Skip to content
Upcoming courses

Walking speed

Tools
From Fire Safety Academy Wiki
(Redirected from Movement speed)
Contents
  1. 1Typical values
  2. 2Effect of crowd density
    1. 2.1The speed-density relation
  3. 3Stairs
  4. 4People with reduced mobility
  5. 5Walking in smoke
  6. 6Use in design and limits
  7. 7See also
  8. 8Notes
  9. 9References

Walking speed is the speed at which people move along an escape route during an evacuation, measured along their line of travel in metres per second. Together with the capacity of doors and stairs, it determines the movement part of the evacuation time.[1] Introduction to Fire Safety Engineering gives about 1.25 m/s for unhindered walking on a level floor, at fewer than 1 person per m².[1] The SFPE Handbook gives the same average unrestricted speed for densities below about 0.5 persons/m².[2]

Walking speed is not a fixed property of a person. It falls as a crowd becomes denser, it is lower on stairs than on a level floor, and it differs widely between people with and without mobility impairments.[3][4] Smoke slows people down further, especially when it irritates the eyes.[5] In the World Trade Center towers in 2001, the median evacuee moved at only about 0.2 m/s along the stairs.[6] Engineers use walking speeds to calculate the travel part of the required safe egress time.[1] Design methods in Australia and New Zealand cap the horizontal speed at 1.2 m/s.[7][8]

Typical values

For a single person on a level floor without crowding, the sources give values between 1.2 and 1.25 m/s. Introduction to Fire Safety Engineering uses about 1.25 m/s for level, unhindered evacuation.[1] Proulx gives 1.25 m/s as the average unrestricted walking speed at densities below about 0.5 persons/m².[2] The Australian Fire Safety Verification Method (FSVM) data sheets note that walking speed varies with age and gender. They consider a typical unimpeded speed of 1.2 m/s a reasonable upper limit for adults.[8]

Simple hand calculations often use a round value. Introduction to Fire Safety Engineering allows 1 m/s for the walk from the last bottleneck to the place of safety, and its worked example uses 1 m/s throughout.[9] Proulx gives an example in which the first person walks 18 m to a door at 61 m per minute, which is about 1 m/s.[10] Evacuation software lets the user choose. The EXIT89 model used by NIST offered an "emergency" horizontal unimpeded speed of 1.36 m/s and a "normal" speed of 0.91 m/s.[11]

At low density, the person matters more than the crowd. Proulx notes that occupant characteristics such as age, limitations and grouping then decide the speed. A family group, for example, is likely to move at the speed of its slowest member, such as a child or an older person.[2][12] The FSVM data sheets add other limits: congestion, the dimensions of the route, poor visibility and fatigue.[13]

Effect of crowd density

When people walk close together, they slow down. Introduction to Fire Safety Engineering bases this on research by the London Transport Board in 1958, whose values are still used.[14] The capacity of an escape route keeps rising up to about 4.5 persons/m², because more of the available space is used even though each person walks more slowly. Above that density, capacity falls sharply and congestion develops.[14]

Nelson and Mowrer describe three regimes, derived from the work of Fruin, Pauls, and Predtechenskii and Milinskii.[3] Below about 0.54 persons/m², people move at their own pace, independent of others. Above about 3.8 persons/m², no movement takes place until the crowd has thinned. Between these limits, speed falls in a straight line as density rises.[3] Proulx gives similar limits, with a standstill at 4 to 5 persons/m², like a fairly crowded lift.[2]

The speed-density relation

The straight-line relation between speed and density is the basis of the hydraulic method for egress calculations.[3] The International Fire Engineering Guidelines describe this methodology as based on hydraulic flow and effective width.[15] It starts from a constant k for the type of route and subtracts a part that grows with the density:[3]

S=k−akD

Symbol Description Unit
S walking speed along the line of travel (on stairs, along the line of the treads) m/s
k speed constant for the route element: 1.40 for corridors, aisles, ramps and doorways; 1.00 to 1.23 for stairs m/s
a constant, 0.266 when speed is in m/s and density in persons/m² m²/person
D density of people on the route persons/m²

The relation is valid for densities between 0.54 and 3.8 persons/m².[3] At 0.54 persons/m², it gives about 1.2 m/s in a corridor, which is the maximum unimpeded speed of 1.19 m/s in the SFPE table.[3][16] At 3.8 persons/m², it gives zero. The same equation is prescribed in New Zealand by C/VM2 (second edition, 2025) and recommended in the Australian FSVM data sheets, both with k = 1.4 and a = 0.266.[7][8] C/VM2 adds that a calculated horizontal speed above 1.2 m/s must be taken as 1.2 m/s.[7]

Line chart of walking speed against crowd density: flat up to 0.54 persons per square metre, then falling to zero at 3.8, for a corridor and two stairs.
Walking speed against crowd density for a corridor and two stair geometries, from S = k − akD with a = 0.266. Based on Nelson and Mowrer (2002), p. 3-370.

