Skip to content
Upcoming courses

Travelling fire

Tools
From Fire Safety Academy Wiki
Revision as of 23:14, 1 November 2026 by Mark (talk | contribs) (New FSE wiki article: Travelling fire (written by Claude Opus 5.5))
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Contents
  1. 1Local fire, flashover or travelling fire
  2. 2How a travelling fire moves
  3. 3Worked example
  4. 4The World Trade Center fires
  5. 5Travelling fires in car parks
  6. 6Design approaches
  7. 7Limits and common misconceptions
  8. 8See also
  9. 9Notes
  10. 10References

A travelling fire is a fire that spreads progressively through a large enclosure instead of involving all combustible surfaces at the same time.[1] In small and many medium-sized rooms, a local fire that grows large enough leads to flashover, after which the whole room burns as a post-flashover fire.[2][3] In larger enclosures flashover may not occur, and the fire either stays local or travels through the space.[2] Each area then burns for a limited time until its fuel is consumed, while the fire moves on to fresh fuel.[4]

The concept matters in structural fire engineering, because common design fires such as the parametric fire curve assume uniform burning and one temperature throughout the compartment.[5] Documented examples include the office fires in the World Trade Center towers in 2001 and fires that pass from car to car in indoor car parks.[6][7] Calculation methods for travelling fires are still limited, so design guidance recommends checking a range of fire durations.[8]

Local fire, flashover or travelling fire

A fire in a room starts as a local fire of one object or a few objects. During this growth phase the fire is fuel-controlled, which means that its heat release rate depends on the burning fuel and not on the air supply.[3][9] In small and many medium-sized enclosures, a local fire that becomes large enough causes flashover. This rapid transition produces a fire that involves all exposed combustible surfaces at once, and the fire becomes fully developed.[2] After flashover the whole room is on fire, and the fire is usually ventilation-controlled.[3]

Larger enclosures can behave differently. The Data Sheets of the Australian Building Codes Board describe two other outcomes for them. Where the fuel consists of separate packages, as in the departure hall of a travel terminal, the largest package may be too small to ignite its neighbours or to cause flashover.[10] Where the fuel is more continuous, the fire may instead travel through the enclosure and burn one area after another.[2] For fire models, the SFPE engineering standard S.01 identifies flashover by an upper layer above 600 °C or a heat flux to the floor above 20 kW/m2.[11]

How a travelling fire moves

A travelling fire advances when heat from the burning area raises nearby fuel to ignition. Flames preheat the unburned material next to the burning part, so that flame spread continues over the fuel surface. In an enclosed space, heat accumulated in the hot gases is also radiated to other combustible objects in the room.[12] Between separate items, radiation can carry the fire across gaps. In car parks, for example, long-burning battery packs prolong the heat radiation between vehicles, so the fire keeps travelling for a long time.[7]

Ventilation also steers the fire. In the north tower of the World Trade Center, WTC 1, the fires were generally ventilation-limited. They burned and spread only as fast as fresh air became available, mostly through additional window breakage.[13] Experiments in deep enclosures with a single opening showed a related pattern. The fire first spread towards the opening and then, once the fuel near the opening was consumed, progressively towards the back of the enclosure.[14]

At any spot, a travelling fire burns only until the local fuel is gone. The fire load is the energy that is released when all materials burn, and its value per square metre of floor is the fire load density.[15] Energy equals heat release rate multiplied by time, and the heat release rate of a burning area equals the heat release rate per unit area multiplied by that area.[16][17] Combining these relations gives the local burning time:[16][17]

tb=qfq˙″

Symbol Description Unit
tb local burning time at one location s
qf fire load density per unit floor area kJ/m2
q˙″ heat release rate per unit area of the burning fuel kW/m2

Worked example

For offices, Introduction to Fire Safety Engineering gives a mean fire load density of 420 MJ/m2, taken from the Dutch national annex to EN 1991-1-2.[18] For an average building other than storage, it gives a heat release rate per unit area of 250 kW/m2 in a developed fire.[16] With 420 MJ/m2 equal to 420,000 kJ/m2, the local burning time is 420,000/250 = 1,680 s, or 28 minutes. In New Zealand, C/VM2 (second edition, 2025) uses a design fire load energy density of 800 MJ/m2 for offices.[19] With that value the local burning time becomes 3,200 s, or about 54 minutes when rounded up.

