may not stop the underlying reactions. NIPV describes prolonged cooling with substantial water as necessary for control in its studied scenarios.<ref name="NIPV-22" /><ref name="NIPV-30">Van der Graaf, Hessels, and Rosmuller (2021), p. 30.</ref>		Putting out visible flames is different from stopping thermal runaway. The battery's closed construction can prevent water from reaching the cells directly, so surface application may not stop the underlying reactions. NIPV describes prolonged cooling with substantial water as necessary for control in its studied scenarios.<ref name="NIPV-22" /><ref name="NIPV-30">Van der Graaf, Hessels, and Rosmuller (2021), p. 30.</ref>
			[[File:fse-lithium-ion-battery-fire-v3.svg\|thumb\|upright=1.5\|alt=Flames are knocked down before prolonged cooling, temperature monitoring and salvage.\|Battery-fire response continues after visible flames disappear. Based on Van der Graaf, Hessels, and Rosmuller (2021), pp. 45–47.]]

	Water can enter the pack only where the design provides access or where another method is used. NIPV describes immersion of the vehicle or battery pack as the currently successful approach in its cited fire-service guidance. Effectiveness depends on water reaching the affected cells and on continued monitoring afterwards.<ref name="NIPV-30" /><ref name="NIPV-46" />		Water can enter the pack only where the design provides access or where another method is used. NIPV describes immersion of the vehicle or battery pack as the currently successful approach in its cited fire-service guidance. Effectiveness depends on water reaching the affected cells and on continued monitoring afterwards.<ref name="NIPV-30" /><ref name="NIPV-46" />

Jan: New FSE wiki article: Lithium-ion battery fire (written by Kimi Code CLI)

2026-11-02T04:18:47Z

New FSE wiki article: Lithium-ion battery fire (written by Kimi Code CLI)

New page

A '''Lithium-ion battery fire''' is a fire involving the cells, electrolyte, packaging or surrounding materials of a lithium-ion battery or battery pack. It can be driven or sustained by thermal runaway, a failure mechanism that heats the cell and can propagate to neighbouring cells. Flammable gases, toxic combustion products, re-ignition and difficult access to the cells make these fires different from many ordinary combustible-material fires.<ref name="NIPV-22">Van der Graaf, Hessels, and Rosmuller (2021), p. 22.</ref><ref name="NUREG-B15">U.S. Nuclear Regulatory Commission (2005), p. B-15.</ref>

Battery packs are often sealed to protect them from moisture and dirt, which also makes direct cooling of the cells difficult. Flame extinction alone may not stop the underlying reactions, so repeated ignition can occur. Continued cooling, temperature monitoring and management of contaminated water and residue are therefore part of the response.<ref name="NIPV-22" /><ref name="NIPV-45">Van der Graaf, Hessels, and Rosmuller (2021), pp. 45–46.</ref>

These fires can occur in vehicles, energy-storage installations and other equipment, but their development depends on cell type, state of charge, packaging and installation. Evidence from electric-vehicle studies should not be applied unchanged to every stationary system. The particular battery, enclosure and fire scenario must be assessed.<ref name="NIPV-28">Van der Graaf, Hessels, and Rosmuller (2021), pp. 27–28.</ref><ref name="NIPV-33">Van der Graaf, Hessels, and Rosmuller (2021), pp. 32–33.</ref>

== Thermal runaway and initiation ==

[[Thermal runaway]] is a failure mechanism that causes self-heating in a battery or cell and can result in fire. Each battery pack has a safety window, and exceeding it can lead to further heating. The reported onset range of 70 to 250 °C depends on battery type and is not a universal ignition temperature.<ref name="NIPV-22" />

The mechanism can begin with design or manufacturing errors involving electrodes, separators, electrolyte or related processes. External heat, misuse, collision damage and charging faults can also contribute. A battery management system is intended to monitor abnormalities and intervene, but its presence does not remove every failure path.<ref name="NIPV-22" /><ref name="NIPV-28" />

