Skip to content
Upcoming courses

Trench effect

Tools
From Fire Safety Academy Wiki
Contents
  1. 1Origin of the term
  2. 2Orientation effects in flame spread
  3. 3Wind-aided and upward spread
  4. 4Effect of walls on flames
  5. 5Consequences for fire growth
  6. 6Prediction and limits
  7. 7See also
  8. 8Notes
  9. 9References

The trench effect is an orientation effect in flame spread that was identified in the London King's Cross fire, in which a fire spread up a wooden escalator.[1] The sidewalls of the escalator caused the flames to stay close to the steps, or to "hug" them, as the fire spread upward.[1] The term appears in the title of a 1992 experimental study of the King's Cross fire.[2] The King's Cross fire is also cited as an example of a fire that burned for some time before attempts were made to close the station.[3]

The handbook account that reports the effect attributes it to the sidewalls of the escalator, and none of the literature cited here explains its mechanism.[1] The later sections therefore describe general effects of surface orientation and of walls on flame spread, not an explanation of the trench effect itself. In general flame spread theory, a flame that lies along unburnt fuel heats a long stretch of it, so that the spread can accelerate. A flame that heats only a few millimetres ahead of its base spreads slowly.[4][5] Configuration effects on flame spread, such as spread in corners and on ceilings, have been studied, but they have not yet led to fully established calculation methods.[6]

Origin of the term

Quintiere describes the behaviour of the flames in the London King's Cross fire, which involved fire spread up a wooden escalator, as a notable orientation effect.[1] He attributes this behaviour to the sidewalls of the escalator, which caused the flames to hug the steps. For this observation he cites a 1992 paper in the Fire Safety Journal by Drysdale, Macmillan and Shilitto.[1][2] The term itself appears in the title of that paper, which presents the work as the experimental verification of the "trench effect" in the King's Cross fire.[2]

Side view of an inclined escalator between two sidewalls, with flames lying low along the steps, and a cross-section showing the trench shape.
The configuration of the King's Cross fire: fire spreading up the steps of a wooden escalator between its sidewalls, which caused the flames to hug the steps. Based on Quintiere (2002b), p. 2-254.

Purser refers to the same fire in a review of major fire disasters. A common feature of these disasters was a failure in early detection and effective warning.[3] In many cases the fire was found at an early stage, but there was a failure to tell people to leave while the fire was small. There was also a failure to realise that a small fire can grow with exponential speed into a life-threatening one.[3] In the King's Cross fire, the escalator fire burned for some time before attempts were made to close the station. Purser concludes that by the time people perceive a fire as a threat to their lives, it may already be too late to escape.[3]

Orientation effects in flame spread

Flame spread over a solid depends on the direction of the gas flow relative to the spread and on the orientation of the surface. A surface can be vertical or horizontal, and it can face upward or downward, and each of these cases changes the heat transfer from the flame to the fuel.[6] Spread against the flow is called opposed-flow flame spread, and spread in the same direction as the flow is called wind-aided spread.[6] The difference in speed is large. Lateral or downward spread on thick solids has a typical rate of about 0.1 cm/s, while upward spread on thick solids reaches 1 to 100 cm/s.[6]

Ito and Kashiwagi measured the heat flux from the flame to the surface ahead of a spreading front on polymethyl methacrylate (PMMA). They tested orientations from vertically upward to vertically downward.[5] For upward spread they found a nearly constant heat flux of about 25 kW/m² over a long distance ahead of the front. For downward spread the flux reached a maximum of about 70 kW/m², but only very close to the front.[1]

Orientation also governs fire growth in ordinary rooms. Introduction to Fire Safety Engineering notes that upward spread along a vertical surface transfers a relatively large amount of heat to the material above the burning part. That material is preheated, so the fire can spread faster.[7] The fire dynamics tools of the U.S. Nuclear Regulatory Commission give the example of a curtain. A pilot flame at the bottom causes rapid upward spread, while the same flame at the top gives slow downward spread and much slower fire growth.[8] In a standard fire propagation apparatus, upward propagation on PMMA was much faster than downward propagation and accelerated with time.[9]