These values are of the same order as the observations in the World Trade Center. There, the combustibles at any given location were almost fully burned out in about 20 minutes.[4] Single office workstations reached their peak burning rate in about 10 minutes and burned for about half an hour in total.[20] The calculation also shows why a travelling fire can heat one structural member for a short time while the fire as a whole lasts much longer. The total duration depends on how long the fire needs to cross the whole floor, which the local burning time does not capture.[6]

The World Trade Center fires

The investigation of the World Trade Center towers by the US National Institute of Standards and Technology (NIST) documented fires moving across very large office floors. These fires followed aircraft impacts that spread about 10,000 gallons (about 38 m3) of jet fuel over several floors and removed fire protection from many structural members.[4] In WTC 1 the major fires were on the 93rd to 99th floors. They generally moved both clockwise and counterclockwise from the north to the south of the tenant spaces.[13] Fire activity was at first concentrated near the impact area in the north, then spread around the east and west faces, and was last seen concentrated in the south when the tower collapsed.[6] In the south tower, WTC 2, the fires moved far less and stayed in the east half of the floors, where broken windows supplied enough air.[13]

The reconstruction used the movement of the fires to estimate the fuel. Simulated fires encircled the WTC 1 floors in roughly the same time as the real fires when the combustible load was 20 kg/m2.[6] Higher loads made the simulated fires spread more slowly than the photographs showed. In the oxygen-limited floor interiors, the burn time was roughly proportional to the fuel mass.[6][20] By the time of collapse, the WTC 1 fires had travelled around the entire floor and most of the combustibles had already been consumed.[4]

The moving fires loaded the structure unevenly. In WTC 1 the floors on the north side sagged first and then contracted as they cooled once the fires had moved south.[21][22] The floors on the south side then sagged until the south perimeter columns bowed inward. By 10:23 a.m. the south exterior wall had bowed inward by as much as 55 in., about 1.4 m.[21] As the fires moved from the north to the south side of the core, significant creep strains weakened the south side of the core over time.[23]

The limited local burning time also limited the heating of protected steel. In the simulations, steel with intact insulation rarely exceeded 400 °C in WTC 1 and 500 °C in WTC 2.[4] A simulation of WTC 2 was extended by 2 hours with all windows broken. In it, the truss steel on the west side, where the insulation was undamaged, heated for about 40 minutes and then cooled rapidly as the combustibles were consumed.[4] On a typical floor, temperatures of 700 °C to 760 °C were reached over about 15% of the west floor area, for less than 10 minutes.[4]

Travelling fires in car parks

In the Netherlands, a report by the Netherlands Institute for Public Safety (NIPV) describes travelling car fires. These are fires in indoor car parks that pass from one vehicle to the next and so travel through the car park.[7] Electric vehicles are expected to make such fires more likely. Their battery packs can burn for a long time, which prolongs the heat radiation between vehicles, and explosions and flare flames add to the spread.[7] A cluster of electric vehicles can therefore lead to a travelling car fire. The report calls for research into the effects of such fires on the structure of a car park.[24][25]

The report compares several measures. Fire-resistant cladding at charging points helps when a single car burns for a long time, but not in a travelling car fire. For that scenario the whole structure would need greater fire resistance.[26] Fire compartments of up to 1,000 m2, based on the Dutch Building Decree 2012, or smaller ones can confine a travelling car fire. Smaller compartments, however, heat up faster and may reach flashover conditions.[27]