The battery management system can support active cooling, stop charging or disconnect the pack when threshold conditions are detected. Its functions depend on the condition of the system and the fault being experienced. NIPV describes fail-safe deactivation for its discussed electric-vehicle packs, not as a guarantee for every battery system.<ref name="NIPV-22" /><ref name="NIPV-28" />

Ageing changes battery behaviour. Increased internal resistance can increase heating, while a lower maximum state of charge can reduce the intensity and gas production of a later thermal runaway. Those effects should not be reduced to a simple claim that every older battery is either more or less dangerous.<ref name="NIPV-28" />

A [[Battery energy storage system]] can contain many cells whose arrangement, monitoring and enclosure affect propagation and cooling. The supplied NIPV source addresses electric-vehicle packs rather than providing a complete stationary-storage design standard. No universal stationary-installation rule is therefore stated here.<ref name="NIPV-28" /><ref name="NIPV-33" />

== Gas release and propagation ==

A cell undergoing thermal runaway can release flammable gases as it heats. Those gases may ignite if sufficient oxygen and a suitable ignition condition are present. If gases accumulate in a confined or poorly ventilated space, they can produce a flammable or explosive mixture.<ref name="NIPV-22" /><ref name="NIPV-27">Van der Graaf, Hessels, and Rosmuller (2021), p. 27.</ref>

The heat from one cell can raise the temperature of adjacent cells and cause a chain reaction. The geometry of the pack and the ease of heat transfer influence that propagation. A fire involving one cell can therefore develop into a much larger battery-pack event.<ref name="NIPV-22" />

Pressure can build in a pouch, cartridge or other enclosure and release gases if the enclosure or pressure valve fails. Mechanical damage can also eject burning parts and fragments from the pack. Those hazards differ from a fire that spreads only by ordinary surface flame.<ref name="NIPV-22" />

Unignited gas does not make the situation safe. Continued degassing can create a cloud that later ignites, and an explosion can occur when the mixture and ignition conditions are suitable. Assessment must therefore include gas transport, ventilation and the timing of ignition.<ref name="NIPV-27" />

An [[Electric vehicle fire]] can involve the battery pack, passenger compartment, charging equipment or an external fire exposing the vehicle. NIPV presents separate scenarios for a pack fire, radiation exposure from another vehicle and several packs burning. Those scenarios have different prevention and suppression requirements.<ref name="NIPV-33" />

== Fire development and re-ignition ==

Battery-pack fire development can differ from a liquid-fuel vehicle fire. NIPV describes a slower initial heating period followed by rapid development when thermal runaway occurs. The source compares the pack with hydrocarbon-fuel fire behaviour, not with every possible vehicle interior fire.<ref name="NIPV-25">Van der Graaf, Hessels, and Rosmuller (2021), pp. 25–26.</ref>

The source reports a conventional passenger-car fire of just over half an hour with about 5 MW after 10 to 15 minutes, and an electric-vehicle burnout of about 90 minutes without intervention. These are study-specific examples, not universal fire curves. Vehicle materials, battery condition and ventilation can change the outcome.<ref name="NIPV-25" />

Re-ignition is a defining concern in many battery-pack scenarios. The reactions can continue after visible flames are suppressed, releasing gases that ignite again. NIPV reports re-ignition for several hours after the start in the cited study, which requires a longer assessment than initial flame knockdown.<ref name="NIPV-25" /><ref name="NIPV-45" />

Continuous temperature monitoring helps identify renewed heating and the need for continued cooling. Extinguishing whenever the pack reignites and applying continuous cooling are different operational choices. Both require sustained attention to the battery's temperature and remaining energy.<ref name="NIPV-45" /><ref name="NIPV-46">Van der Graaf, Hessels, and Rosmuller (2021), p. 46.</ref>

The fire can spread through vehicles to other cars and structures. Radiant heat from an adjacent fire can initiate thermal runaway in a pack that was not the original ignition source. Prevention and suppression must therefore consider both internal failure and external fire exposure.<ref name="NIPV-33" />

