Skip to content
Upcoming courses

Tunnel fire

Tools
From Fire Safety Academy Wiki
Revision as of 23:47, 1 November 2026 by Daan (talk | contribs) (New FSE wiki article: Tunnel fire (written by Claude Opus 5.5))
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Contents
  1. 1Why tunnels differ from buildings
  2. 2Design fires
  3. 3Smoke control and critical velocity
  4. 4Fire exposure of the tunnel structure
  5. 5Suppression and detection
  6. 6Modelling
  7. 7Limitations
  8. 8See also
  9. 9Notes
  10. 10References

A tunnel fire is a fire in a road, rail or service tunnel, usually involving one or more vehicles or the installations in the tunnel. Because a tunnel is enclosed, the smoke of the fire stays around the people inside it, and their ability to move away from the fire is restricted.[1] Temperatures can become far higher than in an ordinary room fire: in the Summit rail tunnel fire on British Railways in 1984, the faces of brick-lined ventilation shafts fused.[2]

In the Netherlands, tunnels do not fit well into the building regulations, which are written mainly for buildings in which people stay. Fire safety engineering is therefore applied to tunnels on several aspects, and tunnels are usually fitted with a smoke management system.[3] Tunnel ventilation can keep part of a tunnel free of smoke by blowing air against the smoke, but the same airflow brings oxygen to the fire.[4] In fire tests, the peak heat release rate of vehicles ranged from about 1.5 MW for an older car to about 120 MW for a loaded truck.[5][6]

Why tunnels differ from buildings

Bukowski describes the special hazards of rail tunnels in his chapter on transportation vehicles. In an accident in a tunnel, the fire effluent stays around the train and keeps exposing passengers and crew who have left the vehicles. Tunnels, bridges and elevated track also restrict how far people can move away from the train, which must be addressed for long tunnels.[1] Some fire scenarios are a threat to passengers only when they occur in a tunnel, because the tunnel increases the exposure to smoke.[7]

Most fire scenarios in trains lead to short exposures, mainly to heat. The exceptions are sleeping cars without fire detection and incidents in a long tunnel. For this reason, the operating procedure is that a train in which an emergency occurs keeps moving until it is out of the tunnel, if possible.[8] Trains are long and narrow spaces, so some assumptions of zone models may not be appropriate for them.[1]

The temperatures in a tunnel fire can be extreme. A post-flashover compartment fire typically reaches about 1,100 °C, according to Drysdale as cited by Iqbal and Salley, but temperatures of 1,300 to 1,400 °C occasionally occur.[9] Such temperatures are enough to fuse the surface of bricks, as happened in the ventilation shafts of the Summit tunnel.[2] For comparison, the cellulosic fire curves used for ordinary fire resistance tests reach a maximum of just over 900 °C.[10]

Regulation treats tunnels as special structures. The book Introduction to Fire Safety Engineering lists tunnels among the examples of buildings for which the Dutch building regulations are not effective. Other examples are public transport terminals and towers with a floor above 200 m. For such buildings, fire safety engineering is used.[11]

In New Zealand, Verification Method C/VM2 (second edition, 2025) names tunnels as an example of buildings outside its scope. Their compliance can be shown with an alternative solution.[12] The International Fire Engineering Guidelines (2005) were written for buildings, but state that their concepts may also help in the design of ships and tunnels.[13] In the United States, road tunnels are covered by NFPA 502, listed in its 2001 edition in NUREG-1805.[14]

Design fires

The design fire of a tunnel is usually a vehicle fire. The book gives a peak heat release rate of 8.5 MW for a car, and an energy content of 6,000 to 12,000 MJ per car and 19,000 MJ for an unloaded truck.[15][16] Measured peaks for single cars vary widely. Tests at the Fire Research Station in the United Kingdom gave 8.5 MW for an Austin Maestro and 4.5 MW for a Citroën BX. Finnish tests of cars from the late 1970s gave about 1.5 to 2.0 MW.[5]