Wind-aided and upward spread

Wind-aided spread can accelerate, so it appears more rapid than opposed-flow spread.[4] In this mode, the length of fuel that the flame heats is of the order of the flame length, typically 0.1 to 10 m. In opposed-flow spread it is only 1 to 3 mm.[4] Quintiere relates the flame length to the energy released by the burning area, so that a growing pyrolysis zone lengthens the flame and speeds up the spread. Depending on how quickly the burnt-out region follows, the spread rate can rise towards a limit or grow without bound.[10]

Two fuel strips: left, a flame heating only a few millimetres ahead; right, a flame lying along the fuel and heating a long stretch ahead.
Heated length ahead of a spreading flame. Opposed-flow spread heats about 1 to 3 mm of fuel; in wind-aided spread the heated length is of the order of the flame length, 0.1 to 10 m. Based on Quintiere (2002b), p. 2-251.

For upward flame spread on a vertical wall, the heat flux from a turbulent flame of moderate size to the fuel ahead of it is about 25 kW/m², nearly independent of the material.[5] Orloff and co-workers found that upward turbulent spread on thick PMMA grew almost exponentially. The speed reached roughly 0.5 cm/s after the flame had spread over about 1 m of the material.[5] Quintiere stresses that the outcome depends on the flame length and on burnout. The flame length is controlled by the configuration and the flow conditions, and the burnout by the material.[5]

Wind-aided spread has also been measured on ceiling- and floor-mounted samples under a forced flow. On such PMMA samples 0.3 m long, the spread rate was steady at a given flow speed and increased with the flow speed, over a range of 0.25 to 4.5 m/s.[5] At larger scales, radiation and buoyancy become important, especially for floor-mounted materials. Tests on PMMA 2.4 m and 9.8 m long, at flow speeds of 0.8 to 3.8 m/s, gave spread rates of 1 to 6 cm/s. Smaller samples spread at only about 0.01 to 0.3 cm/s.[5]

Log-scale range bars: lateral or downward near 0.1 cm/s, short samples 0.01 to 0.3, long floor samples 1 to 6, upward spread 1 to 100 cm/s.
Typical and measured flame spread rates. Spread with the flame lying along the fuel is faster, and on long floor-mounted samples under forced flow it reached 1 to 6 cm/s. Based on Quintiere (2002b), pp. 2-247 and 2-253.

Effect of walls on flames

In general, walls change the shape of a flame by limiting the air that it can draw in. According to the fire dynamics tools of the U.S. Nuclear Regulatory Commission, a fire close to a wall or in a corner has a restricted supply of entrained air. This restriction has a significant effect on fire growth and spread.[11] It lengthens the flames and raises the plume temperature at a given height, compared with a fire in the open.[11]

A fire in the open, against a wall and in a corner, with entrained air from all round, from half and from a quarter; the flames get longer.
Walls restrict the air entrained by a flame: a flame against a wall entrains from half of its perimeter, and walls and corners lengthen the flames. Based on Iqbal and Salley (2004), pp. 4-1 and 4-2.

A flame against a single wall can entrain air from only half of its perimeter.[11] A fire in a corner is more severe than a fire against one wall, because the two burning walls exchange radiation, although the physical processes that control the two cases are very similar.[12] The same report notes that a fuel burning close to a wall gives higher temperatures and higher flames, because combustion must take place over a greater distance.[8]

Consequences for fire growth

Fast spread over combustible surfaces can carry a room fire to flashover, the stage at which the whole room suddenly becomes involved in the fire.[13][14] Introduction to Fire Safety Engineering names a hot smoke layer above 500 °C as one condition for flashover. Such a layer radiates about 20 kW/m² to the floor, a level at which many materials pyrolyse rapidly.[13]

In Australia, the data sheets of the Fire Safety Verification Method (print version 1.1, 2020) group wall and ceiling linings by their time to flashover in the ISO 9705 room-corner test. Group 1 applies when flashover does not occur within the 20-minute test.[14] A Group 3 lining that reaches a heat release rate of 1 MW within 5 minutes in this test is given a fast design growth rate.[15] If it takes 5 minutes or more, a medium rate applies, and Group 4 linings are assigned an ultra-fast growth rate.[15] The room-corner test was developed as a realistic test for interior finish materials, and its results have been predicted from cone calorimeter data and flame spread equations.[1]