== Toxic and corrosive products ==

[[Fire toxicity]] includes the effects of gases and other products released by the cells, electrolyte, packaging and surrounding plastics. NIPV lists methane, carbon monoxide, carbon dioxide, hydrogen, hydrogen fluoride and hydrogen chloride among the products in its reviewed battery fires. The actual mixture depends on the battery and the fire.<ref name="NIPV-26">Van der Graaf, Hessels, and Rosmuller (2021), pp. 26–27.</ref>

Hydrogen fluoride is a particular concern in the cited vehicle research. NIPV reports a source estimate of about 20 to 200 mg HF per Wh, implying 2 to 20 kg for a 100 kWh pack depending on state of charge. Another cited comparison found about 1.8 times as much HF for electric vehicles as for conventional vehicles.<ref name="NIPV-26" />

Those numbers belong to the cited tests and assumptions and must not be used as a universal emission factor. Battery chemistry, state of charge and test arrangement affect releases. The report itself emphasises the differences between studies and the uncertainties in those results.<ref name="NIPV-26" /><ref name="NIPV-28" />

HF is reactive and can bind to other substances and particles, forming deposits on floors, walls and ceilings. NIPV reports that this reduced airborne HF concentration in the cited work but increased contamination of firefighting clothing. Deposition therefore changes rather than eliminates the hazard.<ref name="NIPV-26" />

Combustion products can spread into stairwells and escape routes through open doors or building openings. Location and ventilation affect where exposure occurs. The safety assessment must include smoke movement and people who take longer to leave, not only firefighters at the battery.<ref name="NIPV-39">Van der Graaf, Hessels, and Rosmuller (2021), pp. 39–40.</ref>

== Suppression and cooling ==

Putting out visible flames is different from stopping thermal runaway. The battery's closed construction can prevent water from reaching the cells directly, so surface application may not stop the underlying reactions. NIPV describes prolonged cooling with substantial water as necessary for control in its studied scenarios.<ref name="NIPV-22" /><ref name="NIPV-30">Van der Graaf, Hessels, and Rosmuller (2021), p. 30.</ref>

Water can enter the pack only where the design provides access or where another method is used. NIPV describes immersion of the vehicle or battery pack as the currently successful approach in its cited fire-service guidance. Effectiveness depends on water reaching the affected cells and on continued monitoring afterwards.<ref name="NIPV-30" /><ref name="NIPV-46" />

[[Fire classes]] can guide initial agent selection, but a battery fire does not fit every assumption behind an ordinary solid or electrical-equipment label. Chemical reactions, sealed packaging, high voltage, gas release and re-ignition must also be assessed. Class selection cannot replace mechanism-specific evidence.<ref name="NUREG-B16">U.S. Nuclear Regulatory Commission (2005), p. B-16.</ref><ref name="NIPV-22" />

Sprinklers and water mist can reduce the probability of spread to adjacent structures and vehicles, but NIPV says they cannot necessarily extinguish the pack fire. Further research is needed on which systems best control different electric-vehicle scenarios. Their effect should not be equated with stopping thermal runaway.<ref name="NIPV-36">Van der Graaf, Hessels, and Rosmuller (2021), p. 36.</ref>

Covering a vehicle in foam is not supported as a useful method in the cited NIPV discussion because re-ignition can occur under the foam, even with limited oxygen. That is specific evidence about the battery mechanism, not a general claim that foam has no other role in the incident.<ref name="NIPV-47">Van der Graaf, Hessels, and Rosmuller (2021), p. 47.</ref>

== Firefighting and salvage ==

Firefighting assessment must include explosions, toxic gases, high voltage, structure stability and uncertain fire development. NIPV identifies those factors as increasing the risks and uncertainties of a repressive attack. The presence of a battery also extends the period for which cooling and monitoring may be required.<ref name="NIPV-30" /><ref name="NIPV-43">Van der Graaf, Hessels, and Rosmuller (2021), pp. 43–45.</ref>

The attack location and cooling capacity affect which tactics are possible. NIPV's principles ask whether the seat of fire is known, whether it can be reached from outside and whether sufficient cooling capacity exists. The answers determine whether an offensive exterior attack is possible.<ref name="NIPV-44">Van der Graaf, Hessels, and Rosmuller (2021), pp. 44–45.</ref>