Larger vehicles release much more heat. In a European research programme, nearly two dozen vehicles were burnt, including cars, buses, trams, trucks and rail vehicles. Many of the peak heat release rates lay between about 10 and 35 MW. One test involved a truck loaded with 2 tonnes of modern upholstered furniture, which reached about 120 MW.[5][6] That is about 14 times the peak of a car fire of 8.5 MW. Truck tyres can ignite from an overheated axle and release a substantial amount of heat.[5]

In the Netherlands, the car fire curve in the standard NEN 6098 for mechanically ventilated car parks is based on a TNO design fire from 1999. In those tests, fire could spread to neighbouring cars 10 to 15 min after the first car, and the curve assumes three burning cars after 15 min.[17] Modern cars, with their high plastic content, have a larger fire load and a higher heat release rate than older steel models.[18] The book therefore advises checking whether such curves still fit current vehicles.[17]

An electric vehicle fire has a similar peak to a conventional car fire, about 6 MW according to a review by Sun and co-workers. Vehicles with batteries of more than 100 kWh reach about 8 MW.[19] Without intervention, an electric car burns out in about 90 min, and after extinguishing the battery can keep reigniting for several hours.[20] A battery pack of 100 kWh can also release 2 to 20 kg of hydrogen fluoride, a toxic gas.[21]

Smoke control and critical velocity

Smoke can be stopped from flowing through an opening, such as a doorway or a corridor, by an airflow against it.[4] One criterion for the design air velocity is that no smoke flows back during the evacuation.[22] If the airflow is too weak, smoke flows back against it, a phenomenon known as backlayering.[23] The lowest air velocity that prevents this is the critical velocity. Thomas developed an empirical relation for this velocity in a corridor, which Klote gives in a simplified form for normal air properties:[4]

Vk=0.292(Q˙W)1/3

Symbol Description Unit
Vk critical air velocity to prevent smoke backflow m/s
Q˙ heat release rate of the fire kW
W width of the corridor m

The coefficient 0.292 follows from an air density of 1.3 kg/m³, a specific heat of 1.005 kJ/(kg·K), a downstream temperature of 27 °C and a constant of 1. The results are approximate, because the constant is known only approximately, but they show the size of the air velocity needed for different fires.[4] The relation was developed for corridors. Klote's design chart covers widths of 1 to 4 m and fires of 500 to 2,000 kW, far below the tens of megawatts of a truck fire.[24] It is not appropriate for sprinklered fires in which the smoke is barely warmer than the air.[4] For tunnels it therefore illustrates the principle rather than giving design values.

Klote's worked example is a wastebasket fire of 150 kW in a corridor 1.22 m wide. The relation gives 0.292 × (150 / 1.22)1/3 = 0.292 × 4.97 = 1.45 m/s.[24] Doubling the heat release rate raises the critical velocity by only 26 %, because of the cube root. Within the range of the chart, a fire of 2,000 kW in a corridor 4 m wide needs 2.32 m/s. The cube root explains why larger fires need only moderately higher air speeds.

The airflow has a side effect: it brings oxygen to the fire. One cubic metre of air per second at 21 °C contains enough oxygen for a fire of 3,660 kW. In the worked example, the airflow of 4.3 m³/s that stops the smoke could support a fire of 15,800 kW, more than 100 times the wastebasket fire, so Klote advises extreme caution.[4][24] Strong ventilation also affects suppression and detection, as an example from a cable tunnel in the next section shows.[25]

The same principle is used in some Dutch car parks. Displacement ventilation creates a high air speed over the whole cross-section, for example 1.5 m/s according to NEN 6098, which can easily mean 40 air changes per hour. It gives firefighters a clear view of the fire, but it works only in tunnel-shaped car parks and usually starts only after people have escaped.[26]