An automatic sprinkler system or water mist system can contain the fire to a few cars. The fire then stays localised and causes only a short-lived, limited thermal load on the structure.[26] In open car parks, weather makes sprinklers less suitable, even though they are the most obvious way to stop a fire from travelling from one car to the next.[28] For the fire service, suppression is most likely to succeed while only one car is burning, before the fire has turned into a travelling car fire.[29]

Design approaches

Parametric time–temperature relationships assume uniform burning and homogeneous conditions throughout the compartment, with all of the fire load consumed inside it.[5] The Annex A curves of EN 1991-1-2 are valid only for compartments with up to 500 m2 of floor area and up to 4 m in height.[30] For the time equivalence method of Annex F, the floor area should be below 100 m2 when there are no roof openings. The Data Sheets add that the uniform-temperature assumption has been questioned for enclosures much larger than 100 m2 or for long and narrow ones.[31]

Kirby and co-workers addressed this question with large fire tests. Their compartments were about 22.9 m long, 5.6 m wide and 2.75 m high, with openings in one 5.6 m end wall. Ceramic fibre linings kept heat losses low, so that the compartment simulated a slice of about 45 m through a building of unlimited length.[31] One finding was that the conversion factor for the thermal properties of the enclosure needed adjusting. Another was that reducing the compartment area by a quarter reduced the average equivalent fire resistance time by 25%, so calculations based on small compartments gave conservative answers for this method.[31]

Specific methods for travelling fires superpose the effects of a series of local, fuel-controlled fires near the structural element under study. Stern-Gottfried and Rein reviewed the literature and proposed a design methodology on this basis, using several simplifying assumptions, including that the fire is fuel-controlled.[32] Models of this kind are still limited, and so are the data for validating them. The Data Sheets therefore recommend a pragmatic approach: a sensitivity analysis over a range of fire durations.[8]

In the Australian Fire Safety Verification Method, a proposed design is compared with a reference building. It is then generally enough to show that the consequences of each scenario are less severe for the proposed design than for the reference building.[8] More detailed tools are also available. Computational fluid dynamics can give the thermal load on any type of structure in more detail than a zone model, and a finite element model can then calculate how the heat penetrates the members.[33]

Limits and common misconceptions

Design methods for travelling fires rest on simplifying assumptions, and few data are available to validate them.[8] For car parks with electric vehicles, the Dutch report notes that structures are generally designed for a normative fire development. The effect of an anomalous fire development on the structure still needs research.[26][25] The World Trade Center findings describe one event with unusual causes, and NIST limited them to fires similar to or less intense than those of 11 September 2001. They do not apply to an exposure with uniform high temperatures over most of a floor, or with high temperatures lasting longer than about 20 minutes at any location.[34]

A common misconception is that a travelling fire is always less severe than a fully developed fire, because only part of the floor burns at any one time. The deep-enclosure experiments, however, showed significantly higher exposures for elements near the opening than for elements at the rear.[14] In WTC 1 the moving fires first heated the floors on one side and later those on the other, while the core weakened over time.[21][23] The fire on a floor can also last far longer than the local burning time. At the moment of collapse the WTC fires were still vigorous, indicating that the unchecked fires could have burned for well over an hour.[20]

A second misconception is that more compartmentation is always better. Smaller fire compartments can confine a travelling car fire, but they heat up faster, may reach flashover and make the burning vehicle harder to find and fight. From a fire safety point of view, the NIPV report therefore prefers an open, clearly arranged car park with a sprinkler system to further division into compartments.[27]