Worked example. A corridor holds a crowd at a density of 2.0 persons/m². With k = 1.40 m/s and a = 0.266, the speed is 1.40 − 0.266 × 1.40 × 2.0 = 0.655 m/s, or about 0.66 m/s.[16] The travel time for a travel distance of 30 m is the distance divided by the speed. It is 30 / 0.655 = 45.8 s, rounded up to 46 s. At a density of 0.54 persons/m², the same corridor takes 25 s, so the crowd adds about 21 s over 30 m.[16]

Speed and density together set the flow along a route. The product of speed and density is the specific flow, the number of people passing per second per metre of effective width.[17] It peaks at 1.9 persons/m², where the speed in a corridor has fallen to about 0.69 m/s.[17][16]

Stairs

People move more slowly on stairs than on a level floor.[1] The distance travelled is measured along the "climb line", which runs along the front edges of the treads. Its length follows from the horizontal length of the stair and the height between floors.[1] Speed on a stair also falls as density rises, but it varies differently with density than on a level floor. The London Transport Board measured these descent speeds along the climb line.[14] Nelson and Mowrer give maximum speeds along the line of travel for four stair geometries, at densities below 0.54 persons/m²:[3]

Side view of a stair between two floors with a dashed line along the tread edges forming the hypotenuse of a right triangle.
The climb line along the front edges of the treads is the travel distance on a stair; its length follows from the horizontal length and the storey height. Based on Van Rossum (2027), §7.4.
Riser (mm) Tread (mm) k (m/s) Maximum speed (m/s)
191 254 1.00 0.85
178 279 1.08 0.95
165 305 1.16 1.00
165 330 1.23 1.05

Shallower stairs with deeper treads are faster. Within the range of the table, the speed varies roughly with the square root of the ratio of tread to riser. There are not enough data to say whether this holds outside the range.[3] The same values appear in Introduction to Fire Safety Engineering, in C/VM2 and in the FSVM data sheets.[1][18][19] Proulx states that relatively fit people at low density can average about 1.1 m/s down the stair slope, which is about 0.8 m/s horizontally.[12]

Real descents in tall buildings are much slower. In a hypothetical evacuation of a 15-storey office building, Proulx assumes a descent of 3.6 floors per minute, about 0.5 m/s along the stair slope.[20] In the north tower of the World Trade Center, the median occupant descended 1.2 floors per minute. With about 10 m of stair and landings per floor, this is about 0.2 m/s, including all stops for crowding, smoke, water and rest.[6] NIST concluded that the average surviving occupant descended at about half the slowest speed previously measured in non-emergency evacuations.[21]

People with reduced mobility

Average speeds hide large differences between people. Introduction to Fire Safety Engineering lists mean speeds on a level floor for several groups. A person with a walking frame or rollator moves at 0.58 m/s, in a manual wheelchair at 0.69 m/s and with a walking stick at 0.81 m/s.[1] In an electric wheelchair the speed is 0.89 m/s, on crutches 0.94 m/s, and unaided with limited mobility 0.95 m/s, against 1.25 m/s for healthy people.[1]

The SFPE Handbook reports measurements by Boyce, Shields and Silcock on people with disabilities who walked without assistance. Of 107 participants, the mean horizontal speed was 1.00 m/s, with a range from 0.10 to 1.77 m/s.[4] People using a walking frame or rollator averaged 0.57 m/s, and those without a locomotion disability 1.25 m/s.[4]

Horizontal bars of mean walking speed: 1.25 m/s without locomotion disability down to 0.57 m/s with a walking frame or rollator.
Mean horizontal walking speed of unassisted ambulant people by mobility aid, with the measured range. Based on Boyce et al. (1999), as cited in Bryan (2002), p. 3-337.

On stairs the differences were larger. People with a locomotion disability averaged 0.38 m/s going up and 0.33 m/s going down. Those without a disability averaged 0.70 m/s in both directions.[4][22] Some participants also needed rest periods and extra time to change direction.[23] These data are central to the evacuation of people with disabilities.