NIPV gives low-pressure guidance of 450 litres per minute generally sufficient for a heat-release rate up to 10 MW and high-pressure guidance of 125 litres per minute for up to 2.5 MW. These are source-specific rules of thumb, not universal cooling guarantees for battery packs. The report separately stresses prolonged cooling and access to the cells.<ref name="NIPV-45" />

Salvage planning should begin early. A vehicle may need to be removed to a suitable container or tank where cooling can continue until thermal runaway stops. The capabilities of the salvage company and the access conditions in the car park can limit that option.<ref name="NIPV-42">Van der Graaf, Hessels, and Rosmuller (2021), p. 42.</ref><ref name="NIPV-46" />

Firefighting clothing and equipment can become contaminated by HF and other products. NIPV calls for decontamination in its electric-vehicle discussion and notes that this differs from conventional-vehicle incidents. Protective arrangements must address both airborne and deposited contamination.<ref name="NIPV-26" /><ref name="NIPV-30" />

== Prevention, detection and management ==

Prevention includes battery and charging-equipment condition, user instructions and the condition of the electrical installation. NIPV identifies damaged cables and charging stations, battery-management faults, collisions with charging equipment and unnoticed defective packs as additional causes. Those conditions must be considered in the building's risk assessment.<ref name="NIPV-28" />

Clear identification of electric vehicles and charging locations supports response. The source recommends recognisable charging spaces and information at entrances, while warning that clustering can increase travelling-fire exposure. Identification and spacing must therefore be assessed together.<ref name="NIPV-42" />

Early detection supports suppression and evacuation. NIPV describes smoke, heat and carbon-monoxide detection options and stresses prompt alarm follow-up. It also notes that local standardisation and certification may not treat carbon-monoxide detection as a standard fire-detection method.<ref name="NIPV-36" /><ref name="NIPV-40">Van der Graaf, Hessels, and Rosmuller (2021), p. 40.</ref>

User instructions should address fault messages, distance from released gases, stopping charging, alerting the fire service and providing vehicle information. Public and private car parks require different communication arrangements. Instructions must match the people who actually use the space.<ref name="NIPV-40" />

Runoff and residue require management. NIPV distinguishes cooling water in an immersion tank, which may be contaminated and must not go directly into the sewer, from other research on ordinary extinguishing water. The disposal route must be established for the actual water and residue.<ref name="NIPV-46" />

== Assessment and evidence limits ==

Battery fires require scenario-specific assessment rather than a single fire curve. Cell type, state of charge, pack design, external heat, ventilation and fire-service access can all alter the event. Evidence from one pack or vehicle should be compared carefully with the system being evaluated.<ref name="NIPV-28" /><ref name="NIPV-33" />

The likelihood of battery failure is itself uncertain. NIPV reports differing estimates and warns that technologies are still developing. It does not support a general claim that electric vehicles always have a higher or lower fire probability than conventional vehicles.<ref name="NIPV-22" /><ref name="NIPV-28" />

Suppression evidence is also incomplete. The report calls for research into which fixed systems best control incipient electric-vehicle fires and how travelling fires differ between vehicle types. A suppression measure that limits spread cannot be assumed to stop the pack's reactions.<ref name="NIPV-36" /><ref name="NIPV-42" />

Toxicity and environmental assessments must distinguish air, deposition, residue and contaminated water. Each can contain hazardous products, and a neutral surface pH reading does not prove the residue is free of hazardous substances. The complete exposure and disposal path matters.<ref name="NIPV-26" /><ref name="NIPV-46" />

The incident record should preserve those uncertainties alongside the chosen response. It should identify the pack, fire origin, gas and smoke pathways, cooling method, monitoring period and disposal arrangements. That gives a more accurate basis for judging both immediate and later consequences. It also shows which conclusions depend on a particular pack, test or operational assumption and therefore require renewed assessment when those conditions change.<ref name="NIPV-30" /><ref name="NIPV-46" />

== See also ==

* [[Battery energy storage system]]
* [[Electric vehicle fire]]
* [[Thermal runaway]]

== Notes ==

<references />

== References ==

* U.S. Nuclear Regulatory Commission. (2005). ''Fire dynamics tools: Quantitative fire hazard analysis methods for the U.S. nuclear regulatory commission fire protection inspection program'' (NUREG-1805). Author.
* Van der Graaf, J., Hessels, T., & Rosmuller, N. (2021). ''Fire safety of indoor car parks accommodating electrically powered vehicles''. Netherlands Institute for Public Safety (NIPV). English translation published 2023.