Prediction and limits

Quintiere's framework for wind-aided spread can be applied to a particular problem only when the flame heat flux and the flame length can be expressed for that configuration.[10] Attempts to include configuration effects, such as spread in vertical corners and on ceilings, have not led to fully established methods.[6] Quintiere concludes that flame spread in enclosures, mines, ducts and buildings presents a complex array of conditions. Flame spread on materials therefore has to be evaluated in the context of their use.[16]

Field models based on computational fluid dynamics (CFD) can describe fires in complex geometries and include a wide variety of physical processes.[17] Their ability to predict flame spread is limited, however. The technical guide of the Fire Dynamics Simulator notes that modelling the physical and chemical changes of real materials as they burn is still in its infancy.[18] In the validation of the same model for nuclear power plant applications, even a 10 cm grid was not necessarily fine enough to predict the heat release rate or the flame spread along cable trays. Coarser grids still gave reasonable predictions of quantities tied directly to the conservation of mass and energy, such as average temperatures and pressures.[19]

See also

Notes

  1. ↑ 1.0 1.1 1.2 1.3 1.4 1.5 1.6 Quintiere (2002b), p. 2-254.
  2. ↑ 2.0 2.1 2.2 Quintiere (2002b), p. 2-257.
  3. ↑ 3.0 3.1 3.2 3.3 Purser (2002), p. 2-142.
  4. ↑ 4.0 4.1 4.2 Quintiere (2002b), p. 2-251.
  5. ↑ 5.0 5.1 5.2 5.3 5.4 5.5 5.6 Quintiere (2002b), p. 2-253.
  6. ↑ 6.0 6.1 6.2 6.3 6.4 Quintiere (2002b), p. 2-247.
  7. ↑ Van Rossum (2027), §4.3.
  8. ↑ 8.0 8.1 Iqbal and Salley (2004), p. 6-3.
  9. ↑ Tewarson (2002), p. 3-89.
  10. ↑ 10.0 10.1 Quintiere (2002b), p. 2-252.
  11. ↑ 11.0 11.1 11.2 Iqbal and Salley (2004), p. 4-1.
  12. ↑ Iqbal and Salley (2004), p. 4-2.
  13. ↑ 13.0 13.1 Van Rossum (2027), §4.7.
  14. ↑ 14.0 14.1 Australian Building Codes Board (2020a), Data Sheet B4, p. 49.
  15. ↑ 15.0 15.1 Australian Building Codes Board (2020a), Data Sheet B4, p. 50.
  16. ↑ Quintiere (2002b), p. 2-256.
  17. ↑ McGrattan et al. (2026), p. 2.
  18. ↑ McGrattan et al. (2026), p. 1.
  19. ↑ McGrattan (2007), p. 5-2.

References

  • Australian Building Codes Board. (2020a). Data sheets: Fire Safety Verification Method (Handbook annex, Print version 1.1).
  • 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.
  • McGrattan, K. (2007). Verification and validation of selected fire models for nuclear power plant applications: Vol. 7. Fire Dynamics Simulator (NUREG-1824; EPRI 1011999). U.S. Nuclear Regulatory Commission, Office of Nuclear Regulatory Research; Electric Power Research Institute.
  • 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
  • Purser, D. A. (2002). Toxicity assessment of combustion products. 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. 2-83–2-171). National Fire Protection Association; Society of Fire Protection Engineers.
  • Quintiere, J. G. (2002b). Surface flame spread. 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. 2-246–2-257). National Fire Protection Association; Society of Fire Protection Engineers.
  • Tewarson, A. (2002). Generation of heat and chemical compounds in fires. 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. 3-82–3-161). National Fire Protection Association; Society of Fire Protection Engineers.
  • Van Rossum, E. (2027). Introduction to Fire Safety Engineering. Fire Safety Academy.
Retrieved from "https://fire-safety.academy/w/index.php?title=Trench_effect&oldid=2180"