Fire exposure of the tunnel structure

A tunnel lining must resist fires that are more severe than the standard fire of a building test. The book names two less common temperature–time curves that are more severe than the standard fire curve. One is the Rijkswaterstaat curve for tunnels, known as the RWS fire curve; the other is the hydrocarbon curve, which simulates an oil or gas fire.[17] Fire curves are agreed temperature histories used to compare tested constructions; they do not reproduce the heat development of a real fire.[17] Most furnace tests simulate a cellulosic fire, which grows more slowly than a hydrocarbon fire. In the United States, exposure ratings for hydrocarbon fires are given as H ratings under UL 1709.[27][28]

Concrete linings can suffer from spalling, in which layers or pieces of concrete break off the heated surface, sometimes violently. Spalling is caused by a build-up of pore pressure from water vapour during heating, although some researchers attribute it to restrained thermal expansion near the surface. Extensive spalling can lead to early loss of stability and integrity, because it exposes the reinforcement.[29]

The risk grows with a high moisture content, rapid heating, slender members and high stresses during the fire. Most members of normal-weight concrete withstand severe fires without spalling, but high-strength concrete of 50 to 120 MPa is more susceptible.[30] Design methods for concrete assume that the concrete stays intact, which no longer holds if the cover spalls off.[30]

A common remedy is to add polypropylene fibres to the concrete. The fibres melt at about 170 °C and leave channels through which steam can escape. Kodur and Harmathy give about 0.1 to 0.25 % by volume as the amount needed.[31] The International Fire Engineering Guidelines, citing Kodur, give 0.15 to 0.3 % without stating whether this is by volume or by mass.[32]

Suppression and detection

A water mist system uses fine sprays in which no drop is larger than 1.0 mm.[33] Test programmes for water mist have been carried out for high-voltage cable tunnels and for railway tunnels, including full-scale suppression tests for Eurotunnel and an onboard system for its shuttle trains for heavy goods vehicles.[34] In Germany, VdS has test protocols for local application systems in electric cable tunnels.[25]

A cable tunnel of 9 km in Auckland shows the practical problems. Opening all nozzles along the tunnel at once would need very large water mains and pumps, and costly pumping out afterwards. The chosen solution released water in zones upstream and downstream of the fire.[25] The ventilation air of 5 m/s would carry the heat along the tunnel, away from the nozzles closest to the fire, so individually heat-activated nozzles would not work. Turbulence under normal ventilation also made smoke and heat detection difficult. The detection system and the hydraulic system therefore had to be tested together.[25]

Vehicles with lithium-ion batteries raise further questions. A sprinkler or water mist system cannot extinguish a fire in an electric car, because the water droplets cannot reach the cells inside the closed battery pack. It can, however, reduce the chance that the fire grows and spreads to other vehicles and the structure.[35] For firefighting with hoses, the NIPV report gives two rules of thumb. A high-pressure jet of 125 L/min is generally sufficient for fires of up to 2.5 MW, and a low-pressure jet of 450 L/min for fires of up to 10 MW.[36] A truck fire of 120 MW is far beyond these values.[6]

Modelling

Computational fluid dynamics can be used to study how effective the smoke management system of a tunnel is.[3] The book warns that the fire and its heat release rate must be prescribed by the modeller and that CFD is mainly reliable for qualitative comparisons of situations.[37] Cox and Kumar describe the King's Cross underground station fire, where CFD predicted that the flames would lie down in the trench of the escalator instead of rising. Physical tests confirmed this, and the authors stress that simple calculations should always shadow such simulations.[38]

Tunnels are numerically demanding. The Fire Dynamics Simulator notes that its standard pressure solver is excessively slow for tunnels made of dozens of computational meshes placed end to end, and offers an optional method for such geometries.[39] Its documentation also refers to work that couples a three-dimensional model of the fire region with a one-dimensional network model of the ventilation of a tunnel.[40] An older network program, the Subway Environment Simulation from the early 1970s, is used to size and locate ventilation shafts, evaluate tunnel geometry and fan size, and model heat and smoke from fires.[41]