[[Category:Special Hazards & Building Types]]

← Older revision		Revision as of 19:32, 2 November 2026
Line 1:		Line 1:
	A '''lithium-ion battery fire''' is a fire in a lithium-ion battery or battery pack that is driven by [[thermal runaway]]. Thermal runaway is a failure mechanism in which a battery cell heats itself up, and it can result in fire.<ref name="NIPV-22">Van der Graaf et al. (2021/2023), p. 22.</ref> Many electric cars store their energy in lithium-ion packs, which can hold 100 kWh or more.<ref name="NIPV-21">Van der Graaf et al. (2021/2023), p. 21.</ref> Once thermal runaway has started, heat ~~passes~~ from cell to cell, flammable gases ~~are~~ released and the fire keeps re-igniting until the energy in the pack is used up. The pack is sealed against moisture and dirt, so water cannot reach the cells directly and cooling needs a lot of water for a long time.<ref name="NIPV-22" />		A '''lithium-ion battery fire''' is a fire in a lithium-ion battery or battery pack that is driven by [[thermal runaway]]. Thermal runaway is a failure mechanism in which a battery cell heats itself up, and it can result in fire.<ref name="NIPV-22">Van der Graaf et al. (2021/2023), p. 22.</ref> Many electric cars store their energy in lithium-ion packs, which can hold 100 kWh or more.<ref name="NIPV-21">Van der Graaf et al. (2021/2023), p. 21.</ref> Once thermal runaway has started, heat can pass from cell to cell, flammable gases may be released and the fire keeps re-igniting until the energy in the pack is used up. The pack is sealed against moisture and dirt, so water cannot reach the cells directly and cooling needs a lot of water for a long time.<ref name="NIPV-22" />

	In tests, burning electric cars reached temperatures and peak heat release rates similar to those of petrol and diesel cars.<ref name="NIPV-23">Van der Graaf et al. (2021/2023), p. 23.</ref><ref name="NIPV-25">Van der Graaf et al. (2021/2023), pp. 24–25.</ref> Their fires nevertheless develop more slowly, last longer, can involve repeated small explosions and flash fires, and release more hydrogen fluoride, a toxic and corrosive gas.<ref name="NIPV-29">Van der Graaf et al. (2021/2023), p. 29.</ref><ref name="NIPV-30">Van der Graaf et al. (2021/2023), p. 30.</ref> Together with high voltages, these risks are one reason why firefighting in indoor car parks may be limited or impossible.<ref name="NIPV-11">Van der Graaf et al. (2021/2023), p. 11.</ref> A 2021 study by the Netherlands Institute for Public Safety (NIPV) on indoor car parks with electric vehicles found that the probability of such fires could hardly be determined from empirical data.<ref name="NIPV-20">Van der Graaf et al. (2021/2023), p. 20.</ref>		In tests, burning electric cars reached temperatures and peak heat release rates similar to those of petrol and diesel cars.<ref name="NIPV-23">Van der Graaf et al. (2021/2023), p. 23.</ref><ref name="NIPV-25">Van der Graaf et al. (2021/2023), pp. 24–25.</ref> Their fires nevertheless develop more slowly, last longer, can involve repeated small explosions and flash fires, and release more hydrogen fluoride, a toxic and corrosive gas.<ref name="NIPV-29">Van der Graaf et al. (2021/2023), p. 29.</ref><ref name="NIPV-30">Van der Graaf et al. (2021/2023), p. 30.</ref> Together with high voltages, these risks are one reason why firefighting in indoor car parks may be limited or impossible.<ref name="NIPV-11">Van der Graaf et al. (2021/2023), p. 11.</ref> A 2021 study by the Netherlands Institute for Public Safety (NIPV) on indoor car parks with electric vehicles found that the probability of such fires could hardly be determined from empirical data.<ref name="NIPV-20">Van der Graaf et al. (2021/2023), p. 20.</ref>
Line 24:		Line 24:
	In terms of temperature, fires in electric cars do not differ significantly from those in conventional cars. Battery tests reviewed in the NIPV report measured maximum flame temperatures of 798 to 863 °C, rising with how fully the batteries were charged. Full-scale tests on both types of car measured 800 ± 100 °C.<ref name="NIPV-23" /> The heat radiation from an electric car with a fully charged pack was also similar to that from a conventional car.<ref name="NIPV-25" />		In terms of temperature, fires in electric cars do not differ significantly from those in conventional cars. Battery tests reviewed in the NIPV report measured maximum flame temperatures of 798 to 863 °C, rising with how fully the batteries were charged. Full-scale tests on both types of car measured 800 ± 100 °C.<ref name="NIPV-23" /> The heat radiation from an electric car with a fully charged pack was also similar to that from a conventional car.<ref name="NIPV-25" />