Other groups also move more slowly. For people under the influence of alcohol or drugs, the British Standard BS 9999 suggests a 25 % shorter maximum travel distance or a 25 % lower walking speed.[24] The FSVM data sheets state that the speed-density equation applies to the general population and needs further reduction for older people and people with restricted mobility.[8]

Walking in smoke

Smoke reduces walking speed in two ways: it blocks the view, and irritant smoke makes it hard to keep the eyes open. Jin tested people in a 20 m corridor filled with either a highly irritant white smoke from wood cribs or a less irritant black smoke from kerosene.[25] In non-irritant smoke, walking speed fell gradually as the smoke became denser. In irritant smoke, it fell very quickly over the same range, and subjects walked in a zigzag or step by step along the wall.[25]

Purser summarises these results in terms of optical density. In non-irritant smoke at an optical density of 0.5 per metre, walking speed fell from about 1.2 m/s to 0.3 m/s. People then behaved as if in total darkness, feeling their way along the walls.[5] In irritant smoke, speed dropped to that of darkness at an optical density of only 0.2 per metre.[5]

Smoke also affects whether people walk on at all. The FSVM data sheets put the visibility below which people generally start to slow down at about 3 m, for non-irritant smoke and light-reflecting objects.[26] Studies of building fires found that about 30 % of people turned back rather than continue through smoke. On average they turned back at a visibility of about 3 m.[5] This link between smoke and movement is the subject of visibility in smoke.

Jin also tested adults walking into thick smoke with radiant heat near the far end of the corridor. At extinction coefficients of about 0.9 per metre, 14 of the 31 subjects turned back before the furthest point.[27] For those who went on, walking speed fell in proportion to the distance from the entrance, and they reported feeling uneasy as they walked forward.[27]

Proulx reports that in Jensen's tests in dense smoke, people moved at about 0.2 to 0.4 m/s, whatever wayfinding system was used. Survivors of serious fires had moved on average only 10 m in heavy smoke, which at 0.2 m/s means 50 s of exposure.[28] The FSVM data sheets give reduction factors for irritant smoke, derived from Jin. At an extinction coefficient of 0.3 per metre the factor is 0.9, at 0.4 per metre it is 0.7, and at 0.5 per metre it is 0.3.[29] Below a visibility of 10 m with light-emitting signs, occupants are assumed to look for another route or a safer place and wait for help.[29]

Line chart: walking speed in irritant smoke falls from 1.1 m/s at extinction coefficient 0.2 to 0.3 m/s at 0.5 per metre.
Walking speed in irritant smoke against the extinction coefficient, as tabulated from Jin's experiments. Based on Australian Building Codes Board (2020a), p. 282.

Use in design and limits

In design calculations, the walking speed turns a travel distance into a travel time. C/VM2 in New Zealand and the FSVM data sheets in Australia both divide the travel distance by the speed from the speed-density equation.[7][19] Introduction to Fire Safety Engineering uses the same step in its eight-step hand calculation, together with the flow capacity of the governing door or stair.[9] Evacuation simulations go further. Their occupant profiles can include walking speed and mobility, and models such as Simulex and Pathfinder adjust each person's speed to the distance from other people.[30][31]

The hydraulic approach rests on simple assumptions. The International Fire Engineering Guidelines list them: everyone starts at the same moment, nobody interrupts the flow with individual decisions, and nearly everyone can keep up with the group.[15] Movement times calculated this way may therefore be shorter than real ones.[15]

The guidelines also advise basing the movement calculation on the design occupant group with the largest effect on the overall movement. The travel speed of that group is then used as the input to the model.[32] The IFEG suggest testing the result with significantly lower travel speeds as a sensitivity study.[33] Proulx warns against assuming that people behind a fast first walker will keep up a high flow, because they are rarely close behind.[10]

Data on walking in smoke are limited. A review by Fridolf and colleagues, reported in the FSVM data sheets, found no reliable and valid correlations for predicting walking speed in smoke.[34] Some measurements rest on small groups: the stair speeds of people without a disability in the study by Boyce and colleagues come from eight participants.[4][22] Real evacuations can also be far slower than design values, as the World Trade Center towers showed.[6]