See also

Notes

  1. ↑ Australian Building Codes Board (2020a), Data Sheet B1, p. 16.
  2. ↑ 2.0 2.1 2.2 2.3 Australian Building Codes Board (2020a), Data Sheet B1, p. 13.
  3. ↑ 3.0 3.1 3.2 Van Rossum (2027), §4.7.
  4. ↑ 4.0 4.1 4.2 4.3 4.4 4.5 4.6 National Institute of Standards and Technology (2005), p. 147.
  5. ↑ 5.0 5.1 Australian Building Codes Board (2020a), Data Sheet B1, p. 15.
  6. ↑ 6.0 6.1 6.2 6.3 6.4 Gann, Hamins, et al. (2005), p. xlvi.
  7. ↑ 7.0 7.1 7.2 7.3 Van der Graaf et al. (2021/2023), p. 29.
  8. ↑ 8.0 8.1 8.2 8.3 Australian Building Codes Board (2020a), Data Sheet B1, p. 17.
  9. ↑ Van Rossum (2027), §4.11.
  10. ↑ Australian Building Codes Board (2020a), Data Sheet B1, pp. 12–13.
  11. ↑ Society of Fire Protection Engineers (2011), as cited in Australian Building Codes Board (2020a), Data Sheet B1, p. 12.
  12. ↑ Van Rossum (2027), §4.3.
  13. ↑ 13.0 13.1 13.2 Gann, Hamins, et al. (2005), p. l.
  14. ↑ 14.0 14.1 Thomas and Bennetts (2000) and Thomas et al. (2005), as cited in Australian Building Codes Board (2020a), Data Sheet B1, p. 17.
  15. ↑ Van Rossum (2027), §4.4.
  16. ↑ 16.0 16.1 16.2 Van Rossum (2027), §4.5.
  17. ↑ 17.0 17.1 Van Rossum (2027), §4.6.
  18. ↑ Van Rossum (2027), §8.5.
  19. ↑ Ministry of Business, Innovation and Employment (2025b), p. 22.
  20. ↑ 20.0 20.1 20.2 Gann, Hamins, et al. (2005), p. li.
  21. ↑ 21.0 21.1 21.2 National Institute of Standards and Technology (2005), p. 32.
  22. ↑ National Institute of Standards and Technology (2005), p. 150.
  23. ↑ 23.0 23.1 National Institute of Standards and Technology (2005), p. 185.
  24. ↑ Van der Graaf et al. (2021/2023), p. 42.
  25. ↑ 25.0 25.1 Van der Graaf et al. (2021/2023), p. 55.
  26. ↑ 26.0 26.1 26.2 Van der Graaf et al. (2021/2023), p. 36.
  27. ↑ 27.0 27.1 Van der Graaf et al. (2021/2023), p. 35.
  28. ↑ Van der Graaf et al. (2021/2023), p. 49.
  29. ↑ Van der Graaf et al. (2021/2023), p. 47.
  30. ↑ Australian Building Codes Board (2020a), Data Sheet B3, p. 38.
  31. ↑ 31.0 31.1 31.2 Australian Building Codes Board (2020a), Data Sheet B3, p. 36.
  32. ↑ Stern-Gottfried and Rein (2012), as cited in Australian Building Codes Board (2020a), Data Sheet B1, pp. 16–17.
  33. ↑ Van Rossum (2027), §11.2.
  34. ↑ National Institute of Standards and Technology (2005), p. 149.

References

  • Australian Building Codes Board. (2020a). Data sheets: Fire Safety Verification Method (Handbook annex, Print version 1.1).
  • Gann, R. G., Hamins, A., McGrattan, K. B., Mulholland, G. W., Nelson, H. E., Ohlemiller, T. J., Pitts, W. M., & Prasad, K. R. (2005). Reconstruction of the fires in the World Trade Center towers (NIST NCSTAR 1-5). National Institute of Standards and Technology.
  • Ministry of Business, Innovation and Employment. (2025b). Verification Method C/VM2: Framework for fire safety design (2nd ed.).
  • National Institute of Standards and Technology. (2005). Final report on the collapse of the World Trade Center towers (NIST NCSTAR 1).
  • Van der Graaf, J., Hessels, T., & Rosmuller, N. (2023). Fire safety of indoor car parks accommodating electrically powered vehicles. Nederlands Instituut Publieke Veiligheid. (Original work published 2021)
  • Van Rossum, E. (2027). Introduction to Fire Safety Engineering. Fire Safety Academy.
Retrieved from "https://fire-safety.academy/w/index.php?title=Travelling_fire&oldid=115"