	The heat release rate, the amount of heat a fire releases per second, largely determines how intense a fire is.<ref name="VR-4.5">Van Rossum (2027), §4.5.</ref> A review of published tests found similar peaks of about 6 MW for electric and fossil-fuelled cars. The total heat released was 4.7 to 8.5 GJ for electric cars and 3.3 to 10 GJ for the others.<ref name="NIPV-25" /> Larger batteries raise the peak: packs of about 15 to 25 kWh gave about 4 to 6 MW, while packs of more than 100 kWh give about 8 MW. That is ~~above~~ the 6.8 MW of an average mid-range conventional car.<ref name="NIPV-25" /> ''Introduction to Fire Safety Engineering'' gives a peak of 8.5 MW for a car and 2.5 MW for an average sofa.<ref name="VR-4.5" />		The heat release rate, the amount of heat a fire releases per second, largely determines how intense a fire is.<ref name="VR-4.5">Van Rossum (2027), §4.5.</ref> A review of published tests found similar peaks of about 6 MW for electric and fossil-fuelled cars. The total heat released was 4.7 to 8.5 GJ for electric cars and 3.3 to 10 GJ for the others.<ref name="NIPV-25" /> Larger batteries raise the peak: packs of about 15 to 25 kWh gave about 4 to 6 MW, while packs of more than 100 kWh give about 8 MW. That is significantly higher than the approximately 6.8 MW of an average mid-range conventional car.<ref name="NIPV-25" /> ''Introduction to Fire Safety Engineering'' gives a peak of 8.5 MW for a car and 2.5 MW for an average sofa.<ref name="VR-4.5" />

	The course of the fire over time differs more. In a petrol or diesel car, the plastic tank can fail and suddenly release fuel that catches fire, so the temperature rises very rapidly in the first minutes. Battery cells first need time to warm up and give off gas. The temperature therefore rises evenly, at about 0.2 °C per minute, until thermal runaway makes it climb by more than 10 °C per minute.<ref name="NIPV-26" /> A normal passenger car fire generally burns for just over half an hour and peaks at about 5 MW after 10 to 15 minutes. An electric car left to burn without intervention burns out in about 90 minutes, and an extinguished one can keep re-igniting for several hours.<ref name="NIPV-26" />		The course of the fire over time differs more. In a petrol or diesel car, the plastic tank can fail and suddenly release fuel that catches fire, so the temperature rises very rapidly in the first minutes. Battery cells first need time to warm up and give off gas. The temperature therefore rises evenly, at about 0.2 °C per minute, until thermal runaway makes it climb by more than 10 °C per minute.<ref name="NIPV-26" /> A normal passenger car fire generally burns for just over half an hour and peaks at about 5 MW after 10 to 15 minutes. An electric car left to burn without intervention burns out in about 90 minutes, and an extinguished one can keep re-igniting for several hours.<ref name="NIPV-26" />
Line 34:		Line 34:
	The toxic hazard of a fire, the subject of [[fire toxicity]], depends on how quickly and how much toxic products are produced, and on how potent they are.<ref name="Purser">Purser (2002), p. 2-91.</ref> Burning battery packs release corrosive substances from the electrolyte and combustion products from the plastics and packaging. Besides methane, carbon monoxide, carbon dioxide and hydrogen, they release the toxic gases hydrogen fluoride and hydrogen chloride.<ref name="NIPV-26" /> The SFPE Handbook counts these acid gases among the most important irritants in fire smoke. Irritants can impair escape or incapacitate people through pain and breathing difficulty.<ref name="Purser" />		The toxic hazard of a fire, the subject of [[fire toxicity]], depends on how quickly and how much toxic products are produced, and on how potent they are.<ref name="Purser">Purser (2002), p. 2-91.</ref> Burning battery packs release corrosive substances from the electrolyte and combustion products from the plastics and packaging. Besides methane, carbon monoxide, carbon dioxide and hydrogen, they release the toxic gases hydrogen fluoride and hydrogen chloride.<ref name="NIPV-26" /> The SFPE Handbook counts these acid gases among the most important irritants in fire smoke. Irritants can impair escape or incapacitate people through pain and breathing difficulty.<ref name="Purser" />