See also

Notes

  1. ↑ 1.0 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 Van Rossum (2027), §7.4.
  2. ↑ 2.0 2.1 2.2 2.3 Proulx (2002), p. 3-346.
  3. ↑ 3.0 3.1 3.2 3.3 3.4 3.5 3.6 3.7 3.8 Nelson and Mowrer (2002), p. 3-370.
  4. ↑ 4.0 4.1 4.2 4.3 4.4 Boyce et al. (1999), as cited in Bryan (2002), p. 3-337.
  5. ↑ 5.0 5.1 5.2 5.3 Purser (2002), p. 2-118.
  6. ↑ 6.0 6.1 6.2 Averill et al. (2005), p. 129.
  7. ↑ 7.0 7.1 7.2 7.3 Ministry of Business, Innovation and Employment (2025b), p. 32.
  8. ↑ 8.0 8.1 8.2 8.3 Australian Building Codes Board (2020a), p. 278.
  9. ↑ 9.0 9.1 Van Rossum (2027), §7.5.
  10. ↑ 10.0 10.1 Proulx (2002), p. 3-363.
  11. ↑ Averill et al. (2005), p. 246.
  12. ↑ 12.0 12.1 Proulx (2002), p. 3-347.
  13. ↑ Australian Building Codes Board (2020a), p. 276.
  14. ↑ 14.0 14.1 14.2 London Transport Board (1958), as cited in Van Rossum (2027), §7.4.
  15. ↑ 15.0 15.1 15.2 National Research Council of Canada et al. (2005), pp. 2.8-3–2.8-4.
  16. ↑ 16.0 16.1 16.2 16.3 Nelson and Mowrer (2002), p. 3-370; values calculated with their Equation 3.
  17. ↑ 17.0 17.1 Nelson and Mowrer (2002), p. 3-371.
  18. ↑ Ministry of Business, Innovation and Employment (2025b), p. 33.
  19. ↑ 19.0 19.1 Australian Building Codes Board (2020a), p. 279.
  20. ↑ Proulx (2002), p. 3-355.
  21. ↑ National Institute of Standards and Technology (2005), p. 215.
  22. ↑ 22.0 22.1 Boyce et al. (1999), as cited in Bryan (2002), p. 3-338.
  23. ↑ Bryan (2002), p. 3-337.
  24. ↑ British Standards Institution (2017), as cited in Van Rossum (2027), §7.4.
  25. ↑ 25.0 25.1 Jin (2002), p. 2-44.
  26. ↑ Australian Building Codes Board (2020a), p. 281.
  27. ↑ 27.0 27.1 Jin (2002), p. 2-48.
  28. ↑ Jensen (1993), as cited in Proulx (2002), p. 3-362.
  29. ↑ 29.0 29.1 Australian Building Codes Board (2020a), p. 283.
  30. ↑ Van Rossum (2027), §10.1.
  31. ↑ Van Rossum (2027), §10.3.
  32. ↑ National Research Council of Canada et al. (2005), p. 2.8-4.
  33. ↑ National Research Council of Canada et al. (2005), p. 1.2-24.
  34. ↑ Fridolf et al. (2018), as cited in Australian Building Codes Board (2020a), p. 280.

References

  • Australian Building Codes Board. (2020a). Data sheets: Fire Safety Verification Method (Handbook annex, Print version 1.1).
  • Averill, J. D., Mileti, D. S., Peacock, R. D., Kuligowski, E. D., Groner, N., Proulx, G., Reneke, P. A., & Nelson, H. E. (2005). Occupant behavior, egress, and emergency communications (NIST NCSTAR 1-7). National Institute of Standards and Technology.
  • Bryan, J. L. (2002). Behavioral response to fire and smoke. 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. 3-315–3-341). National Fire Protection Association; Society of Fire Protection Engineers.
  • Jin, T. (2002). Visibility and human behavior in fire smoke. 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. 2-42–2-53). National Fire Protection Association; Society of Fire Protection Engineers.
  • 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).
  • 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.
  • Nelson, H. E., & Mowrer, F. W. (2002). Emergency movement. 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. 3-367–3-380). National Fire Protection Association; Society of Fire Protection Engineers.
  • Proulx, G. (2002). Movement of people: The evacuation timing. 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. 3-342–3-366). National Fire Protection Association; Society of Fire Protection Engineers.
  • 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.
  • Van Rossum, E. (2027). Introduction to Fire Safety Engineering. Fire Safety Academy.
Retrieved from "https://fire-safety.academy/w/index.php?title=Walking_speed&oldid=2607"