Limitations

The critical velocity relation was developed for corridors and covers fires up to 2,000 kW. The vehicle data come from a limited number of tests, and the book names the RWS fire curve without giving its values.[24][5][17] Fire curves describe a test exposure, not a real fire, and CFD results depend on the prescribed fire.[17][37] A common misconception is that a tunnel fire can be treated as a room fire. The long geometry, the limited escape routes and the effect of ventilation on both smoke and fire make it a different problem.[1][4]

See also

Notes

  1. ↑ 1.0 1.1 1.2 1.3 Bukowski (2002), p. 5-229.
  2. ↑ 2.0 2.1 Department of Transport (1984), as cited in Iqbal and Salley (2004), p. 13-7.
  3. ↑ 3.0 3.1 Van Rossum (2027), §11.2.
  4. ↑ 4.0 4.1 4.2 4.3 4.4 4.5 4.6 Klote (2002), p. 4-279.
  5. ↑ 5.0 5.1 5.2 5.3 5.4 Babrauskas (2002b), p. 3-27.
  6. ↑ 6.0 6.1 6.2 Babrauskas (2002b), p. 3-28.
  7. ↑ Bukowski (2002), p. 5-231.
  8. ↑ Bukowski (2002), p. 5-230.
  9. ↑ Drysdale (1998), as cited in Iqbal and Salley (2004), p. 13-7.
  10. ↑ Iqbal and Salley (2004), p. F-13.
  11. ↑ Van Rossum (2027), §1.5.
  12. ↑ Ministry of Business, Innovation and Employment (2025b), p. 6.
  13. ↑ National Research Council of Canada et al. (2005), Part 0 (Australia), p. 0.1-3.
  14. ↑ Iqbal and Salley (2004), p. E-8.
  15. ↑ Van Rossum (2027), §4.5.
  16. ↑ Van Rossum (2027), §4.4.
  17. ↑ 17.0 17.1 17.2 17.3 17.4 17.5 Van Rossum (2027), §8.6.
  18. ↑ Van der Graaf et al. (2021/2023), p. 11.
  19. ↑ Sun et al. (2020), as cited in Van der Graaf et al. (2021/2023), p. 25.
  20. ↑ NFPA (2013) and Sun et al. (2020), as cited in Van der Graaf et al. (2021/2023), p. 26.
  21. ↑ Larsson et al. (2017), as cited in Van der Graaf et al. (2021/2023), p. 26.
  22. ↑ Klote (2002), p. 4-286.
  23. ↑ Klote (2002), p. 4-278.
  24. ↑ 24.0 24.1 24.2 24.3 Klote (2002), p. 4-280.
  25. ↑ 25.0 25.1 25.2 25.3 Mawhinney and Back (2002), p. 4-328.
  26. ↑ Van der Graaf et al. (2021/2023), p. 38.
  27. ↑ Iqbal and Salley (2004), p. C-8.
  28. ↑ Iqbal and Salley (2004), p. C-9.
  29. ↑ Kodur and Harmathy (2002), p. 1-163.
  30. ↑ 30.0 30.1 Fleischmann and Buchanan (2002), p. 4-255.
  31. ↑ Kodur and Harmathy (2002), p. 1-171.
  32. ↑ Kodur (1997), as cited in National Research Council of Canada et al. (2005), p. 2.6-9.
  33. ↑ Mawhinney and Back (2002), p. 4-311.
  34. ↑ Mawhinney and Back (2002), p. 4-325.
  35. ↑ Van der Graaf et al. (2021/2023), p. 37.
  36. ↑ Van der Graaf et al. (2021/2023), p. 45.
  37. ↑ 37.0 37.1 Van Rossum (2027), §11.4.
  38. ↑ Cox and Kumar (2002), p. 3-207.
  39. ↑ McGrattan et al. (2026), p. 43.
  40. ↑ McGrattan et al. (2026), p. 111.
  41. ↑ Iqbal and Salley (2004), p. B-83.