	A study cited in the NIPV report found about 20 to 200 mg of hydrogen fluoride per watt-hour of battery capacity. A 100 kWh pack can thus emit 2 to 20 kg, depending on its charge. Comparable tests on electric and conventional cars found hardly any difference in other toxic gases, but the electric cars appeared to produce about 1.8 times more hydrogen fluoride.<ref name="NIPV-26" /> Hydrogen fluoride binds to other substances and smoke particles and settles on floors, walls and ceilings. In one study this more than halved its concentration in the smoke within 10 to 20 minutes, but made firefighting clothing more contaminated.<ref name="NIPV-26" /> ~~The~~ report requires the clothing to be decontaminated because of the toxic and corrosive gases, unlike ~~after~~ a conventional car fire.<ref name="NIPV-30" />		A study cited in the NIPV report found about 20 to 200 mg of hydrogen fluoride per watt-hour of battery capacity. A 100 kWh pack can thus emit 2 to 20 kg, depending on its charge. Comparable tests on electric and conventional cars found hardly any difference in other toxic gases, but the electric cars appeared to produce about 1.8 times more hydrogen fluoride.<ref name="NIPV-26" /> Hydrogen fluoride binds to other substances and smoke particles and settles on floors, walls and ceilings. In one study this more than halved its concentration in the smoke within 10 to 20 minutes, but made firefighting clothing more contaminated.<ref name="NIPV-26" /> According to the report, a fire attack requires the clothing to be decontaminated because of the toxic and corrosive gases, unlike an attack on a conventional car fire.<ref name="NIPV-30" />
	[[File:fse-lithium-ion-battery-fire-v2.svg\|thumb\|upright=1.5\|alt=Battery smoke travels through air and deposits on clothing, walls, floors and residue.\|Battery-fire contamination includes airborne and deposited hazards. Based on Van der Graaf et al. (2021/2023), pp. 26, 29–30, 39 and 47.]]		[[File:fse-lithium-ion-battery-fire-v2.svg\|thumb\|upright=1.5\|alt=Battery smoke travels through air and deposits on clothing, walls, floors and residue.\|Battery-fire contamination includes airborne and deposited hazards. Based on Van der Graaf et al. (2021/2023), pp. 26, 29–30, 39 and 47.]]