References

  • Babrauskas, V. (2002b). Heat release rates. In P. J. DiNenno, D. Drysdale, C. L. Beyler, W. D. Walton, R. L. P. Custer, J. R. Hall Jr., & J. M. Watts Jr. (Eds.), SFPE handbook of fire protection engineering (3rd ed., pp. 3-1–3-37). National Fire Protection Association; Society of Fire Protection Engineers.
  • Bukowski, R. W. (2002). Fire hazard assessment for transportation vehicles. In P. J. DiNenno, D. Drysdale, C. L. Beyler, W. D. Walton, R. L. P. Custer, J. R. Hall Jr., & J. M. Watts Jr. (Eds.), SFPE handbook of fire protection engineering (3rd ed., pp. 5-227–5-233). National Fire Protection Association; Society of Fire Protection Engineers.
  • Cox, G., & Kumar, S. (2002). Modeling enclosure fires using CFD. In P. J. DiNenno, D. Drysdale, C. L. Beyler, W. D. Walton, R. L. P. Custer, J. R. Hall Jr., & J. M. Watts Jr. (Eds.), SFPE handbook of fire protection engineering (3rd ed., pp. 3-194–3-218). National Fire Protection Association; Society of Fire Protection Engineers.
  • Fleischmann, C., & Buchanan, A. (2002). Analytical methods for determining fire resistance of concrete members. In P. J. DiNenno, D. Drysdale, C. L. Beyler, W. D. Walton, R. L. P. Custer, J. R. Hall Jr., & J. M. Watts Jr. (Eds.), SFPE handbook of fire protection engineering (3rd ed., pp. 4-239–4-256). National Fire Protection Association; Society of Fire Protection Engineers.
  • Iqbal, N., & Salley, M. H. (2004). Fire Dynamics Tools (FDTs): Quantitative fire hazard analysis methods for the U.S. Nuclear Regulatory Commission Fire Protection Inspection Program (NUREG-1805). U.S. Nuclear Regulatory Commission, Office of Nuclear Reactor Regulation.
  • Klote, J. H. (2002). Smoke control. In P. J. DiNenno, D. Drysdale, C. L. Beyler, W. D. Walton, R. L. P. Custer, J. R. Hall Jr., & J. M. Watts Jr. (Eds.), SFPE handbook of fire protection engineering (3rd ed., pp. 4-274–4-291). National Fire Protection Association; Society of Fire Protection Engineers.
  • Kodur, V. K. R., & Harmathy, T. Z. (2002). Properties of building materials. In P. J. DiNenno, D. Drysdale, C. L. Beyler, W. D. Walton, R. L. P. Custer, J. R. Hall Jr., & J. M. Watts Jr. (Eds.), SFPE handbook of fire protection engineering (3rd ed., pp. 1-155–1-181). National Fire Protection Association; Society of Fire Protection Engineers.
  • Mawhinney, J. R., & Back, G. G., III. (2002). Water mist fire suppression systems. In P. J. DiNenno, D. Drysdale, C. L. Beyler, W. D. Walton, R. L. P. Custer, J. R. Hall Jr., & J. M. Watts Jr. (Eds.), SFPE handbook of fire protection engineering (3rd ed., pp. 4-311–4-337). National Fire Protection Association; Society of Fire Protection Engineers.
  • McGrattan, K., Hostikka, S., Floyd, J., McDermott, R., Vanella, M., Mueller, E., & Paul, C. (2026). Fire Dynamics Simulator technical reference guide: Vol. 1. Mathematical model (6th ed., NIST Special Publication 1018-1). National Institute of Standards and Technology. https://doi.org/10.6028/NIST.SP.1018
  • Ministry of Business, Innovation and Employment. (2025b). Verification Method C/VM2: Framework for fire safety design (2nd ed.).
  • National Research Council of Canada, International Code Council, Department of Building and Housing, & Australian Building Codes Board. (2005). International fire engineering guidelines (2005 ed.). Australian Building Codes Board.
  • 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=Tunnel_fire&oldid=462"