	For people escaping, the extra hydrogen fluoride changes little. The plastic parts of any burning car already create severely toxic smoke, and the report puts the effect of the extra hydrogen fluoride on escape safety at zero. Systemic health effects seem unlikely, although slight exposure may irritate the skin and severe exposure may cause chemical burns.<ref name="NIPV-30" /> Research cited in the report found that firefighters in protective equipment can attack ~~safely~~ with proper hygiene if their exposure to the smoke lasts no more than 20 to 30 minutes.<ref name="NIPV-26" /> The report sees no reason to assume that they need more protective equipment than at a fire in a conventional car.<ref name="NIPV-47">Van der Graaf et al. (2021/2023), p. 47.</ref>		For people escaping, the extra hydrogen fluoride changes little. The plastic parts of any burning car already create severely toxic smoke, and the report puts the effect of the extra hydrogen fluoride on escape safety at zero. Systemic health effects seem unlikely, although slight exposure may irritate the skin and severe exposure may cause chemical burns.<ref name="NIPV-30" /> Research cited in the report found that firefighters in protective equipment can carry out a "safe" attack with proper hygiene if their exposure to the smoke lasts no more than 20 to 30 minutes.<ref name="NIPV-26" /> The report sees no reason to assume that they need more protective equipment than at a fire in a conventional car.<ref name="NIPV-47">Van der Graaf et al. (2021/2023), p. 47.</ref>

	Smoke with extra hydrogen fluoride can spread from a car park through open doors into the escape routes and stairwells of buildings above.<ref name="NIPV-30" /> People who leave at once are exposed only briefly, but those who cannot escape or take longer are exposed for longer. For this reason the report advises against placing charging stations near pedestrian escape routes.<ref name="NIPV-39">Van der Graaf et al. (2021/2023), p. 39.</ref>		Smoke with extra hydrogen fluoride can spread from a car park through open doors into the escape routes and stairwells of buildings above.<ref name="NIPV-30" /> People who leave at once are exposed only briefly, but those who cannot escape or take longer are exposed for longer. For this reason the report advises against placing charging stations near pedestrian escape routes.<ref name="NIPV-39">Van der Graaf et al. (2021/2023), p. 39.</ref>
Line 50:		Line 50:
	Fixed sprinkler or water mist systems cannot extinguish a fire in an electric car, because the water droplets cannot enter the pack. They can, however, reduce the probability that the fire grows and spreads to nearby structures and vehicles.<ref name="NIPV-37">Van der Graaf et al. (2021/2023), p. 37.</ref> Covering the car in foam is not a useful method for the time being. The fire can re-ignite under the foam, because packs in thermal runaway can re-ignite even when little oxygen is available.<ref name="NIPV-47" />		Fixed sprinkler or water mist systems cannot extinguish a fire in an electric car, because the water droplets cannot enter the pack. They can, however, reduce the probability that the fire grows and spreads to nearby structures and vehicles.<ref name="NIPV-37">Van der Graaf et al. (2021/2023), p. 37.</ref> Covering the car in foam is not a useful method for the time being. The fire can re-ignite under the foam, because packs in thermal runaway can re-ignite even when little oxygen is available.<ref name="NIPV-47" />

	[[Fire classes]] sort fires by the type of fuel. In the US scheme of NFPA 10 (2002 edition), summarised in NUREG-1805, fires in energised electrical equipment form Class C, for which the agent must not conduct electricity~~. Fires in combustible metals, among them lithium, form Class D, on which water-based extinguishers should never be used~~.<ref name="NUREG-B">Iqbal and Salley (2004), pp. B-13~~, B-16~~.</ref> The NIPV report, which deals with lithium-ion packs in cars, ~~does not use these classes and~~ relies on water.<ref name="NIPV-30" /> Although electric cars carry up to about 400 V, with peaks of up to 800 V, the probability of electrocution is negligible when the fire is fought from a distance.<ref name="NIPV-11" /><ref name="NIPV-43">Van der Graaf et al. (2021/2023), p. 43.</ref>		[[Fire classes]] sort fires by the type of fuel. In the US scheme of NFPA 10 (2002 edition), summarised in NUREG-1805, fires in energised electrical equipment form Class C, for which the agent must not conduct electricity.<ref name="NUREG-B">Iqbal and Salley (2004), p. B-13.</ref> The NIPV report, which deals with lithium-ion packs in cars, relies on water for cooling and extinguishing them.<ref name="NIPV-30" /> Although electric cars carry up to about 400 V, with peaks of up to 800 V, the probability of electrocution is negligible when the fire is fought from a distance.<ref name="NIPV-11" /><ref name="NIPV-43">Van der Graaf et al. (2021/2023), p. 43.</ref>

	== Firefighting and salvage in car parks ==		== Firefighting and salvage in car